On the night of November 8 2016, the now-infamous Sam Bankman-Fried (SBF, as he is commonly known) was employed at the legendary trading firm Jane Street. He had built a system that was designed to access state-level election data minutes ahead of CNN and others, allowing the firm to call the election before anyone else.
It came as a complete surprise when they realised that Donald Trump would beat Hillary Clinton. Predicting a market crash, Bankman-Fried and his team shorted the market for several billion dollars. As the night wore on and the news of Trump’s victory spread, the market rallied, losing the firm $300m. “It went from single most profitable to single worst trade in Jane Street history,” Bankman-Fried told his biographer, Michael Lewis.
Forecasting macroeconomic trends is a key component for many adherents of fundamental investment analysis. They aim to predict a region’s or country’s growth more accurately than the market. If an asset manager believes that Japan, say, will grow faster than consensus forecasts, she will invest in Japanese equity in the expectation that these companies will capture their share of the unexpected growth. This same process is not limited to the geopolitical sphere but is used to make investment decisions between sectors and between stocks themselves.
The great uncertainty, and the ability by which asset managers attempt to distinguish themselves, is in the accuracy of their forecasts. Do they really have the skill, knowledge and access to information that will allow them to generate superior predictions?
But what if an analyst had “perfect information”, perfect clairvoyance or providence? Surely he would be able to consistently outperform to a tremendous degree. Going back 50 years, such a person would undoubtedly identify technology and tobacco as the industries that would outperform and underperform, respectively, over the next half century.
While innovation in the world of atoms may have stagnated — one used to be able to make a transatlantic flight in three hours — innovation in the world of bits has progressed to a degree almost impossible to imagine.
At the same time, following the landmark report from the US surgeon-general in 1964, tobacco companies have faced a perfect storm. Marketing and packaging restrictions, heavily funded anti-smoking campaigns, restrictions on smoking in public places, extreme increases in sin taxes and greater awareness of the risks involved have led to a drastic drop in the percentage of adult smokers, certainly in the Western world. In the US, adult smoking rates dropped from a record high of more than 42% to about 11%.
In 1985, the top tech stock by market share was IBM and the top tobacco stock was Philip Morris. One hundred dollars invested in IBM at the beginning of January of that year would be worth $2,380 now, representing a return of 24 times. Not bad?
One hundred dollars invested in Philip Morris at the beginning of 1985 would now be worth about $75,000, representing a return of 750 times the original investment. Going back a further decade to 1975 and your Philip Morris shares are worth nearly $1.5m. These two companies are not cherry-picked examples but are representative of the return from these sectors as a whole over the period.
In retrospect, there are many reasons that can explain this. A decrease in the proportion of smokers is not the same as a decrease in the absolute number of smokers; the developing world — which could not previously afford cigarettes — has picked up much of the demand while sin taxes were simply passed on to consumers, and there are surely dozens of other contributors.
The message, though, identical to the experience of Bankman-Fried and Jane Street, is that even with perfect knowledge of the future, our best analysis can still lead to incorrect decisions.
• Freidus, a consulting actuary, is cofounder of Five2two Analytics.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.