Description
The purpose of this paper is to examine the investment characteristics of works by
leading Australian artists.
Accounting Research Journal
Australian fine art as an alternative investment
Andrew C. Worthington Helen Higgs
Article information:
To cite this document:
Andrew C. Worthington Helen Higgs, (2008),"Australian fine art as an alternative investment", Accounting
Research J ournal, Vol. 21 Iss 1 pp. 55 - 66
Permanent link to this document:
http://dx.doi.org/10.1108/10309610810891346
Downloaded on: 24 January 2016, At: 21:06 (PT)
References: this document contains references to 17 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 364 times since 2008*
Users who downloaded this article also downloaded:
Mostafa M. Maksy, Lin Zheng, (2008),"Factors associated with student performance in advanced
accounting and auditing: An empirical study in a public university", Accounting Research J ournal, Vol. 21
Iss 1 pp. 16-32 http://dx.doi.org/10.1108/10309610810891328
Raymond da Silva Rosa, J ennifer Filippetto, Ann Tarca, (2008),"ASIC actions: canaries for
poor corporate governance?", Accounting Research J ournal, Vol. 21 Iss 1 pp. 67-86 http://
dx.doi.org/10.1108/10309610810891355
May H. Lo, Le (Emily) Xu, (2008),"Do analysts mislead investors?: A comparison of analysts' and investors'
weightings of cash components in forecasting annual earnings", Accounting Research J ournal, Vol. 21 Iss 1
pp. 33-54 http://dx.doi.org/10.1108/10309610810891337
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Australian ?ne art
as an alternative investment
Andrew C. Worthington and Helen Higgs
Department of Accounting, Finance and Economics, Grif?th University,
Gold Coast, Australia
Abstract
Purpose – The purpose of this paper is to examine the investment characteristics of works by
leading Australian artists.
Design/methodology/approach – About 35,805 paintings by 45 leading Australian artists sold at
auction are used to construct individual hedonic price indices. The attributes included in each artist’s
hedonic regression model include the size and medium of the painting and the auction house and year
sold.
Findings – The indexes show that average annual returns across all artists range between 4 and
15 per cent with a mean of 8 per cent, with the highest returns for works by Brett Whiteley,
Jeffrey Smart, Cecil Brack and Margaret Olley. Risk-adjusted returns are generally lower, with
reward-to-volatility and reward-to-variability ratios averaging 1.5 and 5.8 per cent, respectively.
The portfolio bs for individual artistic works average 0.41. The willingness-to-pay for perceived
attributes in the artwork show that works executed in oils and gouache, and those auctioned by
Deutscher-Menzies, Sotheby’s and Christies are generally associated with higher prices.
Research limitations/implications – The returns on a buy-and-hold strategy in the Australian art
market are at least comparable to the Australian stock market. While total risk is greater, the very low
market risk found in almost all artistic portfolios is suggestive of the possible bene?ts of portfolio
diversi?cation through art investment. Moreover, a number of artist’s works offer very superior
market and non-market risk-adjusted performance.
Originality/value – This is the ?rst Australian study to construct measures of risk, return, b and
Sharpe and Treynor ratios for individual Australian artists.
Keywords Arts, Prices, Investments, Australia
Paper type Research paper
1. Introduction
For some time, investors have been turning to art (paintings, sculpture, ceramics and
prints, and collectibles such as coins, stamps, antiques and furniture) as an alternative
investment. In Australia too, there is burgeoning interest in art investment,
particularly the work of Australian artists. Of course, Australia has a long history of
world-renowned artists, including Frederick McCubbin, Arthur Streeton, Tom Roberts
and Arthur Boyd, to name just a few. But in the last few decades many modern
and contemporary painters, like Charles Blackman, Brett Whiteley, David Boyd,
Ray Crooke and John Olsen, have also produced internationally reputable works and
thereby raised public awareness of art as a potential investment opportunity.
The purpose of this paper is to inform this process by investigating the risk, return
and asset pricing of Australian art. Hedonic pricing equations are used to capture the
characteristics of artwork by 45 well-known Australian artists publicly auctioned
during the past 30 years. The paper itself is organised as follows. Section 2 outlines
the empirical methodology. Section 3 provides a description of the data employed.
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
Fine art
as an alternative
investment
55
Accounting Research Journal
Vol. 21 No. 1, 2008
pp. 55-66
qEmerald Group Publishing Limited
1030-9616
DOI 10.1108/10309610810891346
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The empirical results are dealt with in Section 4. The paper ends with some concluding
remarks in Section 5.
2. Empirical methodology
The method selected to examine the investment characteristics of Australian art is the
hedonic price method (Buelens and Ginsburgh, 1993; de la Barre et al., 1994; Chanel,
1995; Agnello and Pierce, 1996). In this approach, all sales (including repeat sales) are
considered as single sales for which the objective features are recorded (i.e. name of
the painter, size of painting, medium of execution, etc.). Combining all sales allows the
implicit (or shadow) prices for these characteristics to be estimated separately from a
characteristic-free price including only the effects of time and random error. Assuming
the availability of comprehensive data, the hedonic price method’s main strengths
are that it estimates values based on actual auction sales and captures the
willingness-to-pay for perceived differences in the attributes of the artwork. The
hedonic price equation is written as:
ln p
kt
¼ f ðX
1kt
; . . . ; X
mkt
; . . . ; X
Mkt
Þ þ gðtÞ þ1
kt
ð1Þ
where lnp
kt
is the natural logarithm of the price of painting k(k ¼ 1, . . . , K) sold in year
t (t ¼ 1, . . . ,T), X
mkt
is the measurable characteristics m (m ¼ 1, . . . ,M) of painting k at
time t, g(t) is a function of time, and the error term 1 , N(0, S
k
%I
T
). The measurable
characteristics of the paintings for each artist comprise the physical characteristics of
the work and the characteristics of the auction at which the sale took place. The
regression equation is then speci?ed as:
ln p
kt
¼
X
M
m¼1
a
m
X
mkt
þ
X
T
t¼1
b
t
Z
t
þ1
kt
ð2Þ
where a
m
are parameter estimates of the implicit prices of the speci?ed art
characteristics, Z
t
is a dummy variable which takes the value of one for a sale occurring
in year t and zero elsewhere, b
t
is a parameter estimate, e
bt
gives the characteristic-free
price and all other variables are as previously de?ned. Separate regression equations
are speci?ed for each artist.
The data used comprises 35,805 sales of artworks by 45 leading Australian artists.
Information on sales is obtained from Australian Art Auction Records (2003) and
spans a 30 year period since March 1973. While an update of this data has just become
available, the requirement for extensive and lengthy data entry and manipulation acted
against this in the current analysis. The selection of artists to be included in
the analysis is, of course, highly subjective and was arrived at after discussion with
various art auctioneers, curators and dealers on those artists whose works were most
sought after and frequently sold at auction in the past 30 years. Its construction is also
re?ective, in so far is possible, of the widest number of periods, schools and genres in
Australian art history and is purposively restricted to artists who lived most of their
lifetime in Australia. A list of the artists is provided in Table I (column 1).
The ?rst set of information gathered is the price of each artwork for each artist. This
comprises the dependent variable in the hedonic price regression. Each artwork
included is sold exclusively at public auction and its value speci?ed in Australian
dollars.
ARJ
21,1
56
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
r
i
c
e
s
(
$
)
R
e
t
u
r
n
R
i
s
k
S
h
a
r
p
e
r
a
t
i
o
T
r
e
y
n
o
r
r
a
t
i
o
b
A
r
t
i
s
t
M
e
a
n
S
D
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
A
s
h
t
o
n
,
J
o
h
n
2
,
6
4
9
3
,
7
8
2
0
.
0
6
2
3
3
0
.
2
6
4
7
0
.
0
4
5
2
5
0
.
0
4
4
1
8
0
.
2
7
1
2
8
B
l
a
c
k
m
a
n
,
C
h
a
r
l
e
s
8
,
0
0
6
2
0
,
4
9
5
0
.
0
7
7
2
3
0
.
2
7
7
1
0
0
.
0
9
7
2
2
0
.
0
7
5
1
0
0
.
3
5
8
2
3
B
o
y
d
,
A
r
t
h
u
r
2
0
,
4
2
6
5
7
,
3
0
5
0
.
0
5
7
3
8
0
.
3
1
7
1
7
0
.
0
2
2
3
5
0
.
0
3
0
2
2
0
.
2
3
2
3
0
B
o
y
d
,
D
a
v
i
d
2
,
6
9
3
2
,
7
7
2
0
.
0
6
3
3
0
0
.
2
1
1
4
0
.
0
6
2
3
6
0
.
0
3
5
2
0
0
.
3
7
3
2
1
B
o
y
d
,
J
a
m
i
e
7
9
6
8
7
2
0
.
0
6
6
2
8
0
.
6
1
2
4
0
0
.
0
2
6
2
7
0
.
0
1
4
2
8
1
.
1
5
3
3
B
r
a
c
k
,
C
e
c
i
l
J
o
h
n
3
5
,
0
1
0
7
6
,
5
2
1
0
.
1
3
8
3
0
.
5
6
6
3
8
0
.
1
5
5
4
0
.
2
7
3
5
0
.
3
2
2
2
6
B
u
n
n
y
,
R
u
p
e
r
t
2
3
,
8
3
7
7
5
,
6
4
0
0
.
0
8
2
2
1
0
.
4
4
7
3
2
0
.
0
7
2
2
0
0
.
0
3
6
1
9
0
.
8
9
7
7
C
o
b
u
r
n
,
J
o
h
n
3
,
4
9
7
6
,
3
8
6
0
.
0
8
3
1
9
0
.
2
9
7
1
5
0
.
1
1
1
1
8
2
0
.
1
1
3
3
8
2
0
.
2
9
1
4
2
C
r
o
o
k
e
,
R
a
y
4
,
0
0
0
6
,
4
7
2
0
.
0
5
1
4
1
0
.
2
3
2
5
0
.
0
0
4
4
4
0
.
0
0
3
3
3
0
.
3
4
3
2
4
D
a
r
g
i
e
,
W
i
l
l
i
a
m
1
,
9
3
1
4
,
3
1
7
0
.
0
4
9
4
2
0
.
5
3
3
3
7
2
0
.
0
0
2
3
4
2
0
.
0
0
1
3
4
0
.
9
7
0
6
D
i
c
k
e
r
s
o
n
,
R
o
b
e
r
t
4
,
3
2
6
8
,
1
2
1
0
.
0
9
2
1
6
0
.
2
0
0
2
0
.
2
1
0
1
0
0
.
1
1
6
7
0
.
3
6
2
2
2
D
r
y
s
d
a
l
e
,
G
e
o
r
g
e
R
u
s
s
e
l
l
3
2
,
9
4
0
1
1
5
,
7
3
1
0
.
