It is common knowledge that the best traders have been people who trade without any emotions and execute setups, just like a robot would. With new technologies this has become possible and mundane tasks can be automated, but algorithmic trading also has unclear ethical implications.
Between the 2000 dot.com boom and bust and the “Big Short” of the 2008 credit crunch, we heeded the wisdom of the legendary traders of the 20th century – guys like Jesse Livermore (1920s) and Paul Tudor Jones (1970s). The wisdom was to be dispassionate (no emotion) and trade methodically (like a robot) – executing trades mechanistically every time.
Of course, for the previous generation of traders, the rise of the PC and internet and exchange APIs meant that much of this “mechanistic” work could be automated. Most of the technical and methodical signals can be algorithmically determined and acted upon. The trader is merely the strategist and commander who decides when which algorithm (trading method) should be applied.
What is algorithmic trading
Whether we’ve achieved “AI” is debatable. The trader automates his boring and mechanistic method – that’s closer to the truth. Also, does the term “high volume” imply “high frequency”? This is only possible within a few kilometers radius of the exchange being traders because true HFT operates in the nanosecond timeframe. In saying all that and with regards to the question “whether these algorithms are designed to make the market function better or benefit only a select few (namely the developers)” – One belief is that markets are driven by a subconscious collective consciousness that sometimes wants to rally; sometimes wants to crash and burn, and 60% of the time wastes time deciding what to do.
So algorithmic trade merely reflects the predominant mood of the moment – the fact is that it is humans who are telling machines whether to be aggressive (optimistic positive mood) or cautious (conservative negative mood) in the market. Algorithmic trading can handle high-volume trade and offsetting human biases that may negatively impact traders and financial markets as a whole. It may seem that there are human biases that “negatively impact” traders and markets but in another way of looking this is simply natural behavior: there is always a bigger fish in the deep.
Professional traders and bots
What the retail investors (small fish) are enjoying together now soon becomes a feeding frenzy for larger predator fish. And sometimes the small fish (or medium-sized tuna) are unaware that they are being “herded” by bubble curtains and rumors emanating from the deep. Other times, even the apex predators fall into a trap of their own making. Markets and human psychology are highly complex. They are both tragic and comedic at the same time.
The trader’s job is to take advantage of the moments of the highest probability of a profitable outcome. But that does not circumvent human folly when either the retail collective or one single whale does something that is doomed to fail. In General, bots can help this process by analyzing more setups, taking more decisions, and managing more positions comparing to humans. But this does not mean that the taken decisions have a higher EV than the ones taken by a human with the same conditions and data.
Reasons for concern?
Ethically it remains to be discussed whether there could be reasons for concern: brokers, markets, exchanges all provide public APIs, thus enables everybody to develop automatic tools and platforms, the ones who do it spend and risk money and time for that purpose, then they have an advantage. Today this advantage is just in terms of quantity and not of the quality of decisions against humans.
AI is not yet good enough to beat humans in trading, because trading is far more complicated than other fields where this happened (see chess, backgammon, AlphaGo, Texas Holdem). Whenever AI will beat humans in trading, trading existence itself could be at risk and the first AI agent be able to do such a thing will be the undefeated king of all markets and this will break the meaning of the markets itself.
Thus, algorithmic trading is useful for performing mundane tasks. But these mundane tasks require a lot of computing power and high availability – hence the need for a platform like MachinaTrader. As for a silver bullet in the market, it probably does not exists yet, although most people out there do believe this – and this is why they will come.
*Originally posted at CVJ.CH