Wine investment returns can come from overall market trends or price increases with age. Because of the short wine price histories available, market and maturation effects are difficult to separate. Consequently, researchers often obtain dramatically different estimates of investment returns. We find that data sample bias may be the hidden cause of the disparate estimates. In wine auction data, the sample bias refers to a shift in the distribution of which wines are traded as a function of their age. Such sample bias in panel data sampled across many different wine labels can distort the estimation of price increases versus age and consequently impact the estimation of market trends. This analysis shows that segmenting the analysis such that the data panels contain wine labels with similar trading characteristics can lead to a more stable estimation.

The analysis here looks at data from Bordeaux, Italy, Australia, and California. An Age-Period-Cohort (APC) analysis is applied to data panels from each region. Then the data in each region is segmented by a measure of popularity in order to reduce sampling bias. Data thus segmented is then re-analyzed to demonstrate the difference in estimating price appreciation lifecycles and market trends.

In an attempt to expand the understanding of auction-price dynamics for fine wines, an age-period-cohort (APC) algorithm is applied to a database of 1.5 million auction results to quantify key drivers of these price dynamics. APC algorithms are designed to separate price appreciation with the age of the wine from overall wine-market conditions as well as to adjust for the unique value of specific vintages. In this context, the APC modeling provides a kind of Hedonic modeling, with specific controls regarding specification errors.

The analysis was segmented by Château Lafite Rothschild, Bordeaux excluding Lafite, and Burgundy so that we could test specific events related to Château Lafite Rothschild. The results show price dynamics versus the ages of the wines and allow for the measurement of long-term price-appreciation potential. Environment functions versus auction dates quantify the “Lafite Bubble” and suggest past correlation to Chinese stock-market indices. An analysis of wine ratings versus price quantifies their nonlinear relationship. An analysis across nine auction houses shows a significant price spread for similar wines. (JEL Classifications: C23, D44, G11, G12, Q11)

In the literature, there is no standard approach for estimating the return to wine or testing for a portfolio risk diversification benefit from holding wine. Using auction data for Australian wine, we show that the estimation method has a material impact on the estimated wine return distribution and that the type of diversification benefit test used influences whether or not wine is found to provide a portfolio risk diversification benefit. Our results indicated that a simple modification to the hedonic model, which we call a pooled model, is an appropriate method for estimating the return to infrequently traded heterogeneous assets such as wine. Across the various approaches to testing for a risk diversification benefit, we find direct estimation of the efficient frontier with bootstrapped confidence intervals to be the most transparent and comprehensive method of illustrating wine's potential for lowering portfolio risk. (JEL Classifications: G11, C61, C13)