Anyone ever backtest the Turtle system on the DJIA just for fun? I think I just did back to 1930, and it does NOT beat BUY AND PRAY. If you figure taxes into it, BUY AND PRAY destroys this system. Things that make you go hmmmm..... peace axeman
HI Axeman: You have a lot of posts, so I hesitate to ask this question. When you "tested" the turtle system, did you use the original money management rules? Based on memory, I believe there are several versions of turtle system "out there" and without the correct money management system, they don't produce a good result. It is possible that for specific markets, the system doesen't test well for extended periods of time. Also, the turtle system was designed to wait for the trend (another reason to incorporate the proper money management rules). This implies several things as follows: First, the orginal turtles all traded a minimum of 7 or 8 markets with rules for adding or subtracting markets as time went on. Second, they expected to have extended flat periods, and third, when they found an opportunity (a good example is the "Metelgaselchaft" trade), they were willing to risk most of thier previous profits in order to hit a homerun. Just by chance I happened to look into the subject in some detail. The system is hard to stick with, and is best used by strong hands who can work in syndicate fashion. That is probably why the original traders had to be so well funded by their benefactors. I hope these comments offer some food for thought. Best Regards, Steve46
Yes I realize the "real" turtle system was traded against MANY markets. Thats why I asked if anyone has tested it only against the Dow. Also... I think im using the right money management. I was just curious about what its performance would look like against a single market I had a LOT of data for. The stats were pretty typical of average trend systems. Win% of around 40 and pfactor of around 1.4. Its easy to improve on these numbers. Only problem is, overall profitability goes down. In any case.... It seems that it does not beat buy and hold on a single market. This seems to tell me that its real "edge" is in being in as many trends as possible across many markets at the same time. peace axeman
I could be wrong, but my (limited) testing experience shows that short-term counter-trend systems work better on equity indexes than trend following systems... My guess is that the turtle system would have done amazing in commodities during the inflationary 70's as well as would perform well on interest rate futures and currencies...just a guess.
----- it is my contention that trend and counter trend systems would have near equal accuracy in the short term. best surf
Axeman, It sounds like you've read the pdf on this website regarding the original turtles system? http://www.originalturtles.org/system.htm I just read it and it is very informative regarding the turtles method. I think you are right - the edge of this system is in market selection and diversification. It is choosing the 1 or 2 markets that will have big winning trend moves that will offset all the other losses the system incurs in all the other markets it trades. And because the stock market is a market that generally trends upward, buy and hold is a profitable long term approach. But the turtle method will outperform buy and hold in other commodity markets that don't have a long term upward bias.
The myth of the turtle trading being systematic? The whole point of turtle trading is that it can never be backtested. The turtle system was not an exact system or some type of black box where any idiot can do it and make $. Each trader applied a level of discretion.. however the rules were pretty much the same. Thats why the majority of turtle traders failed. It all came down to the ones that got lucky and capitilized on the big moves. The traders that caught a big move like going long nasdaq past year or short the dollar, long copper... pyramided their winnings and have huge size. Those are the guys that cleaned up. Such a system can never be backtested because its not the actual buy or sell rules that made them money. What it came down to is how much money they had on the line when they nailed a huge move... and how they handled their positions. When you position trade its all about money management and scaling massive size into your winner. The losers are easily taken care off by using a stop loss or fixed loss.. but the money is made by handling the winners. Just imagine if u traded 10 markets.. and happened to pyramid heavily into the S&P, Copper, Metals, and shorted the Dollar. The money that could have been made is astronomical. Whereas another turtle trader could have not pyramided and their profits could of only been a fraction of yours. Point is.. it all comes down to money mgt and position size not whether a 50 day breakout works better than a 100 day breakout. --MIKE
I have tested the system as outlined in the .pdf file at originalturtles.org and it worked well. Just because a system doesn't work on DJIA doesn't make it a bad system. I works well in bonds and currencies specifically, and as previously mentioned, their money management is what made the system work. A lot of traders on here seem to think that the stock indices are the only markets traded. I think that either the original poster either didn't read the .pdf file outlining the specifics of the system and therefore didn't program the WHOLE system in or didn't take into account that they didn't trade ONLY the S&P. Many of the turtles did well for themselves after the turtle program died.