Success Theater
Why do Companies Promote Their Own Stock?; Earnings Management and The Fall of GE
For a long time, General Electric was number one. GE was the most valuable company in America for the majority of the period between 1993 and 2005, at least as evaluated by the stock market. It accumulated accolades and awards as the most admired company in America; Jack Welch, its legendary CEO from 1981 to 2001, was a celebrity. An industrial conglomerate best known for manufacturing jet engines and power turbines that later expanded into finance and entertainment, GE employed over 300,000 people and gained a reputation as a training ground for top corporate executives.
Today, GE is a shadow of its former self. The stock is down nearly 80% from its peak in 2000, and the company has shrunk considerably, hit by the financial crisis, operational issues, and bad investments. As shown in this graph from the WSJ ($), the reversal in the stock has been enough to wipe out all of the outperformance of the 1980s and 1990s and then some; by 2018, the dividend had been all but eliminated, and since 1980, GE stock has lagged the S&P 500 index four to one.
In the past year, two books have been published to try to explain what happened at GE. The first, Lights Out, was written by a pair of reporters who covered GE for the Wall Street Journal; the second, Hot Seat, is a memoir by former GE CEO Jeff Immelt, who presided over the company from 2001 to 2017. They paint a picture of the rise and fall of a great American company, but they also provide an interesting case study of how companies choose to promote their stock to Wall Street.
“Jack, how do you do it? How do you get a 50 P/E with that bag of shit you’ve got?”
GE was best known for designing and manufacturing industrial equipment, but they were no slouch when it came to sales and marketing, particularly when it came to marketing their own stock to investors.
There are, roughly speaking, two schools of thought when it comes to investor relations and financial reporting.
One school of thought is that if you run the business to maximize long-term value, and clearly and cogently communicate your strategy and results to your investors, the stock price will take care of itself.
As an example, when Amazon CEO Jeff Bezos published his famous first annual letter to investors shortly after their $54 million IPO in 1997, he laid out his investment philosophy, which stated (in part):
We will continue to make investment decisions in light of long-term market leadership considerations rather than short-term profitability considerations or short-term Wall Street reactions.
When forced to choose between optimizing the appearance of our GAAP accounting and maximizing the present value of future cash flows, we’ll take the cash flows.
The other school of thought reverses the causal relationship: Convince investors to give you a high stock price by telling a good story, manufacturing steady and growing financial results, and consistently hitting short-term profit targets, and this will give the company momentum and publicity that will help lead to business success.
This latter view, although rarely explicitly voiced, was (and is) the most common strategy in Corporate America, at least to some degree. In its heyday, GE was one of the most aggressive and successful practitioners of this tactic. The story of GE in that era, as we will see, is of a company that used every trick in the book to create the appearance of steady growth and hit short-term accounting targets, regardless of the effect on future cash flows.
Jeff Immelt, like many corporate executives, even today still scoffs at the notion that the Bezos strategy is even a viable option:
I sometimes laugh when people say, “Just do the right thing! Don’t worry about the stock.” Sorry, but that is naive.
Still, Amazon did ok. Bezos was true to his word, reinvesting in the business for years, and not reporting a significant accounting profit until 2017. Amazon grew from $148 million in revenue in 1997, when he wrote that letter, to $386 billion in revenue in 2020.
Amazon has plenty of cash flow now - $31 billion in 2020 - and today the company is now worth $1.6 trillion, making it the third most valuable company in America, fifteen times more valuable than GE.
In Immelt’s defense, GE’s strategy seemingly worked to perfection for a long time. Year in and year out, quarter after quarter, they would report steady annual earnings growth of 10% or slightly more, just ahead of the targets they set, while other companies rose and fell with the economy. They were not shy about trumpeting their results, either. Here is an excerpt from their annual report in 2001, Immelt’s first year as CEO, showing steady double digit annual growth in earnings per share (EPS) over the previous two decades:
Immelt starts his book by recounting a dinner in 2001 where a British executive jokingly said to Jack Welch, then still the CEO of GE, “Jack, how do you do it? How do you get a 50 P/E with that bag of shit you’ve got?”
A price to earnings (P/E) ratio describes how much the company is earning each year relative to the market price of the stock. A normal P/E ratio for an industrial company at the time might be around 16 - meaning investors would be earning about 6% (1/16) that year on a dollar invested. A high P/E means that investors think that profits are likely to grow significantly in the future, or that the company is very safe, or both, to compensate for the premium paid.
