It’s a fun fact that one of the best performing investors in the world is the S&P 500 index committee . Over the past 20 years, their picks have now beaten 97 per cent of all US equity funds.
However, this trivia obscures a broader phenomenon: indexing has become a huge, profitable and influential business , utterly dominated by a small cabal of companies. What were once fairly rough measures of financial markets — often calculated with slide rules by a few niche newspapers as a dull but worthy service to readers — now exerts power over them.
The “Big Three” of financial benchmarks are MSCI, S&P Dow Jones Indices and FTSE Russell. The first is a standalone public company — albeit with a lot of other businesses bolted on — while the latter two are part of broader listed companies, S&P Global and LSEG respectively.
That means we actually have fairly detailed information about just how lucrative the industry is. In total, the indexing bits of the three companies generated revenues of $6.5bn last year, broken down this way:
NB, those figures are for LSEG’s investment solutions business, which houses FTSE Russell. Indices accounts for the “bulk” of the unit’s revenues, but it is unlikely that it generates more revenues than MSCI and S&PDJI’s larger indexing franchises.
But how profitable are they? LSEG hasn’t split out earnings in its latest accounts, but S&P Global says its indices generated net profits at $925mn in 2023, and MSCI said they clocked in at $1.1bn. In other words, profit margins in the 60-70 per cent range. That’s almost three times Apple’s margins, and up there with stock market darling Nvidia.
How do they make so much money? The revenues mostly come from licensing indices out as benchmarks for investment managers, reference points for financial derivatives, for index funds to track, and all the resulting data and analytics you can then sell to everyone else.
However, the money doesn’t come from the sheer volume of indices, which is admittedly staggering.
The Index Industry Association (really, there is one) doesn’t seem to have done its annual round-up for 2023 yet, and has in recent years tried to avoid freaking people out by not giving an absolute number at all. But given its estimate of 3.7mn indices in 2018 and reported annual growth rates of ca 3–5 per cent since then, there are now well over 4mn indices of different stripes.
Basically, if you can measure it, you can index it. But the reality is that less than 1 per cent of them probably account for well over 99 per cent of the industry’s revenues.
The top 1,000 most referenced stock and bond market indices in the US fund industry have about $13.5tn of assets benchmarked against them, and are tracked by $18.6tn, according to Morningstar.
But of those, the top 20 — eg, the Nasdaq, the Russell 2000, the MSCI Emerging Markets or the Bloomberg Aggregate family — account for $7.6tn of benchmarked and $12.2tn of indexed assets. At the top of the pile is the S&P 500, which is an absolute beast. Morningstar estimates that it is now tracked by over $4tn of assets, and serves as the benchmark for another $2.9tn of investment funds.
For anyone interested, here are the top 20:
As you can see, the only three outliers from the MSCI/FTSE Russell/S&PDJI industry stranglehold are the Nasdaq, the Center for Research in Security Prices — only used by index fund behemoth Vanguard in a cost-cutting measure — and the Bloomberg Agg indices. These were formerly known as the Barclays Agg family (and before that the Lehman Agg) and are dominant in the bond market.
The exact mix of indexing revenues varies quite a bit. For example, S&P Global’s results say that almost two-thirds of index revenues come from more volatile asset-linked fees (ie typically index fund licensing), and the remainder is equally split between subscriptions (for real time data and investment fund benchmarks) and derivatives licensing.
But MSCI says that asset-based fees only accounted for 38 per cent of its indexing revenues, despite being probably the biggest benchmark provider for BlackRock ( for long-standing historical reasons ). Larry Fink’s empire accounted for almost 10 per cent of MSCI’s overall revenues in 2023, according to the company’s latest 10-Q.
So how long can the industry continue to milk these money cows for near-pure profit? Longer than previously expected, it seems.
A lot of people predicted wider aftershocks when Vanguard, in 2013, shifted its giant Total Stock Market Fund from tracking an MSCI index to a new one created at its behest by the Chicago-based CRSP (as well as switching to FTSE Russell’s cheaper EM indices in other funds).
The move made sense: as the index fund price war has intensified, more and more of the cost of a fund is actually consumed by licensing brand-name indices like the S&P 500. This infuriates many index fund providers, who have seen their margins crimped even as the index providers themselves continue to enjoy fabulous profits from selling benchmarks that are often actually pretty easy to construct.
Vanguard’s move was expected to herald a broad shift to cheaper options, or self-indexing — where the index fund provider themselves create and manage the underlying index. Charles Schwab and Fidelity, for example, use some self-constructed indices or cheaper brand variants for a lot of their suites of super-cheap ETFs and index funds.
This is a big enough danger that MSCI discloses it prominently in its annual report’s risk factors:
. . . Our clients, including our largest clients, may seek to renegotiate existing asset-based fee models with the objective of achieving lower fees, either on a rate basis or in aggregate, which may have a negative impact on our operating revenues. Moreover, clients that have licensed our indexes to serve as the basis of indexed investment products are generally not required to continue to use our indexes and could elect at any time to cease offering the investment product or switch to using a non-MSCI index. Clients that license our indexes to serve as the basis for listed futures and options contracts might also discontinue such contracts.
However, aside from a lot of smaller, nicher and often gimmickier new ETFs, it hasn’t really happened. There have been surprisingly few big funds changing their index, with Vanguard’s long-in-the-works shift still a big outlier. Fidelity and Schwab’s largest passive funds mostly use one of the boldfaced brands. The big three therefore remain the big three, and their margins are just as juicy as before.
This probably reflects the fact that both customers and regulators are (understandably) instinctively wary of any asset manager using what might be perceived to be a second-rate index, or worse, one of its own construction. It might allow an index fund to cut costs even further, but can look a bit like grading your own homework.
Instead, many asset managers have embraced direct indexing — when individual investors create their own bespoke benchmarks — by either constructing their own custom indexing platforms or simply buying a provider .
However, this is likely to remain a sideshow, at least in terms of denting index company revenues. How many investors actually want to fiddle around with their weightings? Cerulli Associates estimates that assets in direct indexing might hit $800bn by 2026, but that’s still pretty modest compared to the size of the global investment industry, or even just the passive bit of it.
Instead, the big index companies are consolidating their positions, building their own direct indexing platforms, and even starting to indexify opaque, untraded private markets. As MSCI’s chief executive Henry Fernández said at an RBC Capital Markets conference earlier this month:
. . . We’re in a journey of becoming a very large, giant data gathering, data building company. It started with float data on equity indices, then it moved to ESG, then it moved to climate, then into private assets. We have the largest databases of private asset classes in the world. It’s about $15 trillion of underlying values, in those databases of private assets and counting. And, according to our estimates in a few years’ time, even though we have an incredible amount of data today, we’re going to have 1,000 times more data inside MSCI than we have today.
Maybe at some point we might even have an “ Everything Index ” that financial theorists have been fantasising about for generations?