just finished my Friday night backtests - risk parity does seem to perform pretty well a lot of the times. So when does it not?
-
but thats a lazy answer. a bad year in risk-parity is characterized by a repeating pattern of sudden crashes followed by quick recoveries and unstable correlations forcing de-levering before sharp recoveries and crashing again right after a lever-up signal
Only if you use quite short-term estimates for volatility and correlation! Risk parity does perfectly well with long-run (eg 10y) vol estimators and correlation doesn’t need to be an input at all if you have a working model of the factors driving risk premiums in each asset class
-
-
This might be a dumb question, but do you think leverage is required for risk parity? And if so, can short term vol burn you on things like margin/liquidation risk even if you’re using longer periods for measuring allocations? Or is that more of a friction than risk of strategy?
-
It’s not required but (depending on implementation) you would see 3-5% vol and annual returns of 2-4% + tbills on unlevered risk parity, so you practically always see leverage used. As with any leveraged strategy you can get burned on big short-term drawdowns.
- Show replies
New conversation -
-
-
Would love if you could expand on this! I can understand using slower vol signals, but how exactly do you substitute out correlations from the optimisation?
-
You have a rough model of the factors that drive each asset class (eg equities long real growth, nominal bonds are long real rates and short inflation, commodities long real growth and long inflation, mortgages long real rates, short inflation and rates vol etc)
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.