Van K. Tharp's book "Trade Your Way to Financial Freedom". Page 20, "Trying To Beat Random Entry". Van K. Tharp claim that he and Tom Basso tested a random entry system which made money 100% of the time. Rules - 1 contract per futures market - Always in the market Entry - Coin flip Stop - 3 times volatility - Volatility determined by a 10-day exponential moving average of the average true range Risk - 1 percent IMHO thats pretty close to bulls**t. Nonetheless, anyone tested it or actually USING IT? It is even possible to make money with this system?? Comments?
Remiraz, you forget a few things to mention... 1) It only works if you start with $1 million dollars. 2) You use only liquid markets. The markets they used where Gold, Silver, Bonds, Eurodollars, Crude Oil, Soybeans, Sugar, D-Mark, Pound and Live Catlle. 3) Exit is a 3 X ATR(19days) subtracted!!! from the Entry price 4) Trailling stop is also a 3 X ATR(19 days) subtracted from the close when there is a profit. The traillingstop only can go up... If the new calculation gives a lower number, then the trailingstop will not change but stand at his highest point. 5)For best result you must pyramid with money management. 6) Risk is indeed 1% of equity but also max 25% of total port. will be invested at one time in one market. You're statement that it will make money 100% of the time is not true; The probability rate is +-36%. This system is to make you understand that Entry is never important, but what counts is your exit strategy. Besides that, the system makes money for more than 20 years now.
Thanks for the in-depth info. Those are not present in the book. Also, I never said that it makes money 100% of the time. It says so on Van K. Tharp's book "Trade Your Way to Financial Freedom" page 200/201, section - "Trying To Beat Random Entry". By the way, what does probability rate 36% mean? Does it mean it profits 36% of the time and is wrong the rest of the time? From what I deduced (information from the book and only the book) is that it makes successful trades 36% of the time but stop loss so well that on the overall it made money 100% of the times they tested it.
I read that several people have tried to reproduce these results and haven't done much better than break even over the longer term. Runningbear
...it ain't random entry. Setting a volatility requirement makes it NOT random entry. BUT IMO, a system of that general type works. I have one volatility system that has only 50% reliability, but entering in times of potentially high volatility makes it work with trade management. Doesn't make alot, but it overcomes slippage and commish. Try truly random entry with NO other rule and you will get kilt! Daid!
...once in a rare while a condition will arise which the system which is currently in control is not coded to handle. In that case, that's exactly what I do. But it's no ordinary coin, it has to be a lucky coin, one found on the street. I have an envelope full of those. But then the problem is, WHICH lucky coin to pick? Do you have a system for picking the coin you flip?
To Hypostomus: The Entry has no volatility requirement, its the exit only that makes use of ATR... The only rule that is used for Entry is that the market must be liquid. And of course the flip coin. To Runnigbear: The problem with reproducing a system is that not everybody understands the rules... This thread is one prove of that. To Remiraz: Yes, it means that the system is only correct for 36% of the time. For the statement in the book: I think it has been used to give you an example that Entryrules or much less important then Exitrules and that it is in no way recommended to trade it. (Because the book doesn't set all the rules that has been used for this system...)
daniel1c & hypo: I got a question for you guys. From reading Tharp's book and some crude backtesting, we know that most trend following system is probably only right 30+% of the time. Or something like that. So, if your average gain per win is LARGER than your average loss, that seems good. But what happens if all those small losses adds up to something nontrivial though so you end up still net down. That's what I don't get. I guess that's where position sizing comes in. But how do you know which trade to double up or triple up or whatever. What happens if you double up and then the market corrects wouldn't you have lose even more.. I think great trend system has more than just the 30% win and a positive expected value...