Collectibles as an investment: what art, wine, watches and classic cars really return
Collectibles as an investment (art, wine, watches, classic cars) promise rising values and the pleasure of ownership. Their prices are set at auctions and by dealers, with high fees, and they are only measured when something is sold. Here you will see how price indices are built, what studies find about returns and correlations, and why a price increase of 4.8% a year can shrink to less than 1% after the buyer's premium, commission, insurance and storage.
What sets collectibles apart from securities
A painting, a case of Bordeaux or a classic car pays neither dividends nor interest. The financial return comes solely from the change in price, minus high costs. In return, ownership brings pleasure, prestige or the joy of driving. This emotional dividend is part of the return that no account statement shows. Mandel (2009) uses it to explain why the financial returns on art are low: if you value the benefit of ownership, you accept less appreciation; the risk premium can be small or even negative.
Each piece is unique, or nearly so. There is no continuous price, buyers and sellers rarely meet, and authenticity, condition and ownership history (provenance) help determine the price.
Markets, prices and costs
Art
According to Art Basel and UBS (2026), the global art market had sales of US$59.6 billion in 2025, 4% more than in the previous year: 34.8 billion through dealers and galleries and 20.7 billion at public auctions, across around 41.5 million transactions. At auction, the buyer pays a buyer's premium on top of the hammer price. At the major international houses, in spring 2026 it was 27 to 29% of the hammer price in the lowest price band and 15% above US$8 million (Grossman LLP 2026). The seller also pays a commission, the amount of which depends on the house and the object.
Wine
Trading focuses mainly on fine wines suitable for aging, such as the Bordeaux first growths (premiers crus), through auctions, merchants and trading platforms. Wine has to be properly stored and insured; provenance and storage history affect the price. Dimson et al. (2015) also show how age plays a part: young, maturing wines from good vintages delivered the highest financial returns; once mature, famous châteaux provide their owners with growing non-financial benefits.
Watches
The secondary market for luxury watches runs through auctions, dealers and online platforms. According to Ulmer et al. (2024), the factors that determine value include authenticity, rarity, type of watch, brand and famous previous owners. Papers, original condition and an authenticity check therefore count too.
Classic cars
Classic cars are sold at auctions, through dealers and privately. Laurs and Renneboog (2019) analyzed 29,000 auction sales from 1998 to 2017, from veteran cars built between 1888 and 1907 to modern classics from 1975 to 1990. Running costs include garaging, insurance, maintenance and restoration.
How returns are measured: repeat sales, hedonics, selection
Prices exist only when something is sold. Two methods turn them into an index:
- Repeat-sales index: it uses only items that have been sold at least twice and compares the price of the same item over time. The method comes from the property market (Bailey et al. 1963); Mei and Moses (2002) used it to build an art index going back to 1875.
- Hedonic price index: it explains every price from characteristics such as artist, size, technique, brand or year of manufacture, and measures what remains afterward as the time effect. This way it can use all sales (Renneboog and Spaenjers 2013; Laurs and Renneboog 2019).
ln(P_i,s / P_i,b) = Σₜ βₜ × D_i,t + ε_i- P_i,b, P_i,s
- purchase and sale price of item i
- D_i,t
- 1 for each period between purchase and sale, otherwise 0
- βₜ
- estimated log return of the index in period t
- ε_i
- deviation of the individual item
Both methods see only items that were sold. That creates a bias: what has risen in value is more likely to come up for auction, while items that would not sell stay in storage. Korteweg et al. (2016) corrected for this for 20,538 paintings that were auctioned more than once between 1972 and 2010: the index return falls from 10% to 6.5% a year, and the Sharpe ratio from 0.24 to 0.04. Their conclusion: in purely financial terms, a passive investment in a painting index is not a viable strategy. On top of this comes survivorship bias: indices show what is still in demand today.
