Trading with AI: déjà vu!?

I remember the times when the Metatrader4 platform became popular - you could easily and simply create bots there. Perform backtests, optimize - watch how they work, etc.

I remember starting the tests myself - creating simple Moving Average (MA) cross ideas, and so on: when to buy and when to sell based on aggressive moving averages (MA). Until we’ve got to grid trading strategies (which, by the way, are now popular in the crypto trading world), Martingale strategies, and other wonders.

In case you are not familiar with the Martingale strategy, it is a high-risk trading technique where a trader doubles their position size after every loss. As for the grid trading strategies, grid trading is an automated, non-directional strategy that places a series of buy and sell orders at incremental intervals (a "grid") above and below a set price.

So, my point here is that it all evolved very quickly - a lot of forums and new websites appeared daily. Each trying to demonstrate how quickly you can achieve dizzying results...

But, over time, one specific thing stood out very clearly: none of these strategies worked in the long run by themselves. From time to time, improvements were needed... Some of these strategies worked only in the short term, some strategies worked only in the direction of the trend, or were suitable only in flat situations. In other cases, the risks were too high, etc.

Then came the “copy trading era”. Various platforms that offer to copy trades from real, successful traders or copy bots. It was the era that screenshots showing off profits of x10 or strategies written by NASA engineers, which became an instant “classic”. It all picked up a huge pace, and the trend was obvious.  

Backtests – always showed cosmic results, but real-time actions in the market were very lame. Of course, with the caveat or disclaimer that it is technically difficult to copy trades in real time, too large spreads for your broker, etc.

However, despite all of this, huge copy trading platforms, “signal providers,” and other beasts emerged. But the statistics are quite sad. For example, those brokers who use CFD products are forced by regulation to indicate retail client statistics, which is: “Between 74-89% of retail investor accounts lose money when trading CFDs.”! Those brokers who do not use CFD products do not indicate such information. Other service providers – to avoid conflict with the regulator – write rather ambiguous sentences: “Did you know that 73% of our investors make a profit when copying top leaders correctly?” And so on...

Most of the profiles of “signal/copy traders” are not older than a few months. Besides, the rating system in these platforms themselves is designed so that you often change the "signal provider". To jump from strategy to strategy. The ratings are designed to reward the most aggressive strategy with the largest number of trades. Simply because it benefits the service provider. And even if there is a good long-term trader/signal provider in this mess, the rating system does not reward him. Again, simply because it is not profitable. Mostly, platforms make a profit from the number of trades, thus rewarding those who were engaged in aggressive day trading. And on top of that, often signal providers get a "volume commission", which also motivates them to make a large number of trades.

Why am I contemplating about all this when the title clearly suggests trading with AI?

Well, simply because all this current hype around it gives me this feeling of déjà vu. Pardon my French, but in case you don’t know the term, déjà vu in French means a feeling of having already experienced the present situation.

So, here are just a few headlines for you: “This algorithm analyzes millions of data points”, “Backtest results +340% in 18 months”, “Our artificial intelligence system can notice these things before other market participants realize it”. I have seen all this somewhere before… The same thing, just in a different outfit. Of course, the devil is not as black and purely evil as he is painted, but you should not fall for these “golden mines” marketers with their get-rich-quick schemes.

A tool is only as good as the person using it. The problem isn’t in the robots or AI in that matter. The problem was, still is, and most likely will be a lack of market understanding. AI can process data faster than any human, no doubt about that. It can spot correlations that we most likely wouldn’t notice. But it can’t replace the context, judgment, and experience gained from years of observing the irrational behavior of the market – and what is even more important: overcoming it.

Even if you’re fully prepared to trade with AI, where do you start? What questions will you ask? And more importantly, how will you know the answers are worth considering? How will you know that the AI ​​isn’t hallucinating at this particular moment? How will you know that it’s “not lying”? How will you verify the veracity of the data sources it’s using? How will you control the strategy it generates? Will you simply “blindly trust” every trade idea it offers? These are not rhetorical questions. These are the exact questions that separate the person using the tool from the person being used by the tool.

AI is powerful. But it is only powerful in the hands of those who have the basic knowledge. Knowledge of how the markets work. How to read the data. How to recognize when something doesn't add up. How to ask the right questions – because without it, even the best answer is useless. How do you know that the artificial intelligence is using reputable sources? How do you know that it is not hallucinating at a given moment? How will you verify the information provided? What is the chance that you will create a long-term, reliable strategy? How will you control it? What will be the decision-making algorithm - will you simply "blindly trust" the trading idea provided? The hype will reach its peak, eventually. Disappointment will follow, inevitably. And those who actually understand what they are doing will still be relevant: there, on the other side. That is the only advantage that does not end.

If you want to be on the right side of that equation, start with the foundation. My course KNOW before you DO! is built exactly for this: understanding how markets work, how to read data, and how to make decisions you can actually stand behind – with or without AI. And if you want to go deeper, the Personal Helpline is there for one-on-one sessions directly with me. See you on the other side!

Stay tuned!

Agris