The Holy Grail and Picking Stocks
People keep asking us, why is it that we don’t take views and pick stocks, for instance? Surely, there must be some value in analyzing historical data and seeing trends or cycles. There is, and we do use quantitative models in order to optimize our trading and risk management systems. But as far as making predictions on the direction of future asset prices is concerned, well let us explore the subject matter in a bit more detail and everyone can make up their own mind.
Let us start with the universe: there are more than 40,000 individual stocks to choose from and we can buy them across hundreds of different exchanges and in dozens of different currencies and time zones. There are many different sectors, ranging from healthcare to autos, natural resources or technology. There are many inherent business models, tax, and legal systems to understand and take into account. It’s a lot of information to consider and ultimately numbers to analyse. Lots of them: fundamental data on balance sheets, margins and income, technical factors based on the historical performance of each particular stock, both in absolute and relative terms and a vast amount of macroeconomic data which can have a direct or indirect impact on each or a particular sector of stocks. There are also qualitative factors to contemplate, such as the quality of the management and governance of the company or the behaviour of the employees, clients, and service providers.
How do you model all of that to pick one stock over another? Well, we have quantitative analysts who can crunch a lot of numbers on superfast computers using neural networks or space time continuum models to advance our ability to predict the likely movement of individual stocks. We can also use an army of financial analysts to calculate (stock) valuations and we can literally spend billions on Nobel laureate winning economists, field medal winners or other masters of the known world.
But even the biggest and greatest team of superstars, using the best technology, will never solve one simple problem: no one can predict the future. It is simply impossible to analyse the trillions and trillions of bits of data in real time, maybe in several dimensions and as part of a very random and ever-expanding infinite universe. Moreover, by picking one stock over the other, we inherently don’t get the exposure to the greater market and other stocks, which can have a huge (opportunity) cost. Say, you really like the utilities sector, because they have stable earnings and a sound business model. By buying these very stocks, we are also not buying the Apple’s, Google’s or biotech stocks that have outperformed by hundreds of percent over the last decades.
Which is why it is so difficult to beat the market and the reason we don’t take views. But good luck to those picking stocks.