At Ardevora, we believe successful stock picking requires an understanding of how three groups of people interact: company managers, financial analysts and investors. Each group is potentially subject to bias, and the biases affecting each group are different.
Investors are prone to overreaction, placing too much weight on a narrow range of information and arriving at binary conclusions. We seek to find signs investors are focusing on old irrelevant traumas. This takes us to October’s market panic. From our perspective, the behaviour of only one of the three blocks of people we try to track has changed: investors.
When you have lived through a lot of ups and downs of stock markets, you tend to trawl your memory for reference points when something extraordinary happens. It is called the Representative Heuristic. It is a good way of dealing with complexity if you have a big bank of relevant experience, but it can also be a profound source of bias if your experience is shallow – or you rely on a irrelevant memory.
Whether the Representative Heuristic is a source of bias or help is essentially a debate about the usefulness of history: does history repeat itself or do we just try to make sense of inherently unpredictable current events by over-fitting them to the past, regardless of logic?
We believe in the usefulness of history, as long as one approaches it carefully, logically, and you do not get dazzled by the Affect Heuristic. The Affect Heuristic is the tendency to focus on similarity of outcome, rather than similarity of causes, as a reference.
With all of this in mind, what is going on now feels a lot like 1987, so let us explore this Representative Heuristic a little and see if it is or is not just a bad case of Affect Bias.
The end of the world?
What happened in 1987? The first episode of the Simpsons aired, and construction started on the Channel Tunnel. We also saw the first criminal was convicted on DNA evidence and the first drug for AIDS was discovered – with DNA research now spawning an explosion of R&D in the drug industry. A US frigate was attacked by Iraqi missiles, which reminds us how long the Middle East has been a source of angst. Finally, stock markets crashed around the world.
My memory of the crash was it seemed to come out of nowhere. Afterwards, metaphorical fingers were pointed at a small rise in German rates and the role of automated trading, called portfolio insurance. No obvious economic trauma was building, just an overheating housing market, a slight pick-up in inflation and gentle monetary tightening in a few places. This is why my Representative Heuristic draws me to 1987.
Back then, I did not think in terms of investor, analyst and management behaviour. I was still young and naïve – believing in dividend discount models, accurate forecasts, macroeconomics and rational man. Now I know better.
Herding and overconfidence
Most problems in markets, in my experience, stem from the herding of management behaviour. Company managers get over-confident and reckless in the pursuit of growth, while assuming recessions have been banished. They overpromise, under-deliver and mislead analysts and investors. However, sometimes investors operate in a bubble.
In 1987, investors drove the stock market up on the same ‘news’ they drove it down on. It was blamed on a belief stock markets could become a largely riskless enterprise through use of portfolio insurance. The result was a strangely strong market in the first three quarters of the year, followed by a violent late correction.
As stocks fell, a feedback loop of anxiety kicked in: investors worried whether the falling market was either signalling an oncoming recession, which no-one had seen coming, or would trigger a recession through negative wealth effects. Such is the circularity of ad-hoc causation. A year later, stock markets crawled back to the highs, before climbing for another three years. On a long-term chart of stock markets, 1987 looks like a blip. But it did not feel like it at the time.
October echoes 1987 crash
A lot seems familiar now. No-one is concerned about an imminent recession. In fact, most worries are focused on the potential for inflation to accelerate, hence some tentative rate rises – just like 1987. Company managers do not seem to be over-exuberant, or overconfident. Most companies are not delivering surprisingly awful results triggering large forecast misses from analysts.
Investment robots also seem to lurk. Portfolio insurance trading has been replaced by a market dominated by ‘smart’ passives in thematically-driven ETFs, smart beta and risk parity strategies. Pockets of stocks burst into accelerating momentum and then out-of-nowhere short sharp selloffs.
We do not subscribe to the view history endlessly repeats itself. We do believe it can repeat, but in different clothes. Our instincts tell us we are in a rough, violent period for stock markets. It seems likely we will have some more terrible feeling days or weeks, but we would expect to feel a lot better in six months’ time. Investors will have been scared, scarred and we can get back to a healthy environment of scepticism for everyone.
Jeremy Lang, partner and co-founder of Ardevora Asset Management