Description
The forest frontier – the next area of intact forest to be degraded or converted to a non-forest land use – is the front line of global deforestation. Deforestation of the world’s intact tropical forests is driven by many factors.
Carbon Finance for Reduced Emissions from
Deforestation & Degradation at the Forest Frontier
Financial Analysis of Alternate Land Uses in the Amazon,
Congo and Papua, Indonesia
Authors:
Marisa Meizlish
1
Dan Spethmann
2
Michael Barbara
3
New Forests
Sydney, Australia and Washington, D.C.
Paper Presented at UNFCCC COP13 and Forest Day
Dec. 3-14, 2007
Bali, Indonesia
1
Manager, Advisory Services, [email protected]
2
Manager, Investment Programs, [email protected]
3
Analyst, Environmental Markets, [email protected]
2
New Forests
New Forests (www.newforests.com.au) is a forestry investment management and
advisory services firm. The company’s investment thesis is unique in seeking
investment opportunities that deliver traditional timber returns, as well as returns from
emerging environmental markets and preferences, such as certified timber, carbon,
biodiversity and water. In addition to world-leading forestry investment expertise in
the areas of acquisitions, modeling, operations, silviculture and ecosystem services,
New Forests is at the forefront of developing and commercializing environmental
products and developing opportunities in emerging environmental markets. The
company is based in Sydney, Australia, with offices in Washington, D.C. and San
Francisco. New Forests holds an Australian Financial Services License (license
#301556).
New Forests Pty Limited
The Zenith Center, Tower A
Level 19, Suite 1905
821 Pacific Highway
Chatswood 2067
Australia
+61-2-9406-4100
New Forests Inc.
1025 Connecticut Ave, NW
Suite 1206
Washington DC 20036
United States
+1-202-715-1700
###
Copyright
This document may be copied for individual use but approval is required for multiple
copies and distribution. An electronic copy is available at www.newforests.com.au.
For citations, please use:
Meizlish, M., Spethmann, D. and Barbara, M. (2007) “Carbon Finance for Reduced
Emissions from Deforestation & Degradation at the Forest Frontier” New Forests Inc.,
Washington, D.C.
3
Table of Contents
Executive Summary ....................................................................................................... 4
1 Introduction ................................................................................................................. 5
2 Background on Drivers of Deforestation at the Forest Frontier ................................. 6
2.1 The Brazilian Amazon ................................................................................... 6
2.2 Papua, Indonesia .......................................................................................... 10
2.3 Democratic Republic of Congo ................................................................... 12
3 Methodology and Assumptions ........................................................................... 14
3.1 Framework for Analysis .............................................................................. 14
3.2 Definition of Baseline Activities ................................................................. 15
3.3 Discount Rates ............................................................................................. 16
3.4 Net Present Value of Baseline Activities ..................................................... 17
3.5 Carbon Quantification .................................................................................. 20
3.6 Carbon Value ............................................................................................... 21
4 Discussion ............................................................................................................ 23
4.1 Land Use Comparisons ................................................................................ 24
4.2 Carbon Price Scenarios ................................................................................ 25
4.3 Policy Uncertainty ....................................................................................... 26
4.4 Government Fees ......................................................................................... 27
5 Conclusions .......................................................................................................... 28
References .................................................................................................................... 31
Appendix A – Production cost schedule for natural tropical forest concession .......... 35
Appendix B – Oil Palm Model Assumptions .............................................................. 39
List of Figures
Figure 1 – World’s remaining areas of intact rainforest ................................................ 6
Figure 2 – Nine states of the Brazilian Amazon and patterns of deforestation ............. 7
Figure 3 – Deforestation in Rodonia, Brazil from 2000 to 2006 .................................. 8
Figure 4 – Papua Province, Indonesia .......................................................................... 10
Figure 5 – Congo Basin spanning five central African nations * ................................ 12
List of Tables
Table 1 – Carbon trading scenarios and prices ............................................................ 22
Table 2 – NPV per hectare of baseline and carbon scenarios ...................................... 23
Table 3 – Carbon scenarios applying a 35% discount rate .......................................... 26
Table 4 – Carbon project NPVs assuming 50% chance of accreditation .................... 27
4
Executive Summary
Background on
Carbon Investment
& Forest
Conservation
Deforestation in tropical rainforests contributes approximately
20% of global greenhouse gas emissions annually. Options
for expanding the inclusion of forest conservation in global
carbon markets are currently being explored. Policies that
support a price mechanism for reducing emissions from
deforestation and degradation (REDD) could lead to
significant new investment in forest conservation. A key
question in designing such market policies is how a carbon
price for REDD would impact land use dynamics, in particular
whether carbon investment may compete effectively with
large-scale activities that are driving deforestation, such as
commercial agribusiness expansion.
Scope of Financial
Analysis
The objective of this study is to evaluate the viability of
carbon finance as a mechanism for forest conservation. This
is explored via a generalized framework that compares the net
present value (NPV) of carbon-financed forest conservation
with the value of land uses resulting in forest degradation and
deforestation. The study conducts this analysis from the
perspective of a private investor making investment decisions
on land use options for existing forested areas.
Current Investment
Scenarios
This analysis is conducted across all of the world’s largest
remaining areas of tropical rainforest. The major drivers of
deforestation in each region are the baseline scenarios against
which conservation is considered. These are:
• Cattle ranching in the Brazilian Amazon
• Oil palm plantations and logging in Papua, Indonesia
• Logging in theDemocratic Republic of Congo
Methodology to
Assess Carbon
Investment
Scenarios
The NPV of the cashflows from these activities are compared
to the NPV of carbon revenue generated from a potential
REDD regime. A range of carbon prices and crediting periods
is applied to consider sensitivities to these variables. The
impact of policy uncertainty is also discussed.
Results Results indicate that carbon-financed forest conservation is
generally competitive with current land uses at the forest
frontier. More frequent crediting periods and the ability to sell
some credits into a voluntary market in the short term improve
the ability of carbon-finance scenarios to compete with
baseline activities. Policy uncertainty is an important variable
that can undermine the ability for carbon investment to
compete with alternate land uses. Further analysis of the cost
structure associated with carbon investments and refinement at
a more localized scale can build on the generalized framework
developed here to inform decision making in greater detail.
5
1 Introduction
The forest frontier – the next area of intact forest to be degraded or converted to a
non-forest land use – is the front line of global deforestation. Deforestation of the
world’s intact tropical forests is driven by many factors. These include the recent
price increases in global agribusiness commodities, such as soybeans, palm oil and
cattle, and increasing global timber demand. At the same time as the forest frontier is
being pushed deeper into previously intact forest areas, the rapid growth in carbon
markets is creating new avenues to support forest conservation and rehabilitation.
This paper assesses whether conservation through carbon finance can compete with
land uses currently driving large-scale deforestation at the forest frontier. Focusing on
the rainforest regions of the Brazilian Amazon, Papua, Indonesia, and the Democratic
Republic of Congo, we have sought to develop a generalized financial model of a
range of land uses and carbon market scenarios. The focus of this study is on whether
carbon investors may compete effectively with the current large-scale actors in
deforestation, such as commercial agribusiness ventures. Ultimately, if the price of a
forest for carbon conservation is greater than the soil expectation value for other uses,
then forest conservation management will be attractive to investors over other land
uses.
Private investment flows that currently finance alternate land uses, such as converting
forests to oil palm, could shift toward investment in forest conservation if risk
adjusted returns are competitive. Consequently, the paper uses a Net Present Value
(NPV) methodology to compare the relative attractiveness of forest conservation
versus agribusiness from the perspective of a private investor. This analysis is
conducted under the assumption that credits generated by reducing emissions from
deforestation or degradation (REDD) are eligible in an operational global carbon
market. However, a range of crediting scenarios and carbon credit prices are
considered to show sensitivity to these factors. A discussion of discount rates and
policy uncertainty is also included.
Our findings indicate that there is a case to be made for investment in carbon-financed
conservation. The generalized approach developed here suggests that forest
conservation can compete with land uses driving deforestation, if policies can provide
market certainty around credit eligibility. This paper is intended to serve as a
launching point for more detailed economic analysis and debate around forest
conservation carbon policy. In particular, further analysis of the cost structure
associated with carbon investments and refinement at a more localized scale can build
on the generalized framework developed here to inform decision making in greater
detail.
Section 2 of this paper provides background on the land use dynamics of the three
study regions. Modelling methodologies and assumptions and a discussion of results
are presented in the following sections. The paper concludes with a discussion of
policy considerations that could generate large-scale investment in carbon-financed
forest conservation.
6
2 Background on Drivers of Deforestation at the
Forest Frontier
The three areas considered in this study are the Brazilian Amazon, the Indonesian
province of Papua and the Congo Basin in the Democratic Republic of Congo (DRC).
These areas represent the largest remaining areas of intact tropical rainforest in the
world (Figure 1). This section briefly describes regional land use dynamics as context
for the financial analysis in the following sections. The main drivers of deforestation
in each region are identified to define the baseline, or “business as usual,” activities
that are most likely to occur in the frontier regions.
Figure 1 – World’s remaining areas of intact rainforest
4
2.1 The Brazilian Amazon
The largest area of remaining intact forest is the Amazon Basin, which covers 5.5
million square kilometers across nine countries in northwest South America.
Approximately 60% of the Amazon rainforest is located in Brazil, which is home to
19% of the world’s remaining intact forests.
5
Around 67 million hectares – 17% of the original Amazon area – have been converted
to other land uses since the 1970s.
6
Seventy-six percent of deforestation occurs in the
4
Source: Mongabay.com http://rainforests.mongabay.com/0102.htm
5
Brazilian Vegetable Oil Industry Association (2006) “Understanding the Moratorium: Responsible
Production” http://www.abiove.com.br/english/sustent/ms_cprodutor_24jul07_us.pdf
6
Ibid
7
“arc of deforestation” from the Atlantic Ocean through southern Para and Mato
Grosso to Acre,
7
which is the frontier area considered in this study (Figure 2). It is
estimated that current trends in agribusiness expansion could convert 40% of the
Amazon rainforest by 2050.
8
This would result in the release of 32 billion tonnes of
carbon to the atmosphere, equivalent to approximately 4 years of global annual
greenhouse gas emissions.
9
Figure 2 – Nine states of the Brazilian Amazon and patterns of deforestation
10
The annual rate of loss of primary forest in Brazil from 2000-2005 was 0.8%,
representing a total loss of 17.3 million hectares over these 5 years. This rate is a 35%
increase over the rate of primary forest loss recorded from 1990-2000.
11
Two of the
largest deforestation years on record were 2002 and 2003 (23,000 square kilometers
7
Margulis, S. (2004) “Causes of Deforestation of the Brazilian Amazon” World Bank Working Paper
No. 22, Washington, D.C.
7
Walker, R. and Moran, E. (2000) “Deforestation and Cattle Ranching in the Brazilian Amazon:
External Capital and Household Processes” World Development Vol. 28, No. 4, p. 683-699
8
Soares-Filho, B. et al (2006) “Modelling Conservation in the Amazon Basin” Nature Vol. 440, March
9
Ibid
10
Global Forest Watch (2007) “Human Pressure on the Brazilian Amazon Forests” (online)
http://www.globalforestwatch.org/english/interactive.maps/Brazil_Datasets.htm#Study_area
11
FAO (2005) “FAO Forest Resource Assessment – Global Tables”
http://www.fao.org/forestry/site/28679/en/
8
and 24,000 square kilometers per year, respectively
12
), and an all-time high of 27,400
square kilometers was cleared in 2004, 50% above the long-term average.
13
Deforestation rates tend to be closely correlated with prices for cattle, timber and
crops.
14
With the decline of soy and beef prices in 2005 and 2006 and the
strengthening of the Real against the US dollar, deforestation rates slowed to 18,800
and 13,100 square kilometers per annum, respectively
15
.
A rise in annual deforestation is expected, as prices increase for these commodities in
response to global growth in demand.
16
This reactive variability suggests that land
practices could change quickly through the introduction of a price signal for carbon-
financed forest conservation. Figure 3 gives one snapshot of how quickly the
conversion of the frontier is advancing.
Figure 3 – Deforestation in Rodonia, Brazil from 2000 to 2006
17
12
Woods Hole Research Center (2007a) “Agriculture Frontier Explosion in Brazil”
http://www.whrc.org/southamerica/agric_expans.htm
13
Nepstad, D., Moutinho, P. & Soares-Filho, B. (2005) “The Amazon in a Changing Climate: Large-
Scale Reductions of Carbon Emissions from Deforestation and Forest Impoverishment” Amazon
Institute for Environmental Research), Woods Hole Research Center (WHRC) and Federal University
of Minas Gerais paper released during COP 12, Nairobi
14
Mongabay.com (2007) “South American development plan could destroy the Amazon”
http://news.mongabay.com/2007/1003-ci_amazon.html
15
Ibid
16
Martino, D. (2007) “Deforestation in the Amazon: Pressures and Outlook” Third World Resurgence
No. 200, April http://www.southdevelopment.org/environment/MartinoAmazonDeforestation.pdf
17
NASA Earth Observatory
http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=17600
9
The most serious threat to the Amazon forests and the main driver of deforestation is
cattle ranching.
18
Government subsidies and tax incentives granted to large-scale
ranchers in the 1970s and 1980s spurred the growth of the industry,
19
and the herd
size nearly doubled from 26.2 million head of cattle in 1991 to 51.6 million in 2001.
20
It is estimated that as much as 88% of deforested areas are occupied by ranches,
21
and
the industry continues to grow.
22
In general, cattle ranching is not the highest economic land use when compared with
other regimes, such as soy or logging. However, landowners tend to favor ranching
because it provides a relatively liquid asset that can provide annual cash-flow and is
also relatively low risk. Converting land to a cattle ranch is also a way of
guaranteeing land possession and securing property rights.
23
Soybean production, which grew approximately 60% between 1998 and 2002,
24
is
contributing to this expansion by moving onto old ranching lands in the southern arc
of deforestation, pushing cattle further north into untouched areas. Soybean
cultivation is becoming more competitive as soy prices rise, driven by a growing
demand for bio-diesel and global food production, and as transportation costs
decrease through increased river usage and expanded road networks.
25
An additional driver of deforestation is selective logging, which occurs over an
average area of 15,000 square kilometers a year.
26
Recent studies indicate that rather
than contributing to deforestation by providing access to areas that are subsequently
converted to agricultural use, selective logging appears to be a driver of deforestation
on its own and is accretive to deforestation caused by conversion for cattle and soy.
27
However, 96% of the logging in the Amazon occurs in old and intermediate frontiers
(10 –30 years since first selective harvest) with the remaining 4% occurring in new
frontiers, principally Novo Progresso and northern Mato Grosso.
28
There are multiple land uses at the Brazilian forest frontier and the land use dynamics
are not fully understood.
29
However, conversion to pasture for cattle ranching is the
largest driver and the land use studied here as the baseline scenario in Brazil.
18
Fearnside, P. (2007) “Deforestation in Amazonia” Encyclopedia of Earth Eds. Cutler J . Cleveland,
Washington, D.C.: Environmental Information Coalition, National Council for Science and the
Environment http://www.eoearth.org/article/Deforestation_in_Amazonia; Margulis, S. (2004); Woods
Hole Research Center (2007)
19
Walker, R. and Moran, E. (2000)
20
Woods Hole Research Center (2007)
21
Margulis, S. (2004)
22
Walker, R. and Moran, E. (2000)
23
Margulis, S. (2004)
24
Woods Hole Research Center (2007)
25
Ibid
26
Foley et al (2007) “Amazonia Revealed: Forest Degradation and the loss of ecosystem goods and
services in the Amazon Basin” Frontiers in Ecological Environment, Issue 5(1), p. 25-32
27
Foley et al (2007) citing Nepstad, D. et al (1999) “Large scale impoverishment of Amazonian forests
by logging and fire” Nature Issue 398, April 8, p. 505-508 and Anser, G. et al (2005) “Selective
Logging in the Brazilian Amazon” Science Issue 310, p. 480-82
28
Woods Hole Research Center (2007b) “Logging and Family Forests”
http://www.whrc.org/southamerica/logging_fam_forests.htm
29
Foley et al (2007)
10
2.2 Papua, Indonesia
Papua is the largest province in Indonesia and accounts for almost a quarter of
Indonesia’s land area (Figure 4). The percentage of forest cover in Papua and West
Papua (now a separate province, shown in pink in Figure 4) was 85% in 1985
30
with
satellite imaging analyses reporting an estimated reduction to approximately 73% by
2005.
31
The rate of loss of primary forest in Indonesia from 2000-2005 was 2.6% per
annum totalling a loss of 7.2 million hectares over the five-year period.
32
This rate
was a 26% increase over the period from 1990-2000. Data specific to Papua is
difficult to obtain, but the rate of loss of primary forest for neighbouring Papua New
Guinea has been 0.95% annually from 1990-2005. This is a loss of 250,000 hectares
per year.
33
Figure 4 – Papua Province, Indonesia
Timber companies are increasingly turning their attention to Papua’s natural forests,
and large scale timber extraction concessions have led to significant clearing of the
province’s primary forests.
34
Forest concessions allowing harvesting of natural
30
Global Forest Watch et al (2001) “The State of the Forest: Indonesia” Global Forest Watch, World
resources Institute and Forest watch Indonesia www.globalforestwatch.org
31
Sumantri, H. and Wijayanto, I. (2005) “Land cover mapping using satellite imagery and GIS in the
Mamberamo Basin and Raja Ampat Islands, Papua, Indonesia” www.mapindia.org/2005/papers
32
FAO (2005)
33
Ibid
34
Tokede, M., William, D., Widodo, Gandhi, Y., Imburi, C., Patriahadi, Marwa, J ., Yufuai, M. (2005)
“The impacts of Special Autonomy in Papua’s Forestry sector: Empowering Customary Communities
(Masyarakat Adat) in decentralised Forestry Development in Manokwari District” Centre for
International Forestry Research, Bogor, Indonesia
11
forests have been granted over 6.5 million hectares of the province’s 42 million
hectares.
35
While concession requirements include preparing environmental impact
assessments and forest management and monitoring plans, logging practices typically
result in over-exploitation and loss of the forest structure.
36
Inventories of harvested
areas in Papua found that 58% of the remaining trees with diameters of 20 cm or
larger were damaged, compromising the ability of the forest to recover and provide
sustainable yields of timber.
37
In addition to these pressures, illegal logging activities
are increasing, particularly as forest resources diminish in the Indonesian provinces of
Kalimantan, Sulawesi and Sumatra.
38
Increasing international demand for green energy sources has led to a demand for new
land to produce palm oil, a feedstock for bio-diesel. Most of the global expansion of
the oil palm plantation estate has occurred in Indonesia and Malaysia.
39
Predictions
are that about half of the new plantation land – 3 out of 6 million hectares – needed to
supply the global palm oil market by 2020 will be established in Indonesia.
40
National government plans for Papua to absorb much of this expansion would dwarf
the province’s current plantation estate of 58,000 hectares (as of 2002).
41
The
provincial government is actively pursuing alternate investment strategies related to
carbon credit generation to avoid this massive land conversion.
42
In Papua New
Guinea, the oil palm plantation estate doubled between 1990 and 2000 to 73,000
hectares, and this growth is expected to accelerate.
43
Irrespective of whether this land use change occurs in Papua or elsewhere in
Indonesia, it would represent a significant loss of frontier forest in the Asia Pacific
region. The land use pressures in Papua are representative of challenges faced across
the region’s remaining tropical forests. Unsustainable logging and oil palm plantation
establishment are the two activities studied here as the baseline scenarios for Papua.
While regional variables will influence the financial viability of logging and palm oil
development across the region, the models for Papua can generally be applied
elsewhere.
