Most obvious test is the bootstrap, (in absense of a "draw" distribution, in which case you should use the Student-T test). There are more complex and smart ways of doing it, but i would recon these results don't even pass that.
A few years ago I switched from a CPO to a CTA for a variety of reasons. It was the best decision I have ever made. My clients accounts are linked to mine so whatever I do in my account get duplicated in the client's account And as far as raising money from Institutional Investor...Jf they they think you are a good risk then you stand a good chance of landing a client like that. Speak to your broker. He will help you with this. Qualified Investors are out there but they are difficult to find w/o some help. Be prepared to pay 2% of any client a broker lands for you and be prepared to answer some very difficult questons.
Curious about the reasons of switching from a CPO to a CTA ? Seems a CPO would be easier to manage than individual accounts.
OP, as for whether you can get additional capital, your numbers make it certainly a possibility and obviously worth a shot if nothing else. Would be tough getting a significant amount of money, as others have said, but that's not a terrible thing if scalability is doubtful. Ah this made me laugh. Generating a p-value is basic statistics. Snag an introductory stat book and flip to the section on p-values. Monthy returns are the numbers he presented, and knowing the probability that he achieved those returns with a negative expected return would be one of the most objective ways to determine if it was luck.
Regarding scalability - Most of my trades are in FX, Gold and Crude. In FX for example my typical trade is for 500,000. On IB most of the times I see bid/ask for upto 10M(depending on the pair) within a few pips. I think If I was managing upto10 times what I am right now - about $6M, I can trade the same strategies without any loss because of the increase in size. Anything more I am not sure, till I have a track record to prove the same strategies work on an account size of $5M-$10M
Why do you think so? Basic statistics? Bootstrapping was mentioned in this thread and it cannot be found, at least in a way that can be used easily, in basic statistics books. I do not use bootstrapping, at least I did not for the numbers presented here. Bootstrapping is an overly optimistic test. I would prefer to use MC simulation if I had the daily returns of the system, for each day in the performance period. But this is a lot of work. I do that rarely for my own systems only. Here a challenge for you since it is so basic for you. Calculate the p-value for these performance results: 2 -1 3 4 -3 2 2 -1 1 2 -3 5 6 4 -2 7 -2 -3 4 7
Your return is excellent. Get yourself a CTA license and have your statements audited by an accountant. Put your record on iasg and if your real time result resembles your past results for at least 6 months you should be able to attract some investors. The -20% month is a bit ugly... if you have multiple accounts you should probably consider leaving out the worse accounts for that month to "massage" your records a bit. You can mix and match any way you want as long as your records are real... (a good reason why most investors would wait for at least 6 months of real time results) When I say any way you want I dont mean you can cherry pick the best accounts per month. You can only cherry pick the start/stop period of the accounts instead of by month. Well not unless your accountant lets you do that and still would certify your audit...