Lesson learnt

One of the most challenging parts of being an investment professional is the lack of a fixed way to manage money. But even if there was one, it would not work because not everyone can beat the market. In aggregate, we are the market. So, in the absence of a blueprint for success, we all have to make it up as we go along. For many, that is a labour of love; a career-long hunt for the holy grail. Indeed, the best thing about working in investments is the fact that there is no prescribed way to manage money.

That is certainly how it has been for me. I have voraciously assimilated anecdotes, facts, and theories wherever I could source them, some of which were so fundamental that I wished I had discovered them sooner. Of these, my biggest regret is that I did not find quantitative analysis until 2011. That said, a lack of data and limited computing power would have been serious obstacles as recently as the 1990’s. Even today, few data series go beyond the 1970’s, and the situation is worse outside the United States.

Our challenge is to learn the lessons of the past to give us a fighting chance of navigating what is a very uncertain future. Although every investment manager does that to a greater or lesser extent, quants are just that bit more thorough. Furthermore, the trading models that they build help them to navigate the minefield of behavioral finance.

Early in my career, I placed heavy emphasis on valuation (buying assets that are ‘cheap’) and contrarian-type strategies (selling into strength and buying into weakness). Quantitative analysis revealed the error of my ways.

Everyone likes a bargain (buying an asset below fair value). However, not only is value hard to define, but even when you can, the mean reversion that drives excess returns will usually play out over years rather than months. Nevertheless, investors like valuation because it is more tangible than fundamentals and more cerebral than simple trend following. But the reality is that markets tend to regularly overshoot the runway, often by big margins. This means that even the best valuation metrics, like the Shiller (cyclically adjusted) price-earnings ratio only correlates well with future returns over a decade (or more), a length of time that is well beyond the patience and pain threshold of most investors. That is why I pay little heed to valuation, no matter how scary the numbers. I would add that markets rarely (if ever) fall because they are expensive. Rather, they fall because events on the ground remind us that every trend has a cycle. Incidentally, that works in both directions.

I used to be very sniffy about momentum (or trend following). Buying something just because it was going up seemed overly simplistic and went against my instinct to ‘buy low, sell high.’ However, back-testing soon revealed that buying assets that had fallen a lot was not a profitable thing to do. Quite the opposite. This tendency for markets to trend flies in the face of the efficient market hypothesis, which holds that all available information is reflected in the current level of markets. It should not be possible to make money in this way. And yet, there it is. Furthermore, momentum has been detected across multiples geographies and asset classes. Its main weakness, by definition, is that it will not get you out at highs or in at lows. On the other hand, it is a useful stop-loss that can give you a heads up when things are not going according to plan, either because ‘this time it is different’ or (more likely) something key has been left out of the investment process. In an ideal world, I would cover all the bases, thereby removing the need for momentum, but for now, it continues to feature in everything I do.

Looking to the future, I would like to dedicate time to understanding market cycles. Investment people will often talk about patterns that play out repeatedly. Perhaps the most famous is the seasonal cycle for equities. According to the old saying we should “sell in May and go away. Don’t come back until St Ledger’s Day.” Back-testing reveals that October is also usually a bad month, which means that St Ledger’s Day (which is in September) might be too soon to buy equities. Other patterns include the four-year (US Presidential election) cycle and the decennial cycle. Why do markets behave in this way (political shenanigans around elections?), and how dependable are they? The voyage of discovery continues.