0
5
9
3
4
0
.
4
0
1
2
6
0
.
0
2
2
3
2
0
.
0
2
2
2
5
0
.
4
1
8
1
9
F
a
i
r
w
e
a
t
h
e
r
,
I
a
n
1
9
,
6
9
9
2
9
,
3
1
6
0
.
1
0
1
1
4
0
.
6
5
9
4
1
0
.
0
7
7
1
6
2
0
.
3
3
3
4
2
2
0
.
1
5
3
3
8
F
i
z
e
l
l
e
,
R
e
g
i
n
a
l
d
C
e
c
i
l
1
,
5
6
4
4
,
7
6
2
0
.
0
6
7
2
7
0
.
4
4
7
3
2
0
.
0
3
8
2
8
0
.
0
3
3
2
1
0
.
5
2
2
1
7
F
o
x
,
E
t
h
e
l
1
0
,
1
7
2
2
0
,
1
9
7
0
.
0
7
4
2
4
0
.
4
8
4
3
6
0
.
0
5
0
2
5
0
.
0
2
4
2
3
1
.
0
1
7
5
F
r
i
e
n
d
,
D
o
n
a
l
d
4
,
2
7
2
8
,
5
0
1
0
.
0
3
7
4
5
0
.
3
4
8
1
9
2
0
.
0
3
7
4
5
2
0
.
0
2
3
3
6
0
.
5
6
3
1
3
F
u
l
l
b
r
o
o
k
,
S
a
m
u
e
l
8
,
0
4
2
1
0
,
5
7
5
0
.
1
0
1
1
4
0
.
7
7
5
4
2
0
.
0
6
6
1
7
2
0
.
3
9
2
4
4
2
0
.
1
3
0
3
7
G
l
e
e
s
o
n
,
J
a
m
e
s
T
i
m
o
t
h
y
3
,
3
1
0
7
,
2
2
5
0
.
0
7
4
2
4
0
.
2
9
4
1
4
0
.
0
8
2
2
4
0
.
0
9
9
8
0
.
2
4
2
2
9
G
l
o
v
e
r
,
J
o
h
n
1
0
,
5
7
2
4
8
,
5
8
0
0
.
0
6
4
2
9
0
.
4
7
2
3
5
0
.
0
3
0
2
9
2
0
.
3
2
6
4
1
2
0
.
0
4
3
3
6
G
r
u
n
e
r
,
E
l
i
o
t
h
1
1
,
1
9
5
1
4
,
8
2
4
0
.
0
4
8
4
3
0
.
4
2
3
2
8
2
0
.
0
0
5
3
9
0
.
0
0
7
3
2
2
0
.
3
0
4
4
3
H
a
r
t
,
K
e
v
i
n
C
h
a
r
l
e
s
(
P
r
o
)
1
,
4
4
2
2
,
6
7
4
0
.
0
5
8
3
7
0
.
1
8
9
1
0
.
0
4
2
4
2
0
.
0
7
0
1
4
0
.
1
1
4
3
3
H
o
d
g
k
i
n
s
o
n
,
F
r
a
n
k
1
,
5
2
6
2
,
5
0
9
0
.
1
0
7
8
0
.
6
0
0
3
9
0
.
0
9
5
1
4
2
0
.
0
9
8
3
7
2
0
.
5
8
2
4
5
H
e
y
s
e
n
,
H
a
n
s
8
,
5
7
1
1
6
,
6
5
3
0
.
0
5
9
3
4
0
.
2
3
5
6
0
.
0
3
8
4
0
0
.
0
1
4
2
9
0
.
6
6
2
9
H
e
y
s
e
n
,
N
o
r
a
3
,
1
5
8
5
,
3
8
3
0
.
0
7
4
2
4
0
.
8
4
7
4
3
0
.
0
2
8
2
6
0
.
0
7
2
1
2
0
.
3
3
4
2
5
J
a
c
k
s
o
n
,
J
a
m
e
s
R
a
n
a
l
p
h
5
,
8
9
4
9
,
6
6
2
0
.
0
5
9
3
4
0
.
2
7
4
9
0
.
0
3
3
3
7
0
.
0
1
6
2
7
0
.
5
6
8
1
1
L
i
n
d
s
a
y
,
N
o
r
m
a
n
5
,
8
2
2
1
3
,
6
5
7
0
.
0
5
7
3
8
0
.
2
0
1
3
0
.
0
3
5
4
3
0
.
0
1
2
3
0
0
.
6
0
6
1
0
L
o
n
g
,
S
y
d
n
e
y
4
,
0
7
3
8
,
7
0
2
0
.
0
9
2
1
6
0
.
3
0
1
1
6
0
.
1
4
0
1
3
0
.
3
2
8
4
0
.
1
2
8
3
2
M
c
C
u
b
b
i
n
,
F
r
e
d
e
r
i
c
k
5
5
,
2
4
5
1
7
1
,
0
1
4
0
.
1
0
3
1
0
0
.
4
3
4
3
0
0
.
1
2
2
1
1
0
.
6
9
7
2
0
.
0
7
6
3
4
(
c
o
n
t
i
n
u
e
d
)
Table I.
Prices, return and risk,
performance measures
and bs by artist
Fine art
as an alternative
investment
57
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
r
i
c
e
s
(
$
)
R
e
t
u
r
n
R
i
s
k
S
h
a
r
p
e
r
a
t
i
o
T
r
e
y
n
o
r
r
a
t
i
o
b
A
r
t
i
s
t
M
e
a
n
S
D
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
N
a
m
a
t
j
i
r
a
,
A
l
b
e
r
t
7
,
3
3
9
6
,
1
3
2
0
.
0
8
5
1
8
0
.
2
7
1
8
0
.
1
2
9
1
5
0
.
0
7
1
1
3
0
.
4
9
4
1
8
N
o
l
a
n
,
S
i
d
n
e
y
1
1
,
1
8
2
4
2
,
8
5
2
0
.
0
5
4
4
0
0
.
3
3
0
1
8
0
.
0
1
2
3
8
0
.
0
0
7
3
1
0
.
5
6
2
1
4
O
l
l
e
y
,
M
a
r
g
a
r
e
t
1
2
,
5
2
9
1
5
,
9
3
0
0
.
1
3
5
4
0
.
4
0
4
2
7
0
.
2
1
0
3
0
.
1
5
9
6
0
.
5
3
3
1
5
O
l
s
e
n
,
J
o
h
n
9
,
1
1
8
2
4
,
8
2
1
0
.
1
0
9
7
0
.
3
9
1
2
2
0
.
1
5
1
9
2
0
.
3
8
3
4
3
2
0
.
1
5
4
3
9
P
e
r
c
e
v
a
l
,
J
o
h
n
1
4
,
1
3
3
3
8
,
2
5
6
0
.
0
6
3
3
0
0
.
3
9
8
2
4
0
.
0
3
3
3
1
0
.
0
7
3
1
1
0
.
1
7
7
3
1
P
r
e
s
t
o
n
,
M
a
r
g
a
r
e
t
1
2
,
4
7
0
2
6
,
2
4
4
0
.
1
0
2
1
2
0
.
4
4
4
3
1
0
.
1
1
7
1
2
0
.
0
5
0
1
6
1
.
0
4
7
4
P
r
o
c
t
o
r
,
A
l
t
h
e
a
1
,
8
6
7
3
,
0
7
2
0
.
1
1
2
6
0
.
3
6
1
2
1
0
.
1
7
2
5
2
0
.
1
3
8
3
9
2
0
.
4
5
0
4
4
R
e
e
s
,
L
l
o
y
d
9
,
6
1
7
2
0
,
6
6
9
0
.
0
7
9
2
2
0
.
3
6
0
2
0
0
.
0
8
1
2
1
0
.
0
5
4
1
5
0
.
5
3
3
1
5
R
o
b
e
r
t
s
,
T
h
o
m
a
s
W
i
l
l
i
a
m
2
4
,
1
6
8
4
8
,
8
4
8
0
.
0
8
3
1
9
1
.
9
3
3
4
5
0
.
0
1
7
2
3
0
.
0
8
0
9
0
.
4
1
2
2
0
R
u
s
s
e
l
l
,
J
o
h
n
P
e
t
e
r
4
5
,
1
6
7
1
0
0
,
0
7
9
0
.
1
0
7
8
1
.
4
7
7
4
4
0
.
0
3
9
1
9
0
.
0
1
7
2
6
3
.
3
3
9
1
S
m
a
r
t
,
F
r
a
n
k
J
e
f
f
r
e
y
3
6
,
5
4
4
5
1
,
7
7
4
0
.
1
4
6
2
0
.
4
5
6
3
4
0
.
2
1
1
2
2
0
.
5
9
3
4
5
2
0
.
1
6
2
4
0
S
m
i
t
h
,
G
r
a
c
e
C
o
s
s
i
n
g
t
o
n
1
7
,
2
0
4
3
0
,
3
2
3
0
.
1
1
6
5
0
.
4
2
8
2
9
0
.
1
5
4
7
0
.
0
4
6
1
7
1
.
4
2
5
2
S
t
r
e
e
t
o
n
,
A
r
t
h
u
r
3
1
,
8
0
0
6
1
,
5
8
7
0
.
1
0
3
1
0
0
.
2
8
3
1
2
0
.
1
8
7
6
2
.
3
0
4
1
0
.
0
2
3
3
5
T
u
c
k
e
r
,
A
l
b
e
r
t
1
4
,
7
6
4
3
8
,
7
9
1
0
.
0
6
3
3
0
0
.
3
9
8
2
4
0
.
0
3
3
3
0
0
.
0
2
3
2
4
0
.
5
6
6
1
2
W
h
i
t
e
l
e
y
,
B
r
e
t
t
2
3
,
9
2
7
8
2
,
4
6
5
0
.
1
4
7
1
0
.
2
8
1
1
1
0
.
3
4
5
1
0
.
3
3
9
3
0
.
2
8
6
2
7
W
i
l
l
i
a
m
s
,
F
r
e
d
e
r
i
c
k
2
1
,
3
0
5
4
9
,
7
7
9
0
.
1
0
2
1
2
0
.
2
9
0
1
3
0
.
1
7
9
8
2
0
.
2
4
3
4
0
2
0
.
2
1
4
4
1
W
i
t
h
e
r
s
,
W
a
l
t
e
r
1
2
,
0
9
7
3
0
,
5
0
2
0
.
0
4
3
4
4
0
.
3
9
2
2
3
2
0
.
0
1
8
4
1
2
0
.
0
0
9
3
5
0
.
8
0
7
8
Table I.
