The wine world was taken by surprise in 1976: Steven Spurrier, the British wine merchant, had a jury of experts taste a series of French and American wines blind. To everyone’s surprise, the United States won gold twice in the categories of Chardonnay (California vs. Burgundy) and Cabarnet Sauvignon (California vs. Bordeaux).
Since then, the United States has been a notorious player in the world of wine. Over the years, however, wine has not only grown into an important means of consumption, it is also gaining an increasingly important place in an investment portfolio.
From an academic point of view, wine is also very interesting for testing theories. For example, there is disagreement about the driving force behind momentum - the finding that stocks that performed well over the past year will do significantly better over the next month than stocks that performed poorly. One set of academics argues that this has to do with risk: stocks with positive momentum are stocks that are correlated with each other, and correlation is something you do not want (when “the stock market” goes down, your whole portfolio goes down together).
Momentum is, at least this stream of researchers, a compensation for the risk taken. Another stream states that it is a behavioural characteristic: people under-react to bad news and over-react to good news about shares that are firmly in the green.
That is why we look for other investments where the characteristics change less quickly: wine. When we look at wine not as a consumer product, but as an investment, we can better try to understand which of the two camps is right: risk or behaviour. How do we proceed? We will look at the price offered at auctions for bottles of wine, specifically Acker - the third most important auctioneer after Sotheby’s and Christies. Which wines are we looking at specifically? All (normal) 75cl bottles: over 300,000 different bids. We proceed in the same way as with a normal momentum strategy:
- We group all wines by domain, such as Chateau Montalena - the winner of the 1976 competition.
- We calculate the return (per domain) from 11 months to 1 month ago.
- We sort all the domains by the calculated yield and look at the best and worst 20%.
- And then we look at the difference in returns for the next month, quarter and year.
Shares
In the case of shares, the result is clear: the best performing shares still have a higher return in the month to 1 year than the worst performing ones. With wine, we see just the opposite: the best performing stocks are the ones that perform worst in the following months (and vice versa, the worst performing stocks are the ones that perform best up to one year after our analysis). This difference is -1.5 percent the following month to -3.5 percent after 1 year. This leads us to believe that behaviour plays a more important role than what other academics dare to claim and that therefore “momentum” is not a compensation for risk, but is more driven by luck.
Which wine domain experienced the biggest decline? Bryant Family, an American producer from Napa Valley, California. This domain consistently had the best performance the year before we applied the strategy (37% of the time it was among the best-performing wine domains).
So should we invest in Bryant Family wine? Unfortunately, it is too good to be true. The difference between Bryant Family and Burgundy’s Dujac, the worst performing by cumulative returns from the previous year, is more than 5 per cent - in France’s favour. If we were to put our strategy into practice, France clearly wins from the US this time. So the world has clearly changed since 1976.
Prof. Gertjan Verdickt is a lecturer at KU Leuven and an Investment Officer Belgium knowledge expert.