Does MARRVIN have the answer?

I think we can safely say that the fine wine market as measured by the Liv-ex 1000, the broadest measure available, bottomed in August 2014 and trended with an upward bias for just over a year before taking off in November 2015.

It is interesting for chart analysts to note that in August 2014 the market bounced of an earlier bottom from 2012, and the theory goes that the length of time spent around these levels forms a base of such significance as to significantly energise the subsequent rally.

The key question, now that the above index is in all-time high ground, is: how does a fine wine investor play it from here?

One of the most important tools in the Amphora locker helps us understand and identify relative value in the market place. Clearly the concept of value is elusive in a world absent of dividends and cash flows, however we must find some way in so complex a market of figuring out what to buy and what to sell. In order to do this you have to have some idea as to the key determinants against which fine wine prices are priced, and over the years we have put an enormous amount of work into understanding this.

Traditionally people might have bought the best wine from the best vintage with the best Parker score, but the problem with this approach, as any stock market investor will confirm, is that the quality is often already reflected in the price. You don’t engineer the best investment returns by paying up the whole time; you have to find areas of the market where the quality is not reflected in the price.

According to our analysis there are some six crucial issues which govern fine wine pricing, but they don’t all have exactly the same weight. Like it or not, for example, Robert Parker’s ranking of any wine has significantly more effect on its price than that of any other critic; in fact, than that of all the other critics combined.

Cru classé status, or regional equivalent, is important, as is regional vintage quality. Production levels can’t be ignored. When you know there are only going to be 400 cases of DRC produced every year that immediate scarcity will impact the price far more than for a wine producing 20,000 cases a year. Current popularity, as expressed by the number of times a wine is sought out on Google and wine-searcher, considerations of brand power, proximity to end of drinking window, all of these elements need to be brought to the table before anyone can establish whether good relative value is on offer, or not.

At Amphora we have created an algorithm to take account of all of these variables, which we call ‘MARRVIN’. (Mathematically automated risk-adjusted relative value investment numeriser.) Happily what it does is more important than what it is called! When we view the characteristics of wines in the market through the prism of the algorithm, along with the current price, we get a very clear impression of what represents good or bad value at any given time, on a relative basis, relative, that is, not just to other vintages from the same producer, but to the whole market.

At times like this it is particularly helpful. The temptation to take profits after such a good run, and to reinvest perhaps in wines that haven’t moved, is a perfectly valid reaction, but it should be tempered by a sense of whether a winner has run its course, and whether there is any good reason to buy a laggard other than the fact that it hasn’t moved.

In its blog today Liv-ex mentions Ausone 2009, and it is quite right to do so. This is not dissimilar to the Latour anomaly we have been banging on about in the past. Ausone 2009 and Ausone 2010 both score 98+ but whilst the 2010 trades at £13,025, the 2009 is languishing at £8,800. A closer inspection reveals that the St Emilion vintage score for 2010 was 94T (where T is “still tannic, youthful and slow to mature”), and for 2009 was 93E (where E is “early maturing and accessible”).

To our way of thinking that might suggest that the 2010 is gaining an awful lot of credit for its extra regional vintage point. Parker says its drinking window is 2023 to 2073, so perhaps it is getting credit too for its longevity? By contrast the 2009 has a window of 2020 to 2060. That’s a great deal of credit for those extra 13 years.

So what happens when we put the scores through the algorithm? The final score is given in terms of algorithmic points per £, where the larger the number of points, the better the relative value. Ausone 2010 scores 1.97, whereas the 2009 scores 2.82. By any measure, if you could short the 2010 and go long the 2009, you would do precisely that.

Would you “go naked” though, as it were, and simply buy the 2009 on the strength of the discount to the 2010? In the current environment where the overall market is on solid foundations and is rising nicely we probably would. What MARRVIN tells us though, is that the best bet is the 2008. As in other areas of Bordeaux 2008 is almost an “on vintage”, indeed in St Emilion it scores a perfectly respectable 92E. Parker cites a drinking window of 2018 to 2066 so it even edges the 2009 on that front.

The 2008 only costs £4,600 though, so its algorithmic score is a princely 4.56.

Over the coming weeks we shall be exploring the market in greater detail through the eyes of the algorithm, and we are very confident it will help create significantly enhanced returns for our investing clients.

Philip Staveley is head of research at Amphora Portfolio Management. After a career in the City running emerging markets businesses for such investment banks as Merrill Lynch and Deutsche Bank he now heads up the fine wine investment research proposition at Amphora Portfolio Management