Over the past century, collectibles have appreciated in value, and they are gaining popularity in high-net-worth investor (HNWI) portfolios. The authors examine the long-term financial returns on high-quality collectible real assets, popularly known as “emotional assets,” and review the unique risks and different types of costs that are associated with these assets. They also compare the investment performance of emotional assets with the returns on financial assets and make some basic inferences that have implications for investment portfolio decision making.

There is aesthetic return from owning a fine painting or a famous stamp. The consumption value from owning emotional assets can be considered a form of an income stream. Given the growth in interest in such items and the increasing number of HNWIs who spend a relatively large proportion of their income in this sector, it is interesting to analyze whether this strategy is optimal from an investment perspective. The consumption value of these items as collectibles may be larger than currently perceived, and investors are often willing to give up financial returns for emotional motives.

Since the onset of the credit crisis, investors have questioned the investment merits of a range of asset classes; emotional assets are emerging as a viable and mainstream asset class. Although their returns are lower than financial asset returns, collectibles produced higher average returns (2.4–2.8% in real terms) than government bonds and T-bills during 1900–2012. But the specific risks and high costs of collectibles need to be considered before they are included in an investor’s portfolio. Smart investors should consider emotional assets to be safe havens and a long-term store of value. The analysis the authors present is useful for investment professionals in portfolio decision making.

The authors analyze geometric returns on art, stamps, and musical instruments from 1900 to 2012. A lot of historical data for these collectibles are available from UK sources because these items have been trading at large auction houses and dealers for a long time. For other collectibles, data are less widespread. The authors also briefly discuss seashells, diamonds, and wine.

They construct an art index using sale and purchase price data from Reitlinger (The Economics of Taste 1961). They then chainlink the index with five-year returns on the UK art market index found at Artprice.com. The result is an index of data through 2012. For stamps, the authors again chainlink their index values with Stanley Gibbons’s (a British stamp dealer) Great Britain 30 Rarities Index.

Because of the small number of observations available for string instruments, specifically violins, returns for this category are estimated for decade-long intervals. The authors construct an index by applying a repeat sales regression method. They then update the index with yearly returns based on the work of Graddy and Margolis (Economic Inquiry 2011).

Correlation analysis between equities, government bonds, gold, and emotional assets indicates that emotional assets can be a legitimate part of a diversified investment portfolio. The authors also examine return volatility, transaction costs, and risks associated with emotional assets. But their conclusions are not without limitations; for example, five issues are usually encountered with art indices. First, auction prices do not fully represent the art market; second, there is a high risk of survivorship bias; third, despite high liquidity in art market research data, changes over time are not clear; fourth, there is time lag between events and index publication data; and finally, it is difficult to estimate transaction costs.

The research substantiates the idea that collectibles have long attracted higher prices and are an interesting asset class from an investment perspective. But the study is a simple analysis of returns and correlations that does not consider the impact of other constraints on investment portfolios. For example, to analyze more formally, the authors could have done performance evaluation tests to determine the statistical significance of the difference in Sharpe ratios between the various mean–variance portfolios. In addition, art market indices are not investable (i.e., these are only indicative).