Our Strategies Ranked by Aggressiveness
A quick note about the part of MarketSci that actually pays the bills: our own independently-audited trading strategies.
Starting this week, I’ve begun sorting all of the summary stats on our Strategies page from the most to the least aggressive program to help folks better understand the difference between them.
Summary stats updated through 03/2009:
Note that I’m measuring “aggressiveness” using the annualized standard deviation of monthly real-time returns. As we’ve talked about before, this is a flawed metric and doesn’t necessarily reflect future volatility, but at least it gives a general idea.
For the statistically-minded, below I’ve listed this annualized volatility stat for each program from inception through 03/2009, compared to the S&P 500.
Two observations:
(1) Eyeballs tend to gravitate towards programs with the highest absolute return – a wrong-headed approach. Returns should always be viewed relative to risk taken. Not just annualized volatility like we’ve presented here, but other metrics, like max exposure on any given day, upside vs downside vol, etc. as well.
From that perspective, I consider our lowest returning strategy (MarketSci Unleveraged Indices) to be one of our best (again, relative to the risk it has taken).
(2) I always say that YK is a volatility monster that must be kept on a tight leash – I think the annualized vol. stats above demonstrate pretty clearly why.
Thanks for the commercial break. Now, back to the geekery.
Happy Trading,
ms
To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.
Filed under: My Performance | 2 Comments





Why don’t you consider the RH strategies the best? They have a Sharpe ratio > 2 after a history of almost 3 years. By the way, would you explain the next to last column, the sorting/sortino(?) ratio. Thanks.
RE to Blue/Russ: I’m assuming you’re referencing the unleveraged indices comment – note that I said “one of the best”. I don’t have a “best”…they’re kind of like children…I have to love them all equally =)
In all seriousness, I consider the combination of all the programs to be far superior to any individual program for all the reasons I touch on in: http://marketsci.wordpress.com/2009/04/29/five-ways-to-create-better-risk-adjusted-returns/
Sortino ratio is similar to a Sharpe ratio but only considers downside volatility rather than all volatility.
Michael