35
Papua Ministry of Forestry (2006) List of Forest Concession Right (Hak Pengusahaan Hutan) and
Timber Forest Product Utilisation Permit (Izin Usaha Pemanfaatan Hasil Hutan Kayu) holders, August
36
J arvis and J acobsen (2006) “Working paper – Incentives to promote forest certification in Indonesia”
Project: Motivating Sustainability, International Finance Corporation
37
Tokede et al (2005)
38
Ibid
39
Wilkinson, M. (2007) “Green Fuel Gets a Black Name” Sydney Morning Herald October 13; Painter,
J . (2007) “Losing Land to Palm Oil in Kalimantan” BBC News, August 3
40
WWF Germany (2002) “Oil palm plantations and deforestation in Indonesia: What role to Europe
and Germany Play?” http://www.fire.uni-freiburg.de/se_asia/projects/wwf_oil.htm
41
Friends of the Earth (2005) “Greasy Palms. The social and ecological impacts of large-scale oil palm
plantation development in southeast Asia”
http://www.foe.co.uk/resource/reports/greasy_palms_impacts.pdf
42
Tedjasukmana, J . (2007) “Heros of the Environment: Leaders & Visionaries” Time Magazine
http://www.time.com/time/specials/2007/article/0,28804,1663317_1663319_1669895,00.html
43
Ibid
12
2.3 Democratic Republic of Congo
The Congo Basin is the world’s second largest tropical rainforest covering 2 million
square kilometers across the central African nations of the Democratic Republic of
Congo (DRC), the Central African Republic, Cameroon, the Republic of Congo,
Equatorial Guinea and Gabon (Figure 5). Fifty-three percent of the rainforest is
within the DRC, the area of focus in this study.
Figure 5 – Congo Basin spanning five central African nations
44
*
* Encircled areas are priority landscapes based on high biodiversity value and intact forest
Compared to many of the world’s tropical rainforests, the Congo Basin is relatively
healthy.
45
This is primarily a result of the low population density and limited
commercial activity. The annual rate of deforestation from 2000-2005 was 0.24%,
which actually represented a 37% decrease from the period of 1990-2000.
46
This may
be a result of decreased economic activity during Congo’s civil war from 1998-2003,
and the country is likely to open to global markets as political stability improves.
Nearly 60% of the DRC’s total forest area is thought to
be productive or commercially
valuable,
47
and logging is currently the main commercial activity contributing to
deforestation.
48
Approximately 20 million hectares of the DRC’s 145 million hectare
forest estate are allocated as timber concessions to about 60 companies, although only
44
Congo Basin Forest Partnership (2005) “The Forests of the Congo Basin: A Preliminary
Assessment” (online) http://carpe.umd.edu/products/PDF_Files/FOCB_APrelimAssess.pdf
45
Congo Basin Forest Partnership (2005)
46
FAO (2005)
47
Congo Basin Forest Partnership (2005)
48
Bavier, J . (2007) “Congo to Cancel Logging Deals to Protect Forests: Reuters, J une 21; Congo Basin
Forestry Partnership (2005)
13
about a dozen are in operation.
49
Three main companies – Siforco (owned by
Germany’s Danzer Group), Sodefor (a Portuguese-owned unit of holding company
NST) and Safbois (an American and Belgian owned conglomerate) – account for over
two-thirds of the country’s current timber production capacity.
50
The government is currently carrying out a World Bank-led legal review of 156
logging licenses covering 21 million hectares, half of which are in intact forest
landscapes. Around three million hectares of illegal logging concessions have already
been cancelled, and the government claims it is willing to cancel an additional 12-15
million hectares.
51
The outcome of these processes and the enforcement of
government policies remain uncertain.
Currently, logging in the DRC is largely driven by demand from Europe for certified
sustainable timber
52
(e.g. Forest Stewardship Council certification). The general
practice in the region is selective logging of high-value species for export. One to two
trees are removed per hectare (ha)
53
with an average of 75-80% of the canopy cover
remaining intact.
54
Felling cycles (in a properly managed tropical forest) rarely
exceed 30–40 years.
55
Estimates of production range from 5-6 cubic meters (m
3
) /ha
56
to 8-12m
3
/ha
57
.
Economic drivers such as increased timber demand from Asia, which is more
accepting of non-certified wood, and increasing population pressures suggest that
there could be an increase in unsustainable logging practices in the near term. For
example, the ITTO reported in late 2006 that China has replaced Europe as the price-
setter for African timber.
58
With only a dozen of the 60 companies holding logging
concessions currently in operation,
59
there is scope for increased production capacity
as non-active concession holders become operational or relinquish licenses that are
acquired by operational companies.
60
In nearby Ivory Coast, Ghana and Uganda,
49
Forests Monitor (2007) “The Timber Sector in the DRC: A Brief Overview”
http://www.forestsmonitor.org/uploads/2e90368e95c9fb4f82d3d562fea6ed8d/Description_of_the_Tim
ber_Sector_in_the_DRC.pdf
50
Bavier, J . (2007)
51
Ibid
52
Forests Monitor (2007)
53
Karsenty, A., and Gourlet-Fleury, S. (2006) “Assessing sustainability of logging practices in the
Congo Basin’s managed forests: the issue of commercial species recovery. Ecology and Society Issue
11(1), p. 26 (online) http://www.ecologyandsociety.org/vol11/iss1/art26/
54
Nasi, R. (2005) “Potential Methodological Flaw in the Examination of the Effects of Logging”
Ecology and Society Issue 10(2) (online) http://www.ecologyandsociety.org/vol10/iss2/resp2/
55
Karsenty, A., and Gourlet-Fleury, S. (2006)
56
Ruiz-Perez, M. et al (2005) “Logging in the Congo Basin: A Multi-Country Characterization of
Timber Companies” Forest Ecology and Management Volume 214, Issues 1-3
57
Karsenty, A. and Gourley-Fleury, S. (2006)
58
ITTO (2006) “Asia Takes Price Leadership from Europe in Africa” Tropical Forest Update Volume
16, No. 4 http://www.itto.or.jp/live/Live_Server/3163/tfu.2006.04.e.pdf
59
Forests Monitor (2007)
60
Greenpeace suggests that this could be an outcome of the World Bank led concession review
currently underway. (See “Carving Up the Congo”
http://www.greenpeace.org/raw/content/international/press/reports/carving-up-the-congo-exec.pdf)
14
favorable economic markets for both timber and agricultural products, such as cacao,
also contributed to the deforestation of old growth forests in those countries.
61
Should Congo’s forests come under greater threat of degradation from increased
logging as the country opens to global markets, a price signal for carbon conservation
could help the country develop land uses that incorporate a full range of forest values.
Studying logging as the main driver of deforestation in this region is therefore more
forward looking and hypothetical than the other regions, but it provides an early
assessment of how carbon finance could mitigate the rapid expansion of logging
activities, as market demand for timber inevitably increases.
3 Methodology and Assumptions
This study evaluates the viability of carbon finance as a mechanism for forest
conservation via a comparison of the net present value (NPV) of carbon-financed
forest conservation versus the land uses driving deforestation. This analysis is
considered from the perspective of private investors making investment decisions on
land use options for existing forested areas. This section describes the methodology
used to conduct the analysis and outlines assumptions underlying the financial
modelling.
3.1 Framework for Analysis
There is no accepted regulated market standard for defining and accrediting avoided
deforestation credits. Many ideas are currently being discussed for what is being
referred to as reduced emissions for deforestation and degradation (REDD). A
position favoured by the UNFCCC, World Bank and others takes a national
accounting approach in which countries establish a baseline deforestation scenario
and receive credit for reducing emissions below the baseline.
62
Other accounting
approaches have taken a more project-level perspective or combined national and
project-level mechanisms (the “nested approach”
63
).
The second major question is whether credits from REDD will be fungible with other
carbon offsets. Some proponents suggest a completely fungible market. A dual
markets approach has also been proposed in which avoided deforestation credits are
sold in a parallel market but are not fungible with emission reduction offsets as
defined under markets of the Kyoto Protocol.
64
61
Laporte et al (2004)
62
World Bank Carbon Finance Unit (2007) “Forest Carbon Partnership Facility”
http://carbonfinance.org/Router.cfm?Page=FCPF&FID=34267&ItemID=34267
63
Pedroni, L. and Streck, C. (2007) “Mobilizing Public and Private Resources for the Protection of
Tropical Rainforests” CATIE and Climate Focus
http://www.climatefocus.com/newspubs/downloads/PolicyBriefonREDD_000.pdf
64
Centre for Clean Air Policy (2007) “Reducing Emissions from Deforestation and Degradation: The
Dual Markets Approach” http://www.ccap.org/international/FINAL%20REDD%20report.pdf
15
Defining baseline scenarios is fundamental in all of these policy debates and is the
fundamental first step in evaluating activities that are seeking carbon finance.
65
A
baseline scenario explains the likely land use in the absence of the emission reduction
activities. Avoided deforestation as a general project type assumes that the baseline
activity results in forest degradation or conversion to a non-forest activity. The
emissions “avoided” by pursuing a forest conservation management strategy rather
than the baseline activities represent the volume of saleable carbon.
For this study, a framework for analysis was designed to compare the NPV of baseline
activities to the conservation scenario. This framework includes the following steps:
• specify baseline activities for each region
• determine the NPV of the cashflows of the baseline activity
• estimate carbon stock loss associated with the baseline activity
• quantify the NPV of maintaining carbon stock by forgoing the baseline
activity
3.2 Definition of Baseline Activities
As discussed in Section 2, the baseline scenarios considered in this study are:
• Cattle ranching in the Amazon
• Unsustainable logging in Papua
• Oil palm plantation development in Papua
• Unsustainable logging in the DRC
Modelling is based on a hypothetical frontier area of 250,000 hectares. This
approximately represents the size of logging concessions in Papua and the Congo but
is much larger than individual landholdings for ranching in the Amazon and oil palm
plantations. This scale is used to approximate land use dynamics across the frontier,
whether that is driven by a single landowner or the cumulative impact of multiple
landowners.
The entire concession area is assumed to be impacted over a 20-year period. This
represents a deforestation rate of 5% per annum, totaling a conversion of 12,500
hectares annually. This rate is reasonable in terms of the operational capabilities of
the land uses in the baseline scenarios.
Ideally, each scenario would apply a deforestation rate based on the country’s
historical and projected rates of deforestation. Although many data points could be
used, the best method for determining these baseline loss rates is an important point of
65
See, for example, the Voluntary Carbon Standard “Guidance for Agriculture, Forestry and Other
Land Use Projects” www.v-c-s.org/docs/AFOLU%20Guidance%20Document.pdf and California
Climate Action Registry “Forest Sector Protocol”
http://www.climateregistry.org/docs/PROTOCOLS/Forestry/Forest_Sector_Protocol_Version_2.1_Sep
t2007.pdf
16
debate in policy discussions around the accreditation of REDD activities. FAO
deforestation statistics presented for each region in Section 2 suggest lower rates
(0.8% in Brazil, 2.6% in Indonesia and 0.24% in Congo), but this data is aggregated
nationally and underestimates rates at the forest frontier. Rates of deforestation also
appear low when considered over a large intact forest land base, such as those in the
Brazilian Amazon and the Congo Basin. A rate of 5% is therefore generalized and is
applied as a simplified approach, which would need further investigation for specific
areas.
Definitions of baseline activities are as follows:
Cattle Ranching Forested land in the hypothetical 250,000 hectare study area is
converted to pasture at a rate of 12,500 hectares per annum. Once a hectare is
converted, it is assumed to become productive in the following year and remains
productive through the project life (to 2040). It is assumed that the net cashflows per
hectare over the project life incorporate start-up costs and operating costs. Any initial
revenue from the liquidation timber sales upon conversion is assumed to be included
in the average net cashflows per hectare over the project life, although in many cases
trees are burned onsite rather than taken to market.
66
Operating costs and cattle prices
are assumed to be static over the project life and do not take into account changing
market conditions.
Logging Logging scenarios assume that reduced impact logging is not applied,
resulting in degradation of logged areas. This results in little or no regrowth of
merchantable timber; therefore a hectare is logged only once. At a rate of 12,500
hectares per annum, the 250,000 hectare area is logged over in 20 years. Land uses
and revenues post-harvesting, such as conversion to agribusiness, are not included in
the model. It can be assumed that population growth would result in small-scale
cultivation and urbanization of these areas in the short term, which is beyond the
scope of the commercially driven activities considered in this study. The assumed
yield of merchantable timber from the initial harvest is 20m
3
/ha.
Oil Palm For oil palm, the same rate of conversion (12,500 hectares per annum) is
applied, resulting in complete conversion of the study area over 20 years. The model
does not factor in the sale of timber from conversion activities under the assumption
that plantation development often occurs on degraded secondary forest, where limited
timber revenues are available, with remaining timber burned on site. The degradation
of the forest for timber as an intermediary process is captured in the land purchase
price of the palm oil investment.
3.3 Discount Rates
Discount rates are a crucial component in determining the NPV of cashflows of these
baseline activities. A discount rate represents the investors cost of capital over time,
or hurdle rate. It ultimately determines the investor’s expected financial return over
66
Bowman, M. (2007) pers comm Project Assistant, Woods Hole Research Center, November 2
17
the project life (internal rate of return, IRR) based on the amount of risk ascribed to
the activity. Investors will ascribe higher discount rates to projects that are assumed
to involve more risk. Risk considers many factors, including market knowledge,
market certainty, country sovereign risk (i.e. defensibility of property rights and
contracts), currency risk, timing of cash flows and availability of skilled labour and
physical capital.
For this study, a 20% real discount rate was used. High performing cattle ranches
deliver a real IRR of up to 20%,
67
and 20% real IRR is likely the minimum return
forest investors would expect in these countries. Anecdotal evidence suggests that
returns from oil palm plantations are currently higher than this and are increasing with
the rising price of crude palm oil. This reflects the rationale for the rapid expansion of
this land use in many places.
The stability and size of beef, timber and palm oil markets provides investors with a
relative level of certainty, although risk remains related to project execution and
market volatility. Discount rates applied to current carbon finance projects would
likely be higher given the current carbon market uncertainty. Nevertheless, this study
is designed to test the impact of establishing a set of rules and market price signals for
carbon offsets from REDD. Assuming such a market is established, investors would
be increasingly likely to use similar discount rates for carbon-related investments as
are applied for agribusiness investments. Carbon investors may even apply lower
discount rates, since the relative establishment and operational costs are minimal,
involving less project delivery risk.
However, it is acknowledged that there is currently little certainty around how REDD
credits will be incorporated into carbon markets. A major outstanding issue is how
project-based forest conservation may be incorporated into national baseline
accounting policies. National accounting requires countries to adopt a country-level
deforestation baseline, and credits would only be saleable if the country as a whole
reduces its emissions below the accepted baseline.
68
This creates a problem for
investors who may reduce emissions over a project baseline but are then precluded
from selling in international markets because of host country non-performance.
With this level of uncertainty, investors need to consider the risk-return profile of
carbon projects, and the implications of this are discussed in Section 4.
3.4 Net Present Value of Baseline Activities
The NPV of the cashflows from baseline activities represents the total cashflows over
the project lifetime discounted back to today’s dollars. This methodology enables a
comparison across investment options using a single metric of present value derived
from the activity (effectively the current purchase value that the land or forest should
67
Margulis, S. (2004)
68
Pedroni, L. and Streck, C. (2007)
18
be worth to an investor contemplating those activities). All modelling is in real terms
(i.e. values are in current dollars and do not account for inflation over time).
A range of approaches was used to determine the NPV per hectare based on available
information and market knowledge. Cashflows are modelled through 2040 for cattle
ranching and palm oil. Cashflows for logging end in 2028, when the resource is
considered to be depleted. (Carbon scenario cashflows are modelled through 2040.)
Cattle Ranching Generalizing the NPV of cattle ranching is difficult because costs
and revenues are highly variable. Costs are dependent on whether landowners are
converting large forest areas directly or buying smaller plots already converted
through intermediary processes. However, as ranching is the final productive activity,
it can be said to underwrite all prior costs and revenues, such as speculative land
purchases and selling timber from initial clearing (i.e. investors in intermediate
logging use land sale to ranching as a “terminal value” for their activities).
69
Income
also varies by type of ranch activity (rearing, fattening, etc.) and regional variations,
such as soil quality and distance to slaughterhouses.
Published data is sparse but one major World Bank study estimates a NPV of $500/ha
applying an 8% discount rate.
70
This is corroborated by another study in which the
NPV per hectare is estimated at $567 (6% discount rate) adjusted to $197 per hectare
at a 12% discount.
71
These discount rates generally reflect low risk projects involving
land that has already been cleared. Projects at the frontier would likely involve higher
risk, and therefore discount rates closer to 20% would be reasonable, as discussed.
The World Bank study reports profits per hectare of $77 per year.
72
Using this
estimate, the NPV of cashflows from the baseline scenario (12,500 hectares converted
to production annually over 20 years with cashflows to 2040) yields a per hectare
NPV of $111/ha, applying a 20% discount rate. Preliminary results from a third study
indicate that NPVs range from $0-1150/ha across the entire Brazilian Amazon.
73
For
the purposes of this study, a median value of this range ($575) was averaged with
$111/ha to yield an NPV of $343/ha.
Logging For logging, cashflows were derived from profits per hectare per year based
on literature and our access to data from regional forestry consultants. The NPV was
then derived based on cashflows over 20 years, at which point the forest is assumed to
be degraded and no longer viable for commercial logging.
In Papua, cashflows were based on estimated average log prices (on barge) minus
production costs. Based on information from industry sources operating in Papua, log
prices are assumed to be $150/m
3
, which is an average of prices achieved for high-
grade logs, medium-grade logs (i.e. Meranti and Agathis) and mixed species. Costs
69
Margulis, S. (2004)
70
Ibid
71
Mattos, M. and Uhl, C. (1994) “Economic and Ecological Perspectives on Ranching in the Eastern
Amazon” World Development Vol. 22, No. 2, p. 145-158
72
High productivity ranches in the best soil regions yield profits of R$138/hectare (Margulis 2004)
multiplied by the conversion rate of 1.76 to convert to US$.
73
Merry, F. (2007) pers comm. Discussing work in review. Woods Hole Research Center
19
were derived from an International Finance Corporation working paper
74
and
incorporate information provided by experienced foresters working in the region in
addition to New Forests’ operational experience in tropical forests. Total production
costs are approximately $120/m
3
assuming logs are loaded on a barge for transport to
Asian markets or to a local processing site near the coast.
This cost estimate represents natural forest management at a level that would meet
reputable sustainable forestry certification standards, i.e. reduced impact logging.
Concessions managed to these standards are rare. Specifically in Papua, there are no
certified natural forest extractive management operations. Therefore, the cost
structure was adjusted to remove activities that would likely be forgone in
unsustainable management regimes (silviculture treatments, planning, rehabilitation
and official taxes). A cost structure of $80m
3
was used to reflect unsustainable
practices. (Line item costs and further discussion are provided in Appendix A.)
Log prices per cubic meter ($150) minus operational costs under an unsustainable
logging scenario ($80) yield a profit of $70/m
3
. With 20m
3
of merchantable timber
per hectare and 12,500 hectares logged annually, this generates an annual cashflow of
$17.5 million. The NPV of project cashflows over 20 years is $350 per hectare,
applying a 20% discount rate.
In the DRC, log prices are reported at $220
75
to $300
76
per cubic meter corroborated
by ITTO average log (on barge) prices for Khaya (African mahogany) and Meranti of
$300.
77
Taking an average ($260) and using the same operational costs assumed in
the Papua scenario, the profit margin is assumed to be $160/m
3
.
Given that increased production in the DRC would necessitate harvesting less
valuable species, it was assumed that $160/m
3
profit was achieved for the first 10m
3
.
The remaining 10m
3
were assumed to yield a profit of $100/m
3
. This resulted in an
average profit of $130/m
3
. The NPV of the cashflows over 20 years is $633 per
hectare, applying a 20% discount rate.
It is important to note that the availability and value of 20m
3
of merchantable timber
in the Congo requires further investigation. In particular, if market pressures result in
a shift toward the unsustainable logging practices assumed here (e.g. shift to Chinese
markets), price premiums for certified sustainable timber would be lost. Deriving
profits from current prices may over-value actual prices in an unsustainable logging
scenario, and this should be taken into consideration.
