Brandon Koepke

Lots of commentary and some incredible looking charts on the low correlation / high rotation regime in which we live. The behavior of the pod shops and long/short equity strategies that are implemented largely market neutral but with a good deal of leverage and run with very tight stop outs is often cited as a factor. I think you can model out the trading activity in the context of how a dealer would hedge an outperformance option. Suppose a dealer sold an investor an option on the outperformance of asset A over asset B. If the trade is struck ATM, there's no initial hedge. Suppose A rallies relative to B. The dealer needs to buy delta in A and sell in B. That's a correlation dampening impact. Suppose the option is in the money and the dealer's hedge has left him/her long shares in A versus short B. If A declines, the dealer needs to sell A and buy B. Again, the impact here is to sell one, buy the other, reducing the extent to which they move together. There aren't actual outperformance options trading in the market of any size, but the pods are essentially manufacturing them through their trading. The tight stop is kind of like option premium being put at risk. If the trade moves against you, you've got to reduce quickly, especially because you have so much leverage on. That reduction is all about the rotation we marvel at. How does it break? First, the leverage does appear substantial. If the cost of or availability of that financing is disrupted, that can matter, quickly. Second, correlation assumptions. Never, ever trust correlation! When you backtest strategies that are contingent on the risk offset coming from especially low correlation, they enable your sizing to get larger, sometimes to a substantial degree. LTCM famously had the "on the run / off the run" trade on in USTs in enormous size because the net risk of that position (based on history) was near zero. If these pod portfolios are really similar on both the long and short side (ie hedge fund hotels), they use lots of leverage, and their "reaction" functions are similar, you could wind up with a large issue if the correlation breaks. If we should learn one lesson from episodes of disruption in markets it is that we underestimate "sameness" and think our portfolios are move diversified than we really are.