GE was hardly a “bag of shit”, but it was also not an especially unique or desirable collection of assets. If investors were giving it a P/E of 50, triple the valuation of a normal company, it implied a strong vote of confidence that management would be able to create value far beyond what could normally be expected from the same businesses operated individually.
By that time, investor enthusiasm was understandable. Blue chip stocks were at the tail end of an unprecedented bull market, and GE really seemed to have built something special - a giant machine that could grow wealth at a blistering pace without suffering much volatility.
The market price of a stock (or any asset, for that matter), is set by the marginal investor buying or selling stock in the market at that time. It can sometimes happen that all of the available shares end up getting hoovered up by true believers, who end up trading amongst each other at high prices while other investors look on skeptically. (Think of GameStop, for example.)1 Which is to say, despite the high stock price, it is definitely not true that all of Wall Street was seduced by GE at that time.
It was quite clear to any experienced investor that GE’s earnings were being “smoothed”. It is not natural for a large financial and industrial conglomerate to report such consistent profit growth year after year.
For example, Berkshire Hathaway, a financial and industrial conglomerate that does not manage reported earnings in any way, reported that after-tax profits (excluding investment gains) shrank 6% in 2020, grew 1% in 2019, and grew 56% in 2018 (the year the US corporate tax rate changed). By contrast, GE, with a similar basket of businesses, grew earnings in a narrow band between 10% and 15% almost every year for over two decades.
This suspicious reporting did not go unnoticed at the time. In 1994, the Wall Street Journal ran a front page story on the methods GE was using to smooth earnings. In 1999, Carol Loomis of Fortune wrote a story on so-called earnings management, where in response to an announced effort by SEC head Arthur Levitt to crack down on aggressive accounting, an anonymous CEO said skeptically, “When he goes after GE, I’ll know he’s serious.”
The official party line from GE was that smooth earnings growth was a product of good management and a diverse portfolio of businesses, although from time to time they would acknowledge the obvious, that earnings were being managed in some way. Former CEO Jack Welch rationalized the practice, once telling Carol Loomis, “What investor would want to buy a conglomerate like GE unless its earnings were predictable?”
Welch’s question is rhetorical, but it is an interesting one. Investors have a preference for predictable, less risky assets; they will usually pay a premium for stock in a company like Coke, which has a stable franchise and has predictable profits. Conglomerates tend to be opaque and full of questionable and cyclical assets - a bag of shit, if you will. There was a period in the 1960s when they were in vogue, but by the 1990s, they typically traded at a discount, and GE was one of the few remaining.
GE management seemed to believe that if they could find a way to mimic some of the features of a business like Coke, such as earnings stability and growth, after a number of years delivering the numbers, they would gain credibility with enough investors to value them as highly as they valued Coke.
GE had only managed to generate the appearance of stability and consistent growth through aggressive accounting. As they got bigger and bigger, the law of large numbers made it harder and harder to keep growing and find ways to keep the ship on course. The effect of what they were doing, as we will see, was to steal from the future, where the bill would come due at some indeterminate date.
In a way, what management created at GE was like a knockoff designer handbag; sure, they could create a superficial resemblance, but measured by most important dimensions, it was still very much not the real thing. A knockoff bag is not made of the same quality materials, or with the same careful stitching; it lacks the same resale value and it falls apart after a few years of wear. Sure, some investors may have paid the close to full Gucci price and felt hoodwinked, but it said “Guci” on the label, what did they expect?
What happened at GE? To understand it better, we have to delve a bit into everyone’s favorite subject, accounting. Andreesen Horowitz has a nice piece explaining that financial data is simply another type of valuable data businesses must collect and tap into to understand where they are and what to do next, and should be treated no differently from any other product data that gets collected.
This is a good framework for thinking about it - accounting just gives you another set of metrics on your dashboard that helps you understand what is really going on.
In theory, proper accounting gives investors, creditors, and management useful information that can be used to evaluate the performance of the business and make better economic decisions - whether to make a loan, or buy a stock, or build a factory. It is often said that accounting is the language of business, and it is a language that can compress a lot of valuable information about a business into a few short pages.
To use a simple example, imagine you buy an apartment building. Your accountant will record that purchase on a financial statement (specifically, the balance sheet) as an asset, and everyone that understands accounting will be able to read it and immediately understand that this shows how much you paid for it. This includes you, your lender, and any future equity partners, and the relevant tax authorities.
