Week 43: Active Management — SPIVA, Persistence, and What Very Few Managers Actually Do

---

1. Why This Is Important

Almost every investor, at some point, hears the same pitch: *this fund beat the market last year, this manager has a special process, this strategy has worked for a decade, you should pay 1% a year (or 2-and-20) for it.* The pitch is sometimes true. It is, on the data, almost always false. This lesson is the cold-shower week. Before we look at the rare strategies that do generate alpha, we have to look squarely at the base rate — at how rare manager skill actually is, how poorly past performance predicts future performance, and how the survivorship bias built into every fund advertisement systematically lies to you.

You need this material for four reasons.

The base rate is brutal. *The market is right most of the time,

and alpha is the rare gap* — that is not a slogan, it is a numerical claim. The SPIVA scorecards make it concrete. Over 15 years, roughly 90% of US large-cap equity managers underperform the S&P 500. Default to passive is not a preference, it is what the evidence forces you to do unless you can name a specific reason this manager is in the surviving 10%.

Persistence is worse than random. Even the funds that did beat

the market in one five-year window very rarely repeat. SPIVA's persistence reports show only ~15% of top-quartile funds remain top quartile in the following five years. Random would predict 25%. Past performance is not just not predictive — for the top performers it is mildly anti-predictive, because mean reversion and benchmark drift work against last cycle's heroes.

You will be sold survivorship-biased numbers. When Morningstar

tells you the average large-cap fund returned 9% over the last twenty years, that average only counts funds that still exist. The 30-40% of funds that closed or merged because of poor performance are dropped from the calculation. Once you re-add them, the gap to the index widens further. If you don't know to ask "survivorship-adjusted?" you will systematically overestimate manager skill.

A small handful of strategies actually work. SPIVA does not

say no one beats the market. It says the average active mutual fund does not. Inside that average sits a tail of strategies — deep value activism, quantitative systematic, event-driven, high-touch CTAs — that are documentably persistent over decades. They are not accessible through a Vanguard ticker, they require capital, lockups, or specific expertise, and three of the four are not what most retail investors think of when they hire an "active" manager. The alpha sources Horace himself uses are exactly these structural pockets; this lesson maps where the institutional taxonomy puts them.

This week is the bridge between should I be active or passive? (almost always passive) and the lessons that follow on *if I am going to be active in part of my book, what do I actually have to do?*

---

2. What You Need to Know

2.1 The SPIVA Scorecard — How Bad Is It Really?

S&P Dow Jones publishes the SPIVA (S&P Indices Versus Active) scorecard twice a year. It is the closest thing the industry has to an audited, survivorship-adjusted ledger of active manager performance. The methodology is simple: for each Morningstar fund category, compare the asset-weighted (and equal-weighted) net-of-fee return of every fund that existed at the start of the period to the appropriate S&P benchmark, and report the percentage that underperformed.

The April 2026 scorecard, looking at periods ending December 2024, gives the following picture for US large-cap equity funds:

• 1-year window: ~60% of funds underperform the S&P 500.
• 3-year window: ~75%.
• 5-year window: ~80%.
• 10-year window: ~85%.
• 15-year window: ~90%.

The pattern is not subtle. As the window lengthens, the percentage of underperforming managers monotonically rises. Why? Because over a single year, luck dominates. Over fifteen years, fees compound, the high-fee tail gets dragged below the index by sheer expense, and survivorship-adjusted accounting starts to bite. Pulling 100 bps a year out of an 8% gross return is pulling more than an eighth of the total return — and that is what you are paying the average mutual fund to do to you, not for you.

Other categories tell similar stories. Small-cap is slightly better for active (more inefficient market, ~80% underperform at 15y). International developed is similar. Emerging markets is slightly worse for the index, but adjusted for currency hedging the gap narrows. The one place where active consistently wins is, surprise, the most expensive corner of the market: liquid hedge funds and illiquid private credit, both of which have selection bias and mark-to-model issues that flatter their numbers. We will look at those carefully in later weeks.

2.2 The Persistence Problem

A reasonable response to SPIVA is: "Fine, the average manager is bad. But there must be some persistent winners. I will just pick those." This is what every fund-of-funds, every wirehouse advisor, and most pension consultants do for a living. The data says it does not work either.

