Futures vs stocks

If one considers IRA investing, choices are limited and futures are not easy to trade in IRA accounts. Yet there are more futures systems with excellent long-term results compared to the ETF-based systems for sale. My criteria for good performance is 6 month or more existence and Sharpe approaching 3 or higher.



Why futures systems seem to perform better and why not more developers put energies into ETF-based models for sale?



Thank you,



Eugene

I think you’ll see that ETFs are less popular because of the lower leverage. Most visitors to this site seem to want highly leveraged systems where the equity curve goes straight up (until it crashes). High fliers sell, boring non-leveraged systems not as well.



I have an ETF system, but its only a month old and sharpe ratio=1.82. I do trade it in my personal IRA with Vanguard, so at least I believe it in (and wish I would have started it in early 2008!)



Kevin

Kevin, Since you don’t believe in back testing but only forward testing, what do you tell your potential subscriber: To wait 6 month, 1 year or longer before considering any of your systems?

SORRY IN ADVANCE FOR THIS LONG POST, BUT I THINK THE POINTS ARE IMPORTANT.



I tell people exactly how I develop a trading system: I use what is called walkforward testing. Some consider it a backtest, but it is most definitely not a backtest in the well known, traditional sense (shown below). There are plenty of good Internet sites that describe it better than I ever could, but in a nutshell here is an example:



Traditional Backtest

Dataset1: 1/1/2000 - 12/31/2008. Optimize all parameters on this data set, start live trading 1/1/2009 with “best” parameter set. Historical backtest results are based on the optimized dataset1 (1/1/2000-12/31/08). That’s what traditional backtesters show people. It is typically a 45 degree equity curve with minimal drawdown. Many times looks too good to be true.



Walkforward Testing

Dataset1: 1/1/2000 - 12/31/2001. Optimize on this set, get “best” parameters set, then trade it on dataset 2: 1/1/2002 - 6/30/2002.

Dataset3: 6/1/2000 - 6/30/2002. Optimize on this set, get "best"parameters set, then trade it on dataset 4: 7/1/2002 - 12/31/2002.

…And so on until you get to today…



The walkforward test results are then dataset2 + dataset4 + … (all the even dataset numbers). Note that the walkforward datasets were never optimized before they were used - it therefore makes a more realistic assessment of a strategy going forward. It is a decent looking equity curve, but far from perfect, and probably looks “bad” if all you’ve ever seen are traditional backtest reports.



I should point out that this method is very time consuming when done correctly (since you have so many small periods to optimize and walkforward, rather than just one period). Also, most testing software can’t even do this automatically (I use Tradestation for strategy development, which only does traditional backtesting). I have a specialized Walk Forward Processor program to automate the process for me.



I hope this describes the process. As far as when would I subscribe to any C2 system: I personally would wait 3-6 months for a system to develop a track record, possibly be exposed to different periods of trend/congestion, volatility, etc., and see if the developer “melts down” when a losing patch is hit (it happens more often than you think). I realize, though, I am not most people. Many/most people look for a first week gain > 50%, then jump in.



It is called due diligence. Subscribers would lose less money if they followed a due diligence process before subscribing. No free lunch, even if you a just a subscriber.



Hope this helps.

Kevin

If you are in the USA, you can only trade futures in your IRA through a trust, as follows:



talk to futures broker about opening an account. Most of them can do an IRA.



he will direct you to a trust company that handles futures in IRAs, usually it is mtrustcompany.com (Millennium Trust) - they are a big dog in the industry.



You will wind up sending in your IRA funds to the trust company. They will then forward some to most of your funds to the broker. Millennium only holds $500 or so. There was another trust company that held half your funds in the past.



Once in the broker account, you can trade to your heart’s delight. It will look the same to you.



There is an annual fee for the trust company. They will discuss it with you. You will fill out paperwork for the trust and the brokerage companies.

Good info. My broker (Tradestation) told me that I could use only 1/2 of the available funds for margin (the rest was basically inaccessible). On a small acount, this could significantly impact your ability to trade.



Then again, if this "1/2 margin" rule applies to most/all brokers, and it would impact you, you probably are trading with too much leverage (especially for a retirement account!)…

Then leave Tradestation. At other brokers, I was not given this rule.



Also, which trust company did you use?



The holding back of funds should be at the trust company, not the broker.



On the other hand, no one should be using more than half their funds - that is overleveraging.

Yes, descriptions like that are always helpful and appreciated.



One more question though: Why don’t you post the results of these forward walking tests in your System Description or on your web site?

Interesting that Tradestation might be unique in this margin issue. Basically it was a deal killer, so I never got to the trust company stage.



I 100% agree with leverage comment. Too many people using too much leverage has caused too many problems for too long!

Thanks. I certainly will be posting more info. If you notice, I started adding info in the description to one of my systems yesterday. Since I mainly spend time developing systems for my own account, posting info is a time permitting task.



PM me any specific request, and I’ll be glad to provide.



Kevin

The info in the System Description is good stuff and explains the large drawdown. Keep up the good work.

people confuse "$500 day margin" with "more profit" when it REALLY means "Faster Risk of Ruin"