ARJ
21,1
58
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The next two sets of variables are considered to be major determinants of the price of
an artist’s work and are speci?ed as explanatory variables. The ?rst set of explanatory
variables relate to the physical characteristics of the work, while the second set
comprise its sale characteristics. Starting with the physical characteristics of the
works, the ?rst group comprises dummy variables identifying the medium used:
namely, acrylic, charcoal, crayon, etching, the heavy, opaque watercolour paint known
as gouache, mixed media, oil, pastel, pencil and watercolour. The reference category is
all other mediums. Of the mediums used by artists in the sample, the largest numbers
of works (percentage of all mediums in brackets) sold are oils (44 per cent), followed by
watercolours (11 per cent), and ?nally etchings (8 per cent). Oil as a medium, though
dif?cult to work, has excellent visual qualities and is not easily faded by natural light.
It is therefore likely to fetch higher prices at auction. Modern alternatives, including
acrylic and gouache, also command high prices. However, a variety of other potentially
valuable media are found in most ?ne art collections. Australian landscape artists, for
instance, often favour watercolours.
The second group of physical characteristic are the dimensions of the painted work
as represented by surface area in square metres (m
2
) and surface area squared as the
nonlinear component. A positive relationship is generally hypothesised when price is
regressed against area, although it is dif?cult for all but the largest public galleries to
display very large works. On this basis, the expected sign on the coef?cient for area
squared is thought to be negative (Agnello and Pierce, 1996). Of course, there are any
number of other physical characteristics that could be included if data were available.
These include the painting’s genre, providence, date completed, the presence of the
artist’s signature and so on.
The second set of explanatory variables incorporate the sales characteristics of the
work. The ?rst of these are dummy variables identifying in which of Australia’s three
largest auction houses the sale took place (percentage of all sales in brackets): that is,
Christies (16 per cent), Deutscher-Menzies (4 per cent) and Sotheby’s (15 per cent). The
reference category is all other auction houses. In the absence of transaction costs, the
law of one price dictates that no signi?cant price difference should exist for paintings
of a similar quality. However, Pesando (1993), de la Barre et al. (1994) and Renneboog
and van Houtte (2002), amongst others, have found that Christies and Sotheby’s
systematically obtain higher hammer prices through their reputation and market
power. The second group of sales characteristics identi?es the year when the work is
sold. This consists of thirty yearly dummy variables with 1973 as the reference
category.
3. Descriptive analysis
The mean and standard deviation of prices for each artist’s works are included in
Table I (columns 2 and 3). Turning ?rst to the prices of artworks by artist, average
prices range from $796 for paintings by Boyd ( Jamie) to $55,245 for works by
McCubbin. Other artist’s with highly valued works are Russell, Smart and Brack with
means of $45,167, $36,544 and $35,010, respectively. The lowest prices are paid for
works by Hart, Hodgkinson and Fizelle with average prices of $1,442, $1,526 and
$1,564, respectively. The standard deviations of art prices by artist range from $872 to
$171,014. On this basis, works by Boyd ( Jamie), Hodgkinson, Hart and Boyd (David)
are the least volatile with standard deviations of $872, $2,509, $2,674 and $2,772,
Fine art
as an alternative
investment
59
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
respectively, whereas works by McCubbin, Drysdale, Russell and Whiteley are the
most volatile with standard deviations of $171,014, $115,731, $100,079 and $82,465,
respectively.
Now consider each artist’s work categorised according to the media used
(not shown). Of these, oils, watercolours and etchings are the most common medium
sold, while crayons and acrylics are the least common. However, the distribution of all
physical and sales characteristics varies dramatically by artist. For example, sold work
by Nolan, Fullbrook and Bunny are almost exclusively oil, while Namatjira’s sold
works are primarily watercolours. Similarly, the distribution of sold works by auction
house also varies across the artists. For example, 54 per cent of Whiteley’s work was
sold at Christies, 32 per cent of Brack’s at Deutscher-Menzies and 54 per cent of
Fullbrook’s at Sotheby’s. This contrasts markedly to the aforementioned sample
averages.
4. Empirical results
In the interests of brevity, the estimated coef?cients of the hedonic pricing regression
models for each of the 45 Australian artists are not shown. All of the estimated models
are highly signi?cant, with likelihood ratio tests (not shown) of the hypotheses that
the slope coef?cients are zero rejected at the 1 per cent level. The adjusted R
2
range between 0.547 and 0.889 and are reasonably high for what is basically
cross-sectional data.
As hypothesised, the percentage changes in value indicate that works executed in
oil and gouache command higher prices, with average percentage increases over each
artist’s standard work of 6.799 and 6.733 per cent, respectively. Of the 44 artists in the
sample with at least some oil works, all but one have signi?cant and positive increases
in value relative to other work, while 28 of the 30 artists with gouache works have
signi?cant and positive increases with this media. However, the percentage increases
in value for individual artists vary widely. For example, with oils the increase in values
ranges from as little as 1.188 per cent (Fizelle) to more than 21.700 per cent (Bracks) and
for gouache from just 0.684 per cent (Olley) to 68.217 per cent (Smith).
By comparison, media such as etchings, crayon and charcoal are associated with
respective average percentage increases across the sample of just 1.105, 3.020 and 1.787
per cent implying these media are generally more affordable, regardless of all other
characteristics, while mixed media, watercolours and pastels have average price
increases across the sample of 3.466, 3.346 and 3.646 per cent, respectively.
Unfortunately, it is dif?cult to compare these ?ndings because other studies are often
limited to periods or movements when fewer media are generally known (de la Barre
et al., 1994; Renneboog and van Houtte, 2002) or to a single medium (Candela and
Scorcu, 1997; Pesando and Shum, 1999). Nevertheless, Agnello and Pierce (1996) found
a 156 per cent increase in prices for oil works as compared to all other media
(watercolour, gouache, ink, pencil, pastel, etc.).
The remaining physical characteristics included in the regression models concern
the size of the work. These are the area of the work in square metres and its nonlinear
component, area squared. The generally positive and signi?cant signs of the area
coef?cients and the negative and signi?cant signs of its squared term indicate that art
prices tend ?rst to increase with size, then decrease as the paintings become too large
and dif?cult to house. Across the sample, a 1 per cent increase in surface area is
ARJ
21,1
60
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
associated with a 0.206 per cent increase in price, while on average the
price-maximising size is 4.08 m
2
. By comparison, Agnello and Pierce (1996) found
the price-maximising size for American artists’ work to be 6.53 m
2
while de la Barre
et al. (1994) calculated this optimal size to be 5.89 m
2
for old masters and 1.70 m
2
for
modern and contemporary European works.
The ?nal set of variables relates to the sale characteristics of the works. The sales
characteristics show that auctions at Sotheby’s, Christies and Deutscher-Menzies
increase the average standard price by 1.759, 1.697 and 1.869 per cent, respectively,
over the remaining houses. Pesando (1993), de la Barre et al. (1994), Agnello and Pierce
(1996) and Renneboog and van Houtte (2002) also found that “. . .Sotherby’s typically
fetches higher prices than Christies, while both experience higher prices than all other
houses” (Agnello and Pierce, 1996, p. 366). However, while variation in the prices
obtained by the different auction houses are small, and certainly smaller than most
other factors included in the model, care should still be taken in interpreting these
differences as a violation of the law of one price. de la Barre et al., 1994, p. 165), for
example, concluded “. . .the quality of a painting, not captured by our characteristics is
partly picked up by the saleroom coef?cients: a “good” Picasso would go to Christies or
Sotheby’s New York, a less good one would be sold at Drouot’s [a Paris-based auction
house]. . .it is impossible to disentangle the two effects”.
Turning to investment risk and return, the index value for each artist is calculated
as 100e
bt
(not shown). Annual returns are then calculated such that the return for artist
i is represented by the continuously compounded return or log return of the price index
at time t such that Dp
it
¼ log ( p
it
/p
it21
) £ 100 where Dp
it
denotes the rate of change of
p
it
. Table I presents the arithmetic mean and standard deviation (risk) of annual
returns and their ranks for the 45 Australian artists (columns 4 to 7).
Also included in Table I are two external risk-adjusted portfolio performance
measures. The Sharpe ratio and its rank (also known as the reward-to-volatility ratio)
(columns 8 and 9) indicates the excess return per unit of risk and is calculated by
dividing the return in excess of the risk-free rate by the standard deviation of returns.
The proxy used for the risk-free rate is the exponentially smoothed average ?tted yield
for three-year Commonwealth Treasury bonds during the sample period (5 per cent).
In the current context, the Sharpe ratio is the most appropriate performance measure
for an investor whose portfolio is composed wholly of a given artist’s work.
The Treynor ratio and its rank (sometimes called the reward-to-variability ratio;
columns 10 and 11) is identical to the Sharpe ratio except that total risk (standard
deviation) is replaced with systematic (market) risk or b. This ratio may be a better
benchmark of performance for investors who do not invest exclusively in art, but
rather consider its diversi?cation potential. Accordingly, the b of each artist’s portfolio
and its rank (columns 12 and 13) is calculated with respect to an equity market
portfolio. The Australian All Ordinaries index (AOI) is speci?ed as the equity market
measure. Since higher Sharpe and Treynor ratios represent better performance, the
artistic portfolios are ranked in descending order.
In terms of returns, mean returns for the individual artists range between
3.70 per cent for works by Friend to 14.70 per cent for those by Whitelely. Annual
returns across all artists average 8.23 per cent, as compared to mean returns on the
AOI of 7.00 per cent over this same period. Other artists with relatively high returns
include Smart, Brack, Olley, Smith, Proctor and Olsen, with relatively low returns for
Fine art
as an alternative
investment
61
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Withers, Gruner, Dargie, Crooke, Nolan, Lindsay and Boyd (Arthur). The standard
deviation (risk) of returns ranges between 0.189 (Hart) to 1.933 (Roberts) with a mean of
0.449. By way of comparison, the standard deviation of returns on the AOI over this
same period was 0.161. Risk is also relatively higher for works by Russell, Heysen
(Nora), Fullbrook and Fairweather and relatively lower for works by Dickerson,
Lindsay, Boyd (David) and Crooke.
For overall returns on Australian art, it would appear that the market has
performed at a comparable level to other national markets. Renneboog and van Houtte
(2002), for example, found Belgian average returns of 8.4 per cent over the period
1970-1989 with a standard deviation of 19.4 per cent, Agnello and Pierce (1996)
estimated that the returns on American artists averaged 9.3 per cent from 1971-1992,
and Mei and Moses (2001) calculated average returns of 5.3 per cent with a standard
deviation of 9.3 per cent, also on American auctions. Mean returns from other art
studies include 1.6 per cent (Frey and Pommerehne, 1989), 6.8 per cent (Gerard-Varet,
1995) and 5.0 per cent (Goetzmann, 1996). Using a different methodology and an
international sample, Worthington and Higgs (2003, 2004) found mean returns of
3.73 per cent for Contemporary Masters, 2.85 per cent for 20th Century English works,
1.49 per cent for Modern Europeans and 1.25 per cent Surrealists. Of course, the returns
as calculated in this paper do not re?ect the fact that a substantial component of the
return from art investment is derived not from its ?nancial returns, rather from its
intrinsic aesthetic qualities. Equally, they also do not include the many and sizeable
transaction and holding costs associated with art portfolios, the absence of which may
serve to in?ate the ?nancial returns.