You will collect rent, and incur expenses on property taxes and maintenance, and these items will get summarized on another financial statement - an income statement - showing how much you took in, how much you spent, and the difference, how much money you made. This will also increase the amount of cash you have, which gets recorded on a balance sheet as an asset. This is all perfectly intelligible to anyone who reads it.
Now imagine a tenant pays you $60,000 for five years of rent in advance. If you were to record that entire slug of cash received as revenue for this month, that would make your business look much better off than it really is this month, and a bit less well off than it really is for the next few years.
This brings us to the concept of profit, or earnings; the idea behind earnings is to have a metric that creates a more accurate picture of how much better off you are economically in each period, without regard for the random timing of, for example, when you or your customers actually send out cash to pay the bills. It tries to match revenue and expenses as well as possible.
In our example, you are required to recognize that five years of rent received today ratably over the next five years, the period in which you are actually providing the service, so you and your lenders have a more clear picture of your actual economic performance each year. So instead of $60,000 of revenue in month one and zero for the remainder of the period, you record $1,000 of revenue per month.
As a last example, imagine you spend $100,000 to renovate one of your apartments. Accounting rules dictate that you have to spread that amount over the expected useful life of that investment, to give the reader a better picture of true economic profitability.
This requires a bit of judgement - is the useful life five years? Ten years? - and that will have a major impact on the annual profit you report next year. If it’s five years, you will report $20,000 of depreciation expense and if it’s ten years, it will be $10,000. Your total expense is the same, it will just be recognized over a different period of time.
Note here that your choice of depreciation period has no impact whatsoever on the actual economics of your investment decision, nor even on the earnings you report over the long term. You spent $100,000, and you will either get a return on it or not. You are just trying to create a more useful picture of the economics of your business than reporting a $100,000 expense in the first year and zero expense thereafter.
It should be clear from this example that if your goal is to manipulate reported earnings for just this year or next year, these kind of judgement calls are the ones you would tweak to get your desired results without necessarily running afoul of SEC rules, although companies frequently violate the rules as well in pursuit of the numbers they want.
A business as complicated as General Electric has thousands of opportunities to fine tune reported earnings without necessarily breaking the law, if it wishes; while there is a set of guidelines, known as Generally Accepted Accounting Principles (GAAP), initially established after the 1920s to curb some of the worst abuses and establish some semblance of uniformity, there is still necessarily some leeway, as no set of rules can be created that perfectly fits every business at every time.
One of the main ways GE was able to show such stable profit numbers was by using one such judgement call - the discretionary ability to take a non-cash restructuring charge against earnings when it judged there was some kind of future liability - to “bank” earnings from exceptionally profitable periods, lowering earnings in one quarter to be able to increase earnings in a later, less profitable quarter.2
In this example, your financial statements are the map, while your business, the apartment building, is the territory.3 Your financial statements can only offer a flat, limited representation of your business: The carrying value of the land and building based on historical cost, as well as revenues, expenses, and cash in and cash out in the most recent reporting period. This is all crucial information about your business, but hardly the whole story.
The value of your business is by definition the present value of all of the cash flows you will get from the business from now until eternity. Your future cash flows, in turn, are a function of all of the other factors that will affect the economics of your business: location, occupancy rate, rent roll, the age of the building, building classification, the health of the surrounding economy, crime, schools, regulation, taxes, new construction, competitive dynamics, even your acumen as a building manager!
John Maynard Keynes, the great economist and investor, said that “the object of skilled investment should be to defeat the dark forces of time and ignorance that envelop our future”. In this battle, you need all of the help you can get, and accurate and useful historical financial statements are but one important tool at your disposal, allowing you to get a handle on some of the key dimensions of a given business.
The question of whether GE technically broke the law in its reporting somewhat misses the mark, in my opinion. It is quite clear that they were intentionally manipulating their reporting to obscure the true state of the business, which goes against the whole purpose of the exercise.
To paraphrase Warren Buffett, the problem with bad accounting is not just that management is lying to investors, but that they are lying to themselves. If you are a pilot, and you expend a lot of effort in tampering with the output of your control panel, it should be no surprise when you find yourself disoriented and in an uncontrollable tailspin.
