Do you find deploying a strategy challenging? I do. This may be because I have to accept the strategy as it is today, not the perfect, pristine baby first imagined. These last couple of weeks I have spent working on both SSJ and SSK to get them prepared for release.
After trading the SSJ live for a several weeks, I noticed there was a potential issue with its risk control. SSJ has an exit which adjusts to current market conditions. That stop adjustment is effective because it allows the trade to stay in under normal market movement. However, under a very specific scenario, the risk was allowed to increase beyond acceptable levels. So a risk floor needed to be established. I went back to the TradeStation filtering signal to make sure the new risk control scenario didn’t adversely affect returns.
One of the key aspects in the signal generation I like to see are contained draw downs, meaning few fast transitions lower. Drifting lower can be acceptable, as the options premium decay can compensate for some price drift, but fast price movements lower often result in outsized losses.
For the SSK, I collected MAE/MFE (Maximum Adverse Excursion, Maximum Favorable Excursion) data on TLT and GLD trades. Observations that stood out:
- As with many trades which operate on a higher time frame, the periods spent in profit and loss can be substantial. I have no issues with being in profit for a long time, but being in a loss for a month or two may be a challenge.
- GLD had a large MAE (-147% of income) which was cut in half by the luck of timing in the initial ONE back test. Implementing a stop will help control this.
Here are the returns on income graphed on DIT. This shows the power of the premium decay portion of the strategy. Once the trade is in the market for a little over a month, the probably of a winner goes up significantly. Collecting these stats helps define what is normal for the trade. This research will provide a base to evaluate current market performance against historical to help make the decision on when the system is ready to be replaced with another.
I’ve also invested some effort into a prototype tool to evaluate combinations of strategies. As an admittedly skeptical engineer, I have found I am more likely to hold through draw downs if I do the to work to convince myself the approach is valid before committing to trade a strategy.
If you are struggling to stick with your plan, then fight the urge to deviate and go verify and validate the edge with back testing and analysis.
Have a great Weekend! Looking forward to chatting with you in the community.