Are Expert Advisors a Good Investment?
Expert advisors, also called EA’s are automated trading systems built in the mql4 language, Metatrader 4 house-made.
They are basically algorithmic trading strategies. To cover the whole topic of algorithmic trading, I wouldn’t have space enough in this small article.
What I want to focus on here is discussing whether a purchase of an EA is a good investment or not.
Basic problems of algorithmic trading systems
Algorithmic trading sometimes seems to good to be true. You let the robot do the work, and can start to live a lazy life, letting profits run automatically without doing anything.
The most deceiving thing about trading algorithms is often in their”to-good-to-be-true” backtests. Ever heard of hindsight bias or overfitting?
When not, this is a topic to dive in before ever buying a trading robot. Bascially a good looking backtest is not hard to create. It’s not better than picking the best performing stocks and saying: I knew it! Only that robots are trimmed to find the best fit of trading parameters, that worked in the past.
2. More and more trades
Automated trading systems in metatrader are mostly only based on price action. The more trades, the more profit? EA’s that try to compete intraday with high-frequency algorithms and highly specialized day traders (like me) have a hard time beating the market.
The reason being, that these algorithms or traders have access to more information than your expert advisor (spread, tape, orderbook, order-flow, etc.) = more informational edge. An expert advisor that makes more than 2 trades per day on average I consider overtrading.
3. High profit factor
A high profit factor (see this article for more details) is considered a good way to measure a trading system’s performance.
One thing that you will eventually do is to choose systems with a high profit factor. Judging a system based only on this number, however, is not a good idea either.
Also important are modelling quality, spread, relative drawdown% and the number of trades. So, to make sure that an EA is a good fit for trading your money, check the following things when backtesting the demo in MT4/5:
- Profit factor is reasonable 1.2-1.6 is good (not too high)
- Modelling quality is 90% or higher
- Spread reflects reality (check your broker and add average slippage)
- Relative drawdown% is not over 30%
- Trade count is significant (over ~100)
- Bot doesn’t overtrade (not more than ~3 trades per day)
Testing it first
The main pitfall for EA’s to avoid is deceiving yourself from the backtest. How to not get deceived? It’s easy. Just choose your favourit expert advisors, write down the current date and wait for 3-6 months. Then backtest them again with the same settings over the time that you were waiting and make sure that the EA didn’t change dramatically.
If the results align with the long-term backtest than consider it a possible good system to trade. I know it’s hard to wait and avoid jumping into a trading system, but it’s the only way to avoid overfitting/hindsight bias.
Scalping and martingale
These two words should make you highly cautious. Scalping systems are shining with their high winrate and equity curves touching the sky. However, the risk/reward is often so bad that the one trade loss or losing strike that you will have running the system, will result in much larger losses than you can see on the backtest.
Martingale means adding to loosing trades. This is generally a bad behavior in trading. Not a single experienced trader will take you seriously if you average down into losers without a reasonable limit. It might look good in the backtest (again hindisight bias) but it will ultimately ruin a trading account in the long run (and long is already one year).
Make sure that when you see the results of a martingale or scalping system, that you are aware of the risks, check if the expert advisors use stoplosses and make some simulations with my Risk Simulator to know what likelihood of blow-up the system has.
Consider not only forex
MT4 is known for forex trading, but there are also good expert advisors for indices, commodity and bond CFD’s available. Trading only forex is by no means good for diversification. Open your mind to other asset classes.
Seperating wheat from chaff
There are thousands of “Experts” available in the Metatrader 4 marketplace and on other websites.
All having fancy pictures and names, such as “Forex Combo System” or “Smart Scalper” and good looking charts and backtests.
However, what is a good way to find systems that you like? There is no way around looking at many of them, find and try out 100s of system that meet your criteria (the one’s above can be your starting point) and then forward test them exactly how described earlier.
Diversification and monitoring systems
With algorithmic trading, I think it’s not the best to try and bet only on one horse.
Moreover, it is crucial to monitor your systems and have a clear exit strategy, meaning that you turn off your system after certain criterias are hit (drawdown%, loosing strike, winning strike, percentage goal, etc).
A good way to learn about how to do all of this is by reading Kevin Davey’s book “Building Winning Algorihmic Trading Systems.”
This book will get you familiar with the general concepts of algorithmic trading and also teaches you how to monitor and diversify algorithmic trading systems.
A passive investment
Expert advisors are a passive investment. They will not make you rich quick but can help you reach your goal of making reasonable amounts of money as a side business. Expert advisors seem suspicious and deceiving at first, however delving into the topic a bit deeper before purchasing an expert advisor and applying common sense can lead you to having a good passive investment for your future.