Index funds versus active funds: what the evidence says

Two different arguments that get mixed together

When people argue about index funds versus active funds, they usually run two separate arguments together and lose both. Separating them is the most useful thing you can do before forming a view.

The first argument is **arithmetic**. It says that, before costs, the average actively managed dollar must earn exactly the market return, and therefore after costs it must earn less. This is not a claim about skill, markets being efficient, or anything empirical. It is a statement about how averages work, and it is true in every market, in every year, in bull markets and crashes alike.

The second argument is **empirical**. It says that, in practice, the specific fund you are looking at is unlikely to beat its benchmark over long horizons, and that even if it does you probably cannot identify it in advance. This is a claim about the world. It is supported by a large body of data, but unlike the arithmetic it is not logically necessary, and it admits exceptions.

Most bad discussions happen because someone attacks the empirical claim ("but my fund beat the index for six years") as if it refuted the arithmetic one, or defends the arithmetic one as if it settled every case. Take them one at a time.

The arithmetic: everyone eats from the same pie

Here is the argument in full, with no statistics involved.

Take a market. All of it. Every share of every listed company in that market is owned by somebody. Divide the owners into two groups: those who hold the market in proportion to its size, and everybody else.

The first group, the index holders, earns the market return by construction. Their holdings are a scale model of the market, so their return is the market's return.

Now consider the second group as a whole. Together, the two groups own the entire market. If group one earns the market return, then group two, taken as a whole, must also earn the market return. There is nowhere else for the money to come from. Every share that an active manager overweights is a share that another active manager must underweight. Their collective bets net to zero against each other.

So before costs, the average actively managed dollar earns the market return. Not less, not more, on average.

Now introduce costs. Index tracking is cheap because it requires little research and little trading. Active management is more expensive because it requires analysts, and because expressing views means buying and selling, which costs money in commissions, in the spread between buying and selling prices, and in market impact when the trade is large. Since the two groups earn the same return before costs, and one group's costs are higher, **the average actively managed dollar must earn less than the average indexed dollar after costs, by exactly the difference in costs.**

That is the whole argument. Notice what it does and does not say.

• It does **not** say no active manager can beat the market. Some must, because the average is made of winners and losers.
• It does **not** depend on markets being efficient or prices being right. It works even in a wildly irrational market.
• It does **not** say index funds are good investments in an absolute sense. If the market falls forty percent, index funds fall roughly forty percent. The argument is about relative performance among strategies, not about whether you should own the asset class at all.
• It **does** say that the aggregate of active management is a negative-sum game against the index after costs, and that the size of the average shortfall is set by the size of the cost gap.

The arithmetic guarantees the average active investor underperforms after costs. It guarantees nothing about any individual fund. Anyone using it to say "active management cannot work" is overstating it; anyone dismissing it with a story about one successful manager has not engaged with it at all.

The empirical part: what the data has consistently shown

The arithmetic tells you the average outcome. It does not tell you the distribution: whether the underperformance is a small shortfall spread evenly, or a majority losing badly while a few win enormously. That is an empirical question, and it has been studied extensively by index providers, regulators, academics and pension supervisors across many markets.

The consistent findings, stated carefully:

**Over long periods, the majority of active funds trail their stated benchmark.** The proportion that trails tends to rise as the measurement period lengthens. Over one year the split is often close to even; over ten or fifteen years, in most markets and most categories studied, a clear majority has trailed. Cost differences compound, and a single bad stretch is hard to recover from.

**Survivorship makes reported figures look better than reality.** Funds that perform badly get closed or merged into other funds. If you look at the funds available today and check their track records, you are looking at a group that has already been filtered for not having failed. Careful studies correct for this; fund advertisements often do not.

**Past performance has weak persistence.** This is the finding that surprises people most. If you rank funds by performance over one period and check where they land in the next, the ranking scrambles far more than a skill-based model would predict. Some persistence has been found, particularly at the bottom (bad funds, often because they are expensive funds, tend to stay bad). Persistence at the top is much harder to demonstrate.

**Cost is the most reliable predictor available.** Across the studies, the single characteristic that most consistently forecasts relative performance is not the manager's track record, education, or process. It is the fee. Cheaper funds beat expensive funds on average within every category, including within active funds. This is exactly what the arithmetic predicts, which is a satisfying moment of theory and data agreeing.

The reason costs receive this much attention from policymakers is precisely that they compound into large lifetime differences in retirement outcomes, which is why cost and charge disclosure sits inside the standard-setting work of international economic bodies and securities regulatorsSource^sourceSource^source.

How to read a fund's track record without fooling yourself

If you are looking at a specific fund and want to assess the claim it is making, four checks strip out most of the illusion.

1. **Compare it to the right benchmark.** A fund holding mostly medium-sized companies should not be measured against a large-company index. Regulators require disclosure of a benchmark for exactly this reason; check that the one shown actually resembles the holdings.
2. **Look at the full history of the strategy, not the flattering window.** If a chart starts in a particular year, ask what happens if it starts one year earlier or later. Genuine skill should not be sensitive to the start date. Marketing usually is.
3. **Separate return from risk taken.** A fund that beat the index by taking substantially more risk did not add skill, it added leverage in disguise. Ask what the fund did in the worst quarter of the period.
4. **Net of everything.** Performance figures may be shown before some costs. What matters to you is the number after the management charge, after transaction costs inside the fund, after any platform or advice fee, and after any entry charge.

