More than ever before the trades on the financial markets are performed by automated trading robots via trading algorithms. Some very simple, some very complex based on deep learning and/or artificial intelligence. Latency arbitrage is the most used strategy and currently approximately 50% of all high frequency trades on a single exchange are won by 3 major parties.
Many people around the world, developer or amateur, are trying to copy these methods and results via numerous strategies. However, many rely on back testing, which is often based on system optimization or incorrect data and leads to unfortunate results. For example, many people trade 1 min, 5 min or 15 min charts with standard indicators tens or hundreds of trades per day.
More trades, is more money right? Correct, more money lost often…
An indicator or candle pattern often has a very short term effect on the price of the asset. With every second passed the effect becomes less. Especially at short time frames. Where the average trade cost (especially in the retail sector) are between 5 and 10 USD per trade, you don’t have to be a rocket scientist to figure out that your hit rate must be above 60%. Do the math yourself! Not a lot of trading algo’s are able to do that. And if they do, there are not a lot of opportunities, or they do not work in all market types and need manual adjustments. Trading algo’s need maintenance, just as cars do.
Unless you are among the top 10 trading algo deep learning developers in the world, the odds are against you. Why not rely on proven statistics to select top performing stocks and avoid the light speed rat race?