S&P also publishes a Persistence Scorecard. Take all US large-cap funds, rank them on five-year performance, and put the top quartile to one side. Now wait five years and look at where those funds rank in the next five-year window. Random chance says 25% of them should remain in the top quartile. The empirical number, averaged across many vintages, is closer to 15%.

That is not just below random — it is meaningfully below random. The last cycle's stars systematically underperform pure chance going forward. The reasons are mechanical, not mysterious:

• Style drift / mean reversion. A fund that crushed it because it

was overweight US tech from 2015 to 2020 is structurally overexposed to the next reversal of that factor. The very thing that put it in the top quartile in window 1 is what makes it vulnerable in window 2.

• Asset bloat. Success attracts inflows. A fund that ran $200M of

small-cap value with a sharp process can absorb maybe $1B before it has to broaden the universe and dilute the edge. Most successful funds end up running $10B+ and look like the index they once beat.

• Manager turnover. The PM who built the track record retires, or

is poached. The "fund" you are buying is now a different team.

• Fees. Top-quartile funds are routinely the ones that hike fees

after a good run, eating tomorrow's relative performance.

The transition picture is laid out below. The diagonal — top-stays-top, bottom-stays-bottom — is barely above the off-diagonal. Last cycle's quartile rank carries almost no information about next cycle's rank.

2.3 Survivorship Bias — Why the Average Looks Better Than It Is

Every fund advertisement you see suffers from the same accounting trick, often unintentionally. When a fund family looks at "the average return of our equity funds over the last 10 years," they only include funds that still exist. The funds that performed badly enough to be closed, merged into a different fund, or quietly liquidated are dropped from the dataset.

Over a 15-year horizon, roughly 30-40% of US equity mutual funds that existed at the start are no longer there at the end. They didn't disappear because they were too good. They disappeared because they were too bad. Removing them from the average is the equivalent of ranking a graduating class only after dropping out the bottom third.

When SPIVA recomputes the same numbers survivorship-adjusted — i.e. including the dead funds at their last reported NAV — the underperformance gap widens by another 50-150 basis points a year. This is why the SPIVA percentages we quoted in §2.1 are honest where fund-family marketing is not.

Survivorship bias also infects:

• Hedge fund index returns. The major HF databases are voluntary.

Funds report when they are doing well and stop reporting when they are not. Estimated upward bias on industry-average returns: 200-400 bps a year.

• Backtest databases. Stock returns from CRSP are clean.

Strategy-level backtests, less so — surviving signals are published, dead ones go in a drawer.

• Manager track records. A PM who ran two funds, one that died

and one that lived, will only mention the live one on the marketing page.

Whenever you hear an average return from any non-SPIVA source, your first question should be: survivorship-adjusted, or not?

2.4 The Four Tranches That Actually Generate Alpha

If active is, in aggregate, a losing game, why are we even teaching it for the next ten weeks? Because hidden in the SPIVA tail are four distinct, durable, documented strategies that do generate persistent alpha — but they don't look like the stockpicking you imagine when you hear the word "active manager." Here is the institutional taxonomy.

1. Deep-value activism. Take large concentrated positions in mispriced public companies, then engage with management or the board to force a value-unlock — a sale, a spin-off, a buyback, a strategic shift. Berkshire Hathaway's pre-1990s era, ValueAct, Elliott Management (Paul Singer), Pershing Square, Trian. Capacity is inherently limited (you can only fit so many activists in one company), holding periods are 3-7 years, and the persistence is documentable. Buffett's pre-1969 partnership compounded ~30%/yr gross. Singer's Elliott has compounded ~13%/yr net since 1977 through every regime. This is *buying what passive flows have abandoned* with engagement attached — a structural alpha source.

2. Quantitative systematic. Statistical, model-driven trading at scale. Renaissance Technologies' Medallion (closed to outsiders, ~40%/yr net since 1988), AQR's market-neutral and managed-futures suite, Two Sigma, DE Shaw, the modern multi-strat platforms (Citadel, Millennium, Point72). These shops mine thousands of weak signals and combine them. Edge sources: speed, data, infrastructure, headcount, and ruthless risk management. Not available via a 1% mutual fund. The retail-accessible cousins (AQR factor funds, BTAL, MTUM) deliver a fraction of the gross edge after fees and constraints. Even where structural alpha exists, you still need the toolkit — these strategies require capital, infrastructure, and discipline most investors don't have.