What research shows about returns and correlation
| Asset class and study | Period | Return per year | Comment |
|---|---|---|---|
| Art, Renneboog and Spaenjers (2013) | 1957–2007 | 3.97% real in US dollars | similar to corporate bonds, much more risk |
| Art, Korteweg et al. (2016) | 1972–2010 | 6.5% instead of 10% after selection correction | Sharpe ratio 0.04 instead of 0.24 |
| Wine, Dimson et al. (2015) | 1900–2012 | 4.1% real after storage costs | more than bonds, art and stamps; positively correlated with equities |
| Stamps, Dimson and Spaenjers (2011) | 1900–2008 | 7.0% nominal, 2.9% real | more than bonds, less than equities, volatility almost as high as equities |
| Classic cars, Laurs and Renneboog (2019) | 1998–2017 | 3.37% real, 5.63% nominal, before costs | risk-adjusted ahead of equities and art, behind bonds and gold |
| Watches, Ulmer et al. (2024) | 1999–2020 | 5.5% real, 7.7% nominal in US dollars | Sharpe ratio beaten only by gold; positively correlated with inflation and equities |
In their survey, Burton and Jacobsen (1999) found that most collectibles deliver positive real returns, but with more risk and less return than equities. Mei and Moses (2002) placed art ahead of fixed-income investments but well behind US equities, with lower correlation to other assets than in earlier studies; expensive masterpieces did worse than the index. Goetzmann (1993) had already traced the link with financial markets over three centuries. Goetzmann et al. (2011) show that stock returns moved art prices significantly over two centuries and that top incomes and art prices are linked in the long run. That argues against the hope that collectibles stay stable when wealth shrinks.
For watches and classic cars, there are only a few studies, covering short periods that end in 2020 at the latest. Our assessment: the figures depend heavily on which brands and models are in the index, and they can hardly be extrapolated into the future.
Worked example: a work of art over ten years
The internal rate of return is the interest rate at which the purchase price, running costs and proceeds together have a present value of zero.
For advanced readers: costs, holding period and smoothing
High transaction costs are spread over the holding period. If buying and selling together cost a share k of the price, the return per year falls by:
Reduction per year = 1 − (1 − k)^(1/n)- k
- costs of buying and selling as a share of the price
- n
- holding period in years
With k = 28%, that is 6.36 percentage points after five years, 3.23 after ten years and 1.63 after 20 years. The return after costs in the example follows exactly from this: 1.0481 × (1 − 0.0323) gives 1.43% a year. Collectibles are therefore, if at all, an investment for very long periods.
Selection bias can be modeled: Korteweg et al. (2016) jointly estimate how prices develop and how likely a sale is. They stress that this also affects other illiquid assets that trade only at times chosen by their owners. Caution is also needed with volatility and correlation: annual or irregular index points smooth out fluctuations, much as appraisals do for real estate (smoothing). A low measured correlation with equities therefore does not mean that collectibles provide support in a crisis.
Fractions, tokens and platforms
Newer offerings promise shares in a painting, a classic car or a wine collection from small amounts, often via a platform or as a digital token. Different rules apply depending on the structure. Germany's financial regulator BaFin, for example, warns about such offerings outside the regulated capital market (offerings that need no authorization, including direct investments in tangible assets): there is no systematic product supervision, often a considerable risk up to total loss, and you can usually sell only to a limited extent because there is no regulated trading. Even a prospectus approved by the regulator is neither a seal of quality nor a license to do business (BaFin 2026).
So find out who legally owns the item, who holds and insures it, what happens if the platform becomes insolvent, what fees apply and how you can get out again. The platform's costs come on top of the costs of the item itself.
Limits and common mistakes
- The index return is not your return: indices usually include neither the buyer's premium nor commission, insurance or storage.
- Only sales are visible: items that do not sell are missing, so the measured return is too high.
- One item is not an index: individual works vary widely; in studies, expensive masterpieces even did worse.
- Liquidity: a sale needs the right moment and the right buyer; in weak markets, items go unsold.
- Authenticity and condition: without provenance, papers and expert opinions, a high price is quickly lost.
- Fashions change: what is in demand today may be forgotten in 20 years.
This article is not investment, tax or legal advice and does not recommend any objects or platforms.
How Portfolio Lab calculates it
Portfolio Lab has no tool and no return series for collectibles: there are no freely available monthly series for these markets, and auction indices would hardly be comparable with price series for the reasons given above. The taxonomy of the Look-through analysis does include a category “Collectibles & alternative real assets” (art, classic cars, wine & whisky, watches, jewelry & gemstones, coins and stamps) with the illiquidity premium as its risk driver, and you can assign a position to it manually there. However, the app can only calculate with securities that have a price.
If you want to include collectibles in your planning, treat them as described in the article total wealth: assume a cautious sale value after costs, and do not count on it providing support in an equity crisis. For a comparison with gold, equities and bonds, the Backtest is suitable, with the difference that its fund series already include the ongoing costs, whereas the returns from the studies above mostly do not include transaction costs.
Frequently asked questions
Is art a good investment?
