How Much Quarterly Turnover Is Normal in a 13F? The Real Distribution Across 95 Funds

How turnover is measured here

This site measures turnover by position count, not dollars. For each filing: (new positions + reduced positions + positions that exited the book) divided by (current positions + exits), then averaged over every quarter we track for that fund. Dollars move less than names: trimming 2% of a mega-cap barely registers in value but is still a REDUCE row. Counting positions captures how much of the book is being re-decided. The fund turnover ranking shows the latest quarter for every tracked fund; the numbers below are the harder, multi-quarter average.

The real distribution (95 funds)

The median fund turns over 55.6% of its position count in an average quarter. The 10th percentile is 41.7% and the 90th percentile is 68.4% — a narrower spread than most people expect. High churn is the norm in 13F data, not the exception, and it is structural: books are rebalanced, pod teams rotate, small positions are culled every quarter.

The patient end

Even Berkshire churns 29% by name count: a decades-old book still carries small legacy positions that occasionally exit, and every exit counts. Patience in 13F terms is not a frozen book — it is a book whose top names barely change. Scan the Berkshire fund page across years and the top ten are old friends; the churn happens in the tail. That is why concentration and turnover read best together.

The busy end

At the top: Maverick (77.3%), Steadfast (75.3%), Whale Rock (75.2%), Scopia (73.1%), Point72 (71.9%), Balyasny (70.5%) — overwhelmingly multi-strategy or pod-style hedge funds, where independent teams each churn their own sleeve. For these managers the quarterly 13F is a parking photograph of hundreds of parallel decisions, not one thesis. Copying their top ten is copying whoever happened to have the largest sleeve that quarter.

What a fund's turnover tells you

• Low turnover + high active share + high concentration = a conviction book. Its NEW positions are events worth reading; its exits are decisions, not noise.
• High turnover makes the snapshot nearly worthless for copying. By the time you see the filing, half the names have changed. Aggregate it up to sector-level flows instead, where single-name churn washes out.
• Turnover changes are signals too. A normally busy fund going quiet — or a patient fund suddenly churning — often marks a regime change: new pods, a redemption, a shift from conviction to trading.
• Turnover explains backtest failure. Our copy-13F backtest shows why fast books cannot be followed: with 45-day-old data you are mimicking positions the manager has likely already traded out of.