Why wine isn’t an investment
Swiss researchers Philippe Masset and Jean-Philippe Weisskopf have a new paper out claiming to demonstrate that if you add wine to a portfolio of financial assets, that decreases your risk, increases your returns, and helps you out (if you care about such things) on the skewness and kurtosis fronts as well. Leslie Gevirtz writes up the results here, and Reuters graphics supremo Silvio DaSilva has even put together some pretty charts from the paper here.
I’m very skeptical about this result, however, for four main reasons.
- The number of people genuinely investing in wine — that is, buying with an eye to selling it, rather than drinking it — is still tiny, but Masset and Weisskopf are probably right that it’s growing. “The resulting improvement in transparency and liquidity has rendered this market even more attractive for investors,” they write, but I’m not so sure: I suspect that wine’s relatively low asset-price correlation during the financial crisis was entirely a function of the fact that it wasn’t really an asset class in the first place. If and when it becomes an asset class, then correlations are bound to rise, especially in times of crisis.
- Incredibly, Masset and Weisskopf treat wine as a cost-free investment: in the world of their paper, it costs nothing to sell wine, it costs nothing to store and insure wine, and it costs nothing to buy wine, over and above the hammer price at auction. On planet earth, of course, none of these things is remotely true. So far, I have yet to see a story on wine as an investment which takes into account reasonable estimates for these costs. If and when such a study comes along, I’m pretty sure that wine will suddenly look much less attractive as an asset class.
- There’s enormous survivorship bias in the dataset. Masset and Weisskopf looked back at the wine market between 1996 and 2009, and then cherry-picked the regions which, in hindsight, turned out to have the greatest volume at auction. Then they took the wines from those regions and cherry-picked again, choosing only wines which traded at least once a year. So the 1982 Barbaresco Riserva Santo Stefano, for instance, made the cut, as it rose sharply in price between 2002 and 2009. But a neighboring Barbaresco which sold for just as much in 2002 would be ignored if it didn’t appear at auction in 2003 or thereafter. An investor in Barbaresco in 2002 would have no way of knowing which one was going to go up and which would end up impossible to sell, but by the lights of Masset and Weisskopf, the one which went up a lot was entirely representative.
- This isn’t a buy-and-hold market: you have to know exactly when to sell your wine before it becomes passé. Masset and Weisskopf try to spin this as a good thing, saying that they “discard wines that are viewed as antiques and not as wine as such” and that doing so “eliminates wines that are mostly illiquid and are traded infrequently”. Without dwelling on the metaphor of a “mostly illiquid” wine, the problem here is that Masset and Weisskopf seem to think that it’s possible for investors, en masse, to buy wine when it’s young and sell it at a profit when it’s middle-aged, but before it gets old. Of course, they can’t: who’s meant to be buying all that middle-aged wine? This strategy is all well and good so long as there aren’t enough wine investors to move the market. But if wine-as-an-investment ever takes off, that’s certain to significantly increase the supply, and therefore decrease the price, of good middle-aged wines being sold before they get too old. And when that happens, the returns from a wine-investment strategy could easily turn negative.
To get an idea of how Masset and Weisskopf think, check out this chart:
The thing to note here is the way that all the different wine regions have been rebased to 100 in 1998, as though people first decide how much money they’re going to spend on wine and then work out how much wine that will buy. Masset and Weisskopf don’t provide the actual datapoints in their paper, so I don’t know how much the average Rhone wine was going for in 1998 compared to the average Bordeaux. (And, of course, remember that we’re not actually talking about the average Rhone wine here: we’re talking only about the Rhone wines which, in hindsight, turned out to be the ones that wine lovers wanted to buy at auction. If you bought an obscure Rhone wine in 1998, which Robert Parker then started extolling in 1999, you would have made lots of money; if he didn’t, it probably wouldn’t make the index.)
But let’s say that Rhone wines were a quarter of the price of their Bordeaux counterparts in 1998, and a third of the price in 2009. No self-styled wine investor would ever allocate on an equal-investment strategy, investing say $10,000 in each: investors are always going to be overweight the most expensive Bordeaux. If you ran this same chart on the basis of average price per bottle, rather than rebasing everything to 100, it would look very different indeed — and would be much more representative of how wine investors actually view the wine market.
At its heart, this paper is an exercise in highly-theoretical number crunching, and bears little if any relation to the real world. If you want to go out and buy fine wines, that’s great. But don’t kid yourself that you’re making an “investment”. I should know: I still own a case of 1963 Croft which my grandfather bought for me when I was born. I’m not a huge port fan, and haven’t been drinking it. But I wouldn’t be at all surprised to hear that storage costs alone, since purchase, exceed its market value. Of course, if my grandfather had bought me first-growth Bordeaux instead of port, then that might not be the case. But he didn’t have the benefit of Masset and Weisskopf’s hindsight. They shouldn’t assume that he did.
Update: Masset and Weisskopf respond in the comments.