It’s no great secret that efficient automation will result in increased profits for traders. Automation is the process of using machines to execute functions depending on certain criteria. It’s done ‘automatically’, but this trading automation has to be:

• Built on intelligent parameters.

• Successfully implemented.

Otherwise, all you are doing is training a machine to execute bad commands. Like all trading, your core assumptions need to be correct in order for your strategy to yield a profit. 

As you are interested in automation tips, you most likely fall into the day trading or HFT category. Automated trading is not as relevant to long-term investors, who prefer a manual approach. The following are 5 automation tips to help out professional traders.

#1. Build Your Own Strategy

There are many off the shelf automation strategies, sometimes referred to as ‘black box’ strategies. But it is easy to change the parameters of any strategy to make it appear profitable. After all, every trading strategy is profitable, at one point or another. You simply change the time horizon to a period where it is profitable. Developers will always tweak the system so that it looks as good as it can. 

It is far better to develop your own strategy and get a coder to develop it. This will increase your knowledge and skill. You will understand the ins and outs and learn far more. If you try to make use of an automated trading strategy, you may not learn as much. Further, there could be a tendency to simply rely on this system to complete all of the work, which is just not how trading works. It could easily lead to complacency. From there, it is a few simple steps to financial ruin. 

Of course, this is not to say that automation strategies do not have their uses – they play a pivotal role in balancing out the volatile emotional nature of the trader with a logical and automated approach. 

#2. Backtest Extensively

When you are designing a system for trading automation, all of the rules need to be precise with no room for interpretation. This will allow a trader to gauge exactly what profit can be expected over a particular time period using specific models. This also limits your risk, so that you won’t lose any money by failing to properly evaluate a trading strategy before implementation. 

What happens to so many traders is that they implement a trading model, which works for a while. Eventually, the strategy starts to lose, and they conclude that it is no longer profitable, for various reasons. But in both the winning and the losing instances, the fact was that they did not understand the risk/reward ratio involved due to a lack of backtesting. 

It is possible to win certain games by luck alone, such as blackjack or roulette. But unless you understand the game very well, you won't win on a long-term basis. Statistics and math do not lie. Backtest your automation strategy extensively before implementation, and understand your asset in its entirety.  

#3. Understand Backtesting Biases

Backtesting is an art in itself. And traders should not underestimate their own biases and the ways that they subconsciously deceive themselves. One issue with backtesting is that you will tweak the results so they are optimal. This is known as ‘curve fitting’. A backtest should always be considered an ideal upper bound. Here are the 3 major biases that can occur:

• Optimization bias – the most common and insidious. You adjust your initial strategy parameters in order to increase its performance after you have seen the results. Once live, the performance will be very different. 

• Forward bias – this occurs where future data is included in the backtest, that would not be available in the real performance of the strategy. Survivorship bias is a variant of a forward bias, where we look at surviving assets and base our data, optimistically, on assets that are already successful. 

• Psychological tolerance bias – looking at a 5 year backtest, it is easy to think you could withstand a 4-month losing period, as the overall picture is profitable. In practice, this is difficult to stomach, when you are directly within this losing period. 

#4. Stick to Your Strategy

An automation strategy is designed based on certain principles and parameters which will win out over time, once it has been backtested intelligently. Despite this, it still has the potential to lose on a short-term basis. This will have a negative impact on the psychology of the trader who has placed more trades than necessary. 

As a result, the trader might decide to tweak the strategy in order to avoid losing money or might stop taking certain positions. This destroys the entire automation strategy. It was built on parameters that you now decide to change and is no longer fully reliable. 

More significantly, if you change once, you could easily change a thousand times. The entire purpose of the automation strategy has been defeated. Your emotions are getting the better of you, and you need to go back to the drawing board. For this reason, it is better to implement your trading strategy for lesser amounts you can afford, so your emotions don’t overtake logic.  

#5. Select Your Software Carefully

You need to choose a good platform that provides the appropriate execution mechanisms. For example, Microsoft Excel might be a good choice for simple trades and could be a good place to start. 

But for more experienced traders, it is just not cut out for complicated algorithms. The software you choose has to be flexible, fast, and easy to implement. Mathlab and Python can be the perfect solution, but for ultra-quick trades, you will need C++ or similar. 

If you are trading cryptocurrencies, then Superorder can tie up a lot of loose ends. It is designed specifically for the automation of simple and complex trades involving cryptocurrencies. Speed, ease of implementation, and flexibility will not be a problem for experienced traders.