Everest29

Albnd, not sure what you mean by "discard part of the portfolio". I haven't had a chance to watch the presentation you posted yet, but looks interesting - definitely will. As far as my portfolio, the process I described is what I do for each system individually. I then test the portfolio as a whole and make sure that the correlation of the systems' results combined with the position sizings don't give me any extreme historical drawdowns. I run about 15+ system/market combinations, and don't really assess the Kelly of the entire portfolio. I only really reassess the Kelly of a system if I re-optimize for some reason (which changes the Win% and Win/Loss) or if I see a single system or correlation of systems that gives me an unexpected or sharp drawdown (then I might reduce the percent of Kelly). As for BlueHorseshoe's comment earlier, I totally agree that you have to have a large enough sample size and, I would add, as few variables as possible. I like to keep it simple. I test with 7-15 years of data (whatever is available) on 2-60 minute bars, so each system will have at least a thousand plus trades in testing. And all my systems (in Easy Language) will fit on less than 1.5 pages. I'll take a look at that video. Thanks for the link.

Sure, Albnd. I first figure the Kelly number for each system (and sometimes for both longs and shorts separately). Kelly was originally developed as a gambling approach for bet size. It tells you the % of your total capital you should risk on each bet (or trade) to maximize your total gain without going bust, assuming you know the Win % and the Win/Loss ratio, and that they do not change going forward (which of course they will in trading). I then take a percentage of that Kelly number (say 15% of what Kelly says I should risk - this is usually determined by modeling a given system and seeing what the drawdowns would be given certain percentages of Kelly). Then I set the maximum risk of my total capital to that amount on every trade that particular system places. For example: I first caclulate the Kelly for the system, as shown below. Let's say it works out to be 30.2%. Then, I pick a pecentage of that number to test. 15% of 30.2% Kelly would be a 4.53% risk per trade (.15*.302=.0453). I keep testing different percentage of Kelly until I find one that gives me a balance of profitability and drawdowns that I like. (To give you an idea, my total portfolio of systems gives me drawdowns of >40%, but I'm comfortable with that. So choose a drawdown that you can handle and be conservative in this estimate because the drawdown will be larger than in testing or history.) So, then I take total capital times percent risk to get my total $ risk per trade. Say, $100,000 total capital * 4.53% = $4,530 total risk per trade. Then I find my stop loss per contract on the next trade. I always make my stop loss dynamic, rather than a set amount, by setting my stops a certain percentage of recent volatility away from my entry. ATR (or higher timeframe ATR) works well. Let's say in this example that it's $500 per contact. Then I just divide total capital risked by stop per contract and round down: $4,530/$500 = 9 contracts. So, what happens is you continue to risk the same amount of your capital on your system on every trade. If you make profits, i.e. - more capital, then your quanitity will rise; if you lose money then it goes down. I try to always think of profits and losses as percentages rather than dollar amounts. That makes all trades (regardless of equtiy) comparable. Hope that helps. From Investopedia: There are two basic components to the Kelly Criterion: • Win probability - The probability that any given trade you make will return a positive amount. • Win/loss ratio - The total positive trade amounts divided by the total negative trade amounts. These two factors are then put into Kelly's equation: Kelly % = W – [(1 – W) / R] Where: W = Winning probability R = Win/loss ratio

I agree with the above that Win Rate is over empasized. You have to look at Win Rate and Win/Loss. Sometimes (depending on your trading approach) you have to be comfortable with a lower win rate - you have to be willing to let go of being "right" more often to be profitable. To answer the original question, I "day trade" meaning I get out of any postions by US EOD regardless of profit or loss. I also trade fully automated (that's another contentious discussion, but let me assure it's possible). Of course, none of my systems trade every day - some are more active in certain types of markets than others. From Jan '08 through yesterday my numbers are: Win Rate: 53.16% Win/Loss: 1.01 And unless you're trading a fixed number of contracts, all other numbers (including PF, CAGR, Average Trade Return, drawdowns, etc.) come down to position sizing, which is hugely important. I use % risk based on a varying percentage of Kelly for each system. Also, when you're testing for trading on shorter timeframes, you have to be very precise with (or overestimate) slippage and commissions. But they definitely have predictive power.

Also, kingkongtracts, you'd need to consider the psychological toll of maximum leverage. You would have drawdowns, even with good systems. And they will be larger than you've planned for several reasons - future is different than the past, unexpected slippage, errors in execution (missed entries/exits, power outages, connectivity, etc.). These drawdowns, even when you're aggresive and not at "maximum leverage" can be tough to sit through (I use a percentage of Kelly for each system with dynamic stops based on an approximation of recent volatility). I've been trading fully automated stategies since 2008 and have made a great returns. But drawdowns are sometimes very challenging psychologically. So much so that you begin to question your systems and your approach. But if your system development was sound, and your approach was built to be robust, then these times are usually the worst to jump ship. This is also a good reason to have developed your system and approach by working though it yourself, as opposed to using one someone else developed - you know the strengths of your systems and know what to expect in terms of performance and drawdowns, allowing you to ride out those drawdowns. Also, Obsidians, comment on duration is correct. Many systems can work over 3 months. But I like to see mine work consistently in testing for 10+ years back. Then, when a system struggles, you can easily see if it's just par for the course, or something more serious that may need to be addressed. Good luck!