Algorithmic trading has been commonplace for many years, yet the rise of AI has caused a recent shakeup, with everyone on the investment food chain taking notice.
Retail investors, in particular, are targeted with the most content aimed at persuading them to embrace AI trading bots, promising that automating their stock picks and exit strategies will set them on the path to risk-free wealth. It sounds too good to be true, and of course, it is. The trick is to understand the mathematics behind the claims about trading bot prowess and not to get lured in.
So, let’s quickly talk about the concept of ‘edge’, how this can quickly sour the apparent excellence of AI investing tools, and why you still need to have a hand in managing your portfolio manually.
Misleading Stats
The effectiveness of AI trading bots can often be expressed as a percentage, indicating how often a correct decision is made. So, a bot might have a 90% ‘win’ rate in this context, meaning that nine out of 10 times, the trades it makes pay off positively.
This sounds good on paper, but it lacks key context that could leave retail investors in a tricky position. Think of it like playing slots online and how return to player (RTP) is expressed. If a slot has an RTP rate of 90%, that represents the amount of money that it takes in, which will eventually be paid out over a vast number of spins. In other words, you shouldn’t expect the 90% RTP to hold true in a single session of play.
Similarly, if you encounter a trading bot with a 90% win rate, you have to ask yourself ‘What about that final 10%?’ If nine winning decisions result in very modest gains, but the 10th wrong decision causes a huge loss that wipes out all of your earnings for the day and then some, the ‘90%’ effectiveness of the bot is irrelevant and the attractive Planet7 bonus codes that sold you on choosing that casino in the first place are also no longer relevant.
Edge Cases
Calculating the ‘edge’ helps address many of the aforementioned concerns and prevents retail investors from falling for the AI trading bot hype too heavily.
To calculate the edge (E) for a given tool, you need to know the probability of a win, multiplied by the average profit per winning trade. From this figure, subtract the probability of a loss multiplied by the average losing trade. Negative E arises if the former outweighs the latter.
One reason AI trading tools may suffer from making money-losing decisions is that, despite their pattern-recognition prowess, they tend to be overly biased toward relying on past events to predict current market activity. A little like human traders basing assumptions on past successes without basis in the realities of moment-to-moment trading, bots built around data sets from bygone years aren’t always best positioned to plot out winning trades.
In short, there’s certainly a place for AI trading bots. However, retail investors need to familiarize themselves with the mathematics of edge in order to make use of them with minimal risks. They’re a useful part of a broader trading strategy, not a magic bullet to make every decision perfectly.