Analysis of the risk-adjusted returns for each artist’s portfolio of works provides
further insights. Starting with the Sharpe ratio, artists ranked highly on the basis of
returns per unit of (total) risk include Whiteley (0.345), Smart (0.211), Olley (0.210),
Brack (0.155) and Proctor (0.172). The Sharpe index for the AOI over this same period
is 0.124 while that for the average artist included in the sample is just 0.015. As shown,
many artists have low Sharpe ratios (and rankings) suggesting that a policy of holding
high-return-high-risk portfolios of a single artist’s work in isolation may not be an
appropriate investment strategy. As an alternative, the Treynor ratio show the returns
per unit of (systematic) risk and thus yields useful insights on the bene?t of holding
Australian art as part of a diversi?ed portfolio (though, of course, limited in this
analysis to listed equity).
As shown in Table I, the bs of most Australian artist’s work are low (,1), if not
negative, indicating potential diversi?cation bene?ts. For example, the negative bs
calculated on art portfolios composed of works by Hodgkinson, Proctor, Gruner,
Coburn and Williams indicate that their returns move contrary to returns on the
Australian stock market. However, some art portfolios are substantially more risky
(in terms of b) than the market, and move in the same direction, including Russell,
Smith, Boyd ( Jamie), Preston and Fox. The average b across the sample is 0.405 with
25 per cent of artistic portfolios having a b , 0.075 and 25 per cent .0.566. By
comparison, Chanel et al. (1994) calculated that national art bs ranged between 0.028
( London) and 0.368 (Tokyo), while Renneboog and van Houtte (2002) estimated
movement bs with respect to a global stock index of 23.7, 22.9 and 0.8 for
Impressionist, Luminist and Expressionist art, respectively.
ARJ
21,1
62
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The Treynor ratios for the 45 artist portfolios in mean/b-space are graphically
represented in Figure 1. The ?gure includes the security market line constructed using
the risk-free rate (intercept) and the return and b (one) for the market portfolio (slope).
Visual inspection indicates that few artist portfolios are correctly priced in relation to
the security market line (that is, on the line) with most underpriced. That is, artist
portfolios lying above the line indicate superior market risk-adjusted returns and a buy
signal, while those lying below the line indicate inferior market risk-adjusted returns
and a sell signal. Artists ranked highly on the basis of the Treynor index include
Streeton, McCubbin, Whiteley, Long and Brack. A buy signal is indicated for these
artists. Those ranked lowly with a consequent sell signal include Smart, Olsen,
Fullbrook, Fairweather and Glover.
Of course, these buy-and-sell strategies must be quali?ed by the fact that they relate
to historical information averaged over a thirty year investment horizon, not the
immediate past, present or future. It should also be remembered that the Treynor ratio
re?ects only systematic (general or market) risk and thereby re?ects the value of these
assets within a diversi?ed portfolio. The change in rankings of artists between
the Sharpe and Treynor measures indicate that most art portfolios as analysed include
much unsystematic (asset speci?c or nonmarket) risk when held in isolation.
Nonetheless, while the rankings of artists on the Sharpe and Treynor criterion do vary,
there is some deal of correspondence between them with the Spearman (rank)
correlation coef?cient signifying a signi?cant and positive relationship ( r ¼ 0.319,
p-value ¼ 0.033).
A ?nal requirement is to examine the relationship between the returns and values of
works included in each artistic portfolio. This follows the suggestion of Mei and Moses
(2001) amongst others that bidders in art auctions are exposed to a “winner’s curse” so
that the returns on expensive paintings tend to under perform the market as a whole:
referred to as the “masterpiece effect”. Figure 2 plots the returns and mean prices of
paintings for each artist, with a linear trendline added as a simple means of evaluating
the relationship between value and return. As shown, there is a small positive
Figure 1.
Market risk and
return by artist
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
–1.00 –0.50 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
b
R
e
t
u
r
n
Fine art
as an alternative
investment
63
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(1.03 £ 10
-6
) and signi?cant ( p-value ¼ 0.019) relationship suggesting that returns
increase with value, thereby supporting the absence of a “masterpiece effect” effect.
5. Conclusion
This paper investigates risk, return and asset pricing for the works of 45 well-known
Australian artists. Unlike most other work in this area which indicates that the returns
to art investment are much less, and the risks much higher than investment markets,
the results show that return on a buy-and-hold strategy in the Australian art market
are at least comparable to the stock market. While total risk is indeed greater than the
stock market, the very low market risk found in almost all artistic portfolios is highly
suggestive of the possible bene?ts of portfolio diversi?cation through art investment.
That said, a number of artist’s works offer superior market and non-market
risk-adjusted performance over the sample period, above all Arthur Streeton, Frederick
McCubbin, Brett Whiteley, Sydney Long, Cecil Brack, Frank Smart, Margaret Olley
and Althea Proctor. One major quali?cation is that the analysis does not take into
account the (high) transaction costs incurred at the moment of sale nor the (equally
sizeable) insurance and other costs associated with restoring, preserving and
displaying art works. However, neither does it take into account the (equally
substantial) aesthetic returns from owning and displaying ?ne art. The methodology
employed in the paper also identi?es factors associated with higher prices in the
Australian art market. All other things being equal, larger-sized works and those
executed in oils or gouache, and auctioned by Sotheby’s or Christies are associated
with higher prices. Conversely, smaller works, etchings, crayon or charcoal works,
along with those auctioned by other auction houses, are associated with systematically
lower prices.
There are many interesting opportunities to expand upon this work. One possibility
would involve gathering additional information to be included in the hedonic pricing
regression models. For example, the prices (and hence returns) on artists’ work may
also depend on the cumulative number of works auctioned, whether the artist is now
Figure 2.
Mean price and return
by artist
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
$0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,000
Mean price
R
e
t
u
r
n
ARJ
21,1
64
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
deceased or the age of the artist at the time of auction, genres of work, interactions
between medium and size and so on. While these impacts are proxied to some extent by
the variables included in the current analysis, a more de?ned speci?cation would
identify some determinants potentially obscured. There may also be opportunities to
examine art markets along the lines of the market ef?ciency literature. One prospect is
to examine the time-series behaviour of returns to examine whether the art market
fully incorporates all historical market information (weak-form ef?cient).
Finally, the art works on which these indices are based may not re?ect the market
for Australian paintings as a whole: private transactions for example conducted
through art galleries are ignored. Depending on the values found in galleries, indexes
constructed using auction data may understate or overstate the true return. There is
also no recognition that different buyers in the market have differing preferences for
artwork: compare, for instance, works bought by public galleries to those purchased
privately. For this reason, sellers of art to public collections are argued to enjoy
systematically higher rates of return. Future work could take into account these
subtleties.
References
Agnello, R.J. and Pierce, R.K. (1996), “Financial returns, price determinants, and genre effects
in American art investment”, Journal of Cultural Economics, Vol. 20, pp. 359-83.
Australian Art Auction Records Pty. Ltd (2003), Australian Art Auction Records, 1972-2003,
Sydney, CD.
Buelens, N. and Ginsburgh, V. (1993), “Revisiting Baumol’s ‘art as a ?oating crap game’”,
European Economic Review, Vol. 37, pp. 1351-71.
Candela, G. and Scorcu, A.E. (1997), “A price index for art market auctions”, Journal of Cultural
Economics, Vol. 21, pp. 175-96.
Chanel, O. (1995), “Is art market behaviour predictable?”, European Economic Review, Vol. 39,
pp. 519-27.
Chanel, O., Gerard-Varet, L.A. and Ginsburgh, V. (1994), “Prices and returns on paintings:
an exercise on how to price the priceless”, Geneva Papers on Risk and Insurance Theory,
Vol. 19, pp. 7-21.
de la Barre, M., Docclo, S. and Ginsburgh, V. (1994), “Returns of impressionist, modern and
contemporary European paintings 1962-1991”, Annales d’Economie et de Statistique,
Vol. 35, pp. 143-81.
Frey, B. and Pommerehne, W. (1989), “Art investment: an empirical inquiry”, Southern Economic
Journal, Vol. 56, pp. 396-407.
Gerard-Varet, L.A. (1995), “On pricing the priceless: comments on the economics of the visual art
market”, European Economic Review, Vol. 39, pp. 509-18.
Goetzmann, W.N. (1996), “How costly is the fall from fashion? Survivorship bias in the painting
market”, in Ginsburgh, V.A. and Menger, P.M. (Eds), Economics of the Arts: Selected
Essays, Elsevier North-Holland, Amsterdam, pp. 71-84.
Mei, J. and Moses, M. (2001), “Art as an investment and the origin of the masterpiece effect:
evidence from 1875-2000”, paper presented at the 8th Asia-Paci?c Finance Association
Annual Conference, Shangri-La Hotel, Bangkok, 22-25 July.
Pesando, J.E. (1993), “Arts as an investment: the market for modern prints”, American Economic
Review, Vol. 83, pp. 1075-89.
Fine art
as an alternative
investment
65
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Pesando, J.E. and Shum, P.M. (1999), “The returns to Picasso’s prints and to traditional ?nancial
assets, 1977 to 1996”, Journal of Cultural Economics, Vol. 23, pp. 183-92.
Renneboog, L. and van Houtte, T. (2002), “The monetary appreciation of paintings: from realism
to Magritte”, Cambridge Journal of Economics, Vol. 26, pp. 331-57.
Worthington, A.C. and Higgs, H. (2003), “Art as an investment: short and long-term
comovements in major painting markets”, Empirical Economics, Vol. 28, pp. 649-68.
Worthington, A.C. and Higgs, H. (2004), “Art as an investment: risk, return and portfolio
diversi?cation in major painting markets”, Accounting and Finance, Vol. 44, pp. 257-72.
Further reading
Goetzmann, W.N. (1993), “Accounting for taste: art and the ?nance markets over three centuries”,
American Economic Review, Vol. 83, pp. 1370-6.
Corresponding author
Andrew C. Worthington can be contacted at: a.worthington@grif?th.edu.au
ARJ
21,1
66
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Dan-Bi Choi, Taeyoung Chung, Hyung-Deok Shin. 2013. The Effects of Artists' Education Level,
College of Graduation and Gender on Art Sales Possibility and Art Price: Focusing on MANIF Art Fair
Market. Journal of the Korea Academia-Industrial cooperation Society 14, 1582-1588. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_410426582.pdf
The purpose of this paper is to examine the investment characteristics of works by
leading Australian artists.
