Art market while being sporadic and unpredictable can still
be a sound alternative for investments since art auctions tend to show a rise
of about 9-10% per annum provided correct estimates of the returns on
investments is predicted (Korteweg, Kräussl and Verwijmeren, 2013). A question
remains if the aforementioned percentage on return of investment is correct
since often the same artwork undergoes multiple sales. Even if a mid-figure
average of 5% increase be taken, it proves as a positive investment, if not an
outperforming or over-performing portfolio diversification.
Going further it can be observed that financial calculation
on artworks are proportional to the art movements or artists associated with
them (Mei and Moses, 2002, Renneboog and Spaenjers, 2013). Factors which can
change the dynamics of calculation include where the artworks are sold since
same piece of art can fetch separate prices in separate regions (Frey and
Eichenberger, 1995) and the inclusion of fake artworks. If the following raw
data in fig. 1.1. and 1.2 is taken, it can be observed that same artworks have
a varied price per region while overall it does correspond to increase in
regional consumer demand in same regions.
Fig. 1.1 Fig.
Without a linearity in data output and varying regions, it
can be said that art market follows an unpredictable and opaque price formation
pattern. An annual index prepared by Renneboog and Spaenjers (2013) consisting
of artworks sold in past forty years can be used to witness the opaque price
formation. If the annual return on investment on artworks is taken and the
variance-ratio test is used, the price-variance difference of order q would
equal q times the variance of the first difference (Lo and MacKinlay, 1988).
test (q = 2, …9)
0.04 to 0.095
Rejection of the
of order one
Pesando (1993), Ljung-Box test (1978) determinants detect
the auto-correlation of order on net returns. Empirical auto-correlation
average from above table is 0.06 using Ljung-Box test which shows that the net
return on investment remains at a minimum of 60% for a one-year period. If the
same results are noted it would point at an exceptionally high return on
investment. However, from the second year onwards, correlation does not remain same
and using the variance ratio analysis shows rather shows a 30% mark (Erdös and
Ormos, 2010). To recheck, a run test shows the growth at 0.16 which would be a
16% on year one, which finally proves that despite independent tests being run
on same data period, the return on investment ratio keeps changing which proves
that the art-market follows an opaque price formation pattern.