I have an good edge in scalping certain futures - ES and CL for certain and it's probably portable to GC. It doesn't appear to work in the rate complex. Theory suggests it might work in other places, but I don't know for sure. The basic nature of the edge is a custom coded indicator I would describe as "automated tape reading". The actual entry and exit decisions are discretionary but don't involve much real discretion - just things like knowing when news releases are. Automation is a possibility. I'm a good programmer and could do it myself. Here's the last ten trading days' stats - they're pretty representative and are from trading a 1 lot on IB: Wins 33 Losses 20 Total Trades 53 Win Rate 62% Avg. Win $62.39 Avg. Loss $43.64 Profit Factor 2.36 Aprox. Confidence Method > Break Even 99.8% Commissions are included in the win and loss size stats. Those stats are a mix of long and short CL and ES trades. The splits don't look radically different except the CL trades have slightly larger wins and losses than ES (no surprise there). Losses on any given trade are limited by in-the-market stops to ~$150+commission with slippage. Wins are bounded at $150-commission for CL and $125-commission for ES. Holding time is <10 minutes so liquidity might very well be a problem if traded on big size. Overall both the quantity and quality of trades has been improving with time and I could see the PF going as high as 3 and my number of trades doubling as my execution improves. So here's my question: how do I best make money out of this and maximize the chance I'm set up for life? I'm ~30 years old. I've got an engineering job that pays about $200K+benefits, and I would lose a small amount of non-vested stock if I quit. Job outlook is stable to positive, but I don't really love it. I don't have to quit to trade (I currently trade mornings before going to work) but could probably find more trades if I didn't have to go in. I live in a mid-priced suburban area far away from major financial centers. I don't want to move for family reasons. So what's the best way to proceed? Should I just trade my own capital? I could allocate about $13K right now, and quite a bit more over the next year. Or should I try to take this to a prop shop or some other institutional environment? As it stands now with a retail brokerage there's zero chance of anyone stealing my approach, but I would be concerned about it in other settings.
Fifty three trades? LOL. Come back here after you have made 5k to 10k trades. No one is going to steal anything because there is no edge. The fact that variance is working in your favor now is not an edge. Keep the day job. BTW, I'm not trying to be rude, just to the point. I have to admit, it's strange that an engineer, someone who I assumed is well versed in statistics would ever consider 50 trades as significant data.
Those trades are representative, not comprehensive. The question is NOT whether the edge is good. I can evaluate that without your help, thanks very much. The issue is how to convert the edge to money assuming it is good.
And I find it odd that someone setting himself up as an expert on statistics would deny that a 50-item sample could be significant for certain types of tests. A simple confidence test based off a binomial distribution will tell you that data is VERY significant - in fact the results were included to save you the math. I guess we all find different things strange...
Yes, technically when n is greater then 30 you "can" have significant data provided your data is normally distributed, independent and little correlation. That is NOT the case here. Here is some advice, keep trading it. No one is going to back you. And yes, keep the day job, the paycheck will come in handy. And I'm being sincere when I say that.
Just ramp up size and re evaluate when your pnl gets closer to your income. Or is there a reason you can't do that? eg. can't trade during the work day
I doubt what you have is a true edge. Its more likely just market action. I could flip a coin for a month and probably get at least one 10 day period with similar results. Think about it. You found an "edge" trading an index??
The sample universe isnt the number of trades, but the time horizon of those trades. It's not 50 items. it's that 50 items represents 500 minutes of trading time. Not knowing the frequency of those signals (did you do 50 trades over 1 day, 5 days, 1 year, 5 years) is what matters. Keep your day job and trade in the morning until you have several years track record and a substantial saftey net.