Quant Investing: Mathematics

Quant Investing: Mathematics

21 November 2022
By Peter Huber, Investment Writer & Daniel Leveau, VP Investor Solutions

In our second in a series of three blog posts reflecting on the criticisms of quant investing, we consider the role of mathematical modelling in finance and the degree to which it should inform investment decisions.

Writing in Forbes, Peter Andersen - founder of Andersen Capital Management - argues that underlying all quant strategies is the assumption that markets follow mathematical rules akin to the physical sciences. This assumption, he argues, is invalid. This criticism is reasonable. Investment strategies by their very nature are unscientific. Their non-stationarity and their lack of falsifiability ensure that they evade experimentation and that any truth they discover is likely to be temporary. However, whilst we must be mindful of falling into a perverse scientism positing that mathematics is the sole source of truth and understanding, we must be equally mindful of the dangers and costs of resisting progress.

Mathematics may lend systematic investing an opaqueness which renders it inscrutable to many. Quants themselves may be mistrusted due to their willingness to place faith in machines. Yet, scepticism and concern are necessary in both science and finance and those raising concerns are performing a necessary service. We cannot be certain that the quantification of finance and the theories that emerge from this process are not simply the faddish theories of a new technological epoch. As such, the scepticism of discretionary investors should not be dismissed as luddism or being entirely motivated by commercial self interest. However, two questions naturally emerge from their criticism:

  • What logic guides discretionary investment?

  • Why is that logic more plausible than a strictly quantitative or hybrid approach?


It seems disingenuous to argue that discretionary investors fail to employ mathematical methods and models when developing their strategies. Mathematical modelling has been present in finance since the turn of the twentieth century, with each decade witnessing an increase and refinement of its use. Thus, whilst George Box’s 1976 assertion that ‘all models are wrong’ may be sound, the oft-cited addendum to this quote - that ‘some are useful’ - remains equally valid.

It appears that much of the criticism of mathematics in finance stems from a disagreement over the extent of its use, rather than simply its use itself; where do we draw the line? Yet, the technological renaissance we are living through may justify the advancement of that line. Computers in this age of technological progress aren’t under the same cognitive constraints as humans and may be better able to recognise patterns in human behaviour and, ergo, financial markets. How can a human comprehend the various forces, from the various factors, across the various dimensions affecting asset price movements as efficiently or effectively as a machine? This is not to say that the rationality of machines is unbounded but in the interests of fiduciary responsibility, asset managers should realise their limits and adapt accordingly. Indeed, reflecting on the observation made by the legendary hedge fund manager Paul Tudor Jones - that ‘no man is better than a machine, and no machine is better than a man with a machine’ - it seems fair to conclude that neither one is strictly better than the other. Rather, both quantitative and discretionary approaches offer advantages and disadvantages. By concentrating on core competencies you can evolve your systems and processes over time as the needs of your organisation demand it; adapting to quantification at your discretion.

Reach out to learn more about how to accelerate your investment process.

Disclaimer

This document is not, and should not be construed as financial advice or an invitation to purchase financial products. It is provided for information purposes only and is subject to the terms and conditions of our disclaimer which can be accessed at: https://www.sigtech.com/legal/general-disclaimer