Why 85% of Investors Lose Money and How AI Models Fix It

Andrew@CANSLIM RESEARCH

Some people say you can never become a “Market Wizard” unless you have lost a lot of money first. This is partly true, but it is best if you don’t have to go through this painful process to succeed. According to US stock exchange data, out of 10 people who buy stocks, 8.5 lose money and only 1.5 win. Among those 1.5 winners, very few can win consistently.

The reason is simple: the stock market is a game that goes against human nature. Why do so many investors love to follow the crowd? Because humans are social animals. We are used to foraging in groups and feel safer doing things together. Even when they lose money, they comfort themselves by saying, “Everyone else is losing too.” But in the stock market, herd mentality is extremely dangerous. Over the past 50 years, countless Market Wizards learned this the hard way after massive losses. In their later books, they stopped talking about trading skills and focused entirely on psychology and risk management.

It is very hard to overcome human nature. When a stock goes up, retail investors get FOMO (Fear of Missing Out). They feel they bought too little or sold too early, so they blindly chase high prices. When a stock falls, they regret not cutting their losses sooner. These trapped investors create the “resistance levels” we see on stock charts. Many people call me a technical analyst, but I am not. When I look at the price action and trading volume on a chart, all I see is psychology—greed, fear, despair, and anger—repeating over and over again. As the late Jesse Livermore said: there is nothing new under the sun.

Having worked in Public Relations and Investor Relations (IR), I understand how the market really operates and how insider trading happens. I can say with 99% certainty that by the time retail investors hear about good “fundamentals,” the big players are already dumping their shares. Take a recent example: Before the Clarity Act was even voted on, the market dropped 8% in pre-market trading (because insiders already knew the outcome). After the vote, the crypto sector crashed even harder. Yet, to this day, countless retail investors still blindly believe in a stock’s so-called “fundamentals.”

Over the past month, I coded 5 large quantitative trading models. During the recent Nasdaq bloodbath, one of my models only dropped by 0.3%, and another dropped by just 1%. While the SPY and QQQ fluctuated by around 2%, many individual stocks were slaughtered. Some high-beta stocks even crashed 30% to 40% in a single month. Among my models, two have already outperformed the Nasdaq and S&P 500 by nearly 1.5%.

To execute this kind of “anti-human” strategy, you must accept that you might lose on 90 out of 100 trades. Very few people can overcome this mental hurdle, which is exactly why an emotionless AI is much better at it. The theory is actually very simple, and many people understand it, but they just can’t execute strict risk control. They fail to let small wins and small losses cancel each other out, and they fail to “let profits run.” Based on the law of large numbers, as long as you strictly manage your risk, your overall account will naturally keep growing.

The models I developed are based on the proven strategies of successful traders over the past half-century. They are completely different from the “AI agents” pushed by brokers (whose only goal is to make you trade non-stop so they can earn fees). I believe that in 3 months, the real-world results of these models will speak for themselves.