Palm Oil A discounted cashflow model was built to determine plantation
establishment costs and crude palm oil yields over time. (Appendix B provides a full
list of cost and revenue assumptions and sources.) In brief, costs include a $300/ha
74
J arvis, B. and J acobson, M. (2006)
75
Forests Monitor (2007)
76
ITTO (2005) “Annual Review and Assessment of the World Timber Situation” Division of
Economic Information and Market Intelligence, Yokohama, J apan
http://219.127.136.74/live/Live_Server/2151/E-AR05-Text.pdf
77
ITTO (2007) “Market Trends” Tropical Forest Update Volume 17, No. 1
http://www.itto.or.jp/live/Live_Server/3241/tfu.2007.01.e.pdf
20
land purchase price and establishment costs of $2555 per hectare, both incurred as
establishment occurs (12,500 hectares per annum). Operating costs total $500 per
hectare per year, and harvesting costs are priced at $187 per hectare. The model
assumes that beginning at year four, each hectare of plantation produces 0.64 tonnes
of crude palm oil. In a linear fashion, the palm oil yield increases to 4 tonnes of crude
palm oil per hectare at age 7, stays constant to age 12 and steadily declines until age
25, when trees are removed and replanted. Crude palm oil is sold at current market
prices of US$886 per tonne over the project life. The net present value of the
cashflows to 2040 is $757 per hectare, applying a 20% discount rate.
3.5 Carbon Quantification
In addition to determining the NPV of the baseline activities, a carbon accounting
methodology is required to quantify the volume of carbon emissions that are avoided
by forgoing the baseline scenario. The assumed change in carbon stock over time
between the baseline scenario and the conservation scenario represents the quantity of
carbon offsets that can be created and sold. Realistic assumptions are therefore
required on the rate of conversion for each land use and associated emissions.
General assumptions are:
• It is assumed that there is an average of 549 tCO
2
/ha stored in the standing
forest in all three regions. This represents 150 tonnes of carbon (C) per
hectare multiplied by the molecular weight conversion factor of 3.66 to
estimate metric tonnes of carbon dioxide equivalence (tCO
2
e), the standard
unit traded in carbon markets. This figure is the average of multiple data
sources that report approximately 100-200 tC per hectare for tropical
rainforests.
78
• With regard to cattle ranching, it is assumed that 86% of the biomass is
removed from the site immediately upon conversion through burning.
Remaining biomass stays on site and decays over time. The minimum
residual carbon stock is 50 t CO
2
/ha, representing primarily below ground
biomass.
• With regard to logging scenarios, it is assumed that after harvesting 10%
of stored carbonis immediately lost (i.e. removed in the form of logs
79
);
25% decays over time, either on site or when debris, litter and soil carbon
78
FAO (2005) http://news.mongabay.com/2006/1031-deforestation.html; Gibbs, H.K, S. Brown, J . O.
Niles and J . A. Foley (2007) “Monitoring and estimating tropical forest carbon stocks: making REDD
a reality” Environmental Research Letters (in press); Margulis, S. (2004); Saatchi, S. et al (2005)
“Spatial Distribution of Carbon Stock in the Amazon Basin” American Geophysical Union, Fall
Meeting 2005 http://adsabs.harvard.edu/abs/2005AGUFM.B54B..03S
79
The standard approach in carbon trading systems is to assume that the carbon stored in logs is
released upon harvest. While carbon may remain stored in wood products for years, decaying later in
landfills or through other processes, this delay is not currently factored into carbon accounting, and
convention is applied here.
21
is washed into waterways; 65% remains in standing forest with 10% lost
each year from fire, illegal logging, mining, agriculture and reclassification
for agribusiness, etc. The minimum residual carbon stock is 50tCO
2
/ha.
• For palm oil, 100% of the biomass is assumed to be removed from the site
immediately with a minimum residual carbon stock of 50tCO
2
/ha. The
model accounts for the carbon sequestered as the palm plantation grows,
which is subtracted from the volume of saleable carbon in the avoided
deforestation scenario.
Standing carbon stocks are regionally variable, and the minimum carbon stock of
50tCO
2
/ha is an estimate based on the likely percentage of the remaining above and
below ground biomass after burning and degradation. Both require refinement to
improve accuracy at a local scale.
3.6 Carbon Value
As discussed, there is no accepted regulated market standard for defining and
accrediting avoided deforestation credits. For this study, it was necessary to make
assumptions about how avoided emissions may be credited and priced. Crediting
periods were designed based on general current practices in voluntary and regulated
markets. Price points were derived from recent transactions and market data as
explained in the notes to Table 1.
Scenario 1 A once only calculation is made of the total credits generated from
foregoing the baseline scenario. All credits are sold in the voluntary
market from 2008-2012. This is called ex-ante accounting, in which
credits are sold before the emission reduction is assumed to have
occurred. Products applying ex-ante accounting tend to attract a lower
market price.
Scenario 2 Bi-annual calculation of credits representing the carbon that would
have been released in the previous 2 years under the baseline scenario.
In this scenario credits are issued and sold bi-annually. Credits are
assumed to be sold to the voluntary market pre-2012 and then sold
through an international regulated market (i.e. Kyoto second
commitment period). All accounting is ex-post whereby credits are
sold after the emission reduction is assumed to have occurred,
generally attracting higher prices.
Scenario 3 Calculation and sale of credits every five years from 2012 onward.
The first tranche is sold in an international regulated market (i.e. Kyoto
second commitment period) in 2017 and every five years thereafter.
No sale of the credits occurs prior to 2017 or in voluntary markets. All
accounting is ex-post whereby credits are sold after the emission
reduction is assumed to have occurred, generally attracting higher
prices.
22
Table 1 – Carbon trading scenarios and prices
Scenario Volume Market Value
1
1/5 of the total credits generated
are sold each year between
2008-2012
Wholesale credits for
voluntary markets
US$2.95 *
2
credits are sold bi-annually for
the quantity of emissions
avoided over the baseline
Voluntary markets to
2012 & international
regulated market
thereafter
US$10/tCO2 **
US$18/tCO2 after
2012 ***
3
credits sold in 5-year tranches
for emissions avoided over the
baseline starting with the 2012-
2017 tranche
International
regulated market
US$18/tCO2 ***
* Average of approximate World Bank Biocarbon fund prices and a recent transaction in which Rio
Tinto purchased credits associated with avoided deforestation of farmland in Queensland, Australia.
80
These are voluntary prices for ex-ante products.
** An average price in the voluntary market for ex-post forestry products.
81
*** The average price for issued Certified Emission Reductions from the Clean Development
Mechanism in 2006,
82
representing the approximate value of a project-based offset in a regulated
international market.
Investors would likely pursue forward carbon sales, purchase options and other
transaction structures to improve the economics of carbon investments. These more
sophisticated structures are not included in the modeling, likely understating the value
of carbon scenarios. This should be taken into consideration.
It is assumed that REDD credits are permanent and therefore equivalent with other
emission reduction credits, trading at comparable per tonne prices.
Ten percent of the total carbon revenue was deducted from carbon sales to account for
transaction costs and forest management. Transaction costs include developing a
carbon accounting system, project documentation, third-party verification, scheme
accreditation or verification to a voluntary standard and management/brokerage fees.
Forest management costs include forest protection and monitoring over time and
could include payments to local communities for services such as boundary
maintenance and collecting inventory data.
80
See http://www.carbonpool.com/
81
Hamilton, K., Bayon, R., Tunrer, G. and D. Higgins (2007) “State of the Voluntary Carbon Markets
2007: Picking Up Steam” Ecosystem Marketplace & New Carbon Finance
www.ecosystemmarketplace.com
82
J otzo, F. (2007) “CDM Project Design” Centre for Energy and Environmental Markets, University of
New South Wales, Applied CDM Short Course, J uly 18
th
23
4 Discussion
Table 2 summarizes the NPV of the cashflows for each baseline activity against
cashflows from each carbon scenario. Results are reported in NPV per hectare.
Table 2 – NPV per hectare of baseline and carbon scenarios
Brazil - cattle Papua - logging Papua - palm Congo - logging
Baseline $343 $350 $757 $633
Carbon
Scenario 1
$660 $649 $550 $649
Carbon
Scenario 2
$1168 $756 $981 $756
Carbon
Scenario 3
$521 $402 $386 $402
Results indicate that, assuming carbon rules are adopted and implemented, carbon
revenue from avoided deforestation may compete with current alternate land uses at
the forest frontier. Carbon appears to effectively compete with logging in Papua and
cattle ranching in the Amazon. The case is not as clear for palm oil and logging in the
Congo, where high profits make it difficult for carbon to compete in some scenarios.
The quantum of saleable carbon associated with the baseline scenarios is
approximately 125-145 million tCO
2
.
In summary:
• Carbon appears competitive in comparison with cattle ranching in Brazil
• Carbon appears competitive in comparison with unsustainable logging in
Papua
• Carbon competes with palm oil revenues in Papua if the project can sell
credits regularly (bi-annually) and make initial sales into the voluntary
carbon market (carbon scenario 2)
• Carbon can compete with logging in the Congo if crediting occurs
regularly (bi-annually) and some sales are made in the near term in
voluntary markets (carbon scenario 2) or all credits are sold in the
voluntary market in the short term (scenario 1)
These results suggest that under certain policy settings, private financing could flow
toward conservation. This is a generalized model, however, and several points need
to be addressed in interpreting the results, including differences between regional land
use scenarios, variation across carbon scenarios, policy certainty and methodology
considerations.
24
4.1 Land Use Comparisons
Across the regions considered, carbon competes most effectively with cattle ranching
in the Amazon. This is based on high carbon values, which reflect the immediate and
permanent change in carbon stock when land is cleared for pasture. It is also a result
of ranching delivering a relatively low NPV, which is consistent with the fact that
cattle ranching is not considered the highest economic use of the land. As discussed,
ranchers are motivated by factors other than financial returns, including securing land
rights and generating low-risk cash flows. Carbon investments for forest conservation
would need to provide for or compete with these preferences in order to be more
attractive to investors than ranch development.
In Papua results show that a strong case can be made for carbon-financed forest
conservation investment when compared to logging revenue. For palm oil, the case is
not as clear. The results that show that only an investment strategy with
characteristics of scenario 2 – frequent crediting and initial sales into a voluntary
market – can compete with palm oil.
However, it is important to note the impact of discount rates in the case of oil palm
production. There are large initial capital costs in establishing an oil palm plantations
and cash flow only occurs as plantations grow. The result is that small changes in
discount rate have large effects on the net present value and, thus, the relative
competitiveness of carbon finance versus oil palm development. More information
regarding investors’ risk-return expectations is needed to explore this scenario more
fully.
In the Congo, high log prices make carbon less competitive that in the Papua logging
comparison. Timber from the DRC is primarily sold to European markets, which are
prepared to pay higher prices for sustainable, certified wood.
83
The market is
increasingly being driven by demand from China and India, which is currently
pushing prices of African species such as Khaya (African mahogany) and Meranti to
near historical highs.
84
As production increases to meet this demand, less valuable
species will be logged, and there will be a move away from sustainable harvesting that
meets certification standards. These drivers would result in lower log prices.
The model accounts for these market changes in a simplified way by reducing profits
per cubic meter from $160 for the first 10m
3
to $100 for the remaining 10m
3
. This is
an approximation for which more detailed market analysis is required. A reduction in
timber prices would make carbon more competitive in the region. The availability of
20m
3
of merchantable timber per hectare also needs to be verified.
83
Forests Monitor (2007)
84
ITTO (2007) “African Prices Stable on the Back of Steady Demand” Tropical Forest Update
Volume 17, No. 2 http://www.itto.or.jp/live/Live_Server/3453/tfu.2007.02.e.pdf
25
4.2 Carbon Price Scenarios
Scenario 2 is the most competitive of the three scenarios presented. This is the result
of a medium to high carbon price (US$10-18/tCO
2
) delivered every two years. While
Scenario 3 has an overall higher price point (US$18), revenue is delayed until 2017
and every 5 years thereafter, which reduces overall value after discounting.
Scenario 1 delivers all of the revenue in the first 5 years. This is a trade off with the
low carbon price of US$2.95, as market evidence suggests that a product sold upfront
in this way would likely attract a lower price. This scenario is somewhat aggressive
in suggesting that the total project carbon volume (approximately 125-145 million
tCO
2
) would be sold in the voluntary markets before 2012, even at this low price
point. Total transactions in the global voluntary market were estimated at 91 million
tonnes in 2006.
85
Although this market is posed for growth, transactions of the size
suggested by Scenario 1 would likely need to be motivated by future acceptance in a
regulatory scheme, where they could be sold in a secondary market at a higher price.
Policy uncertainty is likely to prevent this in the short term.
Scenario 3 represents the most conservative point of comparison. Revenue is skewed
toward the future, reflecting the reality that it will likely be several years before
policies are clear enough to drive investment in conservation carbon. Additionally,
the price point of US$18tCO
2
e, while the highest of the scenarios, may be lower than
future carbon prices in regulated markets. This price represents the value of issued
Certified Emission Reductions (CER) in the Clean Development Mechanism (CDM)
in 2006, a reasonable indicator of the value of project-based offsets. However, CERs
traded as high as $25/tCO
2
e in 2006, and demand for project-based offsets is expected
to increase.
86
Comparatively, allowances in the EU Emissions Trading Scheme are
currently trading for US$33.10,
87
although forestry credits are not currently eligible in
this market. Prices for project-based offsets may be higher in real terms in the future,
particularly as low-cost opportunities, such as Chinese projects to reduce HFCs, are
no longer available.
Conversely, credits from avoided deforestation may not be equivalent to other project-
based offsets, and this would result in lower prices. REDD credits may be treated as
temporary, as is current practice with afforestation/reforestation projects under the
CDM. There is some discussion of dual markets in which REDD credits are sold in a
parallel market and are not fungible with emission reduction offsets as defined under
Kyoto.
88
Additionally, the Voluntary Carbon Standard, which is emerging as a
leading standard for accrediting voluntary carbon projects, suggests that forestry
conservation projects should reduce the amount of saleable carbon by 5-30% (based
on a project risk assessment) as a buffer against future carbon stock losses. Policies
that result in REDD credits being non-equivalent to other emission reduction credits
or require large buffers reduce the value of carbon investments.
85
Hamilton et al (2007)
86
Carpoor, K. and Ambrosi, P. (2007) “State and Trends of the Carbon Market” World Bank/IETA,
Washington D.C.
87
As of November 15, 2007 according to Point Carbon http://www.pointcarbon.com/
88
Centre for Clean Air Policy (2007)
26
4.3 Policy Uncertainty
Investors must determine the required threshold return for each investment by
considering a range of risks and uncertainties. In voluntary carbon markets, those
uncertainties typically include buyer demand and achievable per tonne prices. In
regulatory markets, major risks are associated with project delivery and accreditation.
Existing carbon funds seek returns in the order of 35% real IRR, often based on
paying a very low price for carbon offsets prior to accreditation and then sharing some
of the upside with project developers. For projects where rules are uncertain or
untested, IRRs of 40-50% or more would be expected.
These return expectations imply discount rates that are higher than the 20% applied
here. However, the objective of this study was to consider whether investment in
forest conservation could compete with other land uses under the assumption that a
functional carbon market existed that incorporated REDD credits. Nevertheless, it is
also useful to consider the impact on investment if policies do not provide investors
with adequate levels of certainty. This point is most salient to the current debate
around national accounting baselines and project delivery, as discussed.
Two methods can be applied to consider this uncertainty. One is to apply a higher
discount rate to carbon projects. Table 3 shows the results for carbon projects when a
35% discount rate is applied. These are still compared to the baseline scenarios at a
20% discount rate because of the greater certainty around the timber, beef and palm
oil projects. This results in favouring conversion activities. As Table 3 shows, this
adjustment results in carbon being uncompetitive in the Congo and Papua for palm oil
and only competes with logging in Papua and ranching in Brazil in certain cases.
Figures in red indicate a lower NPV than the baseline scenario.
Table 3 – Carbon scenarios applying a 35% discount rate
Brazil - cattle
Papua -
logging
Papua - palm
Congo -
logging
Baseline (at
20% discount)
$343 $350 $757 $662
Carbon
Scenario 1
$432 $428 $363 $428
Carbon
Scenario 2
$521 $286 $440 $286
Carbon
Scenario 3
$126 $92 $85 $92
Adjusting the discount rate in this way generally captures uncertainty around project
delivery and accreditation, but it is ambiguous in the specific risks that are being
considered. It also favours policies that are skewed toward the present, such as
carbon scenario 1. The more conservative approach represented in scenario 3 cannot
compete financially, as revenues are skewed toward a more heavily discounted future.
27
Another approach is to assume that there is a 50% chance that credits will be awarded
and reduce the NPV by this probability factor. Using the original results in Table 2
and multiplying carbon NPV by 0.50, it can again be seen that carbon only appears
attractive in a limited number of cases. Figures in red indicate a lower NPV than the
baseline scenario.
Table 4 – Carbon project NPVs assuming 50% chance of accreditation
Brazil - cattle Papua - logging Papua – palm Congo - logging
Baseline $343 $350 $757 $662
Carbon
Scenario 1
$330 $324 $275 $324
Carbon
Scenario 2
$584 $378 $490 $378
Carbon
Scenario 3
$260 $201 $193 $201
In reality, investors that are only 50% certain that credits will be awarded will factor
in the optionality to abandon a carbon project in 5 or 10 years, if credits are not likely
to be accredited, and then convert to another land use. From this perspective, the loss
in returns is the difference in NPV from converting land today versus converting it in
the future, whereas the upside is the opportunity for carbon revenue. An investor
would be more likely to consider the original NPV results presented in Table 2 to
evaluate this trade-off.
Under either method, it is difficult to fully quantify the investment implications if the
fundamental question of whether the carbon credits will be recognized cannot be
answered. If it is accepted that private investment can play a major role in land use
change at the forest frontier, then fostering an attractive investment environment for
forest conservation should be an objective of policy debates.
4.4 Government Fees
Governments generally use a concession licensing system to allocate land use rights,
entitling licensees to conduct a particular activity on the land. Fees, ongoing lease
payments and tax arrangements tend to be variable based on activities, i.e. logging
versus oil palm plantations. For example, an area that is designated as a productive
forest may be reclassified for palm oil establishment if it is degraded after a first
harvest and there is minimal residual timber value. It is likely that concessions
specifically allocated for carbon investments would need to be designed or designated
within existing or new concession licensing systems. The structure of fees and other
payments to government would need to be considered under such a system.
There are no specific assumptions made here regarding how this structuring could
occur. As discussed in Section 3.5, ten percent of carbon revenues are reserved to
28
account for forest management and carbon transaction costs, and a portion of this
could also be assumed to be government fees. Investment revenues could be split in
any proportion between the commercial enterprise, local communities and the
regional or national government, and this will be an important point of negotiation as
markets develop. This could be addressed at a national level by governments taking
ownership of some or all carbon concessions, thereby receiving all associated
revenues, or negotiating with private investors.
This is also an important point in relation to the NPV methodology applied in this
study. Investors evaluate investment options in multiple ways. For example, different
outcomes would result from an IRR analysis, which focuses on rate of return on
invested capital rather than the present value of future cashflows. It is arguable that
forest conservation requires a limited outlay of initial capital in comparison to
baseline scenarios, so the financial risk to investors is relatively low. This would
result in a high IRR, making carbon investment appear even more competitive than
under the NPV analysis. (Note: Logging costs may be minimal if concession holders
are using old, depreciated machinery, but costs will still include employees,
harvesting plans, cutting roads, etc. For oil palm plantations and ranching, in some
cases site preparation costs could be offset by selling timber, but the models here
assume that the terminal value of timber sales is captured in baseline scenario land
prices.)
In both the NPV and IRR analysis, a detailed consideration of the costs associated
with developing and managing a carbon forestry project needs to be included. This
analysis has used a simple method to account for costs, 10% of revenues. A higher
cost structure would decrease the competitiveness of the carbon scenario.
5 Conclusions
This study has determined that conservation projects financed through carbon markets
can generally deliver returns that are competitive with current land uses that are
driving deforestation. These findings are based on the assumption that carbon market
policies will provide certainty to investors that REDD credits will be accepted and
creditable in carbon markets. Without this certainty, conservation will remain a high-
risk investment that cannot compete with conversion activities, resulting in limited
direct private sector investment. This has implications for policies which are
considering national baseline accounting approaches that result in high risk at the
project level.
