Sunday
Jan042009
Woah Nelly!
Sunday, January 4, 2009 at 10:29PM The response to "Christmas Gift" was beyond
what I was expecting. It's time to pull back
and try something a little simpler.
Since this is such a
difficult idea to test I am asking people to slow down a
little bit and try something a tad bit simpler. I
am now publishing the second episode that I referred to
previously. This will be simpler to test and will
provide some insight as to whether the original
Christmas Gift can work in real life.
So here we go...
This is real simple. This works ONLY on the DJIA index. Don't try testing this on other indices. The rules are as follows:
(1) sell short if gap down opening
(2) otherwise go long
The system only switches direction at the market opening.
If this system can be made to work in real life then I suspect the original system will also work. So expend your energy on this first. Here is the equity curve (DJIA points profit) from Jan 2000 to present.
So here we go...
This is real simple. This works ONLY on the DJIA index. Don't try testing this on other indices. The rules are as follows:
(1) sell short if gap down opening
(2) otherwise go long
The system only switches direction at the market opening.
If this system can be made to work in real life then I suspect the original system will also work. So expend your energy on this first. Here is the equity curve (DJIA points profit) from Jan 2000 to present.
I will be deleting the rules in a few days. So copy what you need.
Take care and enjoy,
Steve
Reader Comments (8)
I would run this on the DIA since that is the closest proxy for the DJ. My suspicion is that if it doesn't work on that it won't work in actual trading. I have also not tested it. The fact that it does not work on the S&P500 for example is also a big red flag to me, unless there is a logical explanation as to why it should only work on the DJ.
My thinking is that this system only buys/sells at the opening. It is easy to extract opening price for DIA (or other ETF). So one has to use the DJIA raw index as a cue to buy/sell DIA. If this is profitable in backtest then it is likely that there is some merit in pursuing the more complicated system which is difficult or impossible to backtest.
Walk forward testing will provide some indication whether or not the DJIA system will work in the future. The DJIA has the fewest components of any index (I think) and probably the most liquid stocks. So the index calculation may be more accurate or faster than other indices.
Steve
I agree. My understanding is that you ran the backtest with the actual index numbers from YAHOO. If you run it with the DIA ETF numbers does it give similar results? If, so then I agree there may be something there. If not, then I don't think it will work. Not to sound negative. But, better to challenge results now than to have the market teach us an expensive lesson. Keep up the good work, and interesting BLOG.
Charles - if it works with DIA directly then great. But if it doesn't, then no conclusions can be drawn. One still has to test using the raw index as a cue. DJIA != DIA.
If using the DJIA as a cue, would it not be logical to assume that if the DJIA gaps down (up) the DIA would also gap down (up). Then it would be logical to assume that the DIA opening price is a good proxy for the price of the transaction execution. This is basically the same thing as using the DIA for the cue timing as well. To verify this you can use your exisiting model for timing and use the DIA opening prices for the transaction portion. My feeling is that real world results will look very much like using the DIA for both timing and exectution. I have not had a chance to verify this however.