To win at forex, it ultimately comes down to your net position and final close price. So, every time we enter into an existing position, we modify our net position in hope that our final position will be in profit…. but does it really work?
For the longest time, I believed that scaling in or out of positions is completely useless.
You see, the success of scaling ultimately depends on the probability of the price move. If we are to assume that the price can either move up or move down with a 50% chance, then I believe scaling is of no use at all. It simply does not improve your expectancy at all as each trade is mutually exclusive.
However, if for some reason the probability is not 50/50, then the game changes. Reversion strategies typically leverage this. Reversion strategies believe that the probability of reverting back to mean increases the further away we get from the mean, this means (excuse the pun) at times we do have a higher than 50% probability. If we have a higher than 50% chance, then perhaps scaling in to a losing position can offer some benefits as it will shift our net position in a favourable direction.
For those interested, there is a really good thread on ForexFactory about Expectancy Management that is well worth a read: http://www.forexfactory.com/showthread.php?t=411923
So, going back to Reversion strategies and Grids…. a lot of them employ a Martingale style system and I have been doing some analysis into the effect of a Martingale style betting system.
For those that don’t know it, it originated from the idea of a double-down bet, i.e., double your bet each time you lose such that when you win, you win everything you lost back plus your original bet. It pretty much is a sure-way to win, if you have an unlimited bank roll. Of course no one has an unlimited bank roll but it just seems so tempting because at the end of the day you only need to win one game.
However, I was interested in how our net position changes as we go enter into our positions, so I whipped together a spreadsheet to examine this.
Above are three tables that illustrate how our net position changes, distance to break-even, and PnL changes for different re-entry strategies. All of them are for 10-pip grids, I.e. the distance between re-entries is 10 pips. The current price in each table is the price at which we break-even at (the net price) for the 10th trade.
The first table is your standard Martingale. As we can see, the amounts double each time. When I created this table I was shocked to discover that the distance to break-even is increasing! So, our profit target is getting further and further away. I would have thought this was somewhat undesirable given we are increasing our bet size.
So, I produced the second and third tables. The second table shows a constant profit target of 5 pips (half the size of the grid). The third table shows a constant profit target of 3.33 pips (1/3rd the size of the grid). These tables show how quickly our amounts need to increase to maintain our desired profit target – which we will assume is proportional to our probability of a win.
I think there are two main things to learn from this analysis
– the amount we need to bet after the 5th turn or so is getting too risky. A change in strategy may be ideal, perhaps widening of the grid or a profit target further away.
– while the break-even target for the martingale was increasing, it also increases less and less as we go along and starts to approach the grid size. This may mean that there is not much to gain by continuing in a martingale fashion after a few turns, and perhaps a change in strategy is again required.
– clearly more analysis is required on the benefit of the lower profit targets and the probability of winning.
Looking at the above, I think it is suicide to continue to keep increasing the bet. The strategy only works if we never have to close in a loss. However, our bank accounts are not big enough and we are guaranteed to get margin called. Perhaps it is safer to attempt martingale only until a certain amount, and then perhaps stop entering and hope the market to retraces back?