This study has developed a framework to bring together a wide scope of information.
The generalized models and approach are intended to provide input to investment and
policy decision making. Many of the assumptions are necessarily simplified in order
to consider implications across the forest frontier. There is now some evidence to
consider the opportunities and evaluate carbon conservation investment options at a
more localized scale. If it is accepted that private investment can play a major role in
land use change at the forest frontier, then fostering an attractive investment
environment for forest conservation should be an objective of policy debates.
29
Key points for these ongoing debates are:
• Certainty around accreditation at the project level is paramount. If stable
and predictable policies are in place, investors could apply even lower risk
profiles (discount rates) in comparison to baseline activities, since the
primary capital investments are limited to project accreditation, forest
monitoring and protection and community payments or taxes that provide
for local development and non-extractive forest use. This is significantly
less than the appreciable start-up costs associated with baseline activities.
• The price of carbon credits is often discussed as the crucial variable for
carbon investment. Results presented here suggest that crediting periods
and early market sales interact with prices to deliver a range of outcomes.
It is not necessarily useful to identify the most significant input factor
because policies will rarely be designed with one element in mind. An
exception to this may be crediting periods, as our results suggest that more
frequent crediting is beneficial to investment returns.
• This study has not assessed how carbon-financed forest conservation
projects could provide benefit to communities or governments. This point
is central to the success of conservation activities, as forests cannot be
simply “locked up” or separated from the interests of communities that
rely on forest resources. Investors will want to ensure that local
communities dependent on the forests are supportive of conservation
activities to reduce project risks associated with illegal logging and other
destructive activities. Options may be payments for monitoring and
maintaining conservation forests and ensuring community rights to
develop forest resources for non-timber forest products and local use. The
cost structure of these activities needs to be factored into carbon
investments more clearly.
• While carbon investment costs may not be fully captured in this level of
analysis, there are additional benefits that have also not been quantified in
both the baseline and carbon scenarios. Baseline scenarios presumably
have flow-on benefits to communities in the form of employment and
social services, which are often provided by large-scale operators. Carbon
scenarios, conversely, can sustain non-timber forest product industries and
provide continued access for firewood, hunting, small-scale timber, food
and medicines. Ecosystem benefits from retaining the forest, e.g. water
quality protection, landslide/flood prevention, biodiversity habitat,
maintenance of traditional livelihoods and prevention of social dislocation
of communities are also not included, would require a total economic
assessment and are not included in the financial analysis considered here.
• Although this study was framed as a trade-off between one land use type
and the conservation scenario, the impact of carbon-financed conservation
will likely be to shift land use economics, rather than replace activities.
30
For example, if a standing forest is now valued for its carbon, landowners
would be motivated to increase the efficiency of existing agribusiness
activities or rehabilitate degraded areas rather than move into new forested
areas. This could result in more efficient land uses and a wider range of
benefits at a landscape scale.
31
References
Bowman, M. (2007) pers comm Project Assistant, Woods Hole Research Center,
November 2
Brazilian Vegetable Oil Industry Association (2006) “Understanding the Moratorium:
Responsible Production”
http://www.abiove.com.br/english/sustent/ms_cprodutor_24jul07_us.pdf
Butler, Rhett A. (2007) “Is peat swamp worth more than palm oil plantations?”
www.mongabay.com
Carpoor, K. and Ambrosi, P. (2007) “State and Trends of the Carbon Market” World
Bank/IETA, Washington D.C.
Centre for Clean Air Policy (2007) “Reducing Emissions from Deforestation and
Degradation: The Dual Markets Approach”
http://www.ccap.org/international/FINAL%20REDD%20report.pdf
Congo Basin Forest Partnership (2005) “The Forests of the Congo Basin: A
Preliminary Assessment” (online)
http://carpe.umd.edu/products/PDF_Files/FOCB_APrelimAssess.pdf
FAO (2005) “FAO Forest Resource Assessment – Global Tables”
http://www.fao.org/forestry/site/28679/en/
Fearnside, P. (2007) “Deforestation in Amazonia” Encyclopedia of Earth Eds. Cutler
J . Cleveland, Washington, D.C.: Environmental Information Coalition, National
Council for Science and the Environment
http://www.eoearth.org/article/Deforestation_in_Amazonia;
Foley et al. (2007) “Amazonia Revealed: Forest Degradation and the loss of
ecosystem goods and services in the Amazon Basin” Frontiers in Ecological
Environment, Issue 5(1), p. 25-32
Forests Monitor (2007) “The Timber Sector in the DRC: A Brief Overview”
http://www.forestsmonitor.org/uploads/2e90368e95c9fb4f82d3d562fea6ed8d/Descrip
tion_of_the_Timber_Sector_in_the_DRC.pdf
Friends of the Earth (2005) “Greasy Palms. The social and ecological impacts of
large-scale oil palm plantation development in southeast Asia”
http://www.foe.co.uk/resource/reports/greasy_palms_impacts.pdf
Gibbs, H.K, S. Brown, J . O. Niles and J . A. Foley (2007) “Monitoring and estimating
tropical forest carbon stocks: making REDD a reality” Environmental Research
Letters (in press)
32
Global Forest Watch et al (2001) “The State of the Forest: Indonesia” Global Forest
Watch, World resources Institute and Forest watch Indonesia
www.globalforestwatch.org
Global Forest Watch (2007) “Human Pressure on the Brazilian Amazon Forests”
(online)
http://www.globalforestwatch.org/english/interactive.maps/Brazil_Datasets.htm#Stud
y_area
Hamilton, K., Bayon, R., Tunrer, G. and D. Higgins (2007) “State of the Voluntary
Carbon Markets 2007: Picking Up Steam” Ecosystem Marketplace & New Carbon
Finance www.ecosystemmarketplace.com
Ismail, A., Simeh, M. and M. Mohd Noor (2003) “The Production Cost of Oil Palm
Fresh Fruit Bunches: the Case of Independent Smallholders in J ohor” Oil Palm
Industry Economic Journal Vol. 3(1)
ITTO (2005) “Annual Review and Assessment of the World Timber Situation”
Division of Economic Information and Market Intelligence, Yokohama, J apan
http://219.127.136.74/live/Live_Server/2151/E-AR05-Text.pdf
ITTO (2006) “Asia Takes Price Leadership from Europe in Africa” Tropical Forest
Update Volume 16, No. 4
http://www.itto.or.jp/live/Live_Server/3163/tfu.2006.04.e.pdf
ITTO (2007) “Market Trends” Tropical Forest Update Volume 17, No. 1
http://www.itto.or.jp/live/Live_Server/3241/tfu.2007.01.e.pdf
ITTO (2007) “African Prices Stable on the Back of Steady Demand” Tropical Forest
Update Volume 17, No. 2
http://www.itto.or.jp/live/Live_Server/3453/tfu.2007.02.e.pdf
J arvis, B. and J acobson, M. (2006) “Working paper – Incentives to promote forest
certification in Indonesia” Project: Motivating Sustainability, Prepared for Moray
McLeish, International Finance Corporation PENSA. Indonesia, April
J otzo, F. (2007) “CDM Project Design” Centre for Energy and Environmental
Markets, University of New South Wales, Applied CDM Short Course, J uly 18
th
Karsenty, A., and Gourlet-Fleury, S. (2006) “Assessing sustainability of logging
practices in the Congo Basin’s managed forests: the issue of commercial species
recovery. Ecology and Society Issue 11(1), p. 26 (online)
http://www.ecologyandsociety.org/vol11/iss1/art26/
Laporte, N. T, T. S. Lin, J . LeMoigne, D. Devers, and M. Honzak (2004) “Toward an
Integrated Forest Monitoring System for Central Africa” Land Change Science:
Observation, Monitoring, and Understanding Trajectories of Change on the Earth
33
Surface, Remote Sensing and Digital Image Processing Vol (6), Ed. G. Gutman, p.
97-110
Laporte, N., J . Stabach, R. Grosch, T. S. Lin, S. Goetz (2007) “Expansion of
Industrial Logging in Central Africa” Science Vol. 316, No. 5830, p. 1451-1452
Malaysian Palm Oil Board (2007) Palm Oil Weekly Prices, available online
http://econ.mpob.gov.my/upk/weekly/bh_wk02nov07.htm (accessed Oct 2007)
Margulis, S. (2004) “Causes of Deforestation of the Brazilian Amazon” World Bank
Working Paper No. 22, Washington, D.C.
Martino, D. (2007) “Deforestation in the Amazon: Pressures and Outlook” Third
World Resurgence No. 200, April
http://www.southdevelopment.org/environment/MartinoAmazonDeforestation.pdf
Mattos, M. and Uhl, C. (1994) “Economic and Ecological Perspectives on Ranching
in the Eastern Amazon” World Development Vol. 22, No. 2, p. 145-158
Merry, F. (2007) pers comm. Discussing work in review. Woods Hole Research
Center
Mongabay.com (2007) “South American development plan could destroy the
Amazon” http://news.mongabay.com/2007/1003-ci_amazon.html
Pedroni, L. and Streck, C. (2007) “Mobilizing Public and Private Resources for the
Protection of Tropical Rainforests” CATIE and Climate Focus
http://www.climatefocus.com/newspubs/downloads/PolicyBriefonREDD_000.pdf
Nasi, R. (2005) “Potential Methodological Flaw in the Examination of the Effects of
Logging” Ecology and Society Issue 10(2) (online)
http://www.ecologyandsociety.org/vol10/iss2/resp2/
Nepstad, D., Moutinho, P. & Soares-Filho, B. (2005) “The Amazon in a Changing
Climate: Large-Scale Reductions of Carbon Emissions from Deforestation and Forest
Impoverishment” Amazon Institute for Environmental Research), Woods Hole
Research Center (WHRC) and Federal University of Minas Gerais paper released
during COP 12, Nairobi
Ruiz-Perez, M. et al (2005) “Logging in the Congo Basin: A Multi-Country
Characterization of Timber Companies” Forest Ecology and Management Volume
214, Issues 1-3
Saatchi, S. et al (2005) “Spatial Distribution of Carbon Stock in the Amazon Basin”
American Geophysical Union, Fall Meeting 2005
http://adsabs.harvard.edu/abs/2005AGUFM.B54B..03S
34
Soares-Filho, B. et al (2006) “Modelling Conservation in the Amazon Basin” Nature
Vol. 440, March
Sumantri, H. and Wijayanto, I. (2005) “Land cover mapping using satellite imagery
and GIS in the Mamberamo Basin and Raja Ampat Islands, Papua, Indonesia”
www.mapindia.org/2005/papers
Tedjasukmana, J . (2007) “Heros of the Environment: Leaders & Visionaries” Time
Magazine
http://www.time.com/time/specials/2007/article/0,28804,1663317_1663319_1669895,
00.html
Tokede, M., William, D., Widodo, Gandhi, Y., Imburi, C., Patriahadi, Marwa, J .,
Yufuai, M. (2005) “The impacts of Special Autonomy in Papua’s Forestry sector:
Empowering Customary Communities (Masyarakat Adat) in decentralised Forestry
Development in Manokwari District” Centre for International Forestry Research,
Bogor, Indonesia
Vidal, J . (2007) “World Bank Accused of Razing Congo’s Forests” The Guardian,
October 4
http://www.guardian.co.uk/environment/2007/oct/04/congo.forests?gusrc=rss&feed=1
2
Woods Hole Research Center (2007a) “Agriculture Frontier Explosion in Brazil”
http://www.whrc.org/southamerica/agric_expans.htm
Woods Hole Research Center (2007b) “Logging and Family Forests”
http://www.whrc.org/southamerica/logging_fam_forests.htm
World Bank Carbon Finance Unit (2007) “Forest Carbon Partnership Facility”
http://carbonfinance.org/Router.cfm?Page=FCPF&FID=34267&ItemID=34267
WWF Germany (2002) “Oil palm plantations and deforestation in Indonesia: What
role to Europe and Germany Play?” http://www.fire.uni-
freiburg.de/se_asia/projects/wwf_oil.htm
35
Appendix A – Production cost schedule for natural
tropical forest concession
Hypothetical Natural Forest Concession – applying sustainable forest
management in line with international certification standards and
payment of all official taxes and fees
Assumed harvestable area (ha)
Number of years cut over
Hectares harvested per year
Annual production (m3) - 20 m3/ha
Activity Cost/m3
(USD)
I PLANNING COSTS
1 Block delineation 0.04
2 Pre-harvest inventory 0.22
3 Survey, road planning and construction 3.91
4 Amortisation of concession fee 0.25
5 Amortisation of forest inventory 0.05
6 Amortisation of concession boundary delineation 0.02
7 Amortisation of satellite imagery 0.02
Subtotal 4.51
II REHABILITATION PLANTING
1 Seedling costs 0.17
2 Planting in logged areas, roadsides and bare soil 0.33
Subtotal 0.50
III SILVICULTURAL TREATMENTS
1 Post harvest inventory 0.16
2 Plant maintenance 0.12
3 Reserve area maintenance 0.04
4 Permanent sample plots 0.02
5 General facilities 0.08
Subtotal 0.42
IV HARVESTING COSTS
1 Felling (on contract with equipment provided) 5.05
2 Skidding (on contract with equipment provided) 0.09
3 Rafting/Hauling (on contract with equipment provided) 4.99
4 Maintenance of facilities and equipment 9.15
5 Mechanics workshop 3.70
6 Fuel 6.38
7 Tug boat and pontoon costs 12.00
Subtotal 41.37
V OFFICIAL TAXES AND FEES
1 Land and buildings tax 1.50
2 Vehicular and heavy equipment tax 0.12
3 Log grading tax 0.26
4 Local government fees and contributions 3.20
6 Rehabilitation Fund (DR) 14.00
36
7 Forest Resource Provision Tax (PSDH) 7.00
Subtotal 26.09
VI SOCIAL AND ENVIRONMENTAL COSTS
Community liaison 4.31
Community fees 1.24
Amortisation of Environmental Impact Assessment 0.01
Subtotal 5.56
VII PAYMENT TO CUSTOMARY LAND OWNERS
Subtotal 5.50
VIII GENERAL DEPRECIATION COSTS
1 Of field generators, heavy equipment 5.29
2 Of roads and buildings 4.00
3 Of workshop and camp generator and equipment 0.09
Subtotal 9.38
IX WAGES, BONUSES, ADMINISTRATION
1 Wages and bonus 4.00
2 Field camp consumables and associated costs 3.00
3 Head office and marketing costs 4.00
4 Forest protection and security 0.12
Subtotal 11.12
X UNOFFICIAL COSTS
Forest service officials - honorarium, entertainment 5.18
Forest security and protection 1.96
NGOs - honorarium. Entertainment 1.02
Village officials 0.63
Local government 1.57
Others 5.00
Subtotal 15.36
Grand total - Production cost 120.73
The following should be noted in relation to this estimate:
• These are the costs that would be necessary if natural forest extractive
management was practiced at a level that would meet reputable forest
certification standards. There are no certified natural forest extractive
management operations in Papua.
• All costs have been divided by total yield. This approach does not take into
account the effect of the timing of costs i.e. not all costs will be spread evenly
over all the years of harvesting; there will be some up-front costs. There is
therefore an underestimation of the cost, but it is not significant for analysis at
this level.
Further explanation of costs:
I4 – Amortisation of concession fee
37
The Iuran Hak Pengusahaan Hutan (IHPH) is a one-off payment to the Department of
Forestry for the rights to manage the forest for a set period of time, typically 20 to 50
years. The fee is an area based fee set at $10.00 per hectare.
IV3 – Rafting/Hauling
Note – roading costs are a combination of part of the fuel, mechanics workshop,
heavy equipment depreciation and other costs.
Costs are based on a topographic area is largely flat with large sinuous rivers, seasonal
wet areas and swamps, reflecting the geographies of Papua and parts of the DRC.
Roading intensity has to be higher to allow for these obstructions, and in some places
the roads would need to be elevated on causeways. Roading is difficult and, with
limited supplies of rock, expensive and difficult to maintain. There are few public
roads away from larger centres, most logs will be hauled to large rivers and rafted or
barged to the coast.
V. OFFICIAL TAXES AND FEES (in Papua)
V1 – Pajak Bumi Bangunan - Land tax
Paid annually, this is calculated on a fixed percentage of listed government value per
ha. Sources quoted a wide range of costs, from $0.25 to $3.00 per m
3
of merchantable
logs produced.
V4 – Local government taxes and contributions
This includes ‘Retribusi daerah’, a local government tax. Local government taxes
may vary from region to region.
V5 – Dana Reboisasi – Reforestation Levy
The reforestation levy varies by species grouping and, for Papuan logs is likely to be
from $16/m
3
for merbau (Instia bijuga) and other high grade species to $12/m
2
for
mixed grades.
V6 – PSDH – Forest Resource Provision Tax
This tax is set at 10% of the list price set by the Ministry of Trade. The levey varies
between location, diameter and species. It is currently of the order of US$12.30 for
the highest grade including merbau; US$6.90 for the meranti group grade and
US$4.10 for the mixed grades.
VII – PAYMENT TO CUSTOMARY OWNERS
The benefits received by the community cooperatives vary according to the agreement
made with investors, and, in some cases, the volume of timber harvested.
Compensation may range from USD3.00/m3 to even USD22.00/m3. Where
community cooperatives are in partnership with large scale forest concession holders,
the maximum amount of compensation that the concession holders are required to pay
is around USD5.75 (Tokede et al 2005).
X. UNOFFICIAL COSTS
X1 – Government officials – honorarium, entertainment
38
This includes expenses to obtain the concession license. One source estimates the
unofficial costs at perhaps USD100,000 although may often be many times that.
Getting the annual plan (RKT) accepted may cost between USD12 000 and
USD25,000 each year. Signing of the regular log inspection reports by Department of
Forestry staff costs at least USD300/month. Signing of transport documents costs
around USD per shipment and there are additional charges to be paid to port officials.
Companies will frequently respond to a range of requests from the government
officials it deals with for air tickets and accommodation not related to the tasks they
perform.
39
Appendix B – Oil Palm Model Assumptions
1. Costs
• Land costs are $300 per hectare and are assumed to be incurred in the year
of establishment.
• Due diligence costs of $1.25 million ($5 per hectare) based on New Forests’
investment experience are incurred in year 1.
• Establishment costs are $2,555 per hectare, which includes land clearing, site
assessment, ripping and mounding, weed control, seedlings, planting,
contingencies and refilling.
• Operating costs total $500 per hectare per year for established plantation
areas. Costs include pruning, weed control, roading, labor, fertilizer,
equipment and transportation.
• The model assumes harvesting costs of $187/ha.
• These establishment costs are is in?line with figures provided by the Oil Palm
Research Institute.
89
2. Productivity and revenue generation
• Revenue from timber harvests associated with clearing land for plantations
are not included in this model.
• The model assumes that crude palm oil can be sold at current market prices
of US$886 per tonne.
90
• The model assumes that beginning at year 4, each ha of plantation produces
0.64 tonnes of oil. In a linear fashion, the oil palm yield increase to 4 tonnes
of oil per ha at age 7 where it stays consistent to age 12 before steadily
declining until age 25. Productivity curves derived from the Oil Palm
Research Institute and the World Agroforestry Centre.
91
• The applied oil extraction rate is 16%, indicating the volume of crude palm oil
extracted from fruit bunches grown on trees. For example, in the first year
the model produces 50,000 tonnes of fresh fruit bunches (4 tonnes per
hectare) resulting in 8,000 tonnes of crude palm oil (16% of 50,000 tonnes).
89
Ismail, A., Simeh, M. and M. Mohd Noor (2003) “The Production Cost of Oil Palm Fresh Fruit
Bunches: the Case of Independent Smallholders in J ohor” Oil Palm Industry Economic Journal Vol.
3(1)
90
Malaysian Palm Oil Board (2007) Palm Oil Weekly Prices, available online
http://econ.mpob.gov.my/upk/weekly/bh_wk02nov07.htm (accessed Oct 2007)
91
Butler, Rhett A. (2007) “Is peat swamp worth more than palm oil plantations?” www.mongabay.com
doc_143628646.pdf
The forest frontier – the next area of intact forest to be degraded or converted to a non-forest land use – is the front line of global deforestation. Deforestation of the world’s intact tropical forests is driven by many factors.
