Updated: May 17, 2022
I was presenting at a multi-asset investing conference recently and during the Q&A was asked, “Will you be replaced by an ‘algo’?” My answer was “No”, but there was no time to elaborate.
"I did not need an algorithm to tell me that the -1.5% yield on inflation-linked gilts makes them horribly expensive"
In the financial world, an algorithm is a series of unambiguous rules that, when applied via a computer to market data, generate decisions to buy or sell tradable securities. They can be ‘trading algorithms’ that are designed to help investors buy or sell at the best prices on a particular day, or ‘investing algorithms’ that seek to identify good or bad investments over a particular timeframe.
In many respects, computer-driven algorithms have been producing fabulous investment returns for years. Perhaps the most famous ‘quant fund’ is Renaissance Technologies’ Medallion Fund. It was launched in 1988 and generated annualised returns of 35% over a 20-year time period, prior to being taken private by Renaissance employees in 2005.
Unfortunately, the type of algorithms that Renaissance used are not available to the man on the street. Renaissance was set up by cold war codebreaker Jim Simons and staffed by a bunch of other maths and computer geniuses. The company also had a very big and powerful computer.
Simons’ first recruit was Leonard Baum, co-author of the Baum-Welch algorithm. This algorithm was designed to determine unknown parameters in complex data sets, hence its applicability to financial markets.
Simpler investing algorithms might be those that are based on some interpretation of Graham and Dodd’s value investing framework. For example, an algorithm created by one Peter Poon has just two rules. The first rule selects no more than 15 stocks that satisfy the criteria: P/E less than 12 times, P/B less than 2 times, ROE greater than 15% and market cap greater than $100m. The second rules states that one should hold the selected stocks for a year, then start again.
A promising algorithm is one that when back tested produced good returns. But a successful algorithm is one that works going forward. The two do not necessarily go together. It will always be possible to find patterns in a data set. The danger is in mistaking pattern for randomness, like seeing the Virgin Mary in a slice of toast.
Peter Poon’s algorithm when back tested produced great results. It also makes sense that it should, given that it is based on widely accepted measures of stock valuation and company performance – an algorithm that was based on buying companies that had the colour orange in their logo, a strategy that may have back tested successfully, might not fare so well going forward.
But even Peter Poon’s algorithm might not satisfy the third criteria: persistence. Similar to the ‘observer effect’ in physics, in which observing a phenomenon changes the phenomenon, overuse of an investing algorithm can impact its effectiveness.
One thing that distinguishes humans from computers is imagination. Imagination requires an appreciation of the future. Computers can only at best appreciate the future in one dimension – they can predict, for example, the airflow over an airplane wing.
Humans, on the other hand, have the capacity to predict the future on multiple scales. According to Resolution Group’s chief economist Duncan Weldon: “Machines are less likely to be able to replicate creativity, social interaction, and the need for human-to-human contact anytime soon, and a surprising number of jobs involve these attributes”.
One can imagine algorithms of the future designed to make investment decisions also coming into conflict with each other, with a plethora of inputs telling one to buy and the other to sell.
The simple point I made at the conference was that I did not need an algorithm to tell me that the -1.5% yield on long-term inflation-linked gilts makes them horribly expensive.
It is straightforward observations like this that may give humans the edge over computers and keep them in the game.
Published in What Investment
The views expressed in this communication are those of Peter Elston at the time of writing and are subject to change without notice. They do not constitute investment advice and whilst all reasonable efforts have been used to ensure the accuracy of the information contained in this communication, the reliability, completeness or accuracy of the content cannot be guaranteed. This communication provides information for professional use only and should not be relied upon by retail investors as the sole basis for investment.