Visiting professor gives fine wine market a data-driven shakeup

A bartender looks at bottles in a wine cellar

For hundreds of years, a tiny group of négociants or wine brokers have determined the price that distributors, importers, and eventually consumers will pay for France’s top wines. These prices are based on the barrel scores of elite tasters, along with the brokers’ own expertise—and a generous splash of guesswork about the market.

That tradition-bound system is getting a data-driven shakeup this month with the debut of a new pricing algorithm on London’s Liv-ex fine wine market.

Visiting Prof. Burak Kazaz

The algorithm was developed by Burak Kazaz, a Berkeley Haas visiting professor (and the Steven R. Becker professor at Syracuse University’s Whitman School of Management), and Hakan Hekimoğlu of Rensselaer Polytechnic Institute. They established “realistic prices” for the 2018 en primeur (or wine futures) campaign—wines from last year’s vintage that are aging in the barrel and will hit shops and restaurants next year.

“This is the most important progression in making a transparent market for wine futures since the négociant system was established more than 300 years ago,” says Kazaz, who pioneered the field of wine analytics. “The realistic pricing will tell buyers and consumers whether a wine is underpriced or overpriced, leading to more effective and transparent purchase decisions. It will also tell winemakers how they can determine their own selling price to négociants.”

The algorithm incorporates temperatures, precipitation, market conditions based on the Liv-ex 100 index of top wines, and price trends based on barrel scores for wines from Bordeaux’s leading chateaus determined by tasting experts Lisa Perrotti-Brown (of The Wine Advocate) and James Suckling (formerly of The Wine Spectator). These influential tasters sample young wines a year after they are barreled, and one year before they are released to the market, scoring them on a 100-point scale.

The “realistic pricing” algorithm allows distributors and consumers to translate a Bordeaux’s score into a dollar amount, getting a clear idea of whether they’re getting a good deal or overpaying. For the 2018 vintage, the algorithm predicts a 3% price increase per additional point compared with the 2017 vintage.

The algorithmic pricing system is expected to shake up the wine futures market, but also to make it less risky, since investors will have a data-driven price for the first time and winemakers will know what price to set for their young wines. Kazaz and Hekimoğlu’s study was recently featured in Robert Parker’s influential Wine Advocate.

While the wine futures market doesn’t exist in the U.S., where individual wineries tend to set their prices, Kazaz is exploring the idea of bringing data driven-pricing to the U.S. market.

Kazaz teaches operations and supply chain management in the Berkeley Haas MBA program.