Carbon Finance for Reduced Emissions from
Deforestation & Degradation at the Forest Frontier
Financial Analysis of Alternate Land Uses in the Amazon,
Congo and Papua, Indonesia
Authors:
Marisa Meizlish
1
Dan Spethmann
2
Michael Barbara
3
New Forests
Sydney, Australia and Washington, D.C.
Paper Presented at UNFCCC COP13 and Forest Day
Dec. 3-14, 2007
Bali, Indonesia
1
Manager, Advisory Services, [email protected]
2
Manager, Investment Programs, [email protected]
3
Analyst, Environmental Markets, [email protected]
2
New Forests
New Forests (www.newforests.com.au) is a forestry investment management and
advisory services firm. The company’s investment thesis is unique in seeking
investment opportunities that deliver traditional timber returns, as well as returns from
emerging environmental markets and preferences, such as certified timber, carbon,
biodiversity and water. In addition to world-leading forestry investment expertise in
the areas of acquisitions, modeling, operations, silviculture and ecosystem services,
New Forests is at the forefront of developing and commercializing environmental
products and developing opportunities in emerging environmental markets. The
company is based in Sydney, Australia, with offices in Washington, D.C. and San
Francisco. New Forests holds an Australian Financial Services License (license
#301556).
New Forests Pty Limited
The Zenith Center, Tower A
Level 19, Suite 1905
821 Pacific Highway
Chatswood 2067
Australia
+61-2-9406-4100
New Forests Inc.
1025 Connecticut Ave, NW
Suite 1206
Washington DC 20036
United States
+1-202-715-1700
###
Copyright
This document may be copied for individual use but approval is required for multiple
copies and distribution. An electronic copy is available at www.newforests.com.au.
For citations, please use:
Meizlish, M., Spethmann, D. and Barbara, M. (2007) “Carbon Finance for Reduced
Emissions from Deforestation & Degradation at the Forest Frontier” New Forests Inc.,
Washington, D.C.
3
Table of Contents
Executive Summary ....................................................................................................... 4
1 Introduction ................................................................................................................. 5
2 Background on Drivers of Deforestation at the Forest Frontier ................................. 6
2.1 The Brazilian Amazon ................................................................................... 6
2.2 Papua, Indonesia .......................................................................................... 10
2.3 Democratic Republic of Congo ................................................................... 12
3 Methodology and Assumptions ........................................................................... 14
3.1 Framework for Analysis .............................................................................. 14
3.2 Definition of Baseline Activities ................................................................. 15
3.3 Discount Rates ............................................................................................. 16
3.4 Net Present Value of Baseline Activities ..................................................... 17
3.5 Carbon Quantification .................................................................................. 20
3.6 Carbon Value ............................................................................................... 21
4 Discussion ............................................................................................................ 23
4.1 Land Use Comparisons ................................................................................ 24
4.2 Carbon Price Scenarios ................................................................................ 25
4.3 Policy Uncertainty ....................................................................................... 26
4.4 Government Fees ......................................................................................... 27
5 Conclusions .......................................................................................................... 28
References .................................................................................................................... 31
Appendix A – Production cost schedule for natural tropical forest concession .......... 35
Appendix B – Oil Palm Model Assumptions .............................................................. 39
List of Figures
Figure 1 – World’s remaining areas of intact rainforest ................................................ 6
Figure 2 – Nine states of the Brazilian Amazon and patterns of deforestation ............. 7
Figure 3 – Deforestation in Rodonia, Brazil from 2000 to 2006 .................................. 8
Figure 4 – Papua Province, Indonesia .......................................................................... 10
Figure 5 – Congo Basin spanning five central African nations * ................................ 12
List of Tables
Table 1 – Carbon trading scenarios and prices ............................................................ 22
Table 2 – NPV per hectare of baseline and carbon scenarios ...................................... 23
Table 3 – Carbon scenarios applying a 35% discount rate .......................................... 26
Table 4 – Carbon project NPVs assuming 50% chance of accreditation .................... 27
4
Executive Summary
Background on
Carbon Investment
& Forest
Conservation
Deforestation in tropical rainforests contributes approximately
20% of global greenhouse gas emissions annually. Options
for expanding the inclusion of forest conservation in global
carbon markets are currently being explored. Policies that
support a price mechanism for reducing emissions from
deforestation and degradation (REDD) could lead to
significant new investment in forest conservation. A key
question in designing such market policies is how a carbon
price for REDD would impact land use dynamics, in particular
whether carbon investment may compete effectively with
large-scale activities that are driving deforestation, such as
commercial agribusiness expansion.
Scope of Financial
Analysis
The objective of this study is to evaluate the viability of
carbon finance as a mechanism for forest conservation. This
is explored via a generalized framework that compares the net
present value (NPV) of carbon-financed forest conservation
with the value of land uses resulting in forest degradation and
deforestation. The study conducts this analysis from the
perspective of a private investor making investment decisions
on land use options for existing forested areas.
Current Investment
Scenarios
This analysis is conducted across all of the world’s largest
remaining areas of tropical rainforest. The major drivers of
deforestation in each region are the baseline scenarios against
which conservation is considered. These are:
• Cattle ranching in the Brazilian Amazon
• Oil palm plantations and logging in Papua, Indonesia
• Logging in theDemocratic Republic of Congo
Methodology to
Assess Carbon
Investment
Scenarios
The NPV of the cashflows from these activities are compared
to the NPV of carbon revenue generated from a potential
REDD regime. A range of carbon prices and crediting periods
is applied to consider sensitivities to these variables. The
impact of policy uncertainty is also discussed.
Results Results indicate that carbon-financed forest conservation is
generally competitive with current land uses at the forest
frontier. More frequent crediting periods and the ability to sell
some credits into a voluntary market in the short term improve
the ability of carbon-finance scenarios to compete with
baseline activities. Policy uncertainty is an important variable
that can undermine the ability for carbon investment to
compete with alternate land uses. Further analysis of the cost
structure associated with carbon investments and refinement at
a more localized scale can build on the generalized framework
developed here to inform decision making in greater detail.
5
1 Introduction
The forest frontier – the next area of intact forest to be degraded or converted to a
non-forest land use – is the front line of global deforestation. Deforestation of the
world’s intact tropical forests is driven by many factors. These include the recent
price increases in global agribusiness commodities, such as soybeans, palm oil and
cattle, and increasing global timber demand. At the same time as the forest frontier is
being pushed deeper into previously intact forest areas, the rapid growth in carbon
markets is creating new avenues to support forest conservation and rehabilitation.
This paper assesses whether conservation through carbon finance can compete with
land uses currently driving large-scale deforestation at the forest frontier. Focusing on
the rainforest regions of the Brazilian Amazon, Papua, Indonesia, and the Democratic
Republic of Congo, we have sought to develop a generalized financial model of a
range of land uses and carbon market scenarios. The focus of this study is on whether
carbon investors may compete effectively with the current large-scale actors in
deforestation, such as commercial agribusiness ventures. Ultimately, if the price of a
forest for carbon conservation is greater than the soil expectation value for other uses,
then forest conservation management will be attractive to investors over other land
uses.
Private investment flows that currently finance alternate land uses, such as converting
forests to oil palm, could shift toward investment in forest conservation if risk
adjusted returns are competitive. Consequently, the paper uses a Net Present Value
(NPV) methodology to compare the relative attractiveness of forest conservation
versus agribusiness from the perspective of a private investor. This analysis is
conducted under the assumption that credits generated by reducing emissions from
deforestation or degradation (REDD) are eligible in an operational global carbon
market. However, a range of crediting scenarios and carbon credit prices are
considered to show sensitivity to these factors. A discussion of discount rates and
policy uncertainty is also included.
Our findings indicate that there is a case to be made for investment in carbon-financed
conservation. The generalized approach developed here suggests that forest
conservation can compete with land uses driving deforestation, if policies can provide
market certainty around credit eligibility. This paper is intended to serve as a
launching point for more detailed economic analysis and debate around forest
conservation carbon policy. In particular, further analysis of the cost structure
associated with carbon investments and refinement at a more localized scale can build
on the generalized framework developed here to inform decision making in greater
detail.
Section 2 of this paper provides background on the land use dynamics of the three
study regions. Modelling methodologies and assumptions and a discussion of results
are presented in the following sections. The paper concludes with a discussion of
policy considerations that could generate large-scale investment in carbon-financed
forest conservation.
6
2 Background on Drivers of Deforestation at the
Forest Frontier
The three areas considered in this study are the Brazilian Amazon, the Indonesian
province of Papua and the Congo Basin in the Democratic Republic of Congo (DRC).
These areas represent the largest remaining areas of intact tropical rainforest in the
world (Figure 1). This section briefly describes regional land use dynamics as context
for the financial analysis in the following sections. The main drivers of deforestation
in each region are identified to define the baseline, or “business as usual,” activities
that are most likely to occur in the frontier regions.
Figure 1 – World’s remaining areas of intact rainforest
4
2.1 The Brazilian Amazon
The largest area of remaining intact forest is the Amazon Basin, which covers 5.5
million square kilometers across nine countries in northwest South America.
Approximately 60% of the Amazon rainforest is located in Brazil, which is home to
19% of the world’s remaining intact forests.
5
Around 67 million hectares – 17% of the original Amazon area – have been converted
to other land uses since the 1970s.
6
Seventy-six percent of deforestation occurs in the
4
Source: Mongabay.com http://rainforests.mongabay.com/0102.htm
5
Brazilian Vegetable Oil Industry Association (2006) “Understanding the Moratorium: Responsible
Production” http://www.abiove.com.br/english/sustent/ms_cprodutor_24jul07_us.pdf
6
Ibid
7
“arc of deforestation” from the Atlantic Ocean through southern Para and Mato
Grosso to Acre,
7
which is the frontier area considered in this study (Figure 2). It is
estimated that current trends in agribusiness expansion could convert 40% of the
Amazon rainforest by 2050.
8
This would result in the release of 32 billion tonnes of
carbon to the atmosphere, equivalent to approximately 4 years of global annual
greenhouse gas emissions.
9
Figure 2 – Nine states of the Brazilian Amazon and patterns of deforestation
10
The annual rate of loss of primary forest in Brazil from 2000-2005 was 0.8%,
representing a total loss of 17.3 million hectares over these 5 years. This rate is a 35%
increase over the rate of primary forest loss recorded from 1990-2000.
11
Two of the
largest deforestation years on record were 2002 and 2003 (23,000 square kilometers
7
Margulis, S. (2004) “Causes of Deforestation of the Brazilian Amazon” World Bank Working Paper
No. 22, Washington, D.C.
7
Walker, R. and Moran, E. (2000) “Deforestation and Cattle Ranching in the Brazilian Amazon:
External Capital and Household Processes” World Development Vol. 28, No. 4, p. 683-699
8
Soares-Filho, B. et al (2006) “Modelling Conservation in the Amazon Basin” Nature Vol. 440, March
9
Ibid
10
Global Forest Watch (2007) “Human Pressure on the Brazilian Amazon Forests” (online)
http://www.globalforestwatch.org/english/interactive.maps/Brazil_Datasets.htm#Study_area
11
FAO (2005) “FAO Forest Resource Assessment – Global Tables”
http://www.fao.org/forestry/site/28679/en/
8
and 24,000 square kilometers per year, respectively
12
), and an all-time high of 27,400
square kilometers was cleared in 2004, 50% above the long-term average.
13
Deforestation rates tend to be closely correlated with prices for cattle, timber and
crops.
14
With the decline of soy and beef prices in 2005 and 2006 and the
strengthening of the Real against the US dollar, deforestation rates slowed to 18,800
and 13,100 square kilometers per annum, respectively
15
.
A rise in annual deforestation is expected, as prices increase for these commodities in
response to global growth in demand.
16
This reactive variability suggests that land
practices could change quickly through the introduction of a price signal for carbon-
financed forest conservation. Figure 3 gives one snapshot of how quickly the
conversion of the frontier is advancing.
Figure 3 – Deforestation in Rodonia, Brazil from 2000 to 2006
17
12
Woods Hole Research Center (2007a) “Agriculture Frontier Explosion in Brazil”
http://www.whrc.org/southamerica/agric_expans.htm
13
Nepstad, D., Moutinho, P. & Soares-Filho, B. (2005) “The Amazon in a Changing Climate: Large-
Scale Reductions of Carbon Emissions from Deforestation and Forest Impoverishment” Amazon
Institute for Environmental Research), Woods Hole Research Center (WHRC) and Federal University
of Minas Gerais paper released during COP 12, Nairobi
14
Mongabay.com (2007) “South American development plan could destroy the Amazon”
http://news.mongabay.com/2007/1003-ci_amazon.html
15
Ibid
16
Martino, D. (2007) “Deforestation in the Amazon: Pressures and Outlook” Third World Resurgence
No. 200, April http://www.southdevelopment.org/environment/MartinoAmazonDeforestation.pdf
17
NASA Earth Observatory
http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=17600
9
The most serious threat to the Amazon forests and the main driver of deforestation is
cattle ranching.
18
Government subsidies and tax incentives granted to large-scale
ranchers in the 1970s and 1980s spurred the growth of the industry,
19
and the herd
size nearly doubled from 26.2 million head of cattle in 1991 to 51.6 million in 2001.
20
It is estimated that as much as 88% of deforested areas are occupied by ranches,
21
and
the industry continues to grow.
22
In general, cattle ranching is not the highest economic land use when compared with
other regimes, such as soy or logging. However, landowners tend to favor ranching
because it provides a relatively liquid asset that can provide annual cash-flow and is
also relatively low risk. Converting land to a cattle ranch is also a way of
guaranteeing land possession and securing property rights.
23
Soybean production, which grew approximately 60% between 1998 and 2002,
24
is
contributing to this expansion by moving onto old ranching lands in the southern arc
of deforestation, pushing cattle further north into untouched areas. Soybean
cultivation is becoming more competitive as soy prices rise, driven by a growing
demand for bio-diesel and global food production, and as transportation costs
decrease through increased river usage and expanded road networks.
25
An additional driver of deforestation is selective logging, which occurs over an
average area of 15,000 square kilometers a year.
26
Recent studies indicate that rather
than contributing to deforestation by providing access to areas that are subsequently
converted to agricultural use, selective logging appears to be a driver of deforestation
on its own and is accretive to deforestation caused by conversion for cattle and soy.
27
However, 96% of the logging in the Amazon occurs in old and intermediate frontiers
(10 –30 years since first selective harvest) with the remaining 4% occurring in new
frontiers, principally Novo Progresso and northern Mato Grosso.
28
There are multiple land uses at the Brazilian forest frontier and the land use dynamics
are not fully understood.
29
However, conversion to pasture for cattle ranching is the
largest driver and the land use studied here as the baseline scenario in Brazil.
18
Fearnside, P. (2007) “Deforestation in Amazonia” Encyclopedia of Earth Eds. Cutler J . Cleveland,
Washington, D.C.: Environmental Information Coalition, National Council for Science and the
Environment http://www.eoearth.org/article/Deforestation_in_Amazonia; Margulis, S. (2004); Woods
Hole Research Center (2007)
19
Walker, R. and Moran, E. (2000)
20
Woods Hole Research Center (2007)
21
Margulis, S. (2004)
22
Walker, R. and Moran, E. (2000)
23
Margulis, S. (2004)
24
Woods Hole Research Center (2007)
25
Ibid
26
Foley et al (2007) “Amazonia Revealed: Forest Degradation and the loss of ecosystem goods and
services in the Amazon Basin” Frontiers in Ecological Environment, Issue 5(1), p. 25-32
27
Foley et al (2007) citing Nepstad, D. et al (1999) “Large scale impoverishment of Amazonian forests
by logging and fire” Nature Issue 398, April 8, p. 505-508 and Anser, G. et al (2005) “Selective
Logging in the Brazilian Amazon” Science Issue 310, p. 480-82
28
Woods Hole Research Center (2007b) “Logging and Family Forests”
http://www.whrc.org/southamerica/logging_fam_forests.htm
29
Foley et al (2007)
10
2.2 Papua, Indonesia
Papua is the largest province in Indonesia and accounts for almost a quarter of
Indonesia’s land area (Figure 4). The percentage of forest cover in Papua and West
Papua (now a separate province, shown in pink in Figure 4) was 85% in 1985
30
with
satellite imaging analyses reporting an estimated reduction to approximately 73% by
2005.
31
The rate of loss of primary forest in Indonesia from 2000-2005 was 2.6% per
annum totalling a loss of 7.2 million hectares over the five-year period.
32
This rate
was a 26% increase over the period from 1990-2000. Data specific to Papua is
difficult to obtain, but the rate of loss of primary forest for neighbouring Papua New
Guinea has been 0.95% annually from 1990-2005. This is a loss of 250,000 hectares
per year.
33
Figure 4 – Papua Province, Indonesia
Timber companies are increasingly turning their attention to Papua’s natural forests,
and large scale timber extraction concessions have led to significant clearing of the
province’s primary forests.
34
Forest concessions allowing harvesting of natural
30
Global Forest Watch et al (2001) “The State of the Forest: Indonesia” Global Forest Watch, World
resources Institute and Forest watch Indonesia www.globalforestwatch.org
31
Sumantri, H. and Wijayanto, I. (2005) “Land cover mapping using satellite imagery and GIS in the
Mamberamo Basin and Raja Ampat Islands, Papua, Indonesia” www.mapindia.org/2005/papers
32
FAO (2005)
33
Ibid
34
Tokede, M., William, D., Widodo, Gandhi, Y., Imburi, C., Patriahadi, Marwa, J ., Yufuai, M. (2005)
“The impacts of Special Autonomy in Papua’s Forestry sector: Empowering Customary Communities
(Masyarakat Adat) in decentralised Forestry Development in Manokwari District” Centre for
International Forestry Research, Bogor, Indonesia
11
forests have been granted over 6.5 million hectares of the province’s 42 million
hectares.
35
While concession requirements include preparing environmental impact
assessments and forest management and monitoring plans, logging practices typically
result in over-exploitation and loss of the forest structure.
36
Inventories of harvested
areas in Papua found that 58% of the remaining trees with diameters of 20 cm or
larger were damaged, compromising the ability of the forest to recover and provide
sustainable yields of timber.
37
In addition to these pressures, illegal logging activities
are increasing, particularly as forest resources diminish in the Indonesian provinces of
Kalimantan, Sulawesi and Sumatra.
38
Increasing international demand for green energy sources has led to a demand for new
land to produce palm oil, a feedstock for bio-diesel. Most of the global expansion of
the oil palm plantation estate has occurred in Indonesia and Malaysia.
39
Predictions
are that about half of the new plantation land – 3 out of 6 million hectares – needed to
supply the global palm oil market by 2020 will be established in Indonesia.
40
National government plans for Papua to absorb much of this expansion would dwarf
the province’s current plantation estate of 58,000 hectares (as of 2002).
41
The
provincial government is actively pursuing alternate investment strategies related to
carbon credit generation to avoid this massive land conversion.
42
In Papua New
Guinea, the oil palm plantation estate doubled between 1990 and 2000 to 73,000
hectares, and this growth is expected to accelerate.
43
Irrespective of whether this land use change occurs in Papua or elsewhere in
Indonesia, it would represent a significant loss of frontier forest in the Asia Pacific
region. The land use pressures in Papua are representative of challenges faced across
the region’s remaining tropical forests. Unsustainable logging and oil palm plantation
establishment are the two activities studied here as the baseline scenarios for Papua.
While regional variables will influence the financial viability of logging and palm oil
development across the region, the models for Papua can generally be applied
elsewhere.
