The 28-Year Hunt Behind Your Next Trading EdgeVIEW IN BROWSER
BY MIKE CARR, CHIEF QUANTITATIVE STRATEGIST, TRADESMITH In October 1997, I was an Air Force officer stationed in the middle of nowhere in Egypt. I was assisting in a joint training mission called Bright Star with Egyptian forces. I wasn’t a pilot – I was doing computer programming work. And in the downtime, my commander and I kept busy trading markets. Our command tent probably had more electronic capability than all of Cairo. That helped us stay on top of what was happening in U.S. stocks. One morning, the Asian financial crisis that had started in July with a sudden devaluation of the Thai currency spilled into the U.S. market. The S&P 500 lost more than 10% – in the middle of the dot-com boom, no less. So during the worst of the crash, I found myself in the secure conference room with the base general, a do-not-disturb sign on the door, trading for the next 24 hours straight. Something clicked during that marathon session. The fast, sharp dislocations that happen when everyone panics at once – that’s where the real money was made. The problem was, I couldn’t sit around waiting for a market crash every time I wanted to trade. So I started asking what other conditions tend to produce fast, reliable, tradeable moves. That’s when I started hunting for “signals” – repeating patterns in stock market data that reliably point to upcoming moves in stocks.
Thousands of Ideas… Most of Them BadBy the early 2000s, I was trading full-time. I’d watch the market, notice something, and dig in. Maybe a stock was up from where it was 13 days ago. Or its trading range had compressed by a certain degree while interest rates ticked higher. I’d ask: Is there a pattern here? Does it repeat? Can I trade it? Most of the time, the answer was no. I tested thousands of ideas. Maybe two or three times a week, something would survive testing and get added to my collection. The rest went in the trash. An early one that survived had to do with the Relative Strength Index (RSI) – a metric traders use to see if a stock is overbought or oversold. When a shorter-term version of the RSI dropped below 10, and the next two trading days were up and then down in that order, that was a buy signal. A variation of it still works today. After the 2008 Great Financial Crisis, I co-managed mutual funds and high-net-worth accounts worth as much as $200 million. But the regulatory environment made short-term trading nearly impossible for Registered Investment Advisors – regulators would accuse you of churning to generate excess fees. So instead, we used monthly rotation strategies, swapping into the strongest-performing stocks at the end of each month. With the help of this strategy, our fund made 15% from 2008 to 2010 while the S&P 500 lost 15%. But I knew I could do better. That’s what eventually drew me to TradeSmith in 2024. Seven Times the Market’s ReturnI’d known TradeSmith for years – the development team, the data infrastructure, the scale of data they could process. But when I got here, I realized we had even more data than I anticipated and a talented engineering team to put it to work. I proposed a signals project in my first month. The pitch was to take everything I’d learned over that last quarter-century – the rules that worked, the patterns that repeated, the principles that held up across bull and bear markets – and build a system that made it easy for regular investors to follow. What I didn’t anticipate was how much AI would change what we could do with it. I’ve spent more than two decades building and testing signals by hand. The system that came out the other side of this project is better than what I could have built alone. It now evaluates millions of potential trades every day across more than 2,000 stocks, hunting for the specific combinations of factors that have historically preceded big moves. I ran a live internal beta test in January and February of this year – and the top 100 signals generated an average gain of 2.6% in just nine trading days. Over that same stretch, the S&P 500 returned just 0.4%. That’s roughly seven times the market’s return. If you repeat that nine-day gain over the course of a year, you’re talking about a 73% return. And that was from trading signals directly with stocks. When we traded them with options, the results included:
Caterpillar (CAT): 126% in 72 hours Nvidia (NVDA): 129% in 5 days Lockheed Martin (LMT): 365% in 30 days HCA Healthcare (HCA): 461% in 13 days Generac (GNRC): 1,082% in 33 days That’s when the game changed – and we built automated options strategies into our new AI-powered signals trading system. Our Most Advanced Trading Tool EverI don’t write much in these pages. I spend most of my time in the background building software. But our CEO, Keith Kaplan, debuted this system in our launch presentation last Wednesday. You can catch the replay here. I’ve spent the better part of three decades working toward this. And it’s hands down the most advanced trading tool we’ve built. By focusing on signals in the data, it tunes out all the extra noise when it comes to making money in stocks… and focuses only on what matters. Every day, we’re bombarded with apocalyptic forecasts… biased news headlines… thinly veiled propaganda… and an endless stream of useless information on social media. Our new trading technology ignores all of that. Instead, it produces high-probability trade setups based on statistical odds of certain patterns repeating. If you want to see what it looks like in practice, the replay is still up. Watch the AI Signals Trading Event here. Cheers, 
Mike Carr
Chief Quantitative Strategist, TradeSmith P.S. Remember that rotation strategy I used to post returns of 15% for my clients after the 2008 crisis? A version of it is built into our new Signals software – a three-stock model that outperformed the S&P 500 by 3-to-1 over our six-year backtest. Thanks to the magic of compounding, that simple strategy turned every $10,000 into more than $120,000 in our backtesting. Keith walks through it all in the replay. |
0 التعليقات:
إرسال تعليق