TradeSmith’s Biggest-Ever Breakthrough Is 43 Years in the Making VIEW IN BROWSER By Michael Salvatore, Editor, TradeSmith Daily In This Digest: - What TradeSmith has in common with an all-time-great hedge fund
- How we’re applying machine learning to our most powerful trading software
- The five best AI trades in a nonstop cycle
- Why this melt-up is the perfect time to use it
- The team’s in D.C. – ask them anything
The nerdiest of market nerds instantly recognize the name RenTech… Started by math genius Jim Simons, Renaissance Technologies is the most successful hedge fund of all time. Its flagship Medallion fund returned 66% annually before fees from 1988 to 2021. That didn’t just beat the S&P 500’s long run return of 10%, it crushed it by more than six to one. For an everyday investor, returns like this might seem impossible. But for us here at TradeSmith, it’s just another milestone to hit. Our unending goal is to bring hedge-fund level performance to our subscribers using data and software. The first step to doing this is always to study the market’s biggest success stories. Because if we’re going to level the playing field between elite hedge funds and you, we have to know what those elites do and how they do it. Recently, we asked ourselves the question of whether we could apply the same technique Simons used to help our subscribers beat the market. And we found the answer in his revolutionary bet on AI tracing all the way back to 1982… What Simons did has some fascinating parallels to what we do here at TradeSmith today. And as you’ll learn today, it’s directly related our latest project… the most powerful AI trading strategy you’ve ever seen. This strategy has you hold just five positions at a time… and rotate them out when they hit their profit target. Simple – yet, without an ounce of leverage, it returned more than 42,000% from June 2020 through today. If you don’t believe me, I don’t blame you. I hardly believed it myself at first. But I encourage you to read on. Because the groundwork laid by Simons and RenTech 43 years ago doesn’t just make these returns possible. It makes them ripe for the taking. By the end of today’s issue you’ll learn exactly how… and why we’re sharing a sneak peek of this system for free. Recommended Link | | | | AI can now maneuver fighter jets better than actual veteran pilots… Doctors can now use AI to predict the onset of Alzheimer’s 7 years before the symptoms first appear… AI has now been shown to predict lung cancer up to three years before it even shows up… So it’s no surprise that AI can now identify the timeframe stocks are set to soar… And even how much a stock could return within that timeframe. How is this possible? Click here for all the information. |  | |
When Simons started RenTech, he did the unexpected… Simons had a long career before he started managing money. He got his Ph.D. from the University of California, Berkeley at the age of 23. After, he worked with the U.S. government on cryptography and intelligence. He lost this job for publicly opposing the Vietnam War in the late 1960s, which brought him to Stony Brook University. There, he built the foremost geometry center in the country… and developed the earliest versions of high-level physics ideas like string theory. But in the 1970s, Simons got an appetite for the markets. And he decided to apply his mathematics background to trading. So he started Renaissance. But he didn’t do it the way most people would. He started the fund with a team of top scientists from outside the financial world – mathematicians, physicists, and computer programmers. That’s because RenTech wanted to take a radically different approach to trading the markets. One that involved a bunch of smart people who didn’t have a traditional finance background. Setting aside everything else about markets, the most important thing Simons wanted to find was the patterns in data that humans couldn’t see. And he wanted to build systems with the speed and accuracy to find these patterns and trade them. To do that, he used one of the earliest machine learning models… Machine learning is an early form of AI that’s still important today. It uses raw computing power to look at vast amounts of data and detect patterns that would take a human analyst years to find. (If not a lifetime.) This allowed RenTech to scan the market for opportunities that would last days, hours, minutes and even seconds. That pace is a bit much for us… but it let the fund quickly profit on those opportunities thousands of times over. Every trade reinforced the model. And as the quantity and quality of the data within these models grew, so did the success of the fund. And RenTech constantly updated its models as markets evolved over the fund’s multi-decade reign. There’s an important factor in this that is difficult, but critically important to wrap your head around. It didn’t matter why any given pattern worked – only that it did work. To RenTech and Simons, the market was an imperfect machine that one could exploit to make money… When you’re trading so actively, companies are just tickers, prices are entry and exit levels, and everything else is useless noise. We agreed with Simons, and we thought we could do even better… For the past several months, our team of data scientists, mathematicians, and computer programmers have been hard at work perfecting a machine learning model of our own. Dedicated readers are already familiar with it. It’s the machine learning model I introduced several months back as a way to trade stocks in the tech-heavy Nasdaq 100 index. This model selects for the five stocks with the greatest potential to rise. That’s the only criteria the model has. It’s not concerned with the companies’ business fundamentals or even their risk profile. And, according to our backtest – it works. The early versions of the model outperformed the Nasdaq 100 four to one from 2018 through today. We’ve since updated the model to better balance risk and reward. Not everyone is willing to ride out large drawdowns. We also designed it to apply to any group of stocks we trained it on. It can