35
Papua Ministry of Forestry (2006) List of Forest Concession Right (Hak Pengusahaan Hutan) and
Timber Forest Product Utilisation Permit (Izin Usaha Pemanfaatan Hasil Hutan Kayu) holders, August
36
J arvis and J acobsen (2006) “Working paper – Incentives to promote forest certification in Indonesia”
Project: Motivating Sustainability, International Finance Corporation
37
Tokede et al (2005)
38
Ibid
39
Wilkinson, M. (2007) “Green Fuel Gets a Black Name” Sydney Morning Herald October 13; Painter,
J . (2007) “Losing Land to Palm Oil in Kalimantan” BBC News, August 3
40
WWF Germany (2002) “Oil palm plantations and deforestation in Indonesia: What role to Europe
and Germany Play?” http://www.fire.uni-freiburg.de/se_asia/projects/wwf_oil.htm
41
Friends of the Earth (2005) “Greasy Palms. The social and ecological impacts of large-scale oil palm
plantation development in southeast Asia”
http://www.foe.co.uk/resource/reports/greasy_palms_impacts.pdf
42
Tedjasukmana, J . (2007) “Heros of the Environment: Leaders & Visionaries” Time Magazine
http://www.time.com/time/specials/2007/article/0,28804,1663317_1663319_1669895,00.html
43
Ibid
12
2.3 Democratic Republic of Congo
The Congo Basin is the world’s second largest tropical rainforest covering 2 million
square kilometers across the central African nations of the Democratic Republic of
Congo (DRC), the Central African Republic, Cameroon, the Republic of Congo,
Equatorial Guinea and Gabon (Figure 5). Fifty-three percent of the rainforest is
within the DRC, the area of focus in this study.
Figure 5 – Congo Basin spanning five central African nations
44
*
* Encircled areas are priority landscapes based on high biodiversity value and intact forest
Compared to many of the world’s tropical rainforests, the Congo Basin is relatively
healthy.
45
This is primarily a result of the low population density and limited
commercial activity. The annual rate of deforestation from 2000-2005 was 0.24%,
which actually represented a 37% decrease from the period of 1990-2000.
46
This may
be a result of decreased economic activity during Congo’s civil war from 1998-2003,
and the country is likely to open to global markets as political stability improves.
Nearly 60% of the DRC’s total forest area is thought to
be productive or commercially
valuable,
47
and logging is currently the main commercial activity contributing to
deforestation.
48
Approximately 20 million hectares of the DRC’s 145 million hectare
forest estate are allocated as timber concessions to about 60 companies, although only
44
Congo Basin Forest Partnership (2005) “The Forests of the Congo Basin: A Preliminary
Assessment” (online) http://carpe.umd.edu/products/PDF_Files/FOCB_APrelimAssess.pdf
45
Congo Basin Forest Partnership (2005)
46
FAO (2005)
47
Congo Basin Forest Partnership (2005)
48
Bavier, J . (2007) “Congo to Cancel Logging Deals to Protect Forests: Reuters, J une 21; Congo Basin
Forestry Partnership (2005)
13
about a dozen are in operation.
49
Three main companies – Siforco (owned by
Germany’s Danzer Group), Sodefor (a Portuguese-owned unit of holding company
NST) and Safbois (an American and Belgian owned conglomerate) – account for over
two-thirds of the country’s current timber production capacity.
50
The government is currently carrying out a World Bank-led legal review of 156
logging licenses covering 21 million hectares, half of which are in intact forest
landscapes. Around three million hectares of illegal logging concessions have already
been cancelled, and the government claims it is willing to cancel an additional 12-15
million hectares.
51
The outcome of these processes and the enforcement of
government policies remain uncertain.
Currently, logging in the DRC is largely driven by demand from Europe for certified
sustainable timber
52
(e.g. Forest Stewardship Council certification). The general
practice in the region is selective logging of high-value species for export. One to two
trees are removed per hectare (ha)
53
with an average of 75-80% of the canopy cover
remaining intact.
54
Felling cycles (in a properly managed tropical forest) rarely
exceed 30–40 years.
55
Estimates of production range from 5-6 cubic meters (m
3
) /ha
56
to 8-12m
3
/ha
57
.
Economic drivers such as increased timber demand from Asia, which is more
accepting of non-certified wood, and increasing population pressures suggest that
there could be an increase in unsustainable logging practices in the near term. For
example, the ITTO reported in late 2006 that China has replaced Europe as the price-
setter for African timber.
58
With only a dozen of the 60 companies holding logging
concessions currently in operation,
59
there is scope for increased production capacity
as non-active concession holders become operational or relinquish licenses that are
acquired by operational companies.
60
In nearby Ivory Coast, Ghana and Uganda,
49
Forests Monitor (2007) “The Timber Sector in the DRC: A Brief Overview”
http://www.forestsmonitor.org/uploads/2e90368e95c9fb4f82d3d562fea6ed8d/Description_of_the_Tim
ber_Sector_in_the_DRC.pdf
50
Bavier, J . (2007)
51
Ibid
52
Forests Monitor (2007)
53
Karsenty, A., and Gourlet-Fleury, S. (2006) “Assessing sustainability of logging practices in the
Congo Basin’s managed forests: the issue of commercial species recovery. Ecology and Society Issue
11(1), p. 26 (online) http://www.ecologyandsociety.org/vol11/iss1/art26/
54
Nasi, R. (2005) “Potential Methodological Flaw in the Examination of the Effects of Logging”
Ecology and Society Issue 10(2) (online) http://www.ecologyandsociety.org/vol10/iss2/resp2/
55
Karsenty, A., and Gourlet-Fleury, S. (2006)
56
Ruiz-Perez, M. et al (2005) “Logging in the Congo Basin: A Multi-Country Characterization of
Timber Companies” Forest Ecology and Management Volume 214, Issues 1-3
57
Karsenty, A. and Gourley-Fleury, S. (2006)
58
ITTO (2006) “Asia Takes Price Leadership from Europe in Africa” Tropical Forest Update Volume
16, No. 4 http://www.itto.or.jp/live/Live_Server/3163/tfu.2006.04.e.pdf
59
Forests Monitor (2007)
60
Greenpeace suggests that this could be an outcome of the World Bank led concession review
currently underway. (See “Carving Up the Congo”
http://www.greenpeace.org/raw/content/international/press/reports/carving-up-the-congo-exec.pdf)
14
favorable economic markets for both timber and agricultural products, such as cacao,
also contributed to the deforestation of old growth forests in those countries.
61
Should Congo’s forests come under greater threat of degradation from increased
logging as the country opens to global markets, a price signal for carbon conservation
could help the country develop land uses that incorporate a full range of forest values.
Studying logging as the main driver of deforestation in this region is therefore more
forward looking and hypothetical than the other regions, but it provides an early
assessment of how carbon finance could mitigate the rapid expansion of logging
activities, as market demand for timber inevitably increases.
3 Methodology and Assumptions
This study evaluates the viability of carbon finance as a mechanism for forest
conservation via a comparison of the net present value (NPV) of carbon-financed
forest conservation versus the land uses driving deforestation. This analysis is
considered from the perspective of private investors making investment decisions on
land use options for existing forested areas. This section describes the methodology
used to conduct the analysis and outlines assumptions underlying the financial
modelling.
3.1 Framework for Analysis
There is no accepted regulated market standard for defining and accrediting avoided
deforestation credits. Many ideas are currently being discussed for what is being
referred to as reduced emissions for deforestation and degradation (REDD). A
position favoured by the UNFCCC, World Bank and others takes a national
accounting approach in which countries establish a baseline deforestation scenario
and receive credit for reducing emissions below the baseline.
62
Other accounting
approaches have taken a more project-level perspective or combined national and
project-level mechanisms (the “nested approach”
63
).
The second major question is whether credits from REDD will be fungible with other
carbon offsets. Some proponents suggest a completely fungible market. A dual
markets approach has also been proposed in which avoided deforestation credits are
sold in a parallel market but are not fungible with emission reduction offsets as
defined under markets of the Kyoto Protocol.
64
61
Laporte et al (2004)
62
World Bank Carbon Finance Unit (2007) “Forest Carbon Partnership Facility”
http://carbonfinance.org/Router.cfm?Page=FCPF&FID=34267&ItemID=34267
63
Pedroni, L. and Streck, C. (2007) “Mobilizing Public and Private Resources for the Protection of
Tropical Rainforests” CATIE and Climate Focus
http://www.climatefocus.com/newspubs/downloads/PolicyBriefonREDD_000.pdf
64
Centre for Clean Air Policy (2007) “Reducing Emissions from Deforestation and Degradation: The
Dual Markets Approach” http://www.ccap.org/international/FINAL%20REDD%20report.pdf
15
Defining baseline scenarios is fundamental in all of these policy debates and is the
fundamental first step in evaluating activities that are seeking carbon finance.
65
A
baseline scenario explains the likely land use in the absence of the emission reduction
activities. Avoided deforestation as a general project type assumes that the baseline
activity results in forest degradation or conversion to a non-forest activity. The
emissions “avoided” by pursuing a forest conservation management strategy rather
than the baseline activities represent the volume of saleable carbon.
For this study, a framework for analysis was designed to compare the NPV of baseline
activities to the conservation scenario. This framework includes the following steps:
• specify baseline activities for each region
• determine the NPV of the cashflows of the baseline activity
• estimate carbon stock loss associated with the baseline activity
• quantify the NPV of maintaining carbon stock by forgoing the baseline
activity
3.2 Definition of Baseline Activities
As discussed in Section 2, the baseline scenarios considered in this study are:
• Cattle ranching in the Amazon
• Unsustainable logging in Papua
• Oil palm plantation development in Papua
• Unsustainable logging in the DRC
Modelling is based on a hypothetical frontier area of 250,000 hectares. This
approximately represents the size of logging concessions in Papua and the Congo but
is much larger than individual landholdings for ranching in the Amazon and oil palm
plantations. This scale is used to approximate land use dynamics across the frontier,
whether that is driven by a single landowner or the cumulative impact of multiple
landowners.
The entire concession area is assumed to be impacted over a 20-year period. This
represents a deforestation rate of 5% per annum, totaling a conversion of 12,500
hectares annually. This rate is reasonable in terms of the operational capabilities of
the land uses in the baseline scenarios.
Ideally, each scenario would apply a deforestation rate based on the country’s
historical and projected rates of deforestation. Although many data points could be
used, the best method for determining these baseline loss rates is an important point of
65
See, for example, the Voluntary Carbon Standard “Guidance for Agriculture, Forestry and Other
Land Use Projects” www.v-c-s.org/docs/AFOLU%20Guidance%20Document.pdf and California
Climate Action Registry “Forest Sector Protocol”
http://www.climateregistry.org/docs/PROTOCOLS/Forestry/Forest_Sector_Protocol_Version_2.1_Sep
t2007.pdf
16
debate in policy discussions around the accreditation of REDD activities. FAO
deforestation statistics presented for each region in Section 2 suggest lower rates
(0.8% in Brazil, 2.6% in Indonesia and 0.24% in Congo), but this data is aggregated
nationally and underestimates rates at the forest frontier. Rates of deforestation also
appear low when considered over a large intact forest land base, such as those in the
Brazilian Amazon and the Congo Basin. A rate of 5% is therefore generalized and is
applied as a simplified approach, which would need further investigation for specific
areas.
Definitions of baseline activities are as follows:
Cattle Ranching Forested land in the hypothetical 250,000 hectare study area is
converted to pasture at a rate of 12,500 hectares per annum. Once a hectare is
converted, it is assumed to become productive in the following year and remains
productive through the project life (to 2040). It is assumed that the net cashflows per
hectare over the project life incorporate start-up costs and operating costs. Any initial
revenue from the liquidation timber sales upon conversion is assumed to be included
in the average net cashflows per hectare over the project life, although in many cases
trees are burned onsite rather than taken to market.
66
Operating costs and cattle prices
are assumed to be static over the project life and do not take into account changing
market conditions.
Logging Logging scenarios assume that reduced impact logging is not applied,
resulting in degradation of logged areas. This results in little or no regrowth of
merchantable timber; therefore a hectare is logged only once. At a rate of 12,500
hectares per annum, the 250,000 hectare area is logged over in 20 years. Land uses
and revenues post-harvesting, such as conversion to agribusiness, are not included in
the model. It can be assumed that population growth would result in small-scale
cultivation and urbanization of these areas in the short term, which is beyond the
scope of the commercially driven activities considered in this study. The assumed
yield of merchantable timber from the initial harvest is 20m
3
/ha.
Oil Palm For oil palm, the same rate of conversion (12,500 hectares per annum) is
applied, resulting in complete conversion of the study area over 20 years. The model
does not factor in the sale of timber from conversion activities under the assumption
that plantation development often occurs on degraded secondary forest, where limited
timber revenues are available, with remaining timber burned on site. The degradation
of the forest for timber as an intermediary process is captured in the land purchase
price of the palm oil investment.
3.3 Discount Rates
Discount rates are a crucial component in determining the NPV of cashflows of these
baseline activities. A discount rate represents the investors cost of capital over time,
or hurdle rate. It ultimately determines the investor’s expected financial return over
66
Bowman, M. (2007) pers comm Project Assistant, Woods Hole Research Center, November 2
17
the project life (internal rate of return, IRR) based on the amount of risk ascribed to
the activity. Investors will ascribe higher discount rates to projects that are assumed
to involve more risk. Risk considers many factors, including market knowledge,
market certainty, country sovereign risk (i.e. defensibility of property rights and
contracts), currency risk, timing of cash flows and availability of skilled labour and
physical capital.
For this study, a 20% real discount rate was used. High performing cattle ranches
deliver a real IRR of up to 20%,
67
and 20% real IRR is likely the minimum return
forest investors would expect in these countries. Anecdotal evidence suggests that
returns from oil palm plantations are currently higher than this and are increasing with
the rising price of crude palm oil. This reflects the rationale for the rapid expansion of
this land use in many places.
The stability and size of beef, timber and palm oil markets provides investors with a
relative level of certainty, although risk remains related to project execution and
market volatility. Discount rates applied to current carbon finance projects would
likely be higher given the current carbon market uncertainty. Nevertheless, this study
is designed to test the impact of establishing a set of rules and market price signals for
carbon offsets from REDD. Assuming such a market is established, investors would
be increasingly likely to use similar discount rates for carbon-related investments as
are applied for agribusiness investments. Carbon investors may even apply lower
discount rates, since the relative establishment and operational costs are minimal,
involving less project delivery risk.
However, it is acknowledged that there is currently little certainty around how REDD
credits will be incorporated into carbon markets. A major outstanding issue is how
project-based forest conservation may be incorporated into national baseline
accounting policies. National accounting requires countries to adopt a country-level
deforestation baseline, and credits would only be saleable if the country as a whole
reduces its emissions below the accepted baseline.
68
This creates a problem for
investors who may reduce emissions over a project baseline but are then precluded
from selling in international markets because of host country non-performance.
With this level of uncertainty, investors need to consider the risk-return profile of
carbon projects, and the implications of this are discussed in Section 4.
3.4 Net Present Value of Baseline Activities
The NPV of the cashflows from baseline activities represents the total cashflows over
the project lifetime discounted back to today’s dollars. This methodology enables a
comparison across investment options using a single metric of present value derived
from the activity (effectively the current purchase value that the land or forest should
67
Margulis, S. (2004)
68
Pedroni, L. and Streck, C. (2007)
18
be worth to an investor contemplating those activities). All modelling is in real terms
(i.e. values are in current dollars and do not account for inflation over time).
A range of approaches was used to determine the NPV per hectare based on available
information and market knowledge. Cashflows are modelled through 2040 for cattle
ranching and palm oil. Cashflows for logging end in 2028, when the resource is
considered to be depleted. (Carbon scenario cashflows are modelled through 2040.)
Cattle Ranching Generalizing the NPV of cattle ranching is difficult because costs
and revenues are highly variable. Costs are dependent on whether landowners are
converting large forest areas directly or buying smaller plots already converted
through intermediary processes. However, as ranching is the final productive activity,
it can be said to underwrite all prior costs and revenues, such as speculative land
purchases and selling timber from initial clearing (i.e. investors in intermediate
logging use land sale to ranching as a “terminal value” for their activities).
69
Income
also varies by type of ranch activity (rearing, fattening, etc.) and regional variations,
such as soil quality and distance to slaughterhouses.
Published data is sparse but one major World Bank study estimates a NPV of $500/ha
applying an 8% discount rate.
70
This is corroborated by another study in which the
NPV per hectare is estimated at $567 (6% discount rate) adjusted to $197 per hectare
at a 12% discount.
71
These discount rates generally reflect low risk projects involving
land that has already been cleared. Projects at the frontier would likely involve higher
risk, and therefore discount rates closer to 20% would be reasonable, as discussed.
The World Bank study reports profits per hectare of $77 per year.
72
Using this
estimate, the NPV of cashflows from the baseline scenario (12,500 hectares converted
to production annually over 20 years with cashflows to 2040) yields a per hectare
NPV of $111/ha, applying a 20% discount rate. Preliminary results from a third study
indicate that NPVs range from $0-1150/ha across the entire Brazilian Amazon.
73
For
the purposes of this study, a median value of this range ($575) was averaged with
$111/ha to yield an NPV of $343/ha.
Logging For logging, cashflows were derived from profits per hectare per year based
on literature and our access to data from regional forestry consultants. The NPV was
then derived based on cashflows over 20 years, at which point the forest is assumed to
be degraded and no longer viable for commercial logging.
In Papua, cashflows were based on estimated average log prices (on barge) minus
production costs. Based on information from industry sources operating in Papua, log
prices are assumed to be $150/m
3
, which is an average of prices achieved for high-
grade logs, medium-grade logs (i.e. Meranti and Agathis) and mixed species. Costs
69
Margulis, S. (2004)
70
Ibid
71
Mattos, M. and Uhl, C. (1994) “Economic and Ecological Perspectives on Ranching in the Eastern
Amazon” World Development Vol. 22, No. 2, p. 145-158
72
High productivity ranches in the best soil regions yield profits of R$138/hectare (Margulis 2004)
multiplied by the conversion rate of 1.76 to convert to US$.
73
Merry, F. (2007) pers comm. Discussing work in review. Woods Hole Research Center
19
were derived from an International Finance Corporation working paper
74
and
incorporate information provided by experienced foresters working in the region in
addition to New Forests’ operational experience in tropical forests. Total production
costs are approximately $120/m
3
assuming logs are loaded on a barge for transport to
Asian markets or to a local processing site near the coast.
This cost estimate represents natural forest management at a level that would meet
reputable sustainable forestry certification standards, i.e. reduced impact logging.
Concessions managed to these standards are rare. Specifically in Papua, there are no
certified natural forest extractive management operations. Therefore, the cost
structure was adjusted to remove activities that would likely be forgone in
unsustainable management regimes (silviculture treatments, planning, rehabilitation
and official taxes). A cost structure of $80m
3
was used to reflect unsustainable
practices. (Line item costs and further discussion are provided in Appendix A.)
Log prices per cubic meter ($150) minus operational costs under an unsustainable
logging scenario ($80) yield a profit of $70/m
3
. With 20m
3
of merchantable timber
per hectare and 12,500 hectares logged annually, this generates an annual cashflow of
$17.5 million. The NPV of project cashflows over 20 years is $350 per hectare,
applying a 20% discount rate.
In the DRC, log prices are reported at $220
75
to $300
76
per cubic meter corroborated
by ITTO average log (on barge) prices for Khaya (African mahogany) and Meranti of
$300.
77
Taking an average ($260) and using the same operational costs assumed in
the Papua scenario, the profit margin is assumed to be $160/m
3
.
Given that increased production in the DRC would necessitate harvesting less
valuable species, it was assumed that $160/m
3
profit was achieved for the first 10m
3
.
The remaining 10m
3
were assumed to yield a profit of $100/m
3
. This resulted in an
average profit of $130/m
3
. The NPV of the cashflows over 20 years is $633 per
hectare, applying a 20% discount rate.
It is important to note that the availability and value of 20m
3
of merchantable timber
in the Congo requires further investigation. In particular, if market pressures result in
a shift toward the unsustainable logging practices assumed here (e.g. shift to Chinese
markets), price premiums for certified sustainable timber would be lost. Deriving
profits from current prices may over-value actual prices in an unsustainable logging
scenario, and this should be taken into consideration.