Four questions before you consider an active fund

I am not going to tell you never to hold an active fund. There are situations where the choice is genuinely open. Here is the screen I would apply, in order. If a fund fails an early question, later ones do not rescue it.

**Question one: does a good, cheap index option actually exist for this exposure?**

For large listed companies in major developed markets, the answer is almost always yes, and the index option is very cheap. In those areas the burden on an active fund is high and the case is weak. For some other exposures the answer is less clear. If the only available tracker for a niche market is itself expensive, poorly constructed, or tracks an index that is dominated by three companies, then "index versus active" is not the real question. The real question is whether you want that exposure at all.

**Question two: what is the total cost gap, in percentage points?**

Write down the all-in annual cost of the active option and of the index option. The difference is the head start the active manager must overcome every single year, before doing anything clever. A gap of a fraction of a percent is a hurdle. A gap approaching two percent is a wall. Be honest that the manager must clear it not once but repeatedly.

**Question three: is there a structural reason a manager could add value here that the index cannot?**

Some situations have a plausible mechanism, not just a hope. Examples of a mechanism: the index in this market is constructed in a way that forces mechanical buying of overpriced securities; the asset class is illiquid and cannot be indexed cleanly; the mandate involves constraints an index cannot express, such as excluding certain business activities on religious or ethical grounds; the market has few analysts covering it and information genuinely takes time to spread. Examples of a _hope_, not a mechanism: "the manager is very experienced", "they beat the market last year", "this is a stock picker's market". If you cannot articulate the mechanism in one sentence, you do not have one.

**Question four: could you hold this through three bad years?**

Every active strategy that eventually works spends long periods looking wrong. If you would sell after two years of underperformance, you have not bought an active strategy, you have bought a lottery ticket with a fee attached, and you will systematically buy after good runs and sell after bad ones. This is a question about you, not about the fund, and it is the one most people skip.

The special case of constrained mandates

One category deserves separate mention because the arithmetic argument applies differently. If your requirements exclude parts of the market, for religious, ethical or regulatory reasons, then "the index" for you is not the broad market index. It is the index of the permitted subset. The correct comparison is between an actively managed compliant fund and a rules-based compliant index fund, if one exists. The arithmetic still holds within that universe, and the cost question is still decisive. What changes is only which pie you are dividing.

What the index answer costs you

Fair treatment means naming the trade-offs of the index approach too.

**You will own everything, including the parts you dislike.** A broad index contains companies whose businesses you may find objectionable and companies you think are absurdly overpriced. Buying the index means buying them in proportion to their size, which by construction means owning more of whatever has risen most.

**You are guaranteed to lose in every downturn.** An index fund holds through the entire fall. There is no manager to move to cash. Whether an active manager would in fact have moved to cash correctly is a separate question, and the evidence on market timing is not encouraging, but the psychological point stands: you must accept the full drawdown.

**Not all indices are well designed.** "Index fund" is not a quality guarantee. Some indices are dominated by a handful of enormous companies, or by a single sector, or by one country. Some are constructed by rules that produce heavy turnover, which costs money. Read what the index actually contains rather than trusting the word "index".

**Tracking is not free or perfect.** Even a cheap tracker has a fee, incurs trading costs, and may lag or lead the index slightly. The gap is usually small, but it is not zero, and it is worth checking rather than assuming.

**The wrapper matters as much as the fund.** Two people holding the identical index fund can end up with very different outcomes depending on the platform charge, the currency it is bought in, and the local rules governing the product. In the UAE, funds marketed to investors sit under a local licensing and disclosure framework, and confirming that the product you are being offered is properly authorised is a basic step before comparing performance at allSource^source.

Where this leaves you

The honest position, stated as precisely as I can manage:

• The arithmetic is settled. After costs, active management in aggregate must trail the index by the cost difference. No evidence can overturn this because it is not an empirical claim.
• The empirical record has consistently shown that a majority of active funds trail over long horizons, that the winners are hard to identify in advance, and that fees predict relative outcomes better than anything else on the fact sheet.
• Therefore, for mainstream exposures where a good cheap tracker exists, the index option is the default that a fund has to argue its way past, rather than one option among equals.
• That default is a starting point, not a commandment. Constrained mandates, illiquid asset classes, poorly built indices and genuinely under-researched markets are real exceptions with real mechanisms.
• Whichever you choose, the decisions that dominate your result are still how much you save, how long you hold, and how much you pay in total. The index-versus-active debate is mostly a proxy for the third of those, which is why the cost lens keeps producing the same answer.

If you take one habit from this article, make it this: before comparing any two funds on performance, compare them on total annual cost, and note the gap. Then read the performance figures knowing the size of the head start one of them is spotting the other, every year, forever.

_This article is financial education, not financial advice. It does not recommend any specific fund or product and does not take account of your personal circumstances. Consider seeking advice from a licensed professional before making decisions about your money._