As is often the case with anything that is as hyped up as artificial intelligence, people soon assume it's a cure-all for just about everything. The dot-com bubble led to the assumption that any company with the ".com" suffix equated to a golden business idea, the blockchain bubble of 2017 led to the assumption that any company with an ICO was going to make them rich, and now with "AI for X," we assume X will no longer be a game in which mere mortals can contribute or compete.
Newsflash folks: AI has been around for 50 years, and there are various techniques with various tradeoffs and strengths that suit different use cases within the field. Sure, recent advances in hardware and availability of large data sets have led to incredible gains in certain areas such as image recognition (surpassing human ability), but that hardly equates to the idea of eliminating humans in everything, especially in something as complex and dynamic as the markets.
Unless everyone ends up using the same algorithms/techniques with the same parameters, applying AI to trading will have very little visible impact to the markets since it amounts to substituting human decisions with decisions made by machines programmed by humans. In fact, as an individual trader, one should hope that the machines end up applying he same algos, as that would result in high-level market action that is easy to identify as robo-trading (and thus actionable).
Instead, human programmers are biasing their algos to suit their perceptions of the markets--some choosing to parse sentiment from news feeds and trade accordingly, while others looking at correlation data between sectors for arbitrage opportunities. But much like superimposed tones played by different instruments in an orchestra that create a new sound, what emerges is a new, identifiable market "tone" that combines the ones being played. Your task as a human is to look for these new patterns and styles of trading instead of adhering to the old ones (or else you'll be competing against the machines).
I suspect this is why hedge funds have vastly underperformed the markets as of late--they're trying to identify sounds made by individual instruments instead of identifying the sounds that they create in harmony with each other. Instead, you can be in harmony with the new world order of AI (and avoid becoming obsolete) by ignoring the noise polluting traditional expected patterns and characteristics and instead opening your eyes to what the data is now telling you.
No comments:
Post a Comment