Palm Oil A discounted cashflow model was built to determine plantation
establishment costs and crude palm oil yields over time. (Appendix B provides a full
list of cost and revenue assumptions and sources.) In brief, costs include a $300/ha
74
J arvis, B. and J acobson, M. (2006)
75
Forests Monitor (2007)
76
ITTO (2005) “Annual Review and Assessment of the World Timber Situation” Division of
Economic Information and Market Intelligence, Yokohama, J apan
http://219.127.136.74/live/Live_Server/2151/E-AR05-Text.pdf
77
ITTO (2007) “Market Trends” Tropical Forest Update Volume 17, No. 1
http://www.itto.or.jp/live/Live_Server/3241/tfu.2007.01.e.pdf
20
land purchase price and establishment costs of $2555 per hectare, both incurred as
establishment occurs (12,500 hectares per annum). Operating costs total $500 per
hectare per year, and harvesting costs are priced at $187 per hectare. The model
assumes that beginning at year four, each hectare of plantation produces 0.64 tonnes
of crude palm oil. In a linear fashion, the palm oil yield increases to 4 tonnes of crude
palm oil per hectare at age 7, stays constant to age 12 and steadily declines until age
25, when trees are removed and replanted. Crude palm oil is sold at current market
prices of US$886 per tonne over the project life. The net present value of the
cashflows to 2040 is $757 per hectare, applying a 20% discount rate.
3.5 Carbon Quantification
In addition to determining the NPV of the baseline activities, a carbon accounting
methodology is required to quantify the volume of carbon emissions that are avoided
by forgoing the baseline scenario. The assumed change in carbon stock over time
between the baseline scenario and the conservation scenario represents the quantity of
carbon offsets that can be created and sold. Realistic assumptions are therefore
required on the rate of conversion for each land use and associated emissions.
General assumptions are:
• It is assumed that there is an average of 549 tCO
2
/ha stored in the standing
forest in all three regions. This represents 150 tonnes of carbon (C) per
hectare multiplied by the molecular weight conversion factor of 3.66 to
estimate metric tonnes of carbon dioxide equivalence (tCO
2
e), the standard
unit traded in carbon markets. This figure is the average of multiple data
sources that report approximately 100-200 tC per hectare for tropical
rainforests.
78
• With regard to cattle ranching, it is assumed that 86% of the biomass is
removed from the site immediately upon conversion through burning.
Remaining biomass stays on site and decays over time. The minimum
residual carbon stock is 50 t CO
2
/ha, representing primarily below ground
biomass.
• With regard to logging scenarios, it is assumed that after harvesting 10%
of stored carbonis immediately lost (i.e. removed in the form of logs
79
);
25% decays over time, either on site or when debris, litter and soil carbon
78
FAO (2005) http://news.mongabay.com/2006/1031-deforestation.html; Gibbs, H.K, S. Brown, J . O.
Niles and J . A. Foley (2007) “Monitoring and estimating tropical forest carbon stocks: making REDD
a reality” Environmental Research Letters (in press); Margulis, S. (2004); Saatchi, S. et al (2005)
“Spatial Distribution of Carbon Stock in the Amazon Basin” American Geophysical Union, Fall
Meeting 2005 http://adsabs.harvard.edu/abs/2005AGUFM.B54B..03S
79
The standard approach in carbon trading systems is to assume that the carbon stored in logs is
released upon harvest. While carbon may remain stored in wood products for years, decaying later in
landfills or through other processes, this delay is not currently factored into carbon accounting, and
convention is applied here.
21
is washed into waterways; 65% remains in standing forest with 10% lost
each year from fire, illegal logging, mining, agriculture and reclassification
for agribusiness, etc. The minimum residual carbon stock is 50tCO
2
/ha.
• For palm oil, 100% of the biomass is assumed to be removed from the site
immediately with a minimum residual carbon stock of 50tCO
2
/ha. The
model accounts for the carbon sequestered as the palm plantation grows,
which is subtracted from the volume of saleable carbon in the avoided
deforestation scenario.
Standing carbon stocks are regionally variable, and the minimum carbon stock of
50tCO
2
/ha is an estimate based on the likely percentage of the remaining above and
below ground biomass after burning and degradation. Both require refinement to
improve accuracy at a local scale.
3.6 Carbon Value
As discussed, there is no accepted regulated market standard for defining and
accrediting avoided deforestation credits. For this study, it was necessary to make
assumptions about how avoided emissions may be credited and priced. Crediting
periods were designed based on general current practices in voluntary and regulated
markets. Price points were derived from recent transactions and market data as
explained in the notes to Table 1.
Scenario 1 A once only calculation is made of the total credits generated from
foregoing the baseline scenario. All credits are sold in the voluntary
market from 2008-2012. This is called ex-ante accounting, in which
credits are sold before the emission reduction is assumed to have
occurred. Products applying ex-ante accounting tend to attract a lower
market price.
Scenario 2 Bi-annual calculation of credits representing the carbon that would
have been released in the previous 2 years under the baseline scenario.
In this scenario credits are issued and sold bi-annually. Credits are
assumed to be sold to the voluntary market pre-2012 and then sold
through an international regulated market (i.e. Kyoto second
commitment period). All accounting is ex-post whereby credits are
sold after the emission reduction is assumed to have occurred,
generally attracting higher prices.
Scenario 3 Calculation and sale of credits every five years from 2012 onward.
The first tranche is sold in an international regulated market (i.e. Kyoto
second commitment period) in 2017 and every five years thereafter.
No sale of the credits occurs prior to 2017 or in voluntary markets. All
accounting is ex-post whereby credits are sold after the emission
reduction is assumed to have occurred, generally attracting higher
prices.
22
Table 1 – Carbon trading scenarios and prices
Scenario Volume Market Value
1
1/5 of the total credits generated
are sold each year between
2008-2012
Wholesale credits for
voluntary markets
US$2.95 *
2
credits are sold bi-annually for
the quantity of emissions
avoided over the baseline
Voluntary markets to
2012 & international
regulated market
thereafter
US$10/tCO2 **
US$18/tCO2 after
2012 ***
3
credits sold in 5-year tranches
for emissions avoided over the
baseline starting with the 2012-
2017 tranche
International
regulated market
US$18/tCO2 ***
* Average of approximate World Bank Biocarbon fund prices and a recent transaction in which Rio
Tinto purchased credits associated with avoided deforestation of farmland in Queensland, Australia.
80
These are voluntary prices for ex-ante products.
** An average price in the voluntary market for ex-post forestry products.
81
*** The average price for issued Certified Emission Reductions from the Clean Development
Mechanism in 2006,
82
representing the approximate value of a project-based offset in a regulated
international market.
Investors would likely pursue forward carbon sales, purchase options and other
transaction structures to improve the economics of carbon investments. These more
sophisticated structures are not included in the modeling, likely understating the value
of carbon scenarios. This should be taken into consideration.
It is assumed that REDD credits are permanent and therefore equivalent with other
emission reduction credits, trading at comparable per tonne prices.
Ten percent of the total carbon revenue was deducted from carbon sales to account for
transaction costs and forest management. Transaction costs include developing a
carbon accounting system, project documentation, third-party verification, scheme
accreditation or verification to a voluntary standard and management/brokerage fees.
Forest management costs include forest protection and monitoring over time and
could include payments to local communities for services such as boundary
maintenance and collecting inventory data.
80
See http://www.carbonpool.com/
81
Hamilton, K., Bayon, R., Tunrer, G. and D. Higgins (2007) “State of the Voluntary Carbon Markets
2007: Picking Up Steam” Ecosystem Marketplace & New Carbon Finance
www.ecosystemmarketplace.com
82
J otzo, F. (2007) “CDM Project Design” Centre for Energy and Environmental Markets, University of
New South Wales, Applied CDM Short Course, J uly 18
th
23
4 Discussion
Table 2 summarizes the NPV of the cashflows for each baseline activity against
cashflows from each carbon scenario. Results are reported in NPV per hectare.
Table 2 – NPV per hectare of baseline and carbon scenarios
Brazil - cattle Papua - logging Papua - palm Congo - logging
Baseline $343 $350 $757 $633
Carbon
Scenario 1
$660 $649 $550 $649
Carbon
Scenario 2
$1168 $756 $981 $756
Carbon
Scenario 3
$521 $402 $386 $402
Results indicate that, assuming carbon rules are adopted and implemented, carbon
revenue from avoided deforestation may compete with current alternate land uses at
the forest frontier. Carbon appears to effectively compete with logging in Papua and
cattle ranching in the Amazon. The case is not as clear for palm oil and logging in the
Congo, where high profits make it difficult for carbon to compete in some scenarios.
The quantum of saleable carbon associated with the baseline scenarios is
approximately 125-145 million tCO
2
.
In summary:
• Carbon appears competitive in comparison with cattle ranching in Brazil
• Carbon appears competitive in comparison with unsustainable logging in
Papua
• Carbon competes with palm oil revenues in Papua if the project can sell
credits regularly (bi-annually) and make initial sales into the voluntary
carbon market (carbon scenario 2)
• Carbon can compete with logging in the Congo if crediting occurs
regularly (bi-annually) and some sales are made in the near term in
voluntary markets (carbon scenario 2) or all credits are sold in the
voluntary market in the short term (scenario 1)
These results suggest that under certain policy settings, private financing could flow
toward conservation. This is a generalized model, however, and several points need
to be addressed in interpreting the results, including differences between regional land
use scenarios, variation across carbon scenarios, policy certainty and methodology
considerations.
24
4.1 Land Use Comparisons
Across the regions considered, carbon competes most effectively with cattle ranching
in the Amazon. This is based on high carbon values, which reflect the immediate and
permanent change in carbon stock when land is cleared for pasture. It is also a result
of ranching delivering a relatively low NPV, which is consistent with the fact that
cattle ranching is not considered the highest economic use of the land. As discussed,
ranchers are motivated by factors other than financial returns, including securing land
rights and generating low-risk cash flows. Carbon investments for forest conservation
would need to provide for or compete with these preferences in order to be more
attractive to investors than ranch development.
In Papua results show that a strong case can be made for carbon-financed forest
conservation investment when compared to logging revenue. For palm oil, the case is
not as clear. The results that show that only an investment strategy with
characteristics of scenario 2 – frequent crediting and initial sales into a voluntary
market – can compete with palm oil.
However, it is important to note the impact of discount rates in the case of oil palm
production. There are large initial capital costs in establishing an oil palm plantations
and cash flow only occurs as plantations grow. The result is that small changes in
discount rate have large effects on the net present value and, thus, the relative
competitiveness of carbon finance versus oil palm development. More information
regarding investors’ risk-return expectations is needed to explore this scenario more
fully.
In the Congo, high log prices make carbon less competitive that in the Papua logging
comparison. Timber from the DRC is primarily sold to European markets, which are
prepared to pay higher prices for sustainable, certified wood.
83
The market is
increasingly being driven by demand from China and India, which is currently
pushing prices of African species such as Khaya (African mahogany) and Meranti to
near historical highs.
84
As production increases to meet this demand, less valuable
species will be logged, and there will be a move away from sustainable harvesting that
meets certification standards. These drivers would result in lower log prices.
The model accounts for these market changes in a simplified way by reducing profits
per cubic meter from $160 for the first 10m
3
to $100 for the remaining 10m
3
. This is
an approximation for which more detailed market analysis is required. A reduction in
timber prices would make carbon more competitive in the region. The availability of
20m
3
of merchantable timber per hectare also needs to be verified.
83
Forests Monitor (2007)
84
ITTO (2007) “African Prices Stable on the Back of Steady Demand” Tropical Forest Update
Volume 17, No. 2 http://www.itto.or.jp/live/Live_Server/3453/tfu.2007.02.e.pdf
25
4.2 Carbon Price Scenarios
Scenario 2 is the most competitive of the three scenarios presented. This is the result
of a medium to high carbon price (US$10-18/tCO
2
) delivered every two years. While
Scenario 3 has an overall higher price point (US$18), revenue is delayed until 2017
and every 5 years thereafter, which reduces overall value after discounting.
Scenario 1 delivers all of the revenue in the first 5 years. This is a trade off with the
low carbon price of US$2.95, as market evidence suggests that a product sold upfront
in this way would likely attract a lower price. This scenario is somewhat aggressive
in suggesting that the total project carbon volume (approximately 125-145 million
tCO
2
) would be sold in the voluntary markets before 2012, even at this low price
point. Total transactions in the global voluntary market were estimated at 91 million
tonnes in 2006.
85
Although this market is posed for growth, transactions of the size
suggested by Scenario 1 would likely need to be motivated by future acceptance in a
regulatory scheme, where they could be sold in a secondary market at a higher price.
Policy uncertainty is likely to prevent this in the short term.
Scenario 3 represents the most conservative point of comparison. Revenue is skewed
toward the future, reflecting the reality that it will likely be several years before
policies are clear enough to drive investment in conservation carbon. Additionally,
the price point of US$18tCO
2
e, while the highest of the scenarios, may be lower than
future carbon prices in regulated markets. This price represents the value of issued
Certified Emission Reductions (CER) in the Clean Development Mechanism (CDM)
in 2006, a reasonable indicator of the value of project-based offsets. However, CERs
traded as high as $25/tCO
2
e in 2006, and demand for project-based offsets is expected
to increase.
86
Comparatively, allowances in the EU Emissions Trading Scheme are
currently trading for US$33.10,
87
although forestry credits are not currently eligible in
this market. Prices for project-based offsets may be higher in real terms in the future,
particularly as low-cost opportunities, such as Chinese projects to reduce HFCs, are
no longer available.
Conversely, credits from avoided deforestation may not be equivalent to other project-
based offsets, and this would result in lower prices. REDD credits may be treated as
temporary, as is current practice with afforestation/reforestation projects under the
CDM. There is some discussion of dual markets in which REDD credits are sold in a
parallel market and are not fungible with emission reduction offsets as defined under
Kyoto.
88
Additionally, the Voluntary Carbon Standard, which is emerging as a
leading standard for accrediting voluntary carbon projects, suggests that forestry
conservation projects should reduce the amount of saleable carbon by 5-30% (based
on a project risk assessment) as a buffer against future carbon stock losses. Policies
that result in REDD credits being non-equivalent to other emission reduction credits
or require large buffers reduce the value of carbon investments.
85
Hamilton et al (2007)
86
Carpoor, K. and Ambrosi, P. (2007) “State and Trends of the Carbon Market” World Bank/IETA,
Washington D.C.
87
As of November 15, 2007 according to Point Carbon http://www.pointcarbon.com/
88
Centre for Clean Air Policy (2007)
26
4.3 Policy Uncertainty
Investors must determine the required threshold return for each investment by
considering a range of risks and uncertainties. In voluntary carbon markets, those
uncertainties typically include buyer demand and achievable per tonne prices. In
regulatory markets, major risks are associated with project delivery and accreditation.
Existing carbon funds seek returns in the order of 35% real IRR, often based on
paying a very low price for carbon offsets prior to accreditation and then sharing some
of the upside with project developers. For projects where rules are uncertain or
untested, IRRs of 40-50% or more would be expected.
These return expectations imply discount rates that are higher than the 20% applied
here. However, the objective of this study was to consider whether investment in
forest conservation could compete with other land uses under the assumption that a
functional carbon market existed that incorporated REDD credits. Nevertheless, it is
also useful to consider the impact on investment if policies do not provide investors
with adequate levels of certainty. This point is most salient to the current debate
around national accounting baselines and project delivery, as discussed.
Two methods can be applied to consider this uncertainty. One is to apply a higher
discount rate to carbon projects. Table 3 shows the results for carbon projects when a
35% discount rate is applied. These are still compared to the baseline scenarios at a
20% discount rate because of the greater certainty around the timber, beef and palm
oil projects. This results in favouring conversion activities. As Table 3 shows, this
adjustment results in carbon being uncompetitive in the Congo and Papua for palm oil
and only competes with logging in Papua and ranching in Brazil in certain cases.
Figures in red indicate a lower NPV than the baseline scenario.
Table 3 – Carbon scenarios applying a 35% discount rate
Brazil - cattle
Papua -
logging
Papua - palm
Congo -
logging
Baseline (at
20% discount)
$343 $350 $757 $662
Carbon
Scenario 1
$432 $428 $363 $428
Carbon
Scenario 2
$521 $286 $440 $286
Carbon
Scenario 3
$126 $92 $85 $92
Adjusting the discount rate in this way generally captures uncertainty around project
delivery and accreditation, but it is ambiguous in the specific risks that are being
considered. It also favours policies that are skewed toward the present, such as
carbon scenario 1. The more conservative approach represented in scenario 3 cannot
compete financially, as revenues are skewed toward a more heavily discounted future.
27
Another approach is to assume that there is a 50% chance that credits will be awarded
and reduce the NPV by this probability factor. Using the original results in Table 2
and multiplying carbon NPV by 0.50, it can again be seen that carbon only appears
attractive in a limited number of cases. Figures in red indicate a lower NPV than the
baseline scenario.
Table 4 – Carbon project NPVs assuming 50% chance of accreditation
Brazil - cattle Papua - logging Papua – palm Congo - logging
Baseline $343 $350 $757 $662
Carbon
Scenario 1
$330 $324 $275 $324
Carbon
Scenario 2
$584 $378 $490 $378
Carbon
Scenario 3
$260 $201 $193 $201
In reality, investors that are only 50% certain that credits will be awarded will factor
in the optionality to abandon a carbon project in 5 or 10 years, if credits are not likely
to be accredited, and then convert to another land use. From this perspective, the loss
in returns is the difference in NPV from converting land today versus converting it in
the future, whereas the upside is the opportunity for carbon revenue. An investor
would be more likely to consider the original NPV results presented in Table 2 to
evaluate this trade-off.
Under either method, it is difficult to fully quantify the investment implications if the
fundamental question of whether the carbon credits will be recognized cannot be
answered. If it is accepted that private investment can play a major role in land use
change at the forest frontier, then fostering an attractive investment environment for
forest conservation should be an objective of policy debates.
4.4 Government Fees
Governments generally use a concession licensing system to allocate land use rights,
entitling licensees to conduct a particular activity on the land. Fees, ongoing lease
payments and tax arrangements tend to be variable based on activities, i.e. logging
versus oil palm plantations. For example, an area that is designated as a productive
forest may be reclassified for palm oil establishment if it is degraded after a first
harvest and there is minimal residual timber value. It is likely that concessions
specifically allocated for carbon investments would need to be designed or designated
within existing or new concession licensing systems. The structure of fees and other
payments to government would need to be considered under such a system.
There are no specific assumptions made here regarding how this structuring could
occur. As discussed in Section 3.5, ten percent of carbon revenues are reserved to
28
account for forest management and carbon transaction costs, and a portion of this
could also be assumed to be government fees. Investment revenues could be split in
any proportion between the commercial enterprise, local communities and the
regional or national government, and this will be an important point of negotiation as
markets develop. This could be addressed at a national level by governments taking
ownership of some or all carbon concessions, thereby receiving all associated
revenues, or negotiating with private investors.
This is also an important point in relation to the NPV methodology applied in this
study. Investors evaluate investment options in multiple ways. For example, different
outcomes would result from an IRR analysis, which focuses on rate of return on
invested capital rather than the present value of future cashflows. It is arguable that
forest conservation requires a limited outlay of initial capital in comparison to
baseline scenarios, so the financial risk to investors is relatively low. This would
result in a high IRR, making carbon investment appear even more competitive than
under the NPV analysis. (Note: Logging costs may be minimal if concession holders
are using old, depreciated machinery, but costs will still include employees,
harvesting plans, cutting roads, etc. For oil palm plantations and ranching, in some
cases site preparation costs could be offset by selling timber, but the models here
assume that the terminal value of timber sales is captured in baseline scenario land
prices.)
In both the NPV and IRR analysis, a detailed consideration of the costs associated
with developing and managing a carbon forestry project needs to be included. This
analysis has used a simple method to account for costs, 10% of revenues. A higher
cost structure would decrease the competitiveness of the carbon scenario.
5 Conclusions
This study has determined that conservation projects financed through carbon markets
can generally deliver returns that are competitive with current land uses that are
driving deforestation. These findings are based on the assumption that carbon market
policies will provide certainty to investors that REDD credits will be accepted and
creditable in carbon markets. Without this certainty, conservation will remain a high-
risk investment that cannot compete with conversion activities, resulting in limited
direct private sector investment. This has implications for policies which are
considering national baseline accounting approaches that result in high risk at the
project level.
