Can I keep my Sharpe ratio higher then 4 for the rest of the year?

The kurtosis is not “net of 3”. You mention that a strategy can become horrible after a great year, I fully agree, and I have seen that in the past. The way I plan to handle it is rather simple, as I mentioned before. In any given month if my account hits a 10% loss I am closing all positions. I would re enter only if I am able to figure out if things have changed, what happen, and if it’s time to go look for some new strategy


Quote from CT10Gov:

I have no idea. Those numbers obviously look good, though you only have one year. I've seen strategies that end up being horrible having single years where the numbers were like that.

But if you banked 90% YTD - that's good for you;

If you want an actual conversation about what those numbers might say, you'll have to tell us at least the broadstrokes of what you are doing: are you trend following, scalping, selling options, buying options, etc...

I think you are reluctant to do so - which is fine - but that makes drawing any conclusion from these figures somewhat difficult.

(btw, is the kurtosis net of 3?)
More...
 
Quote from mickson:

I always recommend to people who show me incredible Sharpe ratio's to consider whether they think they are amongst the best in the world. I see it often with "curve fitted" oops I mean backtested results.

2012-03-01_1155.png


These are averages and last year was a tough one for the industry. However, anyone producing a Sharpe above 2 is frankly shooting the lights out.

>4 that is unbelievable or unsustainable.
More...

I made this comment earlier in the year and perhaps I was a bit unfair. I don't know how much money Mac is trading but we currently have 2 traders on our recently launched beta <a href=“http://www.rapacapintro.com”>RAPA Cap Intro</a> with Sharpe's around 3 so perhaps Mac will be able to achieve his goal, but I believe it will always be one of those about to collapse Sharpe's.

Should be an interesting last 6 weeks.
 
Your skew and kurt look quite "normal". That is a good feature of return. If those numbers are stable over the long run, then the strategy is an excellent one.

However skew and kurtosis measure distribution asymmetry and thickness of tails. So it is very difficult to measure it very a small sample size. A relatively calm year makes it even harder.

You can try to re-run/backtest your strategy to older data, hopefully including volatile time periods in the market, to get a better measure of true return distributions.

njrookie

****************************************
I have finally made the time to study and analyze some statistics on my account YTD. All of this is based on daily MTM close.
Daily skew is 0.585
Kurtosis is 2.388976
Stdv 1.4147%
During the 13 day period that I had the DD of 7.33%, the STDV was 2.65%
I would like to hear, if these are some good numbers when taken together with my Sharpe/performance.
 
A drop of 0.35 in sharpe ratio, relative to to your stdev of 1.42% implies a drop of 0.5-0.6% in mean return. That is over a short 3 weeks window. We are about 45 weeks into the year. So you probably just experienced another DD in the 7% neighborhood since last update.

I am not sure how this happened and whether it is related to the current market downturn. However you do not want your strategy to be too related to the market. Maybe you can try to calculate a tail correlation? Something like your average return in days where SPX is down 1% or more and the correlation of your returns during those days? Unconditional correlation can probably reveal something as well.

njrookie
 
You are right I am having a DD, however it is around 4%. Re market correlation I don't think it is correlated alot based on its history (my biggest DD has been when the market was moving up).
Quote from njrookie1:

A drop of 0.35 in sharpe ratio, relative to to your stdev of 1.42% implies a drop of 0.5-0.6% in mean return. That is over a short 3 weeks window. We are about 45 weeks into the year. So you probably just experienced another DD in the 7% neighborhood since last update.

I am not sure how this happened and whether it is related to the current market downturn. However you do not want your strategy to be too related to the market. Maybe you can try to calculate a tail correlation? Something like your average return in days where SPX is down 1% or more and the correlation of your returns during those days? Unconditional correlation can probably reveal something as well.

njrookie
More...
 
You must log in or register to reply here.
Back
Top