I would like to point out that the percentages I am using in this demonstration are based on the ability to hold one whole unit of the emini S&P 500. So, if price is 5554, as it is now, then : 5554 x 50 = $227,700. That's how much capital you would need for these percentages to apply to you. If you have that much capital allocated to just this method, then 6% is approximately the expected max drawdown, looking at the last 336 calls. But 6% of $227,700 is $13,662. If trading futures, that's about how much you would need to have in your account, at all times, at a minimum, just to trade ("maintenance"). Thus, to give yourself enough cushion to handle a 10% move in the underlying, you would need : $13, 662 + $22,700 = $36,332 If that's how much capital you have to allocate, then a 6% drawdown in the underlying represents 37% of your overall capital, or 60% of your margin of flexibility. If you can tolerate a 37% drawdown then that's that. But if you can't, then you have to figure out a way to select, out of all these trades, a "set up" which gives you better odds. Either that or use options. Or maybe get into a prop firm. But even then you have to work your way up to a $227,700 account size. However, if you can get established in a prop firm that allows you a 10% drawdown, and pays out every time you reach a 10% target, and you can get your usable capital up to $227,700, then you are golden. You'ld be looking at a $22,000 payout about every three months, or about $7,000 a month . But then you have to consider the profit split with the house.
If 4558 is touched for a long, the short reverse will be 4549. This is a very narrow channel so this applies to each touch of these limits if price decides to ping pong a bit
Is that an optimized backtest (only) ? Or Out-of-sample performance ? Or live track record ? Because when it is only optimized on past data, I highly doubt you can achieve near that performance in the past over the next future, for sure with this Price Action strategy. Let me know what is that behind the posted equity curve...
The posted equity curve is neither in nor out of sample. It is two years of data manually curated with this method. It's not optimized. It's just this method on price action. Different time frames, but same rules. Time frame of posted equity curve is same as what I am working on live with the same rules. I've not really optimized the rules because that would require computing power. I don't have this fully automated yet. There may be better ways of doing this. That said, manually curated data might appear to be performing, only to find later, live, that it underperforms. However to the degree I have been able to automate a TEST (not a full automation), the test performed similarly. While the equity curve is 336 data points, I have curated perhaps 10,000 data points, using these same rules. The equity curves have all a similar signature, so to speak, which is a generally straight, steady walk upwards, random in appearance, but with a fairly reliable max drawdown which generally only comes after a new high, especially a significant new high. One problem I have with this particular live demonstration is I have to check my position at 3 AM in the morning. I can't always be awake at that time. Typically I will set stops before going to sleep after midnight. Last night I forgot to set stops and it cost me significantly. So this test is now marred. I will simply have to make a few more calls to get the average gain up to where it is expected. At the end I will make a note of which trades were not to specifications, and how much that cost.
At some point I'll post an equity curve of the live demonstration next to an equity curve of manually curated historical according to the rules. Ideally they will track closely.