Why Quant GT's Returns Have Been So High: An Honest Breakdown

Quant GT's historical average of roughly 58% per year over 8 years came from three structural choices: a concentrated 5-stock portfolio, a momentum and relative-strength model backed by decades of academic evidence, and disciplined monthly rebalancing inside a liquid large-cap universe. None of these is secret or exotic. The same concentration that produced those returns also produced higher volatility and deeper drawdowns than an index would have, and past performance does not predict future results. What follows is the engineering explanation, including the parts that work against you.

Key takeaways

• The ~58% average annual return over 8 years was a historical result, not a forecast, and individual years varied widely around that average.
• A 5-stock portfolio lets a winning position move the whole portfolio in a way a 500-stock index cannot, and it amplifies losses by the same arithmetic.
• The model ranks stocks above $10B market cap by momentum and relative strength, a return pattern documented in academic finance since Jegadeesh and Titman's 1993 study.
• Monthly rebalancing matters because momentum signals decay over months; the model re-ranks and swaps holdings on a fixed schedule with no human override.
• Every pick is timestamped and published on the live track record, so the history can be audited rather than taken on trust.

Where does the return actually come from?

From holding a small number of the strongest large-cap stocks and replacing them when they stop ranking near the top. That is the whole machine. There is no earnings prediction, no macro view, no chart-pattern judgment call.

The engine underneath is the momentum factor, and it is one of the most heavily documented anomalies in finance. In 1993, Jegadeesh and Titman published evidence that stocks which outperformed over the prior 3 to 12 months tended to keep outperforming over the next 3 to 12 months. The result held across decades of data and survived replication in markets the original study never touched. By 1997, Carhart had added momentum to the standard academic factor model alongside market, size, and value, which is roughly the point at which an anomaly stops being a curiosity and becomes furniture. If you want the broader context, the writeup on the most influential quant factors covers where momentum sits among them.

Quant GT harvests this with relative-strength ranking. Each month the model scores every stock in its universe on price momentum relative to the rest of the field, then holds the top of the list. It is not predicting which company will invent something. It is measuring which stocks are already being bought with persistence and riding that persistence while it lasts. The bet is statistical, not visionary, and that distinction is why the strategy can be run by a model instead of an analyst.

A documented factor, though, only explains why the model made money. It does not explain why the number was as large as 58%. That part is concentration.

Why only 5 stocks?

Because diversification dilutes. An index holds hundreds of names precisely so that no single stock matters much, which is the right design if your goal is to match the market. It is the wrong design if your goal is to beat it. When a stock doubles inside the S&P 500 at a 0.2% weight, the index gains 0.2%. When a stock doubles inside a 5-position portfolio at a 20% weight, the portfolio gains 20%. Same stock, same move, hundredfold difference in impact.

A momentum model that correctly identifies strong stocks but spreads the result across 100 holdings will produce returns that look like the index with a slight tilt. Concentrating in 5 names lets the best ideas actually move the needle. That is most of the gap between the historical ~58% average and what a diluted version of the same signal would have shown.

Now the part that has to be said plainly. The arithmetic cuts both ways. A 20% position that falls 30% takes 6% off the portfolio in one stroke. Five-stock portfolios historically swung harder than the index in both directions, and the drawdowns ran deeper. Anyone who tells you concentration boosts returns without boosting risk is selling something. Quant GT's design accepts the volatility as the cost of the return profile, and you should only follow the picks if you can accept it too.

Why does monthly rebalancing matter?

Because momentum decays. The Jegadeesh and Titman effect lives in the 3-to-12-month window. A stock that ranked first in January is often unremarkable by June, and a portfolio that holds yesterday's winners out of loyalty is no longer a momentum portfolio. So at the end of each month the model re-ranks the entire universe, keeps the holdings that still score near the top, and swaps out the ones that don't. Some months that means one change. Some months it means several. The model doesn't care, and that's the point.

The schedule does a second job that gets less attention than the signal itself: it removes the human override. Most discretionary underperformance comes from interference, selling winners early because the gain feels fragile, holding losers because exiting admits the mistake, skipping a pick because the news flow feels scary. A fixed monthly cycle makes none of those moves available. The model publishes, the portfolio updates, and nobody's mood is consulted. There is a longer argument for why this discipline compounds in why consistency is the whole game, but the short version is that the rebalance calendar protected the historical returns as much as the ranking math did.

Why limit the universe to stocks above $10B?

Liquidity, mostly. Stock-picking services have a long, ugly tradition of posting spectacular returns in microcaps, where the act of subscribers buying moves the price and the published entry was never achievable by anyone reading it. Restricting the universe to companies above $10B in market cap closes that door. These are names that trade tens of millions of shares a day with tight spreads. The price the model records is a price a real subscriber could have gotten, and no plausible amount of subscriber buying nudges Microsoft.

It also means the historical record is honest in a way that's hard to fake. A 58% average built on thinly traded small caps would deserve heavy skepticism. The same figure built on large, liquid, widely covered stocks is at least measuring something real.

What's the catch?

Concentrated momentum has hard years, and the literature says so as loudly as it documents the premium. Momentum is prone to sharp reversals: the best-known case came in 2009, when the stocks that had fallen hardest in the crash snapped back violently and momentum portfolios, positioned in what had recently been working, got run over. Researchers call these momentum crashes, and they are a recurring feature of the factor, not a one-off. A strategy like Quant GT's should be expected to have stretches that feel terrible while they're happening.

The 58% is an average, not a typical year. Averages over 8 years smooth over the variance that you would actually live through month to month. Rather than quote selective drawdown numbers here, the better move is to look at the live track record, where the full month-by-month pick history is published, bad stretches included. Read the worst runs, not just the headline, and ask whether you would have stayed in the strategy through them. If the honest answer is no, the historical average was never available to you, because the return belonged only to people who held through the drawdowns.

And the standard caveat is not boilerplate here, it is the operative sentence: these results were historical, investing carries risk of loss, and a model that averaged 58% over 8 years made no claim about year 9.

Why should you believe any of this?

You shouldn't have to. Every pick the model has made is timestamped and public, with entry prices recorded at publication, and the current model portfolio is visible the day it goes live. That structure exists so the track record can be audited rather than trusted. Most services ask you to believe a number on a landing page. Quant GT's argument is narrower and easier to check: here is every pick, here is when it was published, here is what happened next. Go count.

FAQ

Why have Quant GT's returns been so high?

The historical average of roughly 58% per year over 8 years came from concentration (5 stocks instead of hundreds), a momentum and relative-strength model documented in academic finance since 1993, and disciplined monthly rebalancing in liquid large-cap stocks. Concentration amplified both the gains and the volatility.

Will I get the same returns going forward?

Nobody can promise that, and Quant GT doesn't. The ~58% figure is a historical average over 8 years, not a forecast. Momentum strategies have documented losing stretches, and investing carries risk of loss, including loss of principal.

Is a 5-stock portfolio riskier than an index fund?

Yes. Five positions at roughly 20% weight each mean a single losing stock hurts far more than it would inside a 500-stock index, so the portfolio historically swung harder in both directions than the market.

How can I verify Quant GT's track record?

Every monthly pick is timestamped and published on the live performance dashboard, including entry prices and outcomes. You can audit the full history yourself instead of taking the headline number on trust.