optimize the Nasdaq 100, the best-performing market benchmark of all time. And it can do the same with the S&P 500 – the “retirement account” of you, me, and most of the other Americans you know. But that’s just the beginning of the possibilities… Earlier this year at TradeSmith, we debuted the newest version of our AI-powered stock projection model, Predictive Alpha Prime. It pinpoints the best holding period for any stock between one and 21 days and where the stock is most likely to go in that time. It doesn’t matter what the company does, its valuation, or whether the business would survive the next year. Like RenTech, Predictive Alpha trades tickers – not businesses. And just like RenTech, each trade reinforces the model and makes it more accurate over time. After we launched the model, we watched it closely. And we quickly learned that some stocks trade better on Predictive Alpha than others. Despite its reputation for being volatile and unpredictable, Tesla (TSLA) is an example of one of those stocks. When the following pattern appears, its historical target accuracy – how often it’s hit the target in the past – is nearly 94%: 
And right now, Prime sees TSLA 13% higher over the next 18 trading days, ending on Nov. 3. And there are plenty more projections to see. In fact, for a limited time, you can test out Predictive Alpha Prime yourself for free. (Learn how here.) But here’s where the world of machine learning and AI collide… Remember that machine learning model I told you about? And how we can apply it to any given group of stocks? Well, as exciting as the results were, the Nasdaq 100 monthly rotation strategy was just the first test. We’ve also applied this model to our Predictive Alpha Prime projections. You take five of the top Predictive Alpha Prime picks, like what I just showed you with Tesla. That’s historical target accuracy at that moment of 85% or higher. And you hold each of them until that pattern ends. (Their “Prime projection period.”) At the end of the period, you sell the stock. Then you replace that stock with the next stock our machine learning model selects. It’s simple but powerful. How powerful? Let me show you the most exciting chart I’ve ever seen… The orange line below is the price performance of the Nasdaq 100, as measured by the Invesco QQQ ETF (QQQ), from June 2020 through today. QQQ has returned about 365%. The blue line is the monthly rotation strategy we’ve been talking about today and for a few months now. In that same time, this strategy got you more than 481%. And the green line is our new strategy: the AI Super Portfolio. Powered by machine learning and drawing from our most accurate AI models… This one would have ballooned your returns to more than 42,750% over the same stretch of time: 
Take special note of the performance in 2022, where the green line breaks away from the pack… That’s because the Nasdaq 100 and our monthly rotation strategy fell. But the AI Super Portfolio kept climbing. Final note: The chart above is in logarithmic scale – smoothing out the exponential rise in the AI Super Portfolio – because it has to be. If it weren’t, the blue and orange lines would barely show. You couldn’t ask for a better melt-up strategy… As regular readers know, we called for a major melt-up in stocks back in February of this year that started in 2023. In the eight months since then, despite the Liberation Day crash, the Nasdaq 100 is up more than 14%. But the uncomfortable truth of melt-ups is that they inevitably melt down. And this is where this strategy shines. When you’re only trading stocks for a period of 21 days at most, you aren’t as exposed to long-term trend shifts as you would be just buying and holding the market. With this system, you ride the best gains in the market on the way up… and you’re just as selective when the broad market falls. There’s no emotion, no second-guessing, and no hesitation. Just pure systematic trading. That’s how this system was able to continue outperforming in the 2022 bear market. In fact, over a backtest that included the pandemic, the 2022 crash, soaring interest rates, the tariff tantrum, and even two wars, it returned an average annual gain of 374%. The AI Super Portfolio is the culmination of everything we’ve been working on at TradeSmith for the last three years. Just as Simons’ success at RenTech traces back to the first use of machine learning to trade the markets in 1982, so, too, does this strategy trace back to our first launch of Predictive Alpha in early 2023. And our CEO, Keith Kaplan, will show you every step of that research in our can’t-miss launch event next Wednesday, Oct. 15. There, you’ll learn how we discovered this breakthrough technology and how we’re applying it to our most powerful AI algorithms today. You’ll also hear of a way to boost your gains from trading this system even further. And from now until then, you’ll have unlimited access to the Predictive Alpha Prime tool. Use it as you see fit to trade the thousands of stocks we track in our system. Go straight here to sign up and get all the info on our latest breakthrough. To building wealth beyond measure, 
Michael Salvatore
Editor, TradeSmith Daily P.S. This week, I’m in Washington, D.C., with the rest of the TradeSmith crew at the annual Big Ideas conference. Each year, we meet to brainstorm new investment ideas and strategies that we aim to bring to you in 2026. I’m right alongside options expert Mike Burnick… Master of big money flows Jason Bodner… Signal study pioneer Lucas Downey… Sentiment analysis experts Andy and Landon Swan… AI researchers John Jagerson and Wade Hansen… Our chief quantitative strategist Mike Carr… And CEO Keith Kaplan. If you were here, too, what would you ask them? Ask Lucas and Jason about your favorite small cap. Ask Andy and Landon about the social sentiment of an up-and-coming retailer or streaming service. Ask John and Wade about the latest advances in AI. Write me at feedback@TradeSmithDaily.com, and I’ll pass your question along. And I’ll share their responses here with you in a future TradeSmith Daily. | 
Post a Comment
Post a Comment