After trying out several trading systems that failed, I started to think that maybe those systems could actually be useful, but only if I use them upside down. If my stoploss is being hit too often, than I should make it as my take profit and reverse my buy to a sell. The same goes for take profit levels.
I’m not too advanced with expert advisors so those failing systems were tested by just looking at historical price movements. However I noticed that there is a simple moving average expert advisor and I took it on a test and the results were awful (or perfect?). Just look at that graph. Thats in a year long period. Sorry for the big picture.
I have a feeling that the expert advisor is based on a simple crossover rule and with added SL and TP levels. What if we would reverse everything? Wouldn’t the chart go the opposite way? I think that it doesn’t make a difference if a trading system is always or most of the time loosing. The most important thing is that it can somewhat succesfully predict the possible direction.
I have attached the expert advisor to this thread. Maybe someone can look in to the details and try to edit it so that it works the opposit way.
And just for fun… Maybe someone could post their worst tradind systems (automated or non-automated) and we could try and see if and how we could try and turn it around and make it work for us.
It depends on some other factors as well, including dealing-costs.
“Paradoxically”(? not really), it’s easily possible for both “the system” and “the inverse of the system” to lose money consistently.
On the other hand, it’s also possible for the inverse of some simple systems, and perhaps especially those based on moving average crossovers, to be profitable, at least in some markets and at some times. Linda Bradford Raschke and her co-author have discussed this, to some extent, in their book “[I]Street Smarts[/I]” (which also includes their “Turtle Soup” method based on fading the entries the original turtles were using, since the original system of the Turtles now loses so consistently. I once backtested their “Turtle soup” method, though, over a couple of years’ more recent data, and I found that that lost pretty consistently, too!).
Looking at the info on Raschke’s Turtle Soup system here. It is very easy to see why her system did not produce the exact opposite results of the turtle system over the same period in the same market. They are NOT actually opposite systems.
The entries in the two systems are at the same level and in opposite directions (this is why it would be said they are opposite systems). But the exits are different. If a long position was taken at the break above a periodic high in the Turtle system, the initial stop loss is an ATR stop and the take profit is the bottom of exit channel (which is groomed in the direction of the trade until it is taken out). The opposite system would go short at that break above the periodic high and have an initial take profit where the ATR stop loss is in the Turtle system. It would also have a take profit at the bottom of the Turtle exit channel which would get groomed against the trade until taken out. There would be no stop loss because there is no take profit in the Turtle system.
While there could be periods wherein that opposite system could profit it should be obvious that any big move against a position would be particularly troublesome as there is no stop loss. I presume this is the reason Raschke did not want to actually trade the opposite system but preferred to create a new system that only faded the entries but included a stop loss. Even that system, not having an open exposure to losses but having added a stop loss should still be concerning because it will not allow trades to win gains in excess of 1R.
Trading the true opposite of a given system should produce the opposite results with only a differential for trading costs. If both systems lose money on the very same data, the amount lost should be limited to trading costs.