This study has developed a framework to bring together a wide scope of information.
The generalized models and approach are intended to provide input to investment and
policy decision making. Many of the assumptions are necessarily simplified in order
to consider implications across the forest frontier. There is now some evidence to
consider the opportunities and evaluate carbon conservation investment options at a
more localized scale. If it is accepted that private investment can play a major role in
land use change at the forest frontier, then fostering an attractive investment
environment for forest conservation should be an objective of policy debates.
29
Key points for these ongoing debates are:
• Certainty around accreditation at the project level is paramount. If stable
and predictable policies are in place, investors could apply even lower risk
profiles (discount rates) in comparison to baseline activities, since the
primary capital investments are limited to project accreditation, forest
monitoring and protection and community payments or taxes that provide
for local development and non-extractive forest use. This is significantly
less than the appreciable start-up costs associated with baseline activities.
• The price of carbon credits is often discussed as the crucial variable for
carbon investment. Results presented here suggest that crediting periods
and early market sales interact with prices to deliver a range of outcomes.
It is not necessarily useful to identify the most significant input factor
because policies will rarely be designed with one element in mind. An
exception to this may be crediting periods, as our results suggest that more
frequent crediting is beneficial to investment returns.
• This study has not assessed how carbon-financed forest conservation
projects could provide benefit to communities or governments. This point
is central to the success of conservation activities, as forests cannot be
simply “locked up” or separated from the interests of communities that
rely on forest resources. Investors will want to ensure that local
communities dependent on the forests are supportive of conservation
activities to reduce project risks associated with illegal logging and other
destructive activities. Options may be payments for monitoring and
maintaining conservation forests and ensuring community rights to
develop forest resources for non-timber forest products and local use. The
cost structure of these activities needs to be factored into carbon
investments more clearly.
• While carbon investment costs may not be fully captured in this level of
analysis, there are additional benefits that have also not been quantified in
both the baseline and carbon scenarios. Baseline scenarios presumably
have flow-on benefits to communities in the form of employment and
social services, which are often provided by large-scale operators. Carbon
scenarios, conversely, can sustain non-timber forest product industries and
provide continued access for firewood, hunting, small-scale timber, food
and medicines. Ecosystem benefits from retaining the forest, e.g. water
quality protection, landslide/flood prevention, biodiversity habitat,
maintenance of traditional livelihoods and prevention of social dislocation
of communities are also not included, would require a total economic
assessment and are not included in the financial analysis considered here.
• Although this study was framed as a trade-off between one land use type
and the conservation scenario, the impact of carbon-financed conservation
will likely be to shift land use economics, rather than replace activities.
30
For example, if a standing forest is now valued for its carbon, landowners
would be motivated to increase the efficiency of existing agribusiness
activities or rehabilitate degraded areas rather than move into new forested
areas. This could result in more efficient land uses and a wider range of
benefits at a landscape scale.
31
References
Bowman, M. (2007) pers comm Project Assistant, Woods Hole Research Center,
November 2
Brazilian Vegetable Oil Industry Association (2006) “Understanding the Moratorium:
Responsible Production”
http://www.abiove.com.br/english/sustent/ms_cprodutor_24jul07_us.pdf
Butler, Rhett A. (2007) “Is peat swamp worth more than palm oil plantations?”
www.mongabay.com
Carpoor, K. and Ambrosi, P. (2007) “State and Trends of the Carbon Market” World
Bank/IETA, Washington D.C.
Centre for Clean Air Policy (2007) “Reducing Emissions from Deforestation and
Degradation: The Dual Markets Approach”
http://www.ccap.org/international/FINAL%20REDD%20report.pdf
Congo Basin Forest Partnership (2005) “The Forests of the Congo Basin: A
Preliminary Assessment” (online)
http://carpe.umd.edu/products/PDF_Files/FOCB_APrelimAssess.pdf
FAO (2005) “FAO Forest Resource Assessment – Global Tables”
http://www.fao.org/forestry/site/28679/en/
Fearnside, P. (2007) “Deforestation in Amazonia” Encyclopedia of Earth Eds. Cutler
J . Cleveland, Washington, D.C.: Environmental Information Coalition, National
Council for Science and the Environment
http://www.eoearth.org/article/Deforestation_in_Amazonia;
Foley et al. (2007) “Amazonia Revealed: Forest Degradation and the loss of
ecosystem goods and services in the Amazon Basin” Frontiers in Ecological
Environment, Issue 5(1), p. 25-32
Forests Monitor (2007) “The Timber Sector in the DRC: A Brief Overview”
http://www.forestsmonitor.org/uploads/2e90368e95c9fb4f82d3d562fea6ed8d/Descrip
tion_of_the_Timber_Sector_in_the_DRC.pdf
Friends of the Earth (2005) “Greasy Palms. The social and ecological impacts of
large-scale oil palm plantation development in southeast Asia”
http://www.foe.co.uk/resource/reports/greasy_palms_impacts.pdf
Gibbs, H.K, S. Brown, J . O. Niles and J . A. Foley (2007) “Monitoring and estimating
tropical forest carbon stocks: making REDD a reality” Environmental Research
Letters (in press)
32
Global Forest Watch et al (2001) “The State of the Forest: Indonesia” Global Forest
Watch, World resources Institute and Forest watch Indonesia
www.globalforestwatch.org
Global Forest Watch (2007) “Human Pressure on the Brazilian Amazon Forests”
(online)
http://www.globalforestwatch.org/english/interactive.maps/Brazil_Datasets.htm#Stud
y_area
Hamilton, K., Bayon, R., Tunrer, G. and D. Higgins (2007) “State of the Voluntary
Carbon Markets 2007: Picking Up Steam” Ecosystem Marketplace & New Carbon
Finance www.ecosystemmarketplace.com
Ismail, A., Simeh, M. and M. Mohd Noor (2003) “The Production Cost of Oil Palm
Fresh Fruit Bunches: the Case of Independent Smallholders in J ohor” Oil Palm
Industry Economic Journal Vol. 3(1)
ITTO (2005) “Annual Review and Assessment of the World Timber Situation”
Division of Economic Information and Market Intelligence, Yokohama, J apan
http://219.127.136.74/live/Live_Server/2151/E-AR05-Text.pdf
ITTO (2006) “Asia Takes Price Leadership from Europe in Africa” Tropical Forest
Update Volume 16, No. 4
http://www.itto.or.jp/live/Live_Server/3163/tfu.2006.04.e.pdf
ITTO (2007) “Market Trends” Tropical Forest Update Volume 17, No. 1
http://www.itto.or.jp/live/Live_Server/3241/tfu.2007.01.e.pdf
ITTO (2007) “African Prices Stable on the Back of Steady Demand” Tropical Forest
Update Volume 17, No. 2
http://www.itto.or.jp/live/Live_Server/3453/tfu.2007.02.e.pdf
J arvis, B. and J acobson, M. (2006) “Working paper – Incentives to promote forest
certification in Indonesia” Project: Motivating Sustainability, Prepared for Moray
McLeish, International Finance Corporation PENSA. Indonesia, April
J otzo, F. (2007) “CDM Project Design” Centre for Energy and Environmental
Markets, University of New South Wales, Applied CDM Short Course, J uly 18
th
Karsenty, A., and Gourlet-Fleury, S. (2006) “Assessing sustainability of logging
practices in the Congo Basin’s managed forests: the issue of commercial species
recovery. Ecology and Society Issue 11(1), p. 26 (online)
http://www.ecologyandsociety.org/vol11/iss1/art26/
Laporte, N. T, T. S. Lin, J . LeMoigne, D. Devers, and M. Honzak (2004) “Toward an
Integrated Forest Monitoring System for Central Africa” Land Change Science:
Observation, Monitoring, and Understanding Trajectories of Change on the Earth
33
Surface, Remote Sensing and Digital Image Processing Vol (6), Ed. G. Gutman, p.
97-110
Laporte, N., J . Stabach, R. Grosch, T. S. Lin, S. Goetz (2007) “Expansion of
Industrial Logging in Central Africa” Science Vol. 316, No. 5830, p. 1451-1452
Malaysian Palm Oil Board (2007) Palm Oil Weekly Prices, available online
http://econ.mpob.gov.my/upk/weekly/bh_wk02nov07.htm (accessed Oct 2007)
Margulis, S. (2004) “Causes of Deforestation of the Brazilian Amazon” World Bank
Working Paper No. 22, Washington, D.C.
Martino, D. (2007) “Deforestation in the Amazon: Pressures and Outlook” Third
World Resurgence No. 200, April
http://www.southdevelopment.org/environment/MartinoAmazonDeforestation.pdf
Mattos, M. and Uhl, C. (1994) “Economic and Ecological Perspectives on Ranching
in the Eastern Amazon” World Development Vol. 22, No. 2, p. 145-158
Merry, F. (2007) pers comm. Discussing work in review. Woods Hole Research
Center
Mongabay.com (2007) “South American development plan could destroy the
Amazon” http://news.mongabay.com/2007/1003-ci_amazon.html
Pedroni, L. and Streck, C. (2007) “Mobilizing Public and Private Resources for the
Protection of Tropical Rainforests” CATIE and Climate Focus
http://www.climatefocus.com/newspubs/downloads/PolicyBriefonREDD_000.pdf
Nasi, R. (2005) “Potential Methodological Flaw in the Examination of the Effects of
Logging” Ecology and Society Issue 10(2) (online)
http://www.ecologyandsociety.org/vol10/iss2/resp2/
Nepstad, D., Moutinho, P. & Soares-Filho, B. (2005) “The Amazon in a Changing
Climate: Large-Scale Reductions of Carbon Emissions from Deforestation and Forest
Impoverishment” Amazon Institute for Environmental Research), Woods Hole
Research Center (WHRC) and Federal University of Minas Gerais paper released
during COP 12, Nairobi
Ruiz-Perez, M. et al (2005) “Logging in the Congo Basin: A Multi-Country
Characterization of Timber Companies” Forest Ecology and Management Volume
214, Issues 1-3
Saatchi, S. et al (2005) “Spatial Distribution of Carbon Stock in the Amazon Basin”
American Geophysical Union, Fall Meeting 2005
http://adsabs.harvard.edu/abs/2005AGUFM.B54B..03S
34
Soares-Filho, B. et al (2006) “Modelling Conservation in the Amazon Basin” Nature
Vol. 440, March
Sumantri, H. and Wijayanto, I. (2005) “Land cover mapping using satellite imagery
and GIS in the Mamberamo Basin and Raja Ampat Islands, Papua, Indonesia”
www.mapindia.org/2005/papers
Tedjasukmana, J . (2007) “Heros of the Environment: Leaders & Visionaries” Time
Magazine
http://www.time.com/time/specials/2007/article/0,28804,1663317_1663319_1669895,
00.html
Tokede, M., William, D., Widodo, Gandhi, Y., Imburi, C., Patriahadi, Marwa, J .,
Yufuai, M. (2005) “The impacts of Special Autonomy in Papua’s Forestry sector:
Empowering Customary Communities (Masyarakat Adat) in decentralised Forestry
Development in Manokwari District” Centre for International Forestry Research,
Bogor, Indonesia
Vidal, J . (2007) “World Bank Accused of Razing Congo’s Forests” The Guardian,
October 4
http://www.guardian.co.uk/environment/2007/oct/04/congo.forests?gusrc=rss&feed=1
2
Woods Hole Research Center (2007a) “Agriculture Frontier Explosion in Brazil”
http://www.whrc.org/southamerica/agric_expans.htm
Woods Hole Research Center (2007b) “Logging and Family Forests”
http://www.whrc.org/southamerica/logging_fam_forests.htm
World Bank Carbon Finance Unit (2007) “Forest Carbon Partnership Facility”
http://carbonfinance.org/Router.cfm?Page=FCPF&FID=34267&ItemID=34267
WWF Germany (2002) “Oil palm plantations and deforestation in Indonesia: What
role to Europe and Germany Play?” http://www.fire.uni-
freiburg.de/se_asia/projects/wwf_oil.htm
35
Appendix A – Production cost schedule for natural
tropical forest concession
Hypothetical Natural Forest Concession – applying sustainable forest
management in line with international certification standards and
payment of all official taxes and fees
Assumed harvestable area (ha)
Number of years cut over
Hectares harvested per year
Annual production (m3) - 20 m3/ha
Activity Cost/m3
(USD)
I PLANNING COSTS
1 Block delineation 0.04
2 Pre-harvest inventory 0.22
3 Survey, road planning and construction 3.91
4 Amortisation of concession fee 0.25
5 Amortisation of forest inventory 0.05
6 Amortisation of concession boundary delineation 0.02
7 Amortisation of satellite imagery 0.02
Subtotal 4.51
II REHABILITATION PLANTING
1 Seedling costs 0.17
2 Planting in logged areas, roadsides and bare soil 0.33
Subtotal 0.50
III SILVICULTURAL TREATMENTS
1 Post harvest inventory 0.16
2 Plant maintenance 0.12
3 Reserve area maintenance 0.04
4 Permanent sample plots 0.02
5 General facilities 0.08
Subtotal 0.42
IV HARVESTING COSTS
1 Felling (on contract with equipment provided) 5.05
2 Skidding (on contract with equipment provided) 0.09
3 Rafting/Hauling (on contract with equipment provided) 4.99
4 Maintenance of facilities and equipment 9.15
5 Mechanics workshop 3.70
6 Fuel 6.38
7 Tug boat and pontoon costs 12.00
Subtotal 41.37
V OFFICIAL TAXES AND FEES
1 Land and buildings tax 1.50
2 Vehicular and heavy equipment tax 0.12
3 Log grading tax 0.26
4 Local government fees and contributions 3.20
6 Rehabilitation Fund (DR) 14.00
36
7 Forest Resource Provision Tax (PSDH) 7.00
Subtotal 26.09
VI SOCIAL AND ENVIRONMENTAL COSTS
Community liaison 4.31
Community fees 1.24
Amortisation of Environmental Impact Assessment 0.01
Subtotal 5.56
VII PAYMENT TO CUSTOMARY LAND OWNERS
Subtotal 5.50
VIII GENERAL DEPRECIATION COSTS
1 Of field generators, heavy equipment 5.29
2 Of roads and buildings 4.00
3 Of workshop and camp generator and equipment 0.09
Subtotal 9.38
IX WAGES, BONUSES, ADMINISTRATION
1 Wages and bonus 4.00
2 Field camp consumables and associated costs 3.00
3 Head office and marketing costs 4.00
4 Forest protection and security 0.12
Subtotal 11.12
X UNOFFICIAL COSTS
Forest service officials - honorarium, entertainment 5.18
Forest security and protection 1.96
NGOs - honorarium. Entertainment 1.02
Village officials 0.63
Local government 1.57
Others 5.00
Subtotal 15.36
Grand total - Production cost 120.73
The following should be noted in relation to this estimate:
• These are the costs that would be necessary if natural forest extractive
management was practiced at a level that would meet reputable forest
certification standards. There are no certified natural forest extractive
management operations in Papua.
• All costs have been divided by total yield. This approach does not take into
account the effect of the timing of costs i.e. not all costs will be spread evenly
over all the years of harvesting; there will be some up-front costs. There is
therefore an underestimation of the cost, but it is not significant for analysis at
this level.
Further explanation of costs:
I4 – Amortisation of concession fee
37
The Iuran Hak Pengusahaan Hutan (IHPH) is a one-off payment to the Department of
Forestry for the rights to manage the forest for a set period of time, typically 20 to 50
years. The fee is an area based fee set at $10.00 per hectare.
IV3 – Rafting/Hauling
Note – roading costs are a combination of part of the fuel, mechanics workshop,
heavy equipment depreciation and other costs.
Costs are based on a topographic area is largely flat with large sinuous rivers, seasonal
wet areas and swamps, reflecting the geographies of Papua and parts of the DRC.
Roading intensity has to be higher to allow for these obstructions, and in some places
the roads would need to be elevated on causeways. Roading is difficult and, with
limited supplies of rock, expensive and difficult to maintain. There are few public
roads away from larger centres, most logs will be hauled to large rivers and rafted or
barged to the coast.
V. OFFICIAL TAXES AND FEES (in Papua)
V1 – Pajak Bumi Bangunan - Land tax
Paid annually, this is calculated on a fixed percentage of listed government value per
ha. Sources quoted a wide range of costs, from $0.25 to $3.00 per m
3
of merchantable
logs produced.
V4 – Local government taxes and contributions
This includes ‘Retribusi daerah’, a local government tax. Local government taxes
may vary from region to region.
V5 – Dana Reboisasi – Reforestation Levy
The reforestation levy varies by species grouping and, for Papuan logs is likely to be
from $16/m
3
for merbau (Instia bijuga) and other high grade species to $12/m
2
for
mixed grades.
V6 – PSDH – Forest Resource Provision Tax
This tax is set at 10% of the list price set by the Ministry of Trade. The levey varies
between location, diameter and species. It is currently of the order of US$12.30 for
the highest grade including merbau; US$6.90 for the meranti group grade and
US$4.10 for the mixed grades.
VII – PAYMENT TO CUSTOMARY OWNERS
The benefits received by the community cooperatives vary according to the agreement
made with investors, and, in some cases, the volume of timber harvested.
Compensation may range from USD3.00/m3 to even USD22.00/m3. Where
community cooperatives are in partnership with large scale forest concession holders,
the maximum amount of compensation that the concession holders are required to pay
is around USD5.75 (Tokede et al 2005).
X. UNOFFICIAL COSTS
X1 – Government officials – honorarium, entertainment
38
This includes expenses to obtain the concession license. One source estimates the
unofficial costs at perhaps USD100,000 although may often be many times that.
Getting the annual plan (RKT) accepted may cost between USD12 000 and
USD25,000 each year. Signing of the regular log inspection reports by Department of
Forestry staff costs at least USD300/month. Signing of transport documents costs
around USD per shipment and there are additional charges to be paid to port officials.
Companies will frequently respond to a range of requests from the government
officials it deals with for air tickets and accommodation not related to the tasks they
perform.
39
Appendix B – Oil Palm Model Assumptions
1. Costs
• Land costs are $300 per hectare and are assumed to be incurred in the year
of establishment.
• Due diligence costs of $1.25 million ($5 per hectare) based on New Forests’
investment experience are incurred in year 1.
• Establishment costs are $2,555 per hectare, which includes land clearing, site
assessment, ripping and mounding, weed control, seedlings, planting,
contingencies and refilling.
• Operating costs total $500 per hectare per year for established plantation
areas. Costs include pruning, weed control, roading, labor, fertilizer,
equipment and transportation.
• The model assumes harvesting costs of $187/ha.
• These establishment costs are is in?line with figures provided by the Oil Palm
Research Institute.
89
2. Productivity and revenue generation
• Revenue from timber harvests associated with clearing land for plantations
are not included in this model.
• The model assumes that crude palm oil can be sold at current market prices
of US$886 per tonne.
90
• The model assumes that beginning at year 4, each ha of plantation produces
0.64 tonnes of oil. In a linear fashion, the oil palm yield increase to 4 tonnes
of oil per ha at age 7 where it stays consistent to age 12 before steadily
declining until age 25. Productivity curves derived from the Oil Palm
Research Institute and the World Agroforestry Centre.
91
• The applied oil extraction rate is 16%, indicating the volume of crude palm oil
extracted from fruit bunches grown on trees. For example, in the first year
the model produces 50,000 tonnes of fresh fruit bunches (4 tonnes per
hectare) resulting in 8,000 tonnes of crude palm oil (16% of 50,000 tonnes).
89
Ismail, A., Simeh, M. and M. Mohd Noor (2003) “The Production Cost of Oil Palm Fresh Fruit
Bunches: the Case of Independent Smallholders in J ohor” Oil Palm Industry Economic Journal Vol.
3(1)
90
Malaysian Palm Oil Board (2007) Palm Oil Weekly Prices, available online
http://econ.mpob.gov.my/upk/weekly/bh_wk02nov07.htm (accessed Oct 2007)
91
Butler, Rhett A. (2007) “Is peat swamp worth more than palm oil plantations?” www.mongabay.com
doc_143628646.pdf