Algorithmic Trading: The New World of TradingGeneral Trading September 13, 2016
In today's world, where everything has gone from superfast to hyperfast, I believe there is nobody who needs an introduction to Algorithmic trading or Algo trading (AT). But for the sake of newbies - AT has taken the trading world by storm. It started long back but the retail and hedge fund guys went 'all in' on this technology sometime in the last decade. And as always retail guys are the last to catch up, not to forget the farmers. Basic AT is very simple, it involves using two or more indicators in combination to generate a signal. And there are more indicators in this world than are traders in this world. The problem with indicators is none of them are always correct, not even a combination of them. In Fact, in my 9 year of trading experience, very rarely I found that a type of indicator is correct more than 50% of the time. The reason is very simple - when a certain type of indicator starts working, more market participants start using it, and once it reaches a stage where people start front running it to beat the competition, it starts to fail. How many times have you seen, heard or read about an video/article, which states that something is working in the market and when you try, it immediately stops working. It is because many people start using that indicator along with us and push it fast to its expiry date. After working with and creating many algorithms, I realised that there are no good or bad algorithms. It is just that you need to know when not to trade. For example, let us say that I have an algorithm that is profitable 40% of the time, then it does not that is loss making for 60% of the time. It means that you are forcing it to trade in the wrong market for 60% of the time. To make such a strategy profitable you need to evaluate the strengths and weaknesses of your algorithm. It is something you get after failing thousands of times. Once you do know in what type of market your strategy does not work, you just have to not run it on such days. I know what you are thinking, "How in the world will I ever know if it is going to lose today?". That is where Predictive Trading (PT) comes into play. First of all, you need to have a lot of backtesting data, and the ability to manually observe the patterns where the algo is losing. For example, if the algo is making loss on range bound days then it means you have a momentum trader or a breakout trader in your hands. When you know this all you have to do is avoid when the market tends to constrict. In the same way you can predict when the market tends to be range bound, and when the bulls and bears are equal in strength. This is where experts like me come into play, we can help you understand where to leave and how to leave the market. Remember - A great trader does not trade all the time, but only when he can make profit.