Most gurus in this space came from Wall Street. Keith Kaplan came from a server room.

He is a software architect who became a CEO. He runs the Financial Technology division at MarketWise (Nasdaq: MKTW), a publicly traded company with 2.6 million subscribers. His product is TradeSmith, a suite of investing tools. His team’s crown jewel is An-E, an AI model that forecasts stock prices 21 trading days ahead — to the penny.

This is not a story about a stock picker. It is a story about a builder.

From Towson to the C-Suite

Kaplan grew up in Maryland. He earned a BS in Computer Information Systems from Towson University in 2001. Then he spent 14 years building enterprise software at RWD Technologies and ANCILE Solutions. He led cloud engineering, DevOps, and advanced technical solutions teams. He managed large-scale infrastructure, security policy, and SaaS platforms.

In 2015, he joined Stansberry Research as Director of Cloud Engineering. That was his entry point into the financial publishing world. He later became Director of Operations Strategy. In 2017, he moved to TradeSmith as President and Chief Product Officer. By 2019, he was CEO.

Kaplan brought a software engineer’s mindset to a financial publishing company. Instead of writing newsletters, he started building platforms.

The MarketWise Machine

MarketWise is the parent company. It is a multi-brand subscription platform offering financial research, software, and education. As of Q1 2026, it had 381,000 paid subscribers and 2 million active free subscribers. The company reported $328.1 million in net revenue for fiscal year 2025. Q2 2026 billings hit roughly $91 million — a 56% year-over-year increase and the highest quarterly figure since 2023 (MarketWise earnings release, July 9, 2026).

The company owns 11 primary brands. Stansberry Research, InvestorPlace, TradeSmith, and Oxford Club are among them. Each operates independently. Each targets a different investing style.

Kaplan’s division handles the technology side. He builds the tools that the other brands sell.

TradeSmith: From Trailing Stops to AI

TradeSmith started in 2005 with a single product called TradeStops. It was a simple program that tracked fixed-percentage trailing stops. The idea was practical: help investors know when to sell.

Kaplan joined the company around 2017. At that point, TradeStops was still the only product. Customers kept asking for the same thing: “Can you tell us when to buy, too?”

That question launched a multi-year software build. Kaplan’s team built massive databases, wrote new algorithms, and created what eventually became Options360, Ideas by TradeSmith, and the Trade Cycles seasonal screener.

Today, TradeSmith offers over a dozen products. TradeStops remains the flagship portfolio tracker. Predictive Alpha is the AI-powered forecasting engine. Options360 turns stock forecasts into options strategies. Trade Cycles uses seasonal patterns to time entries and exits. The AI Super Portfolio is a five-position automated portfolio. TradeSmith Platinum bundles everything into one elite membership.

More than 134,000 users rely on TradeSmith tools to track over $29 billion in portfolio assets (MarketWise, 2026).

Building An-E

An-E stands for “Analytical Engine.” The nickname is a nod to Charles Babbage’s 19th-century mechanical computer, the Difference Engine. But the modern version runs on something far more powerful.

Kaplan’s team built An-E over 26 years. That timeline includes the foundational research, the data accumulation, and the iterative model training. They spent tens of millions of dollars. They logged tens of thousands of man-hours.

The team behind it sounds like a sci-fi cast: 69 engineers, 22 machine learning and AI experts, 23 quants, hundreds of analysts, former Wall Street traders, PhDs, a retired lieutenant colonel who worked on top-secret nuclear missions, a molecular geneticist, and an eye surgeon.

An-E forecasts stock prices for more than 2,300 stocks, funds, and ETFs. It projects 21 trading days ahead. It outputs a price target to the penny. It also produces a confidence score and a projected range.

The model was trained on 1.3 quadrillion data points and more than 50,000 backtests (InvestorPlace, April 2025). It uses machine learning to identify patterns that human analysts cannot see. It is an ensemble model, meaning multiple AI systems work together and cross-check each other.

Here is the plain-English analogy: An-E is like a weather radar for the stock market. Weather radar does not control the storm. It tells you where the storm is likely to go, how fast, and how strong. You still have to decide whether to take cover. An-E works the same way. It does not make trades. It provides probabilities.

How Predictive Alpha Works

Predictive Alpha is the product name for An-E’s forecasts. Kaplan launched it in 2023. The system runs daily scans on thousands of securities. It ranks them by expected move, trend direction, and historical accuracy.

