What is the guidelines for strategy risk versus the total risk mandate for an account? Let's say the strategy's observed max draw down is 5% and the accounts max draw down allowed is 10%. How much cushion do one allow for? double the the strategys draw down? or smaller like 50% extra? If one got to much cushion one leave money on the table. If to little there is a risk to blow though the given risk mandate.
The general thought is that the worst drawdown is ahead of you, due to curvefit and liquidity issues. So you come with heuristics to protect yourself from getting knocked out. Eg if you worst DD is 15% based on the backtests or on some scenario analysis, you size the strategy to 3/4 of the capital (and hope that never lose more than 10%, but be ready for a 15% DD). Couple additional points 1. It all gets much hairier once you have multiple strategies with diverse exposures. You don’t want to be overly conservative and assume that all bad things occur at the same time but also don’t want to be dead if they actually do. Judgement call for the trader/PM and a tricky one at that. 2. Generally, when sizing stuff you want to ignore any stop losses and assume instant loss. The whole thing “oh, I’ll be able to get out” is frequently just wishful thinking
That’s a hard question as it depends on both the strategies and the nature of your mandate. Trading my own funds, I always play things far more conservative than I could get away with. If you’re trading for a fund and the rare big loss means you’re fired, but you’re earning a percentage of all profits until then, it can certainly be rational to push more size.
Obviously the “maximum” drawdown is a guideline for sizing and not an actual physical limit. I’m not sure how you would rationally size trades if not by max expected PTDD.
Thanks for the input, appreciated. Yes, I'm also dealing with this issue as I'm running a multi-strategy portfolio. I've chosen to allocate fixed sizes to each strategy, at least for now. Makes it a lot easier with sizing as I can look at each strategy as a unit. I don't allow for overbooking risk in the portfolio. Pretty standard then you receive external capital. That there is limits one got to contain oneself within. If you exceed you breached the contract and it's game over. They will pull out the remaining capital if they lost trust in you as a manager to do your job.
Generally, you don’t want to overthink the expected PNL quality when allocating to different strategies (again, since you probably overfit etc) but definitely be mindful of liquidity constraints. For example, if you have a strategy that trades some agricultural futures, you don’t want to size it the same as a strategy that trades equity index. This assumes, of course, that your AUM is big enough that you care about capacity. In early stages, your primary concern should be risk, you wanna live to fight another day
If you exceed the maximum drawdown limit, your “investors” hunt you down and break your knees. So once you hit that number, you pack a bag and head to Uruguay.
Tip: With options you can clearly limit your max allowed downside (ie. MDD, risk, VaR), but of course some maths is required... You can add options as support to your current strategy. Study for example this options strategy: the loss is capped, and you can define this allowed max risk yourself by chosing the right Strike and Premium (the parameters of the option(s))...
Not to worry. I trade my own capital. I control risk with a loss limit per trade but not overall drawdown. Haven't had a drawdown greater than 10% in years.