Customer-Base Concentration: Implications for Firm Performance and Capital Markets

Customer-Base Concentration: Implications for Firm Performance and Capital Markets

  • Panos N. Patatoukas
  • A version of the paper can be found here.
Abstract:
“This paper investigates whether and how customer-base concentration affects supplier firm fundamentals and stock market valuation. I compile a comprehensive sample of supply chain relationships and develop a measure (CC) to capture the extent to which a supplier’s customer base is concentrated. In contrast to the conventional view of customer-base concentration as an impediment to supplier firm performance, I document a positive contemporaneous association between CC and accounting rates of return which suggests that efficiencies accrue to suppliers with concentrated customer bases. Consistent with a cause-and-effect link between customer-base concentration and supplier firm performance, analysis of intertemporal changes demonstrates that CC increases predict efficiency gains in the form of reduced operating expenses per dollar of sales and enhanced asset utilization. Using stock returns tests, I find that investors underreact to the implications of changes in customer-base concentration for future firm fundamentals when setting stock prices. A trading strategy that exploits investors’ underreaction yields abnormal stock returns over the thirty-year period studied.”

Data Sources:
This study looks at the 1977 - 2006 period.  Stock return data is from the Center for Research in Securities Prices (CRSP), fundamental data is from COMPUSTAT.

Customer revenue numbers and 2-digit SIC codes come from the COMPSTAT Segment Files. Because COMPUSTAT Segment Files extract information directly from 10-k's, and companies self-report the names of their customers, the author uses various algorithms to match the database of customers with actual public companies--a complete nightmare (we know, we maintain a "live" database of customer/supplier relationships).

So where can your everyday Joe get the necessary customer information? Well, you'll either have to pay up for COMPUSTAT, Revere,  any number of expensive outlets, or start collecting the information off 10-Ks by hand--yikes!
Discussion:
There are two conventional views on customer-base concentration:

  1. concentation=bad.
  2. refer to #1.

Why is customer-base concentration bad?

People typically think of large firms as being abusive towards the smaller supplier from which they buy.  The obvious case-in-point is Wal-Mart (McDonalds is another), which has garnered a reputation for exerting so much buyer power that its suppliers are forced to reduce prices to the point at which the supplier is barely earning its cost of capital.  This may be true in certain instances, and indeed, in the case of Wal-Mart.  But the lesson from this paper is that the customer-supplier relationship is actually positive on net.

So why might customer-base concentration be good?

One can imagine that a tight customer/supplier relationship can lead to efficiencies in marketing, operations, collaboration, and so forth. For example, Hasbro, in their 2006 Annual Report, claims that having a concentrated customer base is beneficial because it lowers the costs of sales and distribution and their production pipeline is more efficient.

Why do we care as an investor?

If it is actually the case that a concentrated customer base is not a 100% bad thing, and the market does not recognize this fact, there may be an investment opportunity to exploit. In other words, if a customer base of a supplier firm becomes more concentrated, and investors systematically punish the stock, there may be some "alpha" on the table.

How does the author test for the "customer-base efficiency" effect?

The author of this paper first compiles a list of customer-supplier relationships over the period 1977-2006.  He then assigns each supplier, each year, a value for Customer Concentration (CC).

To establish that high customer concentration is actually beneficial, the author looks at changes in customer-concentration and subsequent firm operational performance.  The author also controls for a variety of other factors related to CC and rates of return, to pinpoint that changes in CC are the driving force behind any changes in future performance. Consistent with his argument, he shows that increases in CC are followed by improvements in operating performance. Interestingly, he also finds that more-concentrated suppliers report lower gross margins--which actually supports the primary hypothesis that high CC is bad for a firm. However, the reality is that while gross margins decrease with CC, and this may throw off the market as to the benefits of CC, operating margins actually increase (likely because of the efficiencies involved in having a higher CC)!

After establishing that increases in CC are predictably followed by improving operating performance, the next tests see if the stock market prices reflect this predictability. The author groups all supplier companies in to one of ten groups (deciles) based on the most recent change in CC.  On a market-adjusted basis, the concentrated suppliers in the top decile outperform those in the lowest decile by 8.96% per year in the following year.  Even after controlling for other known risk factors, there are abnormal return estimates of 6.24% a year. Not too shabby!

Here is the table with all the important statistics for investors (rank(? CC) is the "money shot"):

"The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained. Indexes are unmanaged, do not reflect management or trading fees, and one cannot invest directly in an index. Additional information regarding the construction of these results is available upon request."

The main findings in this paper are similar to a variety of other investor-underreaction papers.  The information is freely available, yet the market simply does not give it much consideration when arriving at prices.  As with other papers in this stream of literature, the author concludes that investors eventually recognize the value of the underlying value driver, but only once it is revealed in company fundamentals.
Investment Strategy:
  1. Identify the annual change in customer concentration calculated from customer/supplier data.
  2. Long high customer concentration stocks
  3. Short low customer concentration stocks
  4. Make money.

Commentary:
This paper is certainly interesting--my null hypothesis would be to short any company with a concentrated customer base...looks like that is a bad idea--D'oh!

And while I like the trading concept in this paper, it is hard to ever know if the results are being driven by customer-concentration or some omitted variable (what if changes in customer-concentration also mean changes in the risk environment?). Nonetheless, I think the author does the best job he can, given the data available, to identify that customer-concentration matters to economic performance.

Implementing this strategy is going to be difficult for a non-quant, however, there are a few options:
  1. hire some cheap labor to build you a customer/supplier database.
  2. hire a data provider to hire some cheap labor to build you a customer/supplier database.
  3. hire a financial advisor with the capabilities to implement this strategy, and won't charge you an arm and a leg to do so (easier said than done!).
  4. Pass on this strategy, sit in cash, and prepare for the oncoming economic apocalypse.