In the press conference that followed my first major keynote address on behavioural finance, the 2001 CLSA Investors’ Forum , a journalist asked me what I would recommend to him – passive or active investment management. He was justifiably stunned when I said "passive".
I had just spent an hour on stage explaining not only that investors (like everyone else) are prone to robust behavioural biases that could undermine performance, but also that these biases are systematic. This means that investors don’t just act in the same way, they do so at the same time. The result is that these biases do not disappear through aggregation: the bias of the individual predictably manifests itself at the market level, and so is potentially exploitable. All of this argues for active management, I reassured the journalist. The problem is: where do you find the active manager free of bias to cast the first stone?
Up until that time, my job had been to develop algorithmic models for in-house and client trading and hedging activities. The algorithmic approach ought to have allowed users to escape their biases. At least, that was the principle value proposition. Yet, over the five years our team had run these algorithms, there had been several occasions where senior management had pressured us to massively increase the exposures to the models, or to dramatically cut them. In each case, the proposed shift would have proved to be (and actually did once) the worst possible moment for doing so.
Institutional clients acted in a similar fashion. One treasury manager, after having sat on his hands as a new model delivered profitable trade after profitable trade, sheepishly asked: “Do you think the next signal will be profitable too, as I want to get in now?” I was speechless. We had provided the tools to eliminate human bias in active management, but users with their fingers on the ON/OFF switch were reintroducing it into the decision process.
In the years that followed, I have dedicated my energy to helping asset managers, as well as other decision makers, to drive bias out of their decision processes and to optimise the human capital at their disposal. I still show them how behavioural biases, such as the dispositions effect, explain exploitable market anomalies, like momentum or the co-skewness risk premium. The lowest-hanging fruit, though, is still the improvement of one’s own decision-making.
P.S. A more chastened Herman Brodie will again be speaking at the 2019 CLSA Investors’ Forum in Hong Kong this September. Bring those questions on.
Comments