GE management seemed to be genuinely surprised in later years when their Power division had stopped generating cash despite reporting healthy paper profits, due to aggressive assumptions made about the future profitability of service contracts that did not materialize. This is an expected result, though, if the goal of your accounting is to maximize reported profit rather than to understand your business.
Furthermore, it is very dangerous when management has a myopic focus on near-term earnings, which is but one line on one financial statement from one period that itself is a very limited representation of the economic totality of the business.4 To continue our aviation analogy, it might be the equivalent of focusing on airspeed while ignoring angle of attack, another way to crash.
There is a saying in business that “you make what you measure”. If top management is focused on earnings, they will make decisions that maximize that, even to the detriment of the business.
There is ample opportunity to increase near-term earnings in a way that diminishes the economic value of the business; in our apartment building example, you could simply put off maintenance, which will decrease your expenses today but will diminish the value of your building in the future.
In the case of GE, the main lever they had to use to grow earnings at such a high rate was to acquire other profitable businesses - there was simply no other way to do it at their size. They bought over a thousand businesses under Jack Welch in the 1990s, and the pace continued under Immelt in the 2000s, with even larger acquisitions.
Acquisitions are hard. The field is very competitive, and the buyer traditionally has to pay a significant premium to close a deal. Most studies have shown that corporate acquirers in particular have a poor track record. Many companies choose to avoid them altogether, or at least stay in fields close to their core competencies.
GE’s acquisitions pushed them into fields far from their historical core competencies of power and aviation, often with poor results. A series of insurance acquisitions under Welch in the 1990s resulted in huge liabilities tied to toxic policies they wrote in long-term care. In 2018, over a decade after they disposed of the insurance business, they announced a $15 billion cash charge tied to policies they were still on the hook for, and they may owe more in the future.
GE expanded their presence in finance and lending in the 1990s and 2000s, which grew to contribute nearly half of their earnings, but this bit them in 2008 when the financial crisis hit, and without government intervention would have taken down the entire company. It was only last month that they managed to wind down GE Capital, their finance arm.
On the industrial side, they had troubled deals as well. They made a foray into the water business that didn’t work out. They built up a large oil and gas equipment business, just in time for the oil prices to tank and the market to shrivel up. In 2014, they made a disastrous $14 billion acquisition of the power business of the French company Alstom, which had to be written off completely just four years later.
Sometimes, even when they made a good acquisition, it would find a way to turn bad anyway. They bought two successful media companies, NBC and Universal, but were forced to sell them to Comcast in 2009, at the bottom of the market, to raise cash.
In some ways, designing and marketing a stock is like designing and marketing any other consumer product. You build a product that you hope will be appealing, you tell a simple, compelling story that highlights its best features, and you hope that will be enough to convince people to line up to pay a premium.
For example, if you wanted to sell beer, you might design something like Coors Light. You tell people that the water is sourced from the Rockies, and that the can turns blue, and that these are the two things you should care about. As a result, consumers pay a premium for each can, enough to pay for the cost of brewing and advertising and leave a profit besides.
Storytelling is a powerful tool, and it works on investors just as well as it does beer drinkers, particularly in a bull market with a lot of confident, novice investors. A sufficiently talented storyteller can convince public market investors to pay 50x for a bag of shit; in the world today, a subreddit can convince speculators to buy shares of a bankrupt car rental company, and Elon Musk can casually tweet about Dogecoin and people will flock to buy it.
The difference, though, is that a high stock price doesn’t help a company in the same way that a high retail price helps a brewer. After a company goes public, they will probably never raise money again by selling stock; quite the opposite, they will likely be generating enough cash to be buying their own stock in a major way.
As major future buyers of their own stock, they should want their stock to trade at a low price, not a high price. An inflated stock price in the near term merely transfers wealth from long-term owners of the stock to those that are selling tomorrow. Management is the one group that cannot sell tomorrow, so they should theoretically care the least about the short-term stock price.
More to the point though, the whole valuation game is by definition a zero-sum distraction, at least from the company’s perspective. The total wealth gained by all shareholders will be determined by the cash the company is able to generate from its business and return to its shareholders. Most long term studies of the stock market bear this out, showing that total returns eventually track dividends and earnings.
Beyond that, the ups and downs of stock valuation simply represent transfers of wealth from one shareholder to another: If one shareholder has the misfortune to sell at a low price, the shareholder that bought from them benefits by an identical amount, while a shareholder that is lucky enough to sell at a high price does so at the expense of the shareholder who buys them out. Shareholders as a whole can only benefit if the company actually produces cash from the business.
