Open the Wine Bottle, HALPosted: November 30, 2015
Your correspondent is following, with interest, the advances in artificial intelligence and robotics. Now, researchers are applying a type of artificial intelligence called machine learning to the task of predicting the future prices of wines. Wine is an example of an “alternative asset,” something that people can invest in as they would in the stock market, but is, unlike the stock market, harder to price because, among other things, it depends on the taste buds of wine experts. But as it becomes more and more challenging to find good returns in traditional assets, alternatives assets such as wine, collector cars and the like become attractive.
Dr. Tristan Fletcher at University College in London is working to apply artificial intelligence to the task of predicting future wine prices. Previously, the most common approach to trying to predict something like future wine prices was to employ a tool called linear regression, where a “line of best fit” is draw through a scattering of data points. These data points might consist of things like the weather in the region, the vineyard’s history in winning medals for its wines etc. and to establish an equation that best lines-up these variables with the historical prices of that wine. For wines whose prices fluctuate over short periods but settle towards the average quickly, this method is still relatively effective. But for wines with more volatile prices over longer periods of time the AI approach creates a line that snakes and squiggles through the more unruly data points, often in a dynamic fashion, that performs better than traditional methods.
On the University College in London website, Dr, Tristan is quoted:
People have been investing in wine for hundreds of years and it’s only very recently that the way they are doing it has changed. Wine investment is becoming more accessible and is a continually growing market, primarily brokered in London: the world-centre of the wine trade. We’ve shown that price prediction algorithms akin to those routinely used by other markets can be applied to wines.
He has also formed a company, Invinio, that provides a quantitative wine asset management service that, among other things, helps customers to understand and manage their portfolio of wines as they would a portfolio of stocks and bonds.