Accounting Research Journal
Australian fine art as an alternative investment
Andrew C. Worthington Helen Higgs
Article information:
To cite this document:
Andrew C. Worthington Helen Higgs, (2008),"Australian fine art as an alternative investment", Accounting
Research J ournal, Vol. 21 Iss 1 pp. 55 - 66
Permanent link to this document:
http://dx.doi.org/10.1108/10309610810891346
Downloaded on: 24 January 2016, At: 21:06 (PT)
References: this document contains references to 17 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 364 times since 2008*
Users who downloaded this article also downloaded:
Mostafa M. Maksy, Lin Zheng, (2008),"Factors associated with student performance in advanced
accounting and auditing: An empirical study in a public university", Accounting Research J ournal, Vol. 21
Iss 1 pp. 16-32 http://dx.doi.org/10.1108/10309610810891328
Raymond da Silva Rosa, J ennifer Filippetto, Ann Tarca, (2008),"ASIC actions: canaries for
poor corporate governance?", Accounting Research J ournal, Vol. 21 Iss 1 pp. 67-86 http://
dx.doi.org/10.1108/10309610810891355
May H. Lo, Le (Emily) Xu, (2008),"Do analysts mislead investors?: A comparison of analysts' and investors'
weightings of cash components in forecasting annual earnings", Accounting Research J ournal, Vol. 21 Iss 1
pp. 33-54 http://dx.doi.org/10.1108/10309610810891337
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Australian ?ne art
as an alternative investment
Andrew C. Worthington and Helen Higgs
Department of Accounting, Finance and Economics, Grif?th University,
Gold Coast, Australia
Abstract
Purpose – The purpose of this paper is to examine the investment characteristics of works by
leading Australian artists.
Design/methodology/approach – About 35,805 paintings by 45 leading Australian artists sold at
auction are used to construct individual hedonic price indices. The attributes included in each artist’s
hedonic regression model include the size and medium of the painting and the auction house and year
sold.
Findings – The indexes show that average annual returns across all artists range between 4 and
15 per cent with a mean of 8 per cent, with the highest returns for works by Brett Whiteley,
Jeffrey Smart, Cecil Brack and Margaret Olley. Risk-adjusted returns are generally lower, with
reward-to-volatility and reward-to-variability ratios averaging 1.5 and 5.8 per cent, respectively.
The portfolio bs for individual artistic works average 0.41. The willingness-to-pay for perceived
attributes in the artwork show that works executed in oils and gouache, and those auctioned by
Deutscher-Menzies, Sotheby’s and Christies are generally associated with higher prices.
Research limitations/implications – The returns on a buy-and-hold strategy in the Australian art
market are at least comparable to the Australian stock market. While total risk is greater, the very low
market risk found in almost all artistic portfolios is suggestive of the possible bene?ts of portfolio
diversi?cation through art investment. Moreover, a number of artist’s works offer very superior
market and non-market risk-adjusted performance.
Originality/value – This is the ?rst Australian study to construct measures of risk, return, b and
Sharpe and Treynor ratios for individual Australian artists.
Keywords Arts, Prices, Investments, Australia
Paper type Research paper
1. Introduction
For some time, investors have been turning to art (paintings, sculpture, ceramics and
prints, and collectibles such as coins, stamps, antiques and furniture) as an alternative
investment. In Australia too, there is burgeoning interest in art investment,
particularly the work of Australian artists. Of course, Australia has a long history of
world-renowned artists, including Frederick McCubbin, Arthur Streeton, Tom Roberts
and Arthur Boyd, to name just a few. But in the last few decades many modern
and contemporary painters, like Charles Blackman, Brett Whiteley, David Boyd,
Ray Crooke and John Olsen, have also produced internationally reputable works and
thereby raised public awareness of art as a potential investment opportunity.
The purpose of this paper is to inform this process by investigating the risk, return
and asset pricing of Australian art. Hedonic pricing equations are used to capture the
characteristics of artwork by 45 well-known Australian artists publicly auctioned
during the past 30 years. The paper itself is organised as follows. Section 2 outlines
the empirical methodology. Section 3 provides a description of the data employed.
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
Fine art
as an alternative
investment
55
Accounting Research Journal
Vol. 21 No. 1, 2008
pp. 55-66
qEmerald Group Publishing Limited
1030-9616
DOI 10.1108/10309610810891346
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The empirical results are dealt with in Section 4. The paper ends with some concluding
remarks in Section 5.
2. Empirical methodology
The method selected to examine the investment characteristics of Australian art is the
hedonic price method (Buelens and Ginsburgh, 1993; de la Barre et al., 1994; Chanel,
1995; Agnello and Pierce, 1996). In this approach, all sales (including repeat sales) are
considered as single sales for which the objective features are recorded (i.e. name of
the painter, size of painting, medium of execution, etc.). Combining all sales allows the
implicit (or shadow) prices for these characteristics to be estimated separately from a
characteristic-free price including only the effects of time and random error. Assuming
the availability of comprehensive data, the hedonic price method’s main strengths
are that it estimates values based on actual auction sales and captures the
willingness-to-pay for perceived differences in the attributes of the artwork. The
hedonic price equation is written as:
ln p
kt
¼ f ðX
1kt
; . . . ; X
mkt
; . . . ; X
Mkt
Þ þ gðtÞ þ1
kt
ð1Þ
where lnp
kt
is the natural logarithm of the price of painting k(k ¼ 1, . . . , K) sold in year
t (t ¼ 1, . . . ,T), X
mkt
is the measurable characteristics m (m ¼ 1, . . . ,M) of painting k at
time t, g(t) is a function of time, and the error term 1 , N(0, S
k
%I
T
). The measurable
characteristics of the paintings for each artist comprise the physical characteristics of
the work and the characteristics of the auction at which the sale took place. The
regression equation is then speci?ed as:
ln p
kt
¼
X
M
m¼1
a
m
X
mkt
þ
X
T
t¼1
b
t
Z
t
þ1
kt
ð2Þ
where a
m
are parameter estimates of the implicit prices of the speci?ed art
characteristics, Z
t
is a dummy variable which takes the value of one for a sale occurring
in year t and zero elsewhere, b
t
is a parameter estimate, e
bt
gives the characteristic-free
price and all other variables are as previously de?ned. Separate regression equations
are speci?ed for each artist.
The data used comprises 35,805 sales of artworks by 45 leading Australian artists.
Information on sales is obtained from Australian Art Auction Records (2003) and
spans a 30 year period since March 1973. While an update of this data has just become
available, the requirement for extensive and lengthy data entry and manipulation acted
against this in the current analysis. The selection of artists to be included in
the analysis is, of course, highly subjective and was arrived at after discussion with
various art auctioneers, curators and dealers on those artists whose works were most
sought after and frequently sold at auction in the past 30 years. Its construction is also
re?ective, in so far is possible, of the widest number of periods, schools and genres in
Australian art history and is purposively restricted to artists who lived most of their
lifetime in Australia. A list of the artists is provided in Table I (column 1).
The ?rst set of information gathered is the price of each artwork for each artist. This
comprises the dependent variable in the hedonic price regression. Each artwork
included is sold exclusively at public auction and its value speci?ed in Australian
dollars.
ARJ
21,1
56
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
r
i
c
e
s
(
$
)
R
e
t
u
r
n
R
i
s
k
S
h
a
r
p
e
r
a
t
i
o
T
r
e
y
n
o
r
r
a
t
i
o
b
A
r
t
i
s
t
M
e
a
n
S
D
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
A
s
h
t
o
n
,
J
o
h
n
2
,
6
4
9
3
,
7
8
2
0
.
0
6
2
3
3
0
.
2
6
4
7
0
.
0
4
5
2
5
0
.
0
4
4
1
8
0
.
2
7
1
2
8
B
l
a
c
k
m
a
n
,
C
h
a
r
l
e
s
8
,
0
0
6
2
0
,
4
9
5
0
.
0
7
7
2
3
0
.
2
7
7
1
0
0
.
0
9
7
2
2
0
.
0
7
5
1
0
0
.
3
5
8
2
3
B
o
y
d
,
A
r
t
h
u
r
2
0
,
4
2
6
5
7
,
3
0
5
0
.
0
5
7
3
8
0
.
3
1
7
1
7
0
.
0
2
2
3
5
0
.
0
3
0
2
2
0
.
2
3
2
3
0
B
o
y
d
,
D
a
v
i
d
2
,
6
9
3
2
,
7
7
2
0
.
0
6
3
3
0
0
.
2
1
1
4
0
.
0
6
2
3
6
0
.
0
3
5
2
0
0
.
3
7
3
2
1
B
o
y
d
,
J
a
m
i
e
7
9
6
8
7
2
0
.
0
6
6
2
8
0
.
6
1
2
4
0
0
.
0
2
6
2
7
0
.
0
1
4
2
8
1
.
1
5
3
3
B
r
a
c
k
,
C
e
c
i
l
J
o
h
n
3
5
,
0
1
0
7
6
,
5
2
1
0
.
1
3
8
3
0
.
5
6
6
3
8
0
.
1
5
5
4
0
.
2
7
3
5
0
.
3
2
2
2
6
B
u
n
n
y
,
R
u
p
e
r
t
2
3
,
8
3
7
7
5
,
6
4
0
0
.
0
8
2
2
1
0
.
4
4
7
3
2
0
.
0
7
2
2
0
0
.
0
3
6
1
9
0
.
8
9
7
7
C
o
b
u
r
n
,
J
o
h
n
3
,
4
9
7
6
,
3
8
6
0
.
0
8
3
1
9
0
.
2
9
7
1
5
0
.
1
1
1
1
8
2
0
.
1
1
3
3
8
2
0
.
2
9
1
4
2
C
r
o
o
k
e
,
R
a
y
4
,
0
0
0
6
,
4
7
2
0
.
0
5
1
4
1
0
.
2
3
2
5
0
.
0
0
4
4
4
0
.
0
0
3
3
3
0
.
3
4
3
2
4
D
a
r
g
i
e
,
W
i
l
l
i
a
m
1
,
9
3
1
4
,
3
1
7
0
.
0
4
9
4
2
0
.
5
3
3
3
7
2
0
.
0
0
2
3
4
2
0
.
0
0
1
3
4
0
.
9
7
0
6
D
i
c
k
e
r
s
o
n
,
R
o
b
e
r
t
4
,
3
2
6
8
,
1
2
1
0
.
0
9
2
1
6
0
.
2
0
0
2
0
.
2
1
0
1
0
0
.
1
1
6
7
0
.
3
6
2
2
2
D
r
y
s
d
a
l
e
,
G
e
o
r
g
e
R
u
s
s
e
l
l
3
2
,
9
4
0
1
1
5
,
7
3
1
0
.
