Thomas Stridsman quote from his “Trading Systems That Work Book”
The benefits of the RAD contract also become evident when you want to put together a multimarket portfolio…For now we only state that the percentage based calculations do not take into consideration how many contracts you’re trading and, therefore, give each market an equal weighting in the portfolio.
The Stridsman Function I presented in the last post can be used to help normalize a portfolio of different markets. Here is a two market portfolio (SP – 250price and JY -125Kprice contract sizes) on a PAD contract.
Here is the performance of the same portfolio on a RAD contract.
The curve shapes are similar but look at the total profit and the nearly $125K draw down. I was trying to replicate Thomas’ research so this data is from Jan. 1990 to Dec. 1999. A time period where the price of the SP increased 3 FOLD! Initially you would start trading 1 JY to 2 SP but by the time it was over you would be trading nearly 3 JY to 1 SP. Had you traded at this allocation the PAD numbers would be nearly $240K in profit. Now this change occurred through time so the percentage approach is applied continuously. Also the RAD data allows for a somewhat “unrealistic” reinvestment or compounding mechanism. Its unrealistic because you can’t trade a partial futures contract. But it does give you a glimpse of the potential. The PAD test does not show reinvestment of profit. I have code for that if you want to research that a little bit more. Remember everything is in terms of Dec. 31 1999 dollars. That is another beauty of the RAD contract.
Another Stridsman Quote
Now, wait a minute, you say, those results are purely hypothetical. How can I place all the trades in the same market at presumably the same point in time? Well, you can’t, so that is a good and valid question; but let me ask you, can you place any of these trades for real, no matter, how you do it? No, of course not. They all represent foregone opportunities. Isn’t it better then to at least place them hypothetically in today’s marketplace to get a feel for what might happen today, rather in a ten-year-old market situation to get a feel for how the situation was back then? I think wall can agree that it is better to know what might happen today, rather than what happened ten years ago.
That is a very good point. However, convenience and time is import and when developing an algorithm. And most platforms, including my TS-18, are geared toward PAD data. However TS-18 can look at the entire portfolio balance and all the market data for each market up to that point in time and can adjust/normalize based on portfolio and data metrics. However, I will add a percentage module a little later, but I would definitely use the StridsmanFunc that I presented in the last post to validate/verify your algorithm in today’s market place if using TradeStation.
Email me if you want the ELD of the function.