3. Event-driven. Merger arbitrage, distressed credit and post-reorg equity, special situations, capital-structure arbitrage. The classic example is the M&A spread trade: when company A announces it will buy company B at $50 and B trades at $48, the $2 gap is paid for taking deal-break risk. Deal-break frequency is ~5-10% historically; the residual after losses is the arbitrageur's wage. Davidson Kempner, Farallon, the distressed desks at Apollo and Oaktree. These returns are not equity beta — they are insurance premium for warehousing event risk. Sharpe ratios are 0.7-1.2, drawdowns are bond-like in normal times and equity-like in crises.

4. High-touch CTAs and discretionary macro. Trend-followers (Man AHL, Winton in its heyday, Aspect, Lynx) and discretionary macro traders (Soros's old Quantum, Brevan Howard, Caxton). These sit on commodity, FX, and rate trends, scaling in and out with volatility, and provide convex crisis alpha — they make money when equities crash. CTA performance over 1990-2024 averages ~5-8% net with a Sharpe of 0.4-0.6, but with a strongly negative correlation to risk-asset drawdowns, which is what makes them useful as a sleeve, not as a standalone bet.

What unites all four: they have a structural reason for the return. Activists get paid for engaging. Quants get paid for processing data faster. Event-driven gets paid for warehousing deal risk. CTAs get paid for providing convex insurance. None of them are "I am just a smarter analyst." The honest active investor asks: what am I being paid for? If you cannot answer in one sentence with a structural reason, you are paying a fee for noise.

2.5 What This Means for You

Translate the data into action.

• Default to passive. Your beta sleeve — the growth tranche of

the four-tranche framework — should be index funds. Three or four ETFs cover 90% of what most investors need. Anything else has to clear a higher bar.

• If you hire active, demand a structural story. Either it is

one of the four tranches above (and you have access — most retail investors do not), or there is a documentable, persistent reason this specific manager has an edge. "Strong process" is not a reason. "We are concentrated activists with a 5-year horizon and $4B AUM" is a reason.

• Watch fees. Every basis point of fee is a basis point of

guaranteed underperformance. SPIVA's underperformance percentages shrink dramatically when you compare gross-of-fee active returns. The fee is most of the gap.

• Be honest about your own active sleeve. If you trade a small

options or single-stock book — the opportunistic ~10% slot in the four-tranche framework — benchmark it. After three years, if your IR is below 0.3, redirect that capital to the index. The same SPIVA logic applies to you.

• Don't chase last cycle's winner. The fund that crushed it

2020-2024 is statistically more likely to underperform 2025-2029 than a fund picked at random. Buying after a great five-year print is the modal retail mistake.

We will spend Week 44 on alpha decomposition (Brinson attribution), Week 45 on the tools active managers actually use to find the rare edges, and Week 46 on building a simple, evaluable active sleeve at the household level.

---

3. Common Misconceptions

"Past performance must mean something." It does — it tells

you what happened. The persistence data says it tells you essentially nothing about what will happen next, especially for funds that did unusually well.

"Active will beat passive in a bear market." SPIVA reports

underperformance percentages by year, including bears. Active does not systematically outperform in bears. Cash-heavy funds can win a single down year and then drag for ten.

"A 1% fee isn't much." Compounded over 30 years on a 7%

gross return, a 1% fee removes about 25% of terminal wealth. It is the single largest controllable drag on a long-horizon portfolio.

"Sharpe ratios always tell the truth." Selling deep

out-of-the-money puts produces a beautiful Sharpe — until the day it doesn't. Many "skill" tracks are short volatility masquerading as alpha.

"Hedge fund index returns prove the industry adds value."

The voluntary-reporting bias is so large in HF databases that the average industry print is overstated by 200-400 bps a year.

"Passive cannot keep growing forever." Maybe, but passive's

share of US equities has gone from 20% in 2010 to over 50% in 2025 and indexes have not stopped working. The reflexive "passive is a bubble" thesis has been wrong for fifteen years. Only a genuine multi-decade regime change would turn it.

"Closet indexers are the safe choice." They charge active

fees for index-like portfolios. Active share studies (Cremers and Petajisto) find low-active-share funds underperform almost without exception.

