

Related articles
 A Simple Price Action Based System for a Basket of AUD Pairs
 Currency pairs: A Look Through the Fractal Dimension
 Exploiting Order Flow for the Discretionary Quant  Part 1
 Exploiting Order Flow for the Discretionary Quant  Part 2
 Simple Mechanical Trend Following in the Forex Market
 Is a Reward to Risk Ratio Inherently Better Than Another?
 Robots Aren’t What They’re Cracked Up To Be
 Creating a Trading System Using Neural Networks
 Function Based Trailing Stop Mechanisms
 The Seven Deadly Sins of Automated Trading
 Exploiting the Volume Profile
 Building Robust FX Trading Systems
 Know Your Currencies
 Automating FX Trading Strategies
 Grammatical evolution
 Identifying an Edge
 Interview with Salvatore Sivieri

Automation, Diversification, and the “Wrong Way” Approach to Forex Trading
Advertising:
For advertising, contact
By pure statistical probability, a large number of those combinations will be profitable, and statistically some of those will also be profitable ‘out of sample’. In fact, it is a statistical certainty that, if you look at enough parameters, some of them will test well, both ‘in’ and ‘out of sample’. However, without any rationale, the resulting systems would not be reliable trading systems, being solely a product of statistical probability.
Therefore, one has to be very careful and appreciate that just because a system works in simulations, it does not mean that one has discovered a robust, or even remotely reliable, trading system  another one of the countless errors the author has paid an expensive price to learn.
It is better to identify even just a small, quantifiable, edge that you understand and which has a sound rationale. To quote Monroe Trout again,
Make sure you have an edge. Know what your edge is... Basically, when you get down to it, to make money, you need to have an edge and employ good money management.’
The Edge Effect
An oftenused ratio for quantifying whether a system is ‘good’ is ‘Profit Factor’ (PF), being the gross profit divided by the gross loss i.e. if the sum of all the profitable trades for a system, over a given period, was $1.1 mio, and the gross losses of all the losing trades was $1.0 mio, the Profit Factor would be 1.1/1 = 1.10
To put this in perspective, to use the roulette analogy: if a player bet on red each time, the player would win on average, 18 out of 38 spins of a wheel (with a double zero table). The house would win 20 times out of 38 (i.e. every nonred slot). For illustrative purposes, let us assume the payout is equal to the odds. This gives the house a PF of 20/18 = 1.111
With a single zero table, the house PF is just 1.055 (19/18).
The important point is that the house edge is a very small one, though still incredibly profitable. Looked at a different way, the odds of the house winning on any single spin of the wheel are only slightly better than evens, being 20/38x100 = 52.63% for a double zero table and 51.35% (19/37x100) for a single zero table.
Even with only that slight edge, as Albert Einstein said,
‘No one can possibly win at roulette, unless he steals money from the table, when the dealer isn’t looking.’
Trading is no different. All a trader has to do, to be consistently profitable, is to identify an edge, and apply good money management. Unfortunately, that is much easier said than done. Just as the casino’s edge is in knowing certain facts, a truly robust trading system, can only be built on known, quantifiable, nonrandom, market behaviour.
If a trading system enters a losing streak, a statistical certainty that it (often) will, it is then possible to identify whether it is just an expected statistical ‘run’, or whether something has changed fundamentally in market behaviour. A casino knows that each of its tables will have many losing ‘runs’ and it also knows that is a statistical certainty. This is where money management plays a vital role. Without understanding its edge, and without being able to quantify it, a casino would not be able to operate.
As we have seen, with an arbitrary trading system and an arbitrary set of parameters, no matter how good the ‘in’ and ‘out of sample’ results are, a system is very unlikely to be robust. Equally importantly, it would be impossible to know if the system had degraded, without understanding the underlying reason why it worked.