May 14, 2008, by  Karl Storchmann (Journal of Wine Economics)

According to a report in the online magazine TheStreet.Com, 90% of high-end buyers of homes in U.S. major cities consider a wine cellar a very important factor when purchasing their house. (http://www.winetrend.com/press/wt_thestreet.pdf)

The report goes on “Imagine walking your date into a cellar made entirely of limestone and hand forged wrought iron wine racks to look up through the windows of the vaulted ceiling into the water fountain above it. It can be done – for $800,000.”

Why do people have wine cellars? Certainly for many different reason.  But if you have a wine cellar for other reasons than showing it to your date every day you want to read the latest report by Gary Thompson and Steve Mutkoski of Cornell University's Center for Hospitality Research. It is entitled “Optimizing a Personal Wine Cellar.”

Essentially, the authors devised an analytical model that optimizes the utility derived from a wine cellar. This ‘Personal Wine Cellar Optimizer (PWCO)’ goes well beyond any conventional software packages that are only inventory tracking devices.

The PWCO is complex and considers many variables. The size of the wine cellar, the average wine consumption (the authors assume in their examples one bottle per day), the envisioned lifetime of the cellar, the price of wines bought and many more. I addition, the authors take into account that, up to a certain limit, most wines improve when they get older. The quality gain, however, declines the older the more mature the wine is. All these factors are combined in an Excel sheet and the outcome suggests an optimal wine inventory and drinking pattern.





The Personal Wine Cellar Optimizer (PWCO) takes the existing cellar composition as its starting point. The PWCO then builds a purchase and consumption plan that observes all the hard constraints imposed by the budget and cellar capacity. Wines are selected for purchase randomly, but purchases are biased towards better value wines (i.e., higher quality and lower costs). Wine consumption is scheduled in a way that cellar capacity is not violated, but with a preference for drinking the wine at or near its peak, as well as spreading the consumption over time.

Of course, many applications outside of private cellars are possible. Gary Thompson and Steve Mutkoski are already working on restaurant versions.

The report is downloadable at http://www.hotelschool.cornell.edu/research/chr/pubs/reports/abstract-14724.html

You have to log in but it is free (and worthwhile)