44: Intech's Adrian Banner - The Stochastic Portfolio Theory and Klezmer music

Adrian Banner is Chief Executive Officer and Chief Investment Officer at Intech Investment Management, a Janus Henderson business. Adrian is also a former maths lecturer at Princeton University, author of the book "The Calculus Lifesaver and last but not least a recorded musician, playing piano. In this podcast, we talk about the importance of rebalancing, which might explain a large part of the size premium, and look at volatility – two topics that lie at the heart of the Stochastic Portfolio Theory. We also find out what Klezmer music is. Enjoy! Overview of Adrian Banner podcast: 3:00 Studying Klezmer Music 5:00 Why both CEO and CIO? 7:00 Stochastic Portfolio Theory and Robert Fernholz 9:48 Does is solve all the problems of Modern Portfolio Theory? 12:42 The Curse of Compounding 14:30 Rebalancing and the size premium. 17:30 Small caps don't outperform large caps. It is the rebalancing to maintain a small cap portfolio that makes sense. 20:00 High or low volatility? 22:30 How will factor strategies hold up when we move away from an environment dominated by quantitative easing? 27:00 The reason why some research finds that randomly chosen portfolios do better than the market index is because they have more small caps in them. But do they have better Sharpe ratios? 29:00 Should we time factors? 31:30 Picking the top and bottoms of markets is very hard to do, but risk timing is probably more achievable 34:00 Bridging the gap between academia and practitioners. Any views as a former lecturer? 35:00 We have every month graduates in mathematical finance from Princeton come to our office, but not so many economics or business school participants 36:00 I reject all models that say: 'The market is wrong and the model is right'. The market is always right. 37:30 Why is it that the top 5 per cent of the S&P 500 represents about 50 per cent of capital. It has pretty stable over multiple decades…, and it is things like this that drive the size premia. 40:30 Machine learning is a nice research tool, but not as useful as part of an investment process. 44:00 Machine learning and fiduciary duty, do they mix? 47:30 As a scientist, I find it hard to say: 'This is behavioural'. What does that really mean?