Friday, May 13, 2016

On Shareholder Value

In response to this email:
"I need your brain. I very badly want a response to this. Please be nice but honest. I don't know enough context to respond with something defensible.

TL;DR: The article's definition of shareholder value is incorrect - shareholder value is truly a very long-term measure, much more so than any concept of "stakeholder value," which is frequently absurdly easy to game in a really negative way. Short-termism lies instead in the tendency of markets to extrapolate from short-term results to long-term outcomes, and the overall error is to equate near-term stock price with shareholder value. However, per-share long term shareholder value is by far the most efficient and least corrupt measure a system can target, and combating that extrapolation bias should be the focus, not moving away from shareholder value.

If shareholder value were defined even remotely like the article defines it, then the article would be reasonable.

The notion of short-termism in business is nebulous and thus gets badly abused.

What is shareholder value and why do we maximize it?
Shareholder value, at its simplest, is the discounted set of all future profits attributable to a single share of equity. This means that the concept of maximizing shareholder value has almost nothing to do with short-term EPS; a responsible manager tries to manage the per-share results of a company over a span of decades. This obviously incorporates a wide range of outcomes, which is why risk is an important component of thinking. It also incorporates the legal consequences of malfeasance, the operational consequences of being shitty to your employees/suppliers/customers/creditors, etc.

In fact, the whole rationale for using shareholder value - which I think this article misses the point of - is that shareholder value is only maximized when you do everything else right. They're at the bottom of the totem pole, the people who get what's left after employees have been compensated and incented, suppliers and customers have been served, and debtholders have been repaid.

In short, shareholder value maximization is anything but short-term - it is one of the few metrics you can actually look at that ISN'T short term and doesn't screw someone over (I have yet to see any measure of "stakeholder value" that looked dynamically at how incentives change when you put in stakeholder initiatives, nor have I ever seen a policy proposal acknowledge economic gravity when it tries to stop something bad from happening but instead just delays it and makes it worse - think "the euro" and "chinese credit-driven stimulus" type problems). Usually, the one who gets screwed is the equity - if they don't have a legal right to be prioritized, it's absurdly easy to steal from them. Occasionally, in really dire situations (think subprime, we'll come to this), "stakeholder" initiatives can screw the customer and the lower-level employees, as well.

For what it's worth, shareholder value maximization makes it MUCH easier to raise capital for innovative businesses - having done a lot of investing in Europe and Asia, where shareholder value is not a focus, the pace of innovation, the manner of competition and the level of corruption is staggeringly poor - managers and governments essentially loot the businesses under the guise of "social welfare" or whatever, and it makes honest businesses very difficult to fund (and makes the whole system far less trusting, which drives away the honest people, and becomes a pretty insidious vicious cycle).

Even in the US, for what it's worth, having looked at a bajillion businesses, there's far more waste and inefficiency in the direction of not being shareholder-oriented enough than there is in being too shareholder-oriented. Probably by more than one order of magnitude, actually... the exceptions tend to be industries whose regulatory regime is badly designed, and the fix is almost never as simple as a refocusing of company priorities.

Then why are there so many problems?
A) there is no such thing as a perfect system, just one that is better than alternatives
B) That's glib, so let's talk about short-termism.

Short-termism isn't a problem with the theory, it's a problem with human behavioral bias. It's the tendency of people to extrapolate from short-term earnings to long-term earnings.

Investors do this - stocks go up a ton on quarterly earnings not because of three months of earnings but because of what those quarterly earnings signal about the long-term trajectory of the company. It's amazing how different the reactions look like to companies missing a quarter when it's clearly a temporary issue vs. when it's the first manifestation of an issue that permanently impairs the long-run trajectory of the company - if you tell the market you're out of business 6 years from now but the intervening 6 years will be great, your stock is almost certain to be punished (unless that's what everyone was expecting anyway). 

That, incidentally, is why it pays to have an IR team that is both honest and good at communicating - investors, by virtue of being at the bottom of the totem pole (and thus being highly vulnerable), can be very skittish, and the ability to communicate honestly about what types of events are temporary vs permanent is a valuable skill. This doesn't happen often because a lot of management teams know their career isn't that long, and so they care far more about the stock price NOW than they do over time because by pretending the long-term trajectory is better than it is, they are paid more for being smart. Thus, they have their IR teams pretend that negative events are temporary and positive events are permanent, no matter what they actually are. This would be termed an "agency cost", where the management is NOT maximizing shareholder value but instead maximizing their own compensation... it is the antithesis of shareholder value, and investors get pretty pissed when they figure out that's what you're doing.

For what it's worth, Enron and Worldcom didn't "maximize shareholder value", they committed fraud to pretend they were. Subprime was the product of a regulatory environment that incentivized social goals of homeownership (ie, customer welfare) over social goals of a solvent banking system (when the latter was far more consistent with long term shareholder value, as anybody who has ever looked at credit will tell you), exacerbated by a set of customers and private sector businesses who couldn't extrapolate forward to see that the incentives the government offered to distort sources of value would end in disaster. BP oil was a perfect storm of engineering errors that had nothing to do with shareholder maximization and everything to do with engineering system design that I promise you the CEO had nothing to do with and barely understood. 

Lynn Stout's quote is mostly true (the tragedy of the commons example is a bit of a straw man, but i'll let it slide)... it's just really, really obvious - it's unfortunate for the state of the lay discourse that this analysis is what passes for thoughtful. The real mistake the Chicago school made was to assume markets are efficient and thus to claim you can equate short term stock price with shareholder value - the mistake was not in the shareholder value maximization part of the system.