Unlike in our beer example, where successful marketing generates more revenue, any effort the company spends on marketing the stock usually makes the pie smaller for shareholders, one way or another. This is certainly true if the company is making decisions to make the income statement look better at the expense of shareholder wealth, or obscuring their own access to critical data to keep investors in the dark.
The short term earnings game described in this essay seems to finally be going out of style, thankfully, but other simplistic narratives and fanciful metrics are always popping up to replace it. You have “story stocks”, and companies that target volume growth or earnings growth or community-adjusted EBITDA, and they all have the same dangers. As Bezos observed, the only thing that matters is long term cash flows, and any narrative or metric that detracts from maximizing value is likely to be counterproductive.
The idea that management will be motivated by long-term moves in the stock price is entirely theoretical, of course; in practice, hired executives probably feel more comfortable doing what everyone else is doing, and would rather not take any short term hit to the valuation of their stock, even if it is from discouraging demand from speculators.5
To be more cynical, executives are probably more motivated by ego, acclaim and stock-based compensation, than they are by the desire to create long-term wealth for shareholders. Jack Welch got $400 million in severance when he left GE, and Jeff Immelt made around $200 million over his difficult tenure and was best known for taking two private jets on business trips, so he had an empty spare in case of emergency.
Executives often complain that it is the investors that are unduly focused on the short term and generally uninformed about the business, and they feel forced to cater to them. Immelt recalls being surprised when one of his major investors remarked that he had no idea GE was in the insurance business (before the divestiture).
It is difficult to separate cause from effect. Buffett likes to cite Phil Fisher’s analogy to running a restaurant, where management teams can attract like-minded investors through consistent messaging, in the same way a restauranteur can choose to attract diners who like hamburgers or diners who like French food.
We saw earlier that GE, through their promotional messaging and aggressive accounting, had long ago lost their credibility with a large subset of knowledgeable investors. It should be no surprise that they attracted a shareholder base that demanded that they continue to deliver consistent earnings growth and did not bother to learn what businesses they owned.6 As Warren Buffett put it in his 1979 annual letter:7
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.
Companies with good shareholder communication are all too rare, but they are increasingly more common among the largest companies today. It is worth mentioning some examples:
Berkshire Hathaway is always the gold standard, with Buffett’s annual letter and investor meeting where he clearly explains the business and how they see the world, and the 10-K is quite clear as well, for such a complicated business.
Netflix is another good one, if a bit unique, with its thorough quarterly letters and its website sections that lay out their long-term view of the industry, FAQ, and even a primer on content accounting.
Two other good ones are Amazon and J.P. Morgan, both for the clear reporting and the annual letters.
For further reading on GE that isn’t book length: Byrne Hobart has a good summary ($) of this period, and the WSJ ($) has a good series of articles that were the foundation for Lights Out.
Of course, investors can short the stock, but that comes with its own dangers!
To be clear, they used many more aggressive tools to manipulate reported earnings, including selling assets held at GE Capital at historical cost for a quick gain. Before Enron made it illegal, they would often sell the assets to a special purpose entity that they themselves controlled - a transaction with no economic substance but at the time allowed them to book a profit.
Extending the analogy, it would be ridiculous to think that you could widen the Suez Canal by making a tweak on Google Maps. Although it would make a good Borges story.
This is sometimes better known as Goodhart’s Law - any metric becomes less useful if you begin to explicitly target it.
There is also the possibility that hired executives are more likely to have a sales or marketing background - Immelt was a salesman at GE Plastics - and thus more comfortable marketing the stock or to suffer from a case of “man-with-a-hammer” syndrome. Also, getting promoted to the CEO position at a big company requires skillful marketing of one’s own merits, possibly to the detriment of developing other skills.
Founder-led companies seem to be much more likely to have better shareholder communication. A founder might be more likely to have risen based on different skillsets, and they naturally have more of an interest in the long-term health of the company’s stock, having no way to diversify.
Your investors are responsible for electing your bosses, the board of directors; in practice this is not a factor until things go wrong, at which point activist investors may spot an opportunity to buy in with the intent of using their voting power to threaten to replace the directors with the intent of replacing value-destroying management and pocketing the profit once the job is done. Jeff Gramm’s Dear Chairman is a great read on the topic; this is basically what eventually happened at GE, with activist investors Trian.