(Find Agustin on LinkedIn)
1. “Normal” trades may have a specific set of biases like revenge trading, paralysis by analysis, gambler’s fallacy and many others. Quant traders are thought to not have biases, but in fact they do, just different ones, from curve fitting to closing themselves to specific patterns. Which have you seen and how can quant traders recognize and overcome them?
Quant traders are indeed full of biases. Here are some common and not-so-common ones:- Falling in love with your tools. To a hammer, everything looks like a nail, and to someone who likes linear models, that’s all they see. Same for any other quant tool like neural networks or fundamental analysis.- Not thinking carefully about your priors. With so much data available now, it’s easy to tell yourself that’s all you need to do: look at the data.
In fact, prior beliefs and “knowing where to look” are just as important as ever. Doing so usefully restricts the possible alternatives and focuses your mind.- Data dredging without noticing. Every little parameter tweak, every little test you run, every adjustment or filter or squeeze you do is dirtying up the data you’re using to tell you whether your strat has any alpha. Fixing this is as much about the sociology of your company as anything related to the data, math or stats.
2. One of the biggest problems I see in asset manager clients is how to stimulate idea generation among their traders/PMs/analysts. Do you see this a cultural issue in the company (creativity in a corporate environment), does it depend on individual backgrounds, or completely different factors? How does a CIO effectively stimulate it?
It depends on a lot of factors, of course, but the most important one in my experience is the mindset of leaders. People emulate how their leaders think and act. If you seem never to change your mind, if you institute a hierarchical culture where people only speak up when they’re sure they’re right, those are the people you’re going to get. In a competitive undertaking like trading, the best ideas have many parents. The image of a lone genius sitting in a corner dreaming up profitable strats is, in 2020, just wrong. Leaders who participate as equals with their team reap the benefits of that collaborative culture.
3. Promoting analyst-trader collaboration is another subtle topic. Even if you have traders executing and ideas being generated, making the bridge between having an idea to backing it up with research and actually building an investment thesis from that research are completely different things. How can a CIO address that in his team?
The best way to break down that barrier is to have people do each other’s jobs for a while. The training program for a trader should involve months of doing analyst work. And vice versa. And having people rotate for a week every 6 months is an incredibly powerful way to break down those barriers. There are many other approaches that work, and the point is that you have to work at it every day.
4. Many CIOs and fund managers have a finance background but not a lot of talent in managing and/or assessing people. What can one do to determine if someone is a fit with the team and culture versus just analyzing performance and background? Which criteria would you use?
The most important thing, from what I’ve seen in my career, is not necessarily whether leaders have those skills coming in. It’s whether they’re willing to put in the time and effort to learn them. Empathy, listening, and all of the other leadership skills are learnable, but only if the student is willing! Therefore the thing I look for most is a willingness to learn and an openness to other viewpoints. Overconfidence, especially about one’s own beliefs, is the mind-killer for traders, investment professionals, and managers too.
