A culture similar to wine is emerging around table water. There are water competitions, and awards for the world's best bottled waters are given based on blind tastings. Similar to wine, there are guidebooks for fine bottled waters. Many fine dining establishments now offer a wide selection of bottled waters. One Los Angeles venue even has a 45-page “water menu” with 20 different brands of bottled water selected by a “water sommelier.”
This issue of the Journal of Wine Economics opens with a hedonic price analysis of table water. In his study “Fine Water: A Hedonic Pricing Approach,” Kevin Capehart examines the prices of over 100 bottled waters (Capehart, Reference Capehart2015). He regresses the price of each bottled water on various characteristics such as source (e.g., spring, artesian aquifer, well, iceberg), carbonation (e.g., still, effervescent), pH orientation (e.g., acidic, alkaline), minerality, hardness, and virginality. His results suggest that only a small part of the price variation is determined by water-related characteristics. He concludes that “to a large extent, the premium that consumers pay for a more expensive bottled water does not seem to be a premium for its water.”
In the second paper of this issue, entitled “Machine Learning in Fine Wine Price Prediction,” Michelle Yeo, Tristan Fletcher and John Shawe-Taylor analyze machine learning techniques such as Gaussian process regression and multi-task learning to forecast fine wine prices (Yeo et al., Reference Yeo, Fletcher and Shawe-Taylor2015). Drawing on historical price data of the 100 wines included in the Liv-Ex 100 index, they show that Gaussian process regressions outperform simple ARMA and GARCH time series prediction models.
Jeffrey Bodington applies ranking and mixture models (Bodington, Reference Bodington2015a, Reference Bodington2015b) to the 1976 Judgment of Paris and the 2012 Judgment of Princeton wine tastings (Ashenfelter and Storchmann, Reference Ashenfelter and Storchmann2012). In “Testing a Mixture of Rank Preference Models on Judges' Scores in Paris and Princeton” he finds that the group preference orders implied by the mixture model are highly correlated with the orders implied by rank-sum methods. However, the mixture model satisfies choice axioms that rank-sum methods do not. It yields an estimate of the proportion of scores that appear to be assigned randomly, and it also yields a preference order based on nonrandom preferences that tasters appear to hold in common.
In “Fair Revaluation of Wine as an Investment,” Fabian Bocart and Christian Hafner analyze various methods to revaluate an existing wine portfolio (Bocart and Hafner, Reference Bocart and Hafner2015). They present a new quantitative method aimed at achieving compliance with IFRS (International Financial Reporting Standard) 13 for fair valuation. Using auction data of 26,640 lots, they apply their method to compute the current fair value of a basket of 232 different wines.
In the last paper of this issue, Guillaume Coqueret analyzes “Optimal Wine Pricing for Restaurants” with a given wine portfolio (Coqueret, Reference Coqueret2015). In his model, the optimal price is mainly determined by a rating parameter. Coqueret provides a numerical sensitivity analysis of prices to various parameters and examines a realistic large-scale example based on two wine lists with 50 bottles each.
