Investment horizon for private‐value assets: Evidence from the art market

• Published date: 01 June 2022
• Author: Milad Nozari
• Date: 01 June 2022
• DOI: http://doi.org/10.1111/jfir.12273

Received: 29 May 2019

|

Accepted: 19 January 2022

DOI: 10.1111/jfir.12273

ORIGINAL ARTICLE

Investment horizon for private‐value assets:

Evidence from the art market

Milad Nozari

International Center for Finance, Yale School

of Management, Yale University, New Haven,

Connecticut, USA

Correspondence

Milad Nozari, 165 Whitney Avenue,

New Haven, CT 06511, USA.

Email: milad.nozari@yale.edu

Abstract

Over the past two decades, the idea of purchasing art as an

asset class has attracted many collectors with investment

goals to the art market. I examine how collectors trade

art by focusing on their investment horizons. Controlling

for quality and market conditions, I show that collectors'

personal tastes and uncertainty in valuation could affect

how long they keep their artwork. I also find that specific

characteristics of artwork or artists are determinants of

investment horizons. The natural conclusion of the findings

is that artwork with high common market value is traded

more frequently, whereas higher personal valuation can

lead to longer duration ownership.

JEL CLASSIFICATION

Z11

1|INTRODUCTION

The booming art market has attracted wealthy individuals the last few decades. The majority of art collectors trade

through auctions in a mysterious but somewhat visible market. Although some collectors enjoy possessing art for its

aesthetic pleasure, ownership value, or social status, others also consider its potential pecuniary gains. Numerous

studies have focused on the return and price of art, but less attention has been paid to the timing and investment

length, which are important elements for any financial decision. I study the factorsthat influence collectors' decision

to sell art in their possession and examine investor holding periods as a function of investors' preference and art

characteristics.

Art can be considered both a consumption good and an investment asset. One can view art as an asset with

nontradable emotional dividends (Mandel, 2009). Similar to financial assets, an investor's decision to buy art can be

influenced by his/her appraisal of expected profit or capital gain. Therefore, if an investor considers purchasing

J Financ Res. 2022;45:229–246. wileyonlinelibrary.com/journal/JFIR

|

229

© 2022 The Southern Finance Association and the Southwestern Finance Association.

artwork for its pecuniary benefits, she also needs to consider the investment horizon. A more extended holding

period may indicate difficulties in selling or lower quality; it may also imply a strong personal preference for keeping

and enjoying the artwork. Investigating the holding periods of art can help us better understand trade incentives in

the art market.

Art with a shorter investment horizon has a higher probability of being resold in a specified time. I use duration

models and data for artworks auctioned in a representative sample of leading salerooms worldwide from 1990 until

2008 to determine what affects the probability of resale attempts in a given interval. Because I explore trade

incentives and holding periods, I include the resale attempts in addition to the successful auctions. My focus is on

the secondary market and visible auction records; private sales are not included in the analysis due to data

availability.

The decision to sell art is affected by the market as well as the owner's private valuation. Consequently,

I consider the following aspects in my tests. First, auction houses are the primary source of information about art

quality. Their reputation or expertise in providing price estimates could be an important factor affecting trade

decisions. Therefore, I include their estimates in my analysis. Also, to proxy the private value, I compare collectors'

bids to these estimates. Second, because collectors consider the characteristics of artwork and artists when

they trade, I control for these. Finally, I take into account macroeconomic conditions, such as the time or location of

the trade.

Before any auction event, experts evaluate artwork and provide price estimates. Whereas Milgrom and Weber

(1982) and Mei and Moses (2005) argue that these estimates are not biased, Ashenfelter and Graddy (2003)

challenge the assumption that ex ante valuations can be considered unbiased predictors of the market value.

Therefore, I test whether presale estimates reflect the market value of art. Following the method suggested by

McAndrew et al. (2012), I find that the expert appraisals in the data set are not biased.

1

Because presale estimates reflect the market value of art, those who pay higher than the estimates could have a

stronger preference to purchase the artwork. I find that those who bid more than the ex ante appraisals and

possibly purchase art mainly for enjoyment keep the artwork for a more extended period. Also, the presale estimate

includes a price range with low and high values. In line with Ashenfelter and Graddy (2006) and McAndrew et al.

(2012), I interpret a bigger gap between the high and low estimates as reflecting higher uncertainty about market

value.

2

My results suggest that this uncertainty in art appraisal leads to a longer possession interval. The uncertainty

could be due to a lack of experience in evaluating the quality of a specific work or a tiny population of interested

potential buyers (Bruno et al., 2018).

Art owners consider the characteristics of artwork or artists when they trade, and therefore I control for these.

The collecting category, size, price tag, and medium for each piece of artwork are considered in my tests. I also

explore the effect of artists' fame on trade decisions. Specifically, I proxy for each artist's recognition and

importance by looking at his/her origin and the number of his/her works presented in auction events.

3

Based on the

results, higher priced artwork, works on canvas, and contemporary art are more likely to reappear in auction events

than are cheaper artwork, works on paper, and old masters or modern art. Also, the origin of artists and their

presence in the auction market affects resale decisions. The findings are in line with the hypothesis that the same

works from well‐known artists show up more frequently for sale.

Moreover, the transaction costs are high in this market (15% on average) and could be important for those who

pay attention to financial gains. However, after controlling for the price level, I do not find that the transaction cost

has a significant effect on resale decisions. According to Ashenfelter and Graddy (2003), the buyers' premium is

1

Considering all the artworks that are brought to auction, the true value is slightly higher than expert appraisals; however, this difference is less than 0.4%

and negligible compared to the magnitude of the effects I report. The estimation results are robust using various initial guesses and after trimming the

outliers.

2

Ashenfelter et al. (2003) and Ekelund et al. (1998) interpret the estimation gap as an indication of the level of the reserve price relative to the ex ante

predictions.

3

I consider unique sale attempts in the secondary market and exclude repeat sales.

230

|

JOURNAL OF FINANCIAL RESEARCH