Introduction

In my Week 16 newsletter sent out this past Sunday to subscribers, I briefly touched on the importance of understanding industry concentration before performing asset allocation and ultimately, security selection for your portfolio. Asset allocation (i.e. the division between stocks and bonds, as well as the individual sectors in which you invest) will account for the majority of your portfolio’s return, so it’s in your best interest to get this part right. It will require you to take a step back from analyzing P/E ratios, dividend yields, and virtually all the other fundamental metrics which you use to pick stocks. While important, let’s just save that for later.

A favourite show of mine back in my college days was Penn & Teller: Bullsh*t! - a show about two famous Las Vegas magicians debunking everything from 9/11 conspiracy theories to alien abductions. They used humour and of course magic to get their point across, and although they got it wrong a few times, most of their arguments made sense to me. One of the episodes I remember the most was about multi-level marketing in their Season 8, Episode 5 show titled “Easy Money”, and how there really is no such thing as easy money.

Multi-level marketing companies such as Herbalife, Mary Kay and Tupperware all work the same. You start out small and earn a commission on sales, and then you build your own network of employees below you and take a cut of their commissions as well. As you keep building a bigger network (i.e. multiple levels), you end up earning so much money you don't even need to work anymore yourself! And then there’s this quote from the episode:

The market is unlimited.

Factually incorrect, of course. But this is how they sell their business strategies to would-be independent consultants. The fact of the matter is that markets have limits. There will never be an unlimited number of people who you can sell product or offer your service to - eventually, the market becomes saturated.

Porter’s Five Forces

I bring this up because it helps to take a similar point of view with the stock market. Anyone who took a business class before would have likely been introduced to Porter’s Five Forces which analyzes the competitive nature of a business based on the:

threat of new entrants
threat of substitutes
bargaining power of customers
bargaining power of suppliers
competitive rivalry

A company poised for growth wants it to be hard for potential competitors to enter the industry. They want their product or service to be hard to replicate with their customers having little say over the price they’re charged. They want their suppliers to need them to be happy with them rather than the other way around. And they want to crush their competition, either forcing them into bankruptcy or simply buying them out.

In other words, they want to be Apple. Google. Amazon. Walmart. The list goes on and on. A helpful way to assess an industry’s competitive nature is by using a well-known and accepted measure for market concentration: the Herfindahl-Herschman Index (HHI). The HHI merely takes the squares of market share percentages for all companies in a defined industry and adds them up. The lower the value, the more concentrated and competitive it is.

For example, consider a monopoly - a single company with 100% market share. The HHI of that industry would be 10,000 (100 squared). Contrast that with a general clothing company which may have hundreds of competitors. If there are a hundred companies, each with 1% market share, then the HHI is only 100. Far more concentrated and thus, much tougher to stand out. The company itself may be great, but there are going to be limits as to how much market share they can obtain and thus, limits to the share price growth you as an investor are hoping for.

The HHI can be applied to other things as well. The U.S. Bureau of Labor Statistics routinely does studies of occupational concentrations. A 2014 report using 2012 data listed Education, Training and Library as the highest concentrated occupational group. Computers and Mathematical was the least concentrated. Which group would you think is going to be in higher demand and experience higher wage growth in the coming years?

Methodology

For the purposes of this analysis, I wanted to calculate Industry HHI using sales numbers and the S&P 500 as the sample for U.S. companies. In order to classify companies, I chose to use the North American Industry Classification System (NAICS). For those unfamiliar with NAICS codes, I’ve briefly summarized it below:

A NAICS code is six-digits which represents what’s known as the National Industry Code
The first two digits represents the Sector
The first three digits represents the Subsector
The first four digits represents the Industry Group
The first five digits represents the Industry

By using the NAICS, this gave me the flexibility to calculate HHI by any of the groups above. By using just the Sector, this gave me 19 categories to work with and relatively large sample sizes within each group. By using the subsector, this gave me 61 categories to work with and comparatively smaller sample sizes. Naturally as you go into more detail, the sample sizes get smaller and less meaningful, so I will stick to just using the sector and subsector for this analysis.

Results

As mentioned, my analysis resulted in 19 sector groups of the 498 distinct companies trading on the S&P 500. The interpretation of the HHI is relatively straightforward - the higher the HHI value, the less concentrated the industry is (i.e. one or two companies dominate the sector). The lower the HHI value, the less concentrated the industry is. Some economists may refer to such industries as “perfectly competitive”.

Below are the results by sector:

And below are the results by subsector:

Interpretation & Conclusion

As you can see from the Sector table, Transportation and Warehousing seems to be the least concentrated and least competitive sector. Indeed, we can dive into the details and see that FedEx and United Parcel Service (UPS) dominate this sector. Conversely, the Chemical Manufacturing subsector has a very low HHI value of 564, with 36 companies in the category and DuPont de Nemours having that largest market share of just under 7%.

While it may be difficult to classify each and every company properly, I believe this at least provides a framework for industry analysis. I have made the source file available to anyone who would like to do a bit more in-depth analysis themselves and perhaps use it as a way to analyze their current investments from an industry perspective. What you’re looking for is a high level of market share in a relatively uncompetitive industry, making these companies “the only game in town”. I would strongly argue that companies with high HHI scores are more likely to be “wide moat” companies and are good candidates for targeted industry selection by institutional and retail investors alike.

I also would like to point out that fund managers of large ETF’s and mutual funds absolutely need good levels of liquidity. They often make very large purchases and sales in bulk and if a company doesn’t have good liquidity, which often comes along with high market shares and market capitalizations, they simply won’t trade these securities. Think of it as an extra demand boost these types of companies have, which should partially drive share price growth.

What do you think? Is this type of analysis worthwhile doing, or do you prefer to just stick to classic company fundamental analysis? Let me know in the comments or on Twitter @SundayInvestor.

An Industry Concentration Analysis of the S&P 500