Futures and forex trading contains substantial risk. Past performance is not indicative of future results. See full Disclaimers and Disclosures at Legal. Scroll to bottom of page to Unsubscribe at any time.
Well, I crashed 2 servers and the NinjaTrader platform this week. I would say I am making progress.
The theory behind my strategy is the standard deviation bell curve. If price leaves the 1, it has no where to go, except to the 2. And if it gets past the 2, it will usually head straight to the 3. Ok, not rocket science here. The trick is trading it.
This is where the hedge funds come up with their different models. I am using volatility position as far as where it is relative to the 1-2-3 on the higher time frame, like a daily distribution; or even a custom distribution that I create myself for the RTH session off the open. I put myself in the middle of the distribution at 8;30 am central. Clever girl.
Ok, enough of the showing off. I can also create custom distributions of the 1 hour initial balance, and wait for price to break <or> the 3 on that smaller one. Reasoning, if there is going to be a breakout, then price will have to break that 3 on the IB. Or you can extend this logic out to a 5 or 15 or 30 min opening range. Draw a distribution with PT indicator for those periods and wait for the 3 to break.
But for regular trading every day, I use a combination of the daily and RTH distributions. I wait for price to break the 1 on those higher time frames. That has a high probability of follow through in the equity indexes.
Here are recent BACKTEST results over the past 150 days using 15 and 30 min charts breaking the PT1 and moving to the 2 or 3.
It is a high probability strategy. So there is the base for my auto trading strategy. Wait for the 1 to break on these higher time frame volatility envelopes and go to the 2 or the 3. Or even to the tails with momentum and linear regression.
Then, how to trade with the auto sizing unirenko bars I have been working on building? Well, by incorporating pullbacks at key statistical levels within the volatility framework, this will monetize the strategy even further. Theory is to take these backtest results and amplify them 2x, 3x, 4x, 5x the amount.
So I use the threshold solver to capture actual real volatility on the 15 min and 30 min and 60 min charts. These volatility levels are consistent across the instruments. Each instrument has its own unique volatility threshold. So, if price is trading below a volatility threshold on the 15 min chart, the unirenko that auto sizes itself to the 15 min volatility will automatically take the pullbacks. But if the threshold is above a level, that auto size will stop trading and another higher time frame auto size unirenko for the 30 min time frame will start taking the trades. And if the volatility increases again, exceeding the 30 min threshold, that auto size unirenko will stop trading and the auto size unirenko for the 60 min threshold will start taking the pullbacks. I graduate to higher volatility levels automatically as the volatility increases.
So I have the theory. I am trying to get the proof in the pudding report. Have been working on this for several months. It’s in the last 10% that has been getting sticky. I have to optimize my logic so I stop crashing the platform.
All the logic is finished, I just have to make it better. Get rid of anything superfluous and clean the code. See if I can get a real time example running of my volatility threshold envelopes working real time. Hopefully in time for some volatility to hit the market in the next few weeks.
Wish me luck.