0
5
9
3
4
0
.
4
0
1
2
6
0
.
0
2
2
3
2
0
.
0
2
2
2
5
0
.
4
1
8
1
9
F
a
i
r
w
e
a
t
h
e
r
,
I
a
n
1
9
,
6
9
9
2
9
,
3
1
6
0
.
1
0
1
1
4
0
.
6
5
9
4
1
0
.
0
7
7
1
6
2
0
.
3
3
3
4
2
2
0
.
1
5
3
3
8
F
i
z
e
l
l
e
,
R
e
g
i
n
a
l
d
C
e
c
i
l
1
,
5
6
4
4
,
7
6
2
0
.
0
6
7
2
7
0
.
4
4
7
3
2
0
.
0
3
8
2
8
0
.
0
3
3
2
1
0
.
5
2
2
1
7
F
o
x
,
E
t
h
e
l
1
0
,
1
7
2
2
0
,
1
9
7
0
.
0
7
4
2
4
0
.
4
8
4
3
6
0
.
0
5
0
2
5
0
.
0
2
4
2
3
1
.
0
1
7
5
F
r
i
e
n
d
,
D
o
n
a
l
d
4
,
2
7
2
8
,
5
0
1
0
.
0
3
7
4
5
0
.
3
4
8
1
9
2
0
.
0
3
7
4
5
2
0
.
0
2
3
3
6
0
.
5
6
3
1
3
F
u
l
l
b
r
o
o
k
,
S
a
m
u
e
l
8
,
0
4
2
1
0
,
5
7
5
0
.
1
0
1
1
4
0
.
7
7
5
4
2
0
.
0
6
6
1
7
2
0
.
3
9
2
4
4
2
0
.
1
3
0
3
7
G
l
e
e
s
o
n
,
J
a
m
e
s
T
i
m
o
t
h
y
3
,
3
1
0
7
,
2
2
5
0
.
0
7
4
2
4
0
.
2
9
4
1
4
0
.
0
8
2
2
4
0
.
0
9
9
8
0
.
2
4
2
2
9
G
l
o
v
e
r
,
J
o
h
n
1
0
,
5
7
2
4
8
,
5
8
0
0
.
0
6
4
2
9
0
.
4
7
2
3
5
0
.
0
3
0
2
9
2
0
.
3
2
6
4
1
2
0
.
0
4
3
3
6
G
r
u
n
e
r
,
E
l
i
o
t
h
1
1
,
1
9
5
1
4
,
8
2
4
0
.
0
4
8
4
3
0
.
4
2
3
2
8
2
0
.
0
0
5
3
9
0
.
0
0
7
3
2
2
0
.
3
0
4
4
3
H
a
r
t
,
K
e
v
i
n
C
h
a
r
l
e
s
(
P
r
o
)
1
,
4
4
2
2
,
6
7
4
0
.
0
5
8
3
7
0
.
1
8
9
1
0
.
0
4
2
4
2
0
.
0
7
0
1
4
0
.
1
1
4
3
3
H
o
d
g
k
i
n
s
o
n
,
F
r
a
n
k
1
,
5
2
6
2
,
5
0
9
0
.
1
0
7
8
0
.
6
0
0
3
9
0
.
0
9
5
1
4
2
0
.
0
9
8
3
7
2
0
.
5
8
2
4
5
H
e
y
s
e
n
,
H
a
n
s
8
,
5
7
1
1
6
,
6
5
3
0
.
0
5
9
3
4
0
.
2
3
5
6
0
.
0
3
8
4
0
0
.
0
1
4
2
9
0
.
6
6
2
9
H
e
y
s
e
n
,
N
o
r
a
3
,
1
5
8
5
,
3
8
3
0
.
0
7
4
2
4
0
.
8
4
7
4
3
0
.
0
2
8
2
6
0
.
0
7
2
1
2
0
.
3
3
4
2
5
J
a
c
k
s
o
n
,
J
a
m
e
s
R
a
n
a
l
p
h
5
,
8
9
4
9
,
6
6
2
0
.
0
5
9
3
4
0
.
2
7
4
9
0
.
0
3
3
3
7
0
.
0
1
6
2
7
0
.
5
6
8
1
1
L
i
n
d
s
a
y
,
N
o
r
m
a
n
5
,
8
2
2
1
3
,
6
5
7
0
.
0
5
7
3
8
0
.
2
0
1
3
0
.
0
3
5
4
3
0
.
0
1
2
3
0
0
.
6
0
6
1
0
L
o
n
g
,
S
y
d
n
e
y
4
,
0
7
3
8
,
7
0
2
0
.
0
9
2
1
6
0
.
3
0
1
1
6
0
.
1
4
0
1
3
0
.
3
2
8
4
0
.
1
2
8
3
2
M
c
C
u
b
b
i
n
,
F
r
e
d
e
r
i
c
k
5
5
,
2
4
5
1
7
1
,
0
1
4
0
.
1
0
3
1
0
0
.
4
3
4
3
0
0
.
1
2
2
1
1
0
.
6
9
7
2
0
.
0
7
6
3
4
(
c
o
n
t
i
n
u
e
d
)
Table I.
Prices, return and risk,
performance measures
and bs by artist
Fine art
as an alternative
investment
57
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
r
i
c
e
s
(
$
)
R
e
t
u
r
n
R
i
s
k
S
h
a
r
p
e
r
a
t
i
o
T
r
e
y
n
o
r
r
a
t
i
o
b
A
r
t
i
s
t
M
e
a
n
S
D
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
S
t
a
t
i
s
t
i
c
R
a
n
k
N
a
m
a
t
j
i
r
a
,
A
l
b
e
r
t
7
,
3
3
9
6
,
1
3
2
0
.
0
8
5
1
8
0
.
2
7
1
8
0
.
1
2
9
1
5
0
.
0
7
1
1
3
0
.
4
9
4
1
8
N
o
l
a
n
,
S
i
d
n
e
y
1
1
,
1
8
2
4
2
,
8
5
2
0
.
0
5
4
4
0
0
.
3
3
0
1
8
0
.
0
1
2
3
8
0
.
0
0
7
3
1
0
.
5
6
2
1
4
O
l
l
e
y
,
M
a
r
g
a
r
e
t
1
2
,
5
2
9
1
5
,
9
3
0
0
.
1
3
5
4
0
.
4
0
4
2
7
0
.
2
1
0
3
0
.
1
5
9
6
0
.
5
3
3
1
5
O
l
s
e
n
,
J
o
h
n
9
,
1
1
8
2
4
,
8
2
1
0
.
1
0
9
7
0
.
3
9
1
2
2
0
.
1
5
1
9
2
0
.
3
8
3
4
3
2
0
.
1
5
4
3
9
P
e
r
c
e
v
a
l
,
J
o
h
n
1
4
,
1
3
3
3
8
,
2
5
6
0
.
0
6
3
3
0
0
.
3
9
8
2
4
0
.
0
3
3
3
1
0
.
0
7
3
1
1
0
.
1
7
7
3
1
P
r
e
s
t
o
n
,
M
a
r
g
a
r
e
t
1
2
,
4
7
0
2
6
,
2
4
4
0
.
1
0
2
1
2
0
.
4
4
4
3
1
0
.
1
1
7
1
2
0
.
0
5
0
1
6
1
.
0
4
7
4
P
r
o
c
t
o
r
,
A
l
t
h
e
a
1
,
8
6
7
3
,
0
7
2
0
.
1
1
2
6
0
.
3
6
1
2
1
0
.
1
7
2
5
2
0
.
1
3
8
3
9
2
0
.
4
5
0
4
4
R
e
e
s
,
L
l
o
y
d
9
,
6
1
7
2
0
,
6
6
9
0
.
0
7
9
2
2
0
.
3
6
0
2
0
0
.
0
8
1
2
1
0
.
0
5
4
1
5
0
.
5
3
3
1
5
R
o
b
e
r
t
s
,
T
h
o
m
a
s
W
i
l
l
i
a
m
2
4
,
1
6
8
4
8
,
8
4
8
0
.
0
8
3
1
9
1
.
9
3
3
4
5
0
.
0
1
7
2
3
0
.
0
8
0
9
0
.
4
1
2
2
0
R
u
s
s
e
l
l
,
J
o
h
n
P
e
t
e
r
4
5
,
1
6
7
1
0
0
,
0
7
9
0
.
1
0
7
8
1
.
4
7
7
4
4
0
.
0
3
9
1
9
0
.
0
1
7
2
6
3
.
3
3
9
1
S
m
a
r
t
,
F
r
a
n
k
J
e
f
f
r
e
y
3
6
,
5
4
4
5
1
,
7
7
4
0
.
1
4
6
2
0
.
4
5
6
3
4
0
.
2
1
1
2
2
0
.
5
9
3
4
5
2
0
.
1
6
2
4
0
S
m
i
t
h
,
G
r
a
c
e
C
o
s
s
i
n
g
t
o
n
1
7
,
2
0
4
3
0
,
3
2
3
0
.
1
1
6
5
0
.
4
2
8
2
9
0
.
1
5
4
7
0
.
0
4
6
1
7
1
.
4
2
5
2
S
t
r
e
e
t
o
n
,
A
r
t
h
u
r
3
1
,
8
0
0
6
1
,
5
8
7
0
.
1
0
3
1
0
0
.
2
8
3
1
2
0
.
1
8
7
6
2
.
3
0
4
1
0
.
0
2
3
3
5
T
u
c
k
e
r
,
A
l
b
e
r
t
1
4
,
7
6
4
3
8
,
7
9
1
0
.
0
6
3
3
0
0
.
3
9
8
2
4
0
.
0
3
3
3
0
0
.
0
2
3
2
4
0
.
5
6
6
1
2
W
h
i
t
e
l
e
y
,
B
r
e
t
t
2
3
,
9
2
7
8
2
,
4
6
5
0
.
1
4
7
1
0
.
2
8
1
1
1
0
.
3
4
5
1
0
.
3
3
9
3
0
.
2
8
6
2
7
W
i
l
l
i
a
m
s
,
F
r
e
d
e
r
i
c
k
2
1
,
3
0
5
4
9
,
7
7
9
0
.
1
0
2
1
2
0
.
2
9
0
1
3
0
.
1
7
9
8
2
0
.
2
4
3
4
0
2
0
.
2
1
4
4
1
W
i
t
h
e
r
s
,
W
a
l
t
e
r
1
2
,
0
9
7
3
0
,
5
0
2
0
.
0
4
3
4
4
0
.
3
9
2
2
3
2
0
.
0
1
8
4
1
2
0
.
0
0
9
3
5
0
.
8
0
7
8
Table I.
