Three weeks ago, Wells Fargo Chairman Richard Kovacevich blasted the federal government for relying on “asinine” stress tests to make business decisions regarding the billions of dollars in capital that it has invested in banks. Last week, the government used similar analyses to reject the business recovery plans of General Motors and Chrysler as “untenable,” thereby leading to the firing of GM CEO Rick Wagoner and a proposed shotgun marriage between Chrysler and Fiat.
Stress testing models can be quite complicated; as complex, in fact, as the economic models that (supposedly) sophisticated investors relied on when deciding to purchase huge quantities of mortgage backed derivative securities during the past decade. But when applied correctly, with a significant degree of healthy skepticism, stress tests can help us diagnose the financial problems that afflict firms.
In fact, the performance indicators that lie at the heart of stress tests have been in existence for almost a century; we can trace their roots back to the very birth of the modern business corporation. It might be helpful to revisit their origins to gain a bit of perspective on the uses of the stress tests of today.
Eureka! Smokeless Gunpowder!
The turn of the 20th century produced a steady stream of industrial inventions. The portable motion picture camera. The automobile. And smokeless gunpowder.
Smokeless gunpowder? Yes, smokeless gunpowder. It revolutionized blasting techniques and facilitated large-scale construction projects in major urban areas. And, somewhat sadly, it revolutionized the development of tremendously lethal war munitions as well.
This new product supported the rapid growth of DuPont Corporation, a Delaware based firm that contributed mightily to the industrial growth of the United States in the 1900s and 1910s, as well as the militarization of the American economy during the first World War. During that era, DuPont’s Chief of Financial Operations (and later President) Pierre du Pont organized the firm into a modern hierarchical and divisional business structure, and popularized the use of returns as performance indicators of firm success.
Three Indicators For Three Functions
For instance, du Pont believed that modern manufacturing organizations should focus on building three distinct functions. First, they should hire property development specialists to purchase land and construct factories. Second, they should hire salesmen to sell products to customers. And third, they should hire plant managers to produce the products in the factories.
Pretty basic stuff, isn’t it? Pierre du Pont’s great insight, though, is that the managers of each of these three functions should hire very different people with distinctly different skills and backgrounds. In fact, the officers who manage these functions don’t necessarily need to know each other at all in order to tend to their own divisional business. They don’t need to work in the same location, or to mingle at company meetings. As long as the property development team is visiting prospective factory sites, the sales team is backslapping customers, and the plant management team is supervising factory operations, the firm should be able to achieve prosperity.
Thus, the performance of each function within this triumvirate can be judged on the basis of its own outcome statistic. Over the past century, many different versions of outcome statistics have been developed; for instance, here is a “common sense” example of a Return on Equity (ROE) indicator:
> The property development team’s goal is to maximize the amount of property developed in comparison to the size of the firm; thus, its goal is to maximize the value of the fraction Assets / Equity.
> The sales team’s goal is to maximize the amount of products that are shipped and sold from the factories that are constructed by the property development team; thus, its goal is to maximize the value of the fraction Sales / Assets.
> The plant management team’s goal is to minimize the cost and thus maximize the profits that are earned by producing the products that are shipped and sold; thus, its goal is to maximize the value of the fraction Profits / Sales.
What happens if we multiply these three fractions together? Well, the Assets & Sales numbers cancel out of the numerators and denominators, leaving us with the fraction Profits / Equity …
… which you might recognize as the investment community’s favorite statistic, Return on Equity (ROE)!
Pretty clever arithmetic, isn’t it? Because these three fractions are multiplied together to calculate ROE, we can immediately estimate how any changes in future business conditions might impact all three organizational functions, as well as the firm as a whole.
For instance, is the sales team slacking off? That wouldn’t impact the property development team’s fraction, but a decline in sales would obviously harm the sales team’s fraction … and it would likely damage the plant management team’s fraction as well. That’s because, due to a large portion of a manufactuer’s costs being fixed in nature, profits would likely drop more than sales on a percentage basis.
How badly would this impact ROE? Should investors be alarmed? On the back of an envelope, the CFO could easily multiply the three fractions together and estimate future ROE. In other words, by using these performance outcome statistics, the CFO could engage in a fair amount of instant analysis.
What if the members of the property development team protest that they deserve bonuses because they did their jobs well, despite the “screw ups” who are responsible for the other functions? That would necessitate a judgment call on the part of the CEO and CFO, but if the firm decides to reward them for their good work, they would be able to rely on these performance statistics to establish appropriate bonus levels.
Richard Kovacevich might indeed find this process asinine, and Rick Wagoner may not be particularly impressed by it either. Nevertheless, after a century of use in corporate settings, it is doubtful that performance indicators and stress tests will be fading into history any time soon.