What Happens When We Remove Persistfolio's Sector Limits?
Published August 2, 2026 · Updated September 7, 2026
Persistfolio does not simply take the highest-ranked companies and put them into a portfolio. It also limits how much of a portfolio can come from a single sector. The reason is straightforward: sometimes many of the companies that score well in our research come from the same part of the market. Without a limit, a portfolio can become heavily concentrated without that concentration being an intentional investment decision.
But there is an obvious question: does that restriction actually help? By preventing the portfolio from owning more of its highest-ranked companies, perhaps we are giving up returns simply for the sake of diversification. We decided to test it.
Removing one rule
We used the corrected historical experiment from our 3-year research horizon with a 10-company portfolio. We then created a second version in which we removed the sector limit. Everything else remained the same. Both portfolios used the same Persistfolio research, the same historical information available at the time, the same ranking of companies and the same portfolio review dates. The only difference was that one portfolio had Persistfolio's sector limit and the other did not.
This is important because it allows us to test the portfolio-construction rule rather than changing several things at once. If a large number of highly ranked companies came from one sector, the unconstrained portfolio could take them. The constrained portfolio could not.
The two portfolios were different on 30 of the 41 review dates, so the sector limit was not merely a rule sitting in the background. It regularly affected which companies made it into the portfolio.
What happened?
The constrained portfolio produced the better historical result.
| With sector limits | Without sector limits | |
|---|---|---|
| Cumulative return | +355.06% | +308.85% |
| Annualized return | 16.36% | 15.12% |
| Annualized volatility | 20.15% | 20.30% |
| Maximum drawdown | −31.6% | −35.1% |
| Quarters beating the S&P 500 | 57.5% | 52.5% |
| Average holdings changed per quarter | 1.52 | 1.57 |
We expected the sector limit to reduce concentration. What was less obvious was whether that diversification would come at the expense of returns. In this experiment, it didn't. The constrained portfolio returned 355.06% compared with 308.85% for the unconstrained portfolio, while also experiencing a smaller maximum drawdown and slightly lower volatility. Turnover was almost identical.
The difference in concentration was substantial. Without the sector limit, the largest sector typically accounted for 4 of the portfolio's 10 companies. At its most concentrated point, 7 of the 10 companies came from a single sector. That means a portfolio created from individually attractive research results could nevertheless end up making a very large bet on one area of the market.
Why this matters
Stock selection and portfolio construction are not quite the same problem. A research process can identify companies that look attractive individually, but simply putting the highest-ranked names together does not necessarily produce the best portfolio. If many of those companies are exposed to similar economic forces, what looks like ten separate investments may contain considerably less diversification than the number of holdings suggests.
That is what makes this result interesting. Persistfolio's sector limit did what it was intended to do — it reduced concentration — but in this historical test it did so without sacrificing performance. The constrained portfolio actually produced the higher return, shallower drawdown and better rate of outperforming the S&P 500 from quarter to quarter.
We should not read more into one experiment than it can tell us. This test used one research horizon, one portfolio size and one historical period. It does not prove that sector limits will always improve returns, nor does it prove that Persistfolio's particular limit is the only or optimal way to manage concentration. A different period could produce a different result.
What it does tell us is more specific. When we removed Persistfolio's sector limit and changed nothing else, the portfolio became substantially more concentrated and its historical results became worse, not better. For this experiment, diversification was not simply a defensive compromise. It was a useful part of portfolio construction.
These results are hypothetical historical research results based on price returns. They do not include dividends, fees, taxes, trading costs or other implementation costs. Historical performance does not guarantee future results. Persistfolio publishes quantitative investment research and model portfolios for informational purposes and does not provide individualized investment advice.
Learn more about Persistfolio: