Just how concentrated are equity markets?

The increasing importance of a handful of US tech stocks to widely followed market-cap-weighted indices has led to rising concerns around concentration risk.

In this blog we’ll examine whether the US equity market qualifies as concentrated according to accepted academic standards, and how it compares with other major stock markets. In part two, we’ll consider what our findings mean in terms of performance, risk and fundamentals.

We should start by noting that market concentration can occur at different levels, including region, sector and individual securities. The North American market, for instance, has seen a significant increase in value since the Global Financial Crisis. This coincided with the growth of the technology sector, with many leading tech companies being US-based.

But the primary concern today is security-level concentration, where overall market returns are increasingly being driven by the so-called ‘Magnificent 7’.[1] This raises the question of how this type of concentration can be assessed.

Measuring security concentration

We can begin by examining cumulative index weights, with securities ranked in descending order by market weight. This simple measure shows there has been a rapid ascent in the cumulative weight in the bottom decile, meaning that the top 10% of holdings account for nearly 72% to the total weight.[2]A more sophisticated approach, one used by antitrust regulators, is the Herfindahl-Hirschman Index (HHI). In practice, this is often expressed in its reciprocal form, referred to as the ‘effective number of stocks’.[3]

For our analysis, we have adopted a composite measure designed to be as intuitive as possible by representing the effective number of holdings and the actual number of holdings in a single metric. The composite metric we use is the ratio between the effective and actual number of stocks, multiplied by 100. This metric reflects the concentration of holdings in a portfolio, unlike the HHI, which compensates for portfolios with many holdings and varies between 0 (minimum) and 100 (maximum).[4]

An alternative way to assess concentration is by decoupling the effective and actual number of stocks historically, which also reveals a diverging relationship over time. The increase in the concentration metric aligns with jumps in the actual number of holdings, and a corresponding decrease in the effective number of holdings, which indicates growing concentration in fewer stocks, despite more stocks being added to the index.

The table below illustrates different concentration metrics for a ‘Mag 7’ portfolio, ranging from 0 in an equally weighted portfolio to very high levels when the (notional) portfolio is centred around a single stock. The ‘Mag 7’ portfolio with rescaled market cap weights as at July 2024 is in the middle.

By applying this concentration metric to the global index historically, we see that concentration levels have been consistently high (over 80) and started exceeding 90 after the COVID-19 period.

An alternative way to assess concentration is to decouple the effective and actual number of stocks levels historically. This approach essentially tells the same story – the increase in the concentration metric aligns with jumps in the actual number of holdings with decreasing effective number of holdings.

The market is concentrated, but this isn’t new or unique to the US

Based on the above, we can conclude that the US market does exhibit signs of concentration, largely due to the ‘Mag 7’. However, this isn’t just seen in the US. For example, Taiwan Semiconductor Manufacturing Co* accounts for 45% of the overall Taiwanese index, while Samsung Electronics Co* is around 31% of the South Korean market.

Concentration within markets is also not new. For instance, in 2001, the top 10 stocks in Finland and Switzerland comprised over 80% of total market capitalisation in their respective markets. As an example, mobile phone company Nokia* drove around 66% of the market capitalisation of the Finnish market.

In the next instalment of this blog we’ll assess what concentration means for investors in terms of performance, risk and fundamentals.

*For illustrative purposes only. Reference to a particular security is on a historic basis and does not mean that the security is currently held or will be held within an LGIM portfolio. The above information does not constitute a recommendation to buy or sell any security.

 
[1] Comprising Microsoft*, Apple*, Amazon*, Alphabet*, Meta*, Nvidia* and Tesla*. Hereafter referred to as ‘Mag 7’.
[2] Based on analysis of the Stoxx World AC Large and Mid Cap Index as at July 2024.
[3] Effective number of stocks is equal to 1/HHI or 1/sum(squared stock weights).
[4] Our measure was informed by the academic paper Is Index Concentration an Inevitable Consequence of Market-Capitalization Weighting? Our forthcoming whitepaper will provide greater mathematical detail on the methodology behind this metric.