Submitted by Eric Peters, CIO of One River Asset Management
“Whatever investment style you adopt will blow up someday,” said the CIO.
“When that day comes, will you fold or double down?” he continued. We were discussing systematic investing. I see its future dominance and am building my firm accordingly.
“If you’ve surrendered control to a machine, how will you make that decision?” he asked. Before I could answer, he supplied his own. “I’d rather practice making decisions along the way so that I’ll either avoid the blow up or at least understand my strategy in the crisis.”
That’s a credible position to take on the matter; for years I took it myself. But time changes most things, ourselves in particular. Day by day, month by month, we’re different people. Humble, hubristic, stubborn, objective, greedy, fearful, certain, confused, euphoric, depressed, and every imaginable combination thereof.
The two greatest advantages of developing decision-making algorithms are that they allow us to consistently be our finest selves, and they can apply our process across more markets than a single human ever could. But the difficulty of distilling profound complexity into a set of robust rules leads many practitioners to cut corners – which takes the form of choosing rules that worked in the recent past for seemingly arbitrary reasons, and building algorithms without sensible risk-mitigation to avoid its corresponding costs.
Such strategies put their investors at risk of catastrophic loss in exchange for a pile of pennies, and/or tend to make money in every time period except for the future. But such pitfalls are not machine error, they reflect human weakness, and are thus common to both poorly designed discretionary and systematic strategies.
Because ultimately, every conceivable form of successful money management requires the experience to identify rules that tend to make money over time, and the introspection necessary to come to know our finest selves.