OLD SCHOOL VS NEW SCHOOL?
By Alban Ramadani
In light of advancements in machine learning and systematic trading, development of sophisticated computer-driven investment strategies, and probably driven by the success of James Simons’ Medallion Fund over the last 34 years, an increasing number of hedge funds have been allocating a portion of their AUM to systematic trading. However, one aspect of this shift towards quantitative investing has been underestimated: can there be a coexistence of quant traders and traditional discretionary portfolio managers in tomorrow’s hedge fund?
Hired to develop new quantitative models by processing an ever-increasing amount of data and to lead a change to a more systematic approach to investing, quants at fundamental funds have been struggling to gain support from discretionary managers. On one hand, data scientists suggest one of the advantages of data-driven systematic forecasting and statistical probability is a detachment from emotion and personal ego, and complain about finding themselves on the sidelines of the investment process and often seeing their research and trading signals discredited by discretionary Portfolio Managers. In fact, old-school fundamental managers don’t seem to have a propensity for basing their investment decision solely on systematic signals, but they would rather use them as a way to confirm their investment ideas. On the other hand, fundamental investors believe data scientists do not properly ascertain the validity of the analysis behind their trading signals, and often fail to thoroughly explain the margin of error in their price targets.
However, in my opinion, the difficulty of fundamental managers to coexist with data scientists, ultimately comes down to a matter of trust. Since it was first established in 1951 by Graham and Dodd in their initial printing of Security Analysis, fundamental analysis has been the cornerstone of investment analysis and portfolio managers have extensively relied on it. Even though fundamental investors and quants have ultimately the same goal, beat the market, it is hard for discretionary investors to move away from fundamental analysis – an approach that to many of them has proven to be very profitable. In order for fundamental investors to embrace the purely quantitative side of finance and trust its trading signals, it is necessary they become more acquainted with the data sets (traditional and non-traditional), the inputs that go into the models, and understand how to effectively incorporate machine-generated signals into their strategies.
Time and progress won't pause and wait for the laggards to catch up.