Blame the Investors?
Byrne Hobart points us to an interesting tweetstorm by Flexport CEO Ryan Petersen, decrying the alleged obsession of modern finance with return on equity (ROE). Here are some excerpts (but it is worth reading the whole thing):
What caused all the supply chain bottlenecks? Modern finance with its obsession with "Return on Equity."
To show great ROE almost every CEO stripped their company of all but the bare minimum of assets. Just in time everything. No excess capacity. No strategic reserves. No cash on the balance sheet. Minimal R&D.
We stripped the shock absorbers out of the economy in pursuit of better short term metrics.
Big businesses are supposed to be more stable and resilient than small ones. And economy built around giant corporations like America's should be more resilient to shock. However the obsession with ROE means that no company was prepared for the inevitable hundred year storms.
How much of the amazing quarterly earnings we're seeing out of public companies right now is actually just accounting shenanigans?
A well run business seeks to minimize accounting profit to reduce its tax bill, freeing up cash to reinvest in compounding the value creation. Most public companies now do the opposite, overstating profits to trigger bonuses and applause from Wall Street.
…
CEOs hired by committees can never have the security required to challenge [the] dogma of an entire generation of finance professionals, including all the people who hired them.
…
Only founder led companies and family owned businesses can stand up to the immense pressure from the dogmas of modern finance.
Petersen makes some compelling points,1 but Hobart also links us to a good counterpoint by “Skeletor” on Twitter; Skeletor notes that Japanese companies generally do not care about return on equity, they are fairly insulated from the pressures of modern finance, and yet they are generally doing worse than American companies when it comes to coping with Covid.
So what is to blame for supply chain bottlenecks? Is it the high church of modern finance? Or something else?
Here is the second slide of Amazon’s earnings presentation from Thursday (this is not an exception; all of the slides are like this):
Matt Levine frequently points out that people complain that Wall Street is supposedly too focused on short term financial metrics and ratios, but Big Tech is generally entirely focused on the very long term (metaverse!) and Wall Street grants them a premium valuation.
Amazon is one such example: It is worth $1.7 trillion, and its stock barely budged after reporting these quarterly results. Amazon focuses its financial reporting on long term free cash flow instead of short term earnings, and this does not hurt its standing in the investor community. If quarterly free cash flow is down $27 billion from a year ago, well, that’s apparently just fine with Wall Street.
Amazon and the rest of Big Tech make up a very small minority of publicly traded companies by number, but a much larger proportion by value. The S&P 500 (which accounts for around 80% of the value of all publicly traded companies) is worth about $40 trillion today; Amazon alone makes up 4% of that. The ten most valuable companies in the U.S. (which includes all of Big Tech) combine to account for around a quarter of all American corporate market value, and none of these ten companies are known to be at all concerned with quarterly ROE or any similar short term financial metric.2 Clearly the high church of modern finance shows a tolerance, if not a preference, for companies that ignore ROE.
Most of these companies are no longer founder led or family owned. Facebook is still run by its founder, but Apple, Google and Microsoft are now led quite capably by professional managers, and Amazon recently joined their ranks with the retirement of Jeff Bezos. None of them show any sign of concern with the whims of Wall Street. They invest heavily in R&D, they carry plenty of excess cash and eschew debt, and Wall Street cheers them on anyway.
One major part of the disconnect is the popular notion that Wall Street is some sort of monolith. Wall Street is actually made up of a diverse array of investors, from activist hedge funds to conservative mutual fund managers to Robinhood options traders. There is a sorting process where different companies attract different kinds of investors; struggling management teams attract activists, stable companies attract mutual fund managers, and bankrupt movie theater chains attract Robinhood options traders.
Corporate managers tend to believe that Wall Street is disproportionately composed of the particular investors they interact with the most. We studied the case of Jeff Immelt’s tenure at General Electric; Immelt recently wrote (with some frustration) that when GE exited the insurance business, one of his major investors remarked to him that they had no idea GE was in the insurance business in the first place.
