Automated Trading - Introduction

In this section I would like to start the discussion surrounding the topic of automated trading.

This is a large subject and there is plenty of information available either via the internet or published books, however,  much of that information is conflicting and there are many pitfalls along the way for the unsuspecting trader.

The idea that automating your trading will solve all your problems, is not as straightforward as it seems, on the face of it taking away the emotion from your trading is one the best decisions a trader can usually make.

However by removing some issues the trader inevitably replaces them with others and one of the biggest mistakes automated traders can make is by under estimating their trading costs.

Automated Trading – Trading and Costs

Please take a good look at the following results, from a simple trading system I threw up and ran in MT4 using an FXCM Demo account.

Sharp eyed traders will immediately spot the lousy tick modelling percentage of just 25%,  but for this test and example that doesn’t matter, as in the following paragraphs I will be comparing like for like and the method in question buys and sells only at the start of the next bar.

On the face of the results, taken over a ten year period on the four hour time frame of the USDJPY are at Ok, they are a starting point and need adjustment.

Looking at the draw-down which is a lot more than I would personally accept during my own trading, this system is untradeable as it stands.

Be that as it may, what I want to draw to people’s attention is that the back test which I used to run the test used tick data from DUKASCOPY and a variable spread and slippage.

The next test I ran was using the same system but with a  fixed spread of 30 or 3 pips. Please take a look at the difference between the two results and understand that the second test using the 3 pips to include spread and slippage so a total of 6 pips round turn is far more representative of the real world and is my standard spread/slippage for the Euro/USD contract.

My first point then is for traders to understand that anyone who shows you results or worse still tries to sell you a  system which is based on zero spread and slippage is at best spinning you a yarn and at worst pulling the wool over your eyes. Systems do not trade at zero trading costs.

When you develop systems please use a realistic spread and slippage value during your testing so that you do not deceive yourself about the likelihood of your systems success.