Is Market Index Concentration Good or Bad for Index Investors?

In his recent Morgan Stanley post, Michael Mauboussin ponders how the recent increase in concentration of US companies compares to the degree of concentration in the past and whether we can learn something from it.

There are some interesting takeaways. I think it shows that the top companies, unlike the past, is able to create more shareholder returns when they are bigger.

If you wish to read the report in full, you can find the 18-page report here.

Or you can take a look at this summary.

The Degree of Concentration We see Now is Not Unique

1963 was the peak of market concentration where the top 10 companies are 30% of the market. At the end of 2023, it is 27%. We are near but this is not new.

Elroy Dimson, Paul Marsh, and Mike Staunton, studied the concentration of the market dating back to 1900. They found that

1. Concentration in the 1930s was similar to that of the early 1960s in the US.
2. They estimate the top 10 stocks were 38% of the market in the 1900s.

Not a lot of Stocks have Occupied the Top Spot in the Past 74 Years

The chart below shows the top 3 stocks in market capitalization at the end of the year:

What is surprising is that there are just 17 stocks on the list:

1. During this time, 28,000 stocks have been listed at any time since 1950.
2. 11 stocks held a spot in top 3 for greater than 2 years.
3. 5 stocks appear in the top most frequent (Exxon, AT&T, IBM, GE, and Microsoft)

United States Stock Market is Not As Concentrated Relative to Other Countries

We need to have some basis if we say that the US index is very concentrated:

The chart above compares the degree of concentration to other equity markets. There are greater index concentration by the top companies in other regions than the US.

They wonder if we can deduce any fundamental drivers for this concentration.

The Top 10 Companies Earn More Economic Profit Than the Rest, Despite Their Lower Allocation.

The following chart shows the [economic profit of the top 10 companies by market capitalization] and [the rest of the universe]:

They define economic profit as [(ROIC -WACC) x Invested Capital].

Perhaps some explanation is in order. WACC is the weighted average cost of capital, or the equity and fixed income borrowing cost of the business. By right, you should out earn the cost over the long run. WACC can also be seen as the “hurdle rate”, or the rate of return a business of a typical profile needs to do better. If a business can earn a ROIC higher than WACC, then its doing a good job, especially over a long time. Michael Mauboussin writes a lot about quality investing and he is evaluating things this way.

What we see is that the top 10 stocks at the end of 2023 earn 69% of the economic profit but their concentration is “only ” 27%. This means while their numbers are not much, they drive a lot of value added returns.

We can also see this quality aspect happening in the last few years.

The ROIC Gap Between Large and Small Companies Are Damn Wide

They then show us the gap between the ROIC (Return on Invested Capital) of [large cap] and [small cap] companies over the decades:

There has always been a gap between the ROIC of small caps versus the large cap. Well, except in the 1990s that is. But the gap recently has been bigger, which kind of show the great quality of the large-cap.

Michael mentions that the rising disparity in ROIC does not directly address the ROIC for the top companies that have drive concentration higher.

When Smaller Companies Do Better than Larger Companies, More Active Funds Outperform the Market

But this concentration would pose a challenge for the portfolio managers who measure their performance against a large-cap index, perhaps like the S&P 500.

The chart below plots the return of small cap minus large cap (x-axis) against the % of funds that outperform:

We can see a kind of relationship there. When returns of small minus large is high, there seem to be more outperformance.

It does seem that most active managers lean towards non-top stocks and the fund will naturally do well if the non-top stocks do better.

• Most funds outperform the market the most in the 1970s and 2000s.
  • 1970s: 50% of active managers beat the benchmark. S&P 500 return was 5.9% p.a.
  • 2000s: 48% beat. Market index return was -0.9% p.a.
• 1980s, 1990s and 2010s the funds struggle to outperform
  • 1980s: 40% beat. Market return 17.5% p.a.
  • 1990s: 36% beat. Market return 18.2% p.a.
  • 2010s: 34% beat. Market return 13.6% p.a.

Market Have Higher Returns than Historical Average When We Shift into a More Concentrated Regime.

How does the returns change when we go from a period of more concentration to less concentration and vice versa?

The market tends to produce returns above the historical average in periods when concentration is rising and returns below the average when concentration is falling.

Investing in the Top Stock Have always Been a Poor Idea

Michael then wonders if investing in the top stocks is a good idea.

Michael created an index from 1950 till today. The index is made up of the top 3 stocks of the S&P 500. Since the top 3 stocks will change, this index captures the return of the top 3 over time:

Here is his findings:

1. The top stock has historically been a bad investment.
   • Arithmetic average annual returns of top stock from 1950 to 2023 -1.9% p.a. Geometric return -4.3% p.a.
2. 2nd and 3rd largest stocks are far better.
   • 2nd: Arithmetic returns: 2.6% p.a. Geometric returns: 0.8% p.a.
   • 3rd: Arithmetic returns: 1.6% p.a. Geometric returns: 0.3% p.a.

Except in the Past 10 Years…

In 2013 to 2023, the results were very different:

The top 3 stocks produce much better relative returns than they did in the past.

• 1st: Arithmetic returns: 15.9% p.a. Geometric returns: 14.2% p.a.
• 2nd: Arithmetic returns: 9.8% p.a. Geometric returns: 7.5% p.a.
• 3rd: Arithmetic returns: 8.4% p.a. Geometric returns: 5.3% p.a.

This return largely reflect the returns of the top three stocks Apple, Microsoft and Alphabet.

Last Words

The article didn’t make any strong conclusions.

It focus more on presenting the data and let the data do its talking.

Perhaps the data does show the following:

1. When companies grow large, they will reach a point where valuation is not cheap, relative to their earnings. They also find it difficult to grow.
2. The large companies in the past 10-years don’t have this problem. They are able to maintain growth rates despite its size.
3. Since the data set is pretty long, it does show how easily these large companies can slow down. This makes the recent performance all the more impressive. The last few charts on relative returns show us the difference.
4. A deciding factor is whether the size of the companies gives the company its moat.
   • In the past, when a company become large, they are subjected to anti-trust. The current page have not (although MSFT did ten years prior).
   • Given the debt burden of the country, would the government look at higher tax that is specific to these top 3 software-based companies?
5. Betting on the outperformance of even an equal-weighted index is a bet that the top companies will faced the similar issues that the top companies faced in the past.