As Warren Buffett observed in his 1979 annual letter:
In large part, companies obtain the shareholder constituency that they seek and deserve. If they focus their thinking and communications on short-term results or short-term stock market consequences they will, in large part, attract shareholders who focus on the same factors.
GE had long focused on delivering smooth, obviously manipulated short-term accounting results, which naturally repelled any knowledgeable investor who focused on the long term and valued transparency and honest management. GE was left with a self-selected group that wasn’t about to be concerned with trivialities such as “does this company own a big insurance business” or “is this company really making money”.
It follows that any management team that is highly focused on ROE will attract investors who also care about ROE. They might conclude that because their particular investors really care about ROE, Wall Street as whole must really care about ROE, but that is clearly not the case; we have seen that at least a quarter of Corporate America (by market value) does not seem to care about ROE at all, and their investors are perfectly fine with this. And this makes sense; when was the last time you saw a stock pitch or fund manager update touting the ROE of a particular holding?3
ROE is a good idea. However, it is as Benjamin Graham often said: “You can get in way more trouble with a good idea than a bad idea, because you forget that the good idea has limits.”
Andy Grove, in High Output Management, advocates the notion of pairing indicators. His first example, ironically, is inventory: if you want to start tracking inventory as a performance metric (lower being better), you should also start tracking the prevalence of inventory shortages, to make sure you aren’t damaging the business by holding too little inventory. Another general example he cites is pairing measures of output quantity with measures of output quality; if you track the number of widgets produced, you should also track the defect rate.
If you measure profitability, you might want to pair it with a measure of invested capital. Even if a business invests its money poorly, it will still grow profits, but it will not be delivering a return to its investors. For example, if a company invests $10 billion in a manufacturing plant that only ends up making $100 million a year, it will still grow annual profit by $100 million, even if a 1% return on invested capital is wholly inadequate.
A metric like ROE accounts for this problem by pairing profit and invested capital. In this example, earnings goes up by $100 million, equity goes up by $10 billion, and the average ROE for the company will be pulled downward (to a weighted average between 1% and whatever it was previously). In general, looking at the ROE for a company in a traditional industry will give you some idea of how profitable it is. (In an industry that has significant investments that are not capitalized under ordinary accounting rules, such as R&D and customer acquisition costs, you will first have to make some adjustments to get a meaningful measurement.)
There are additional adjustments commonly made to make ROE more useful. One is to adjust for financial leverage. Companies can reduce the amount of equity invested in the company by capitalizing the business with more debt and less equity, perhaps by issuing bonds and returning the proceeds to shareholders. This will increase ROE but will also increase risk, which might be a desirable tradeoff, but is also something that should be controlled for. A measure like ROIC (or ROA, for banks) backs out the impact of financial debt.
Unfortunately, every metric has weaknesses. Many of the weaknesses of ROE are more hidden and thus more difficult to adjust for directly. It is up to the investor to become knowledgeable enough about the industry and the business to understand these issues and find metrics that take them into account.
For example, Petersen observes that reducing inventory also effectively increases leverage, insofar as it leaves a business exposed to losses (or unable to profit) in a situation where it is suddenly cut off from access to the supplies it needs.
His observation is in fact also valid for other line items on the balance sheet; you can increase accounts payable by delaying payment to your suppliers, which reduces your capital deployed, but you run the risk that they will deprioritize you or cut you off.
Of course you can also directly manipulate the numerator, earnings, as GE did. And then there are non-financial assets one can neglect without apparent consequence until the system comes under stress; Petersen also notes that companies that have not invested in employee loyalty are paying the price now.
From the point of view of any individual investor, this is a solvable problem. Recall our case study of Shelby Cullom Davis, who turned $50,000 into $900 million by selecting insurance stocks in the middle of the last century. Insurance is a particularly tricky business to measure because several years can pass between the date the premiums are collected and the date that the final claims are paid out; in the meantime, the insurer has to make estimates about future losses and it is unclear how profitable the business truly is.