ARJ
21,1
58
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The next two sets of variables are considered to be major determinants of the price of
an artist’s work and are speci?ed as explanatory variables. The ?rst set of explanatory
variables relate to the physical characteristics of the work, while the second set
comprise its sale characteristics. Starting with the physical characteristics of the
works, the ?rst group comprises dummy variables identifying the medium used:
namely, acrylic, charcoal, crayon, etching, the heavy, opaque watercolour paint known
as gouache, mixed media, oil, pastel, pencil and watercolour. The reference category is
all other mediums. Of the mediums used by artists in the sample, the largest numbers
of works (percentage of all mediums in brackets) sold are oils (44 per cent), followed by
watercolours (11 per cent), and ?nally etchings (8 per cent). Oil as a medium, though
dif?cult to work, has excellent visual qualities and is not easily faded by natural light.
It is therefore likely to fetch higher prices at auction. Modern alternatives, including
acrylic and gouache, also command high prices. However, a variety of other potentially
valuable media are found in most ?ne art collections. Australian landscape artists, for
instance, often favour watercolours.
The second group of physical characteristic are the dimensions of the painted work
as represented by surface area in square metres (m
2
) and surface area squared as the
nonlinear component. A positive relationship is generally hypothesised when price is
regressed against area, although it is dif?cult for all but the largest public galleries to
display very large works. On this basis, the expected sign on the coef?cient for area
squared is thought to be negative (Agnello and Pierce, 1996). Of course, there are any
number of other physical characteristics that could be included if data were available.
These include the painting’s genre, providence, date completed, the presence of the
artist’s signature and so on.
The second set of explanatory variables incorporate the sales characteristics of the
work. The ?rst of these are dummy variables identifying in which of Australia’s three
largest auction houses the sale took place (percentage of all sales in brackets): that is,
Christies (16 per cent), Deutscher-Menzies (4 per cent) and Sotheby’s (15 per cent). The
reference category is all other auction houses. In the absence of transaction costs, the
law of one price dictates that no signi?cant price difference should exist for paintings
of a similar quality. However, Pesando (1993), de la Barre et al. (1994) and Renneboog
and van Houtte (2002), amongst others, have found that Christies and Sotheby’s
systematically obtain higher hammer prices through their reputation and market
power. The second group of sales characteristics identi?es the year when the work is
sold. This consists of thirty yearly dummy variables with 1973 as the reference
category.
3. Descriptive analysis
The mean and standard deviation of prices for each artist’s works are included in
Table I (columns 2 and 3). Turning ?rst to the prices of artworks by artist, average
prices range from $796 for paintings by Boyd ( Jamie) to $55,245 for works by
McCubbin. Other artist’s with highly valued works are Russell, Smart and Brack with
means of $45,167, $36,544 and $35,010, respectively. The lowest prices are paid for
works by Hart, Hodgkinson and Fizelle with average prices of $1,442, $1,526 and
$1,564, respectively. The standard deviations of art prices by artist range from $872 to
$171,014. On this basis, works by Boyd ( Jamie), Hodgkinson, Hart and Boyd (David)
are the least volatile with standard deviations of $872, $2,509, $2,674 and $2,772,
Fine art
as an alternative
investment
59
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
respectively, whereas works by McCubbin, Drysdale, Russell and Whiteley are the
most volatile with standard deviations of $171,014, $115,731, $100,079 and $82,465,
respectively.
Now consider each artist’s work categorised according to the media used
(not shown). Of these, oils, watercolours and etchings are the most common medium
sold, while crayons and acrylics are the least common. However, the distribution of all
physical and sales characteristics varies dramatically by artist. For example, sold work
by Nolan, Fullbrook and Bunny are almost exclusively oil, while Namatjira’s sold
works are primarily watercolours. Similarly, the distribution of sold works by auction
house also varies across the artists. For example, 54 per cent of Whiteley’s work was
sold at Christies, 32 per cent of Brack’s at Deutscher-Menzies and 54 per cent of
Fullbrook’s at Sotheby’s. This contrasts markedly to the aforementioned sample
averages.
4. Empirical results
In the interests of brevity, the estimated coef?cients of the hedonic pricing regression
models for each of the 45 Australian artists are not shown. All of the estimated models
are highly signi?cant, with likelihood ratio tests (not shown) of the hypotheses that
the slope coef?cients are zero rejected at the 1 per cent level. The adjusted R
2
range between 0.547 and 0.889 and are reasonably high for what is basically
cross-sectional data.
As hypothesised, the percentage changes in value indicate that works executed in
oil and gouache command higher prices, with average percentage increases over each
artist’s standard work of 6.799 and 6.733 per cent, respectively. Of the 44 artists in the
sample with at least some oil works, all but one have signi?cant and positive increases
in value relative to other work, while 28 of the 30 artists with gouache works have
signi?cant and positive increases with this media. However, the percentage increases
in value for individual artists vary widely. For example, with oils the increase in values
ranges from as little as 1.188 per cent (Fizelle) to more than 21.700 per cent (Bracks) and
for gouache from just 0.684 per cent (Olley) to 68.217 per cent (Smith).
By comparison, media such as etchings, crayon and charcoal are associated with
respective average percentage increases across the sample of just 1.105, 3.020 and 1.787
per cent implying these media are generally more affordable, regardless of all other
characteristics, while mixed media, watercolours and pastels have average price
increases across the sample of 3.466, 3.346 and 3.646 per cent, respectively.
Unfortunately, it is dif?cult to compare these ?ndings because other studies are often
limited to periods or movements when fewer media are generally known (de la Barre
et al., 1994; Renneboog and van Houtte, 2002) or to a single medium (Candela and
Scorcu, 1997; Pesando and Shum, 1999). Nevertheless, Agnello and Pierce (1996) found
a 156 per cent increase in prices for oil works as compared to all other media
(watercolour, gouache, ink, pencil, pastel, etc.).
The remaining physical characteristics included in the regression models concern
the size of the work. These are the area of the work in square metres and its nonlinear
component, area squared. The generally positive and signi?cant signs of the area
coef?cients and the negative and signi?cant signs of its squared term indicate that art
prices tend ?rst to increase with size, then decrease as the paintings become too large
and dif?cult to house. Across the sample, a 1 per cent increase in surface area is
ARJ
21,1
60
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
associated with a 0.206 per cent increase in price, while on average the
price-maximising size is 4.08 m
2
. By comparison, Agnello and Pierce (1996) found
the price-maximising size for American artists’ work to be 6.53 m
2
while de la Barre
et al. (1994) calculated this optimal size to be 5.89 m
2
for old masters and 1.70 m
2
for
modern and contemporary European works.
The ?nal set of variables relates to the sale characteristics of the works. The sales
characteristics show that auctions at Sotheby’s, Christies and Deutscher-Menzies
increase the average standard price by 1.759, 1.697 and 1.869 per cent, respectively,
over the remaining houses. Pesando (1993), de la Barre et al. (1994), Agnello and Pierce
(1996) and Renneboog and van Houtte (2002) also found that “. . .Sotherby’s typically
fetches higher prices than Christies, while both experience higher prices than all other
houses” (Agnello and Pierce, 1996, p. 366). However, while variation in the prices
obtained by the different auction houses are small, and certainly smaller than most
other factors included in the model, care should still be taken in interpreting these
differences as a violation of the law of one price. de la Barre et al., 1994, p. 165), for
example, concluded “. . .the quality of a painting, not captured by our characteristics is
partly picked up by the saleroom coef?cients: a “good” Picasso would go to Christies or
Sotheby’s New York, a less good one would be sold at Drouot’s [a Paris-based auction
house]. . .it is impossible to disentangle the two effects”.
Turning to investment risk and return, the index value for each artist is calculated
as 100e
bt
(not shown). Annual returns are then calculated such that the return for artist
i is represented by the continuously compounded return or log return of the price index
at time t such that Dp
it
¼ log ( p
it
/p
it21
) £ 100 where Dp
it
denotes the rate of change of
p
it
. Table I presents the arithmetic mean and standard deviation (risk) of annual
returns and their ranks for the 45 Australian artists (columns 4 to 7).
Also included in Table I are two external risk-adjusted portfolio performance
measures. The Sharpe ratio and its rank (also known as the reward-to-volatility ratio)
(columns 8 and 9) indicates the excess return per unit of risk and is calculated by
dividing the return in excess of the risk-free rate by the standard deviation of returns.
The proxy used for the risk-free rate is the exponentially smoothed average ?tted yield
for three-year Commonwealth Treasury bonds during the sample period (5 per cent).
In the current context, the Sharpe ratio is the most appropriate performance measure
for an investor whose portfolio is composed wholly of a given artist’s work.
The Treynor ratio and its rank (sometimes called the reward-to-variability ratio;
columns 10 and 11) is identical to the Sharpe ratio except that total risk (standard
deviation) is replaced with systematic (market) risk or b. This ratio may be a better
benchmark of performance for investors who do not invest exclusively in art, but
rather consider its diversi?cation potential. Accordingly, the b of each artist’s portfolio
and its rank (columns 12 and 13) is calculated with respect to an equity market
portfolio. The Australian All Ordinaries index (AOI) is speci?ed as the equity market
measure. Since higher Sharpe and Treynor ratios represent better performance, the
artistic portfolios are ranked in descending order.
In terms of returns, mean returns for the individual artists range between
3.70 per cent for works by Friend to 14.70 per cent for those by Whitelely. Annual
returns across all artists average 8.23 per cent, as compared to mean returns on the
AOI of 7.00 per cent over this same period. Other artists with relatively high returns
include Smart, Brack, Olley, Smith, Proctor and Olsen, with relatively low returns for
Fine art
as an alternative
investment
61
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Withers, Gruner, Dargie, Crooke, Nolan, Lindsay and Boyd (Arthur). The standard
deviation (risk) of returns ranges between 0.189 (Hart) to 1.933 (Roberts) with a mean of
0.449. By way of comparison, the standard deviation of returns on the AOI over this
same period was 0.161. Risk is also relatively higher for works by Russell, Heysen
(Nora), Fullbrook and Fairweather and relatively lower for works by Dickerson,
Lindsay, Boyd (David) and Crooke.
For overall returns on Australian art, it would appear that the market has
performed at a comparable level to other national markets. Renneboog and van Houtte
(2002), for example, found Belgian average returns of 8.4 per cent over the period
1970-1989 with a standard deviation of 19.4 per cent, Agnello and Pierce (1996)
estimated that the returns on American artists averaged 9.3 per cent from 1971-1992,
and Mei and Moses (2001) calculated average returns of 5.3 per cent with a standard
deviation of 9.3 per cent, also on American auctions. Mean returns from other art
studies include 1.6 per cent (Frey and Pommerehne, 1989), 6.8 per cent (Gerard-Varet,
1995) and 5.0 per cent (Goetzmann, 1996). Using a different methodology and an
international sample, Worthington and Higgs (2003, 2004) found mean returns of
3.73 per cent for Contemporary Masters, 2.85 per cent for 20th Century English works,
1.49 per cent for Modern Europeans and 1.25 per cent Surrealists. Of course, the returns
as calculated in this paper do not re?ect the fact that a substantial component of the
return from art investment is derived not from its ?nancial returns, rather from its
intrinsic aesthetic qualities. Equally, they also do not include the many and sizeable
transaction and holding costs associated with art portfolios, the absence of which may
serve to in?ate the ?nancial returns.