Subscribers get a Top 5 Watchlist that updates monthly. They can also search any ticker for its 21-day projection. The system includes an Options tab that maps stock forecasts to options strategies. It also produces Top Bullish and Top Bearish lists each trading day.

The accuracy metrics are public. Each stock shows a Historical Directional Accuracy score and a Historical Target Accuracy score. These are not hypothetical. They are the model’s actual track record on that specific stock.

Kaplan demonstrated the system with real examples. In one demo, he showed Applied Digital rising 16% in 6 days. SoFi gained 9% in 3 days. Upstart rose 10% in 1 day. Carvana surged 25% in 2 days. These were not cherry-picked winners. They were live forecasts from the system.

An-E also called Nvidia, Amazon, Meta, and Tesla early. Kaplan’s team identified those stocks before the mainstream AI rally took off.

The Lost Decade Thesis

In November 2025, Kaplan launched a major promotion called “Buy and Hold is Dead.” The hook was a Goldman Sachs and Morgan Stanley forecast predicting a lost decade for U.S. stocks. Goldman’s Peter Oppenheimer projected the S&P 500 would return just 6.5% annualized over the next decade — well below the historical average (Bloomberg, November 2025).

Kaplan’s argument is straightforward. If buy-and-hold returns 3% to 6% annually for the next 10 years, that strategy is dead money. The solution is active trading powered by AI. An-E is designed for exactly that environment — short-term forecasts, defined exit points, and rotation into the strongest names each month.

The logic is not anti-investing. It is anti-passivity. Kaplan is not saying stop investing. He is saying stop holding blindly.

What Sets Kaplan Apart

Kaplan is a genuine engineer. He does not pretend to be a trader who discovered a secret formula. He is a builder who assembled a team and a technology stack to solve a concrete problem: information asymmetry.

Retail investors have always been at a disadvantage. Hedge funds have quantitative models, dedicated data centers, and PhDs. Kaplan’s thesis is that the technology gap is closing. AI models running on consumer hardware can now rival institutional systems. The question is who builds them and who uses them.

He is also a CEO who uses his own product. He sold most of his personal stocks on February 27, 2020 — days before the COVID crash — after seeing TradeSmith’s indicators shift into the Red Zone. He has been trading alongside his subscribers ever since.

The TradeSmith Ecosystem

TradeSmith is not just An-E. The full product suite covers the entire investing lifecycle.

TradeStops handles the exit problem. It syncs with brokerage accounts and sets trailing stops based on each stock’s volatility profile. Ideas by TradeSmith provides seven factor-based strategies for finding stocks. Options360 turns stock forecasts into options trades. Trade Cycles identifies seasonal patterns. The Trade360 bundle combines TradeStops and Ideas by TradeSmith into a single platform.

The newest product is Signals by TradeSmith, launched in April 2026. It scans 2,467 stocks every morning and surfaces high-probability setups more than 90 minutes before the market opens. Kaplan held a live launch event that drew 8,946 attendees.

The Bigger Picture

Kaplan operates at the intersection of two megatrends: the democratization of investing and the commoditization of AI. MarketWise sits on a massive subscriber base. TradeSmith sits on a massive data set. An-E sits on top of both.

The combination is powerful. MarketWise feeds TradeSmith with customer acquisition. TradeSmith feeds An-E with data. An-E feeds back with forecasts that drive product sales. It is a flywheel, not a feature.

The risk is execution. AI models are not magic. They degrade. They overfit. They fail in regime changes. Kaplan’s team has been working on this for 26 years. They have seen market cycles. They have retrained the model. They have released An-E 2.0 and are working on version 3.0. The question is not whether the model works today. It is whether the team can keep it working tomorrow.

Cassandra Close

Kaplan is not the face of a promotion. He is the engineer behind the promotion. He does not give stock picks. He gives a system that generates stock picks. That distinction matters.

The financial publishing industry is full of people who predict. Kaplan is one of the few who build. He built a cloud infrastructure team at Stansberry. He built a product suite at TradeSmith. He built an AI model that forecasts stock prices. He built a company that went public and serves millions.

The lost decade thesis may be wrong. The Goldman Sachs prediction may be too pessimistic. Markets may rip higher for another 10 years. But the strategy Kaplan is selling — short-term, data-driven, AI-powered rotation — does not depend on that thesis being right. It works in bull markets, bear markets, and sideways markets. That is the point.

When the market stops rewarding patience, it starts rewarding precision. Kaplan bet his career on that trade. So far, it is paying off.