In purely financial terms, not really: studies measure real appreciation for art of around 4% a year with high risk, and after correcting for selection bias the Sharpe ratio falls almost to zero. On top of that come the buyer's premium, commission, insurance and storage. The rest of the return comes, if at all, from the pleasure of ownership.
What returns do wine, watches and classic cars deliver?
Studies found 4.1% a year real after storage costs for top Bordeaux wines from 1900 to 2012, 3.37% real before costs for classic cars from 1998 to 2017 and 5.5% real for collectible watches from 1999 to 2020. The periods are short, and the indices contain only items that were sold and no transaction costs.
How high is the buyer's premium at auctions?
At major international auction houses, in spring 2026 it was 27 to 29% of the hammer price in the lowest price band, falling to 15% at very high prices. The seller also pays a commission. Smaller houses have their own rates.
What is the emotional dividend?
The benefit that ownership itself provides: enjoying the painting, the joy of driving, prestige. It is part of the return but does not appear in any return figure. Models such as Mandel's (2009) use it to explain why the financial returns on art are low.
Literature
- Bailey, Martin J.; Muth, Richard F.; Nourse, Hugh O. (1963): A Regression Method for Real Estate Price Index Construction. Journal of the American Statistical Association 58 (304), pp. 933–942. doi:10.1080/01621459.1963.10480679
- Bundesanstalt für Finanzdienstleistungsaufsicht (2026): Grauer Kapitalmarkt (Verbraucherinformation). BaFin, ohne Datum; abgerufen am 25.10.2026. Source
- Burton, Benjamin J.; Jacobsen, Joyce P. (1999): Measuring Returns on Investments in Collectibles. Journal of Economic Perspectives 13 (4), pp. 193–212. doi:10.1257/jep.13.4.193
- Dimson, Elroy; Spaenjers, Christophe (2011): Ex post: The investment performance of collectible stamps. Journal of Financial Economics 100 (2), pp. 443–458. doi:10.1016/j.jfineco.2010.12.005
- Dimson, Elroy; Rousseau, Peter L.; Spaenjers, Christophe (2015): The price of wine. Journal of Financial Economics 118 (2), pp. 431–449. doi:10.1016/j.jfineco.2015.08.005
- Goetzmann, William N. (1993): Accounting for Taste: Art and the Financial Markets over Three Centuries. American Economic Review 83 (5), pp. 1370–1376. Source
- Goetzmann, William N.; Renneboog, Luc; Spaenjers, Christophe (2011): Art and Money. American Economic Review 101 (3), pp. 222–226. doi:10.1257/aer.101.3.222
- Grossman LLP (2026): Spring 2026 Auction Season: Examining Recent Changes in Buyer's Premiums. Grossman LLP (Kanzlei für Kunstrecht), 14.5.2026; abgerufen am 25.10.2026. Source
- Korteweg, Arthur; Kräussl, Roman; Verwijmeren, Patrick (2016): Does it Pay to Invest in Art? A Selection-Corrected Returns Perspective. The Review of Financial Studies 29 (4), pp. 1007–1038. doi:10.1093/rfs/hhv062
- Laurs, Dries; Renneboog, Luc (2019): My kingdom for a horse (or a classic car). Journal of International Financial Markets, Institutions and Money 58, pp. 184–207. doi:10.1016/j.intfin.2018.10.002
- Mandel, Benjamin R. (2009): Art as an Investment and Conspicuous Consumption Good. American Economic Review 99 (4), pp. 1653–1663. doi:10.1257/aer.99.4.1653
- McAndrew, Clare (2026): The Art Basel and UBS Global Art Market Report 2026. Art Basel und UBS (Arts Economics); Zusammenfassung auf artbasel.com, abgerufen am 25.10.2026. Source
- Mei, Jianping; Moses, Michael (2002): Art as an Investment and the Underperformance of Masterpieces. American Economic Review 92 (5), pp. 1656–1668. doi:10.1257/000282802762024719
- Renneboog, Luc; Spaenjers, Christophe (2013): Buying Beauty: On Prices and Returns in the Art Market. Management Science 59 (1), pp. 36–53. doi:10.1287/mnsc.1120.1580
- Ulmer, Simon; Schmid, Patrick; Widenhorn, Andreas (2024): Luxury watches: a viable alternative investment or mere speculative trend? An analysis of two decades before the pandemic. The Journal of Investment Strategies 13 (1), pp. 39–64. doi:10.21314/jois.2024.006
This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.