"I picked good funds — I am in the 15%." Almost everyone

thinks that. By construction, only 15% are right.

"Buffett proves stockpicking works." Buffett is sample size

one, was a deep-value activist before activism was a label, ran capital you cannot replicate (insurance float), and stopped beating the S&P meaningfully in his last fifteen years.

**"If most managers underperform, the rest must outperform —

so I can pick those."** No. Most managers cluster near the benchmark net of fees, with a left tail of bad ones dragging the percentages. The right tail is small, and you cannot identify it ex-ante from past returns.

---

4. Q&A Section

Q: What is SPIVA, exactly? A: S&P Indices Versus Active. A semi-annual scorecard published by S&P Dow Jones since 2002 that compares net-of-fee active mutual fund returns to S&P benchmarks, by Morningstar category, with survivorship adjustment. It is the most credible public dataset on active manager underperformance.

Q: Why does the underperformance percentage rise with horizon? A: Three reasons. Fees compound. The luck component shrinks with sample size, exposing the negative average alpha. And funds that got lucky early often see asset growth that erodes their edge.

Q: Are small-cap or international active managers better? A: Marginally. Small-cap active still has 75-80% underperformance over 15 years. International developed is similar. Emerging-market active does slightly better when measured in local currency, but worse after adding currency-hedging costs. The "inefficient markets, active wins there" thesis is mostly folklore.

Q: How is Renaissance Medallion 40%/yr if alpha is so rare? A: It is closed to outsiders, capacity-limited (~$10B), staffed by PhDs in physics and statistics, and runs at the speed of light. The strategy is structurally inaccessible, which is why it persists. Their public funds (RIEF, RIDA) generate a fraction of that and have had multi-year drawdowns.

Q: What is a "high-active-share" fund? A: A fund whose holdings deviate substantially from its benchmark — by construction, more than ~80% different by weight. Cremers and Petajisto (2009) found high-active-share funds modestly outperformed the index. Closet indexers (active share < 60%) systematically underperformed. Most fund families now publish this number.

Q: If I cannot identify skilled managers ex-ante, what do I do? A: Use index funds for 90%+ of your equity exposure. If you want an active sleeve, prefer (a) low-fee factor ETFs (Week 23) over discretionary funds, (b) any access you have to the four documented alpha categories above, or (c) a small DIY active book that you benchmark honestly against an index. The household SPIVA test applies to your account too.

Q: Is the rise of passive itself going to break SPIVA? A: Doubtful in any meaningful horizon. Passive forces price-makers to be even more concentrated and competitive among themselves; the active universe shrinks but doesn't get easier. The share of underperformers has been stable around 70-90% over fifteen years even as passive has gone from 20% to 50% of US equity AUM.

Q: Does SPIVA include hedge funds? A: No. SPIVA is mutual funds and ETFs only. Hedge fund performance has its own (worse) reporting biases and is not directly comparable.

Q: Is there a household equivalent of SPIVA? A: Yes — DALBAR's QAIB (Week 11). Average individual investor underperforms a buy-and-hold of the S&P 500 by 4-5% a year because of behavioural mistakes. DALBAR is the SPIVA of the household: most self-directed active investors also lose to passive.

Q: How should I think about my own DIY trading book then? A: Treat it as an active sleeve with a clear benchmark (SPY for single-stock US, AGG for bond trades) and a clear capital cap (the barbell rule — typically 5-15% of net worth). Run it for three years, compute IR honestly, and if it doesn't clear 0.3, return the capital to the index sleeve.

Q: Why is persistence below random and not just at random? A: Mean reversion plus fee-driven inflows. The funds that did best are typically overweight whatever style won last cycle; that style mean-reverts. Their AUM doubles, fees often rise, and the strategy gets diluted. The combination is structurally anti-predictive at the top quartile.

**Q: What about top quintile or top decile? Is the persistence better there?** A: It is even worse. The narrower you slice the top, the more luck-driven the original ranking, the more violent the mean reversion. SPIVA's persistence reports show top-decile-to- top-decile persistence in the 5-10% range — well below the 10% random baseline.

Q: Should I just give up on active completely? A: For the bulk of your wealth, yes — that is what *default to passive* means. But understanding the four tranches and how a barbell-style allocation places a small explicit opportunistic sleeve on top of a large passive core lets you participate where edge actually exists, without risking the core.