Davis attacked the problem by first spending years studying the insurance industry by working for the New York regulator, and then going out to talk to management teams, to “separate the bluffers from the doers”. He wanted to assess the competency and honesty of the management teams he was evaluating, and he knew he could do so by comparing what the story they told to what he knew to be true about the industry.
A management team’s communications with shareholders speaks volumes about both, at least to a knowledgeable investor. Every metric is imperfect, but a good manager will figure out a constellation of metrics that makes sense for their particular business and explain clearly why they make sense and their limitations. Jeff Bezos has a classic letter explaining Amazon’s focus on free cash flow, Warren Buffett frequently explains his use of book value and owner earnings, and Netflix goes into depth in every quarterly letter discussing investment in content and net subscriber additions by geography. Compare those examples to companies today that produce increasingly fanciful versions of EBITDA or other “creative” vanity metrics.
The honesty of a management team is also on display when they talk about their metrics. In our GE case study, we looked at “managed earnings” and the “15% delusion”, terms to describe the trend among many blue chip companies in the 1990s to engage in accounting shenanigans to report unnaturally smooth and unnaturally large earnings growth. While those specific trends have thankfully died out, management teams are always finding new ways to test limits when it comes to corporate reporting.
Petersen argues, implicitly, that the root problem is not ROE, but rather the general ignorance of investors. His story is that (ignorant) investors are forcing (expert) managers to focus on ROE to the detriment of corporate value. The solution, then, is to entrench founders and managers through the use of multiple voting classes, in a sort of tenure system.
The alternative argument is that the problem is reversed; it is (ignorant) managers that are the root problem, and bad managers will tend to blame bad investors because a) bad managers blame others instead of fixing the root problem and b) bad managers tend to attract bad investors. If this is the case, entrenching existing managers would be counterproductive; the good managers are in no danger of getting fired anyway, so Corporate America will just end up with more bad managers.
Both investors and managers exist on a spectrum, so it is likely that neither side is entirely correct or entirely wrong. If the issue is a shortage of good investors, that problem will probably be solved by the evolution of the financial landscape. Passive investors are replacing active investors, and the active investors that remain are much smarter than they used to be. Indeed, it is hard to imagine blue chip companies embracing some of the reporting fads from twenty years ago today.4
If the issue is that the current system serves to do a poor job of selecting good managers, this too is a problem that will solve itself, but more slowly. It is surely the case that major companies are better managed today than at any point in the past, at least when weighted by market value. There is no comparison to the world of 2000 when GE was the most valuable company and was engaged in all sorts of accounting games while neglecting its own business. Well managed companies tend to crowd out poorly managed companies over time, either through competition or acquisition.
For the investor, the lesson is that it is crucial to learn as much as possible about the actual business and industry, rather than relying on summary metrics like earnings, ROE, and margins. It is often the dog that didn’t bark that says the most about a business, and the only way to spot that is to gain a deep understanding about what one should be seeing.
This is perhaps worthy of its own essay, but he advocates for a replacement cost accounting framework that harks back to the proposals around inflation-adjusted accounting in the 1970s.
The current list: Microsoft, Apple, Google, Amazon, Tesla, Facebook, Berkshire, NVIDIA, JPMorgan Chase, and Visa. As is typical for a commercial bank, JPMorgan reports ROE, but paired with other counterweight metrics such as leverage, loan quality, and overhead ratio, which would keep them from manipulating ROE in the manner Petersen is concerned about.
Byrne Hobart also makes a good point that low inventory is often a byproduct of an efficient operation, as you might see with Amazon, Walmart, or Costco;. Investors want to own the low-cost producer, and low-cost producers often also have low inventory relative to sales. It might be that some companies get confused and try to “cargo cult” low inventory, but investors don’t want to own a company that chooses to hold minimal inventory, they want to own the low-cost producer.
Remember the debate about whether stock-based compensation is an expense?