Analysis of the risk-adjusted returns for each artist’s portfolio of works provides
further insights. Starting with the Sharpe ratio, artists ranked highly on the basis of
returns per unit of (total) risk include Whiteley (0.345), Smart (0.211), Olley (0.210),
Brack (0.155) and Proctor (0.172). The Sharpe index for the AOI over this same period
is 0.124 while that for the average artist included in the sample is just 0.015. As shown,
many artists have low Sharpe ratios (and rankings) suggesting that a policy of holding
high-return-high-risk portfolios of a single artist’s work in isolation may not be an
appropriate investment strategy. As an alternative, the Treynor ratio show the returns
per unit of (systematic) risk and thus yields useful insights on the bene?t of holding
Australian art as part of a diversi?ed portfolio (though, of course, limited in this
analysis to listed equity).
As shown in Table I, the bs of most Australian artist’s work are low (,1), if not
negative, indicating potential diversi?cation bene?ts. For example, the negative bs
calculated on art portfolios composed of works by Hodgkinson, Proctor, Gruner,
Coburn and Williams indicate that their returns move contrary to returns on the
Australian stock market. However, some art portfolios are substantially more risky
(in terms of b) than the market, and move in the same direction, including Russell,
Smith, Boyd ( Jamie), Preston and Fox. The average b across the sample is 0.405 with
25 per cent of artistic portfolios having a b , 0.075 and 25 per cent .0.566. By
comparison, Chanel et al. (1994) calculated that national art bs ranged between 0.028
( London) and 0.368 (Tokyo), while Renneboog and van Houtte (2002) estimated
movement bs with respect to a global stock index of 23.7, 22.9 and 0.8 for
Impressionist, Luminist and Expressionist art, respectively.
ARJ
21,1
62
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The Treynor ratios for the 45 artist portfolios in mean/b-space are graphically
represented in Figure 1. The ?gure includes the security market line constructed using
the risk-free rate (intercept) and the return and b (one) for the market portfolio (slope).
Visual inspection indicates that few artist portfolios are correctly priced in relation to
the security market line (that is, on the line) with most underpriced. That is, artist
portfolios lying above the line indicate superior market risk-adjusted returns and a buy
signal, while those lying below the line indicate inferior market risk-adjusted returns
and a sell signal. Artists ranked highly on the basis of the Treynor index include
Streeton, McCubbin, Whiteley, Long and Brack. A buy signal is indicated for these
artists. Those ranked lowly with a consequent sell signal include Smart, Olsen,
Fullbrook, Fairweather and Glover.
Of course, these buy-and-sell strategies must be quali?ed by the fact that they relate
to historical information averaged over a thirty year investment horizon, not the
immediate past, present or future. It should also be remembered that the Treynor ratio
re?ects only systematic (general or market) risk and thereby re?ects the value of these
assets within a diversi?ed portfolio. The change in rankings of artists between
the Sharpe and Treynor measures indicate that most art portfolios as analysed include
much unsystematic (asset speci?c or nonmarket) risk when held in isolation.
Nonetheless, while the rankings of artists on the Sharpe and Treynor criterion do vary,
there is some deal of correspondence between them with the Spearman (rank)
correlation coef?cient signifying a signi?cant and positive relationship ( r ¼ 0.319,
p-value ¼ 0.033).
A ?nal requirement is to examine the relationship between the returns and values of
works included in each artistic portfolio. This follows the suggestion of Mei and Moses
(2001) amongst others that bidders in art auctions are exposed to a “winner’s curse” so
that the returns on expensive paintings tend to under perform the market as a whole:
referred to as the “masterpiece effect”. Figure 2 plots the returns and mean prices of
paintings for each artist, with a linear trendline added as a simple means of evaluating
the relationship between value and return. As shown, there is a small positive
Figure 1.
Market risk and
return by artist
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
–1.00 –0.50 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
b
R
e
t
u
r
n
Fine art
as an alternative
investment
63
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(1.03 £ 10
-6
) and signi?cant ( p-value ¼ 0.019) relationship suggesting that returns
increase with value, thereby supporting the absence of a “masterpiece effect” effect.
5. Conclusion
This paper investigates risk, return and asset pricing for the works of 45 well-known
Australian artists. Unlike most other work in this area which indicates that the returns
to art investment are much less, and the risks much higher than investment markets,
the results show that return on a buy-and-hold strategy in the Australian art market
are at least comparable to the stock market. While total risk is indeed greater than the
stock market, the very low market risk found in almost all artistic portfolios is highly
suggestive of the possible bene?ts of portfolio diversi?cation through art investment.
That said, a number of artist’s works offer superior market and non-market
risk-adjusted performance over the sample period, above all Arthur Streeton, Frederick
McCubbin, Brett Whiteley, Sydney Long, Cecil Brack, Frank Smart, Margaret Olley
and Althea Proctor. One major quali?cation is that the analysis does not take into
account the (high) transaction costs incurred at the moment of sale nor the (equally
sizeable) insurance and other costs associated with restoring, preserving and
displaying art works. However, neither does it take into account the (equally
substantial) aesthetic returns from owning and displaying ?ne art. The methodology
employed in the paper also identi?es factors associated with higher prices in the
Australian art market. All other things being equal, larger-sized works and those
executed in oils or gouache, and auctioned by Sotheby’s or Christies are associated
with higher prices. Conversely, smaller works, etchings, crayon or charcoal works,
along with those auctioned by other auction houses, are associated with systematically
lower prices.
There are many interesting opportunities to expand upon this work. One possibility
would involve gathering additional information to be included in the hedonic pricing
regression models. For example, the prices (and hence returns) on artists’ work may
also depend on the cumulative number of works auctioned, whether the artist is now
Figure 2.
Mean price and return
by artist
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
$0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,000
Mean price
R
e
t
u
r
n
ARJ
21,1
64
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
deceased or the age of the artist at the time of auction, genres of work, interactions
between medium and size and so on. While these impacts are proxied to some extent by
the variables included in the current analysis, a more de?ned speci?cation would
identify some determinants potentially obscured. There may also be opportunities to
examine art markets along the lines of the market ef?ciency literature. One prospect is
to examine the time-series behaviour of returns to examine whether the art market
fully incorporates all historical market information (weak-form ef?cient).
Finally, the art works on which these indices are based may not re?ect the market
for Australian paintings as a whole: private transactions for example conducted
through art galleries are ignored. Depending on the values found in galleries, indexes
constructed using auction data may understate or overstate the true return. There is
also no recognition that different buyers in the market have differing preferences for
artwork: compare, for instance, works bought by public galleries to those purchased
privately. For this reason, sellers of art to public collections are argued to enjoy
systematically higher rates of return. Future work could take into account these
subtleties.
References
Agnello, R.J. and Pierce, R.K. (1996), “Financial returns, price determinants, and genre effects
in American art investment”, Journal of Cultural Economics, Vol. 20, pp. 359-83.
Australian Art Auction Records Pty. Ltd (2003), Australian Art Auction Records, 1972-2003,
Sydney, CD.
Buelens, N. and Ginsburgh, V. (1993), “Revisiting Baumol’s ‘art as a ?oating crap game’”,
European Economic Review, Vol. 37, pp. 1351-71.
Candela, G. and Scorcu, A.E. (1997), “A price index for art market auctions”, Journal of Cultural
Economics, Vol. 21, pp. 175-96.
Chanel, O. (1995), “Is art market behaviour predictable?”, European Economic Review, Vol. 39,
pp. 519-27.
Chanel, O., Gerard-Varet, L.A. and Ginsburgh, V. (1994), “Prices and returns on paintings:
an exercise on how to price the priceless”, Geneva Papers on Risk and Insurance Theory,
Vol. 19, pp. 7-21.
de la Barre, M., Docclo, S. and Ginsburgh, V. (1994), “Returns of impressionist, modern and
contemporary European paintings 1962-1991”, Annales d’Economie et de Statistique,
Vol. 35, pp. 143-81.
Frey, B. and Pommerehne, W. (1989), “Art investment: an empirical inquiry”, Southern Economic
Journal, Vol. 56, pp. 396-407.
Gerard-Varet, L.A. (1995), “On pricing the priceless: comments on the economics of the visual art
market”, European Economic Review, Vol. 39, pp. 509-18.
Goetzmann, W.N. (1996), “How costly is the fall from fashion? Survivorship bias in the painting
market”, in Ginsburgh, V.A. and Menger, P.M. (Eds), Economics of the Arts: Selected
Essays, Elsevier North-Holland, Amsterdam, pp. 71-84.
Mei, J. and Moses, M. (2001), “Art as an investment and the origin of the masterpiece effect:
evidence from 1875-2000”, paper presented at the 8th Asia-Paci?c Finance Association
Annual Conference, Shangri-La Hotel, Bangkok, 22-25 July.
Pesando, J.E. (1993), “Arts as an investment: the market for modern prints”, American Economic
Review, Vol. 83, pp. 1075-89.
Fine art
as an alternative
investment
65
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Pesando, J.E. and Shum, P.M. (1999), “The returns to Picasso’s prints and to traditional ?nancial
assets, 1977 to 1996”, Journal of Cultural Economics, Vol. 23, pp. 183-92.
Renneboog, L. and van Houtte, T. (2002), “The monetary appreciation of paintings: from realism
to Magritte”, Cambridge Journal of Economics, Vol. 26, pp. 331-57.
Worthington, A.C. and Higgs, H. (2003), “Art as an investment: short and long-term
comovements in major painting markets”, Empirical Economics, Vol. 28, pp. 649-68.
Worthington, A.C. and Higgs, H. (2004), “Art as an investment: risk, return and portfolio
diversi?cation in major painting markets”, Accounting and Finance, Vol. 44, pp. 257-72.
Further reading
Goetzmann, W.N. (1993), “Accounting for taste: art and the ?nance markets over three centuries”,
American Economic Review, Vol. 83, pp. 1370-6.
Corresponding author
Andrew C. Worthington can be contacted at: a.worthington@grif?th.edu.au
ARJ
21,1
66
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Dan-Bi Choi, Taeyoung Chung, Hyung-Deok Shin. 2013. The Effects of Artists' Education Level,
College of Graduation and Gender on Art Sales Possibility and Art Price: Focusing on MANIF Art Fair
Market. Journal of the Korea Academia-Industrial cooperation Society 14, 1582-1588. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
0
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_410426582.pdf