Where does the value of stocks come from?
Breaking down dividends, buybacks, mergers, acquisitions and iterative predictions
If we open up a stock screener and filter companies to those with a market value over $10 billion (so roughly those in the S&P500), we’ll see that out of 614 American stocks that fit that criteria as of December 2021, 220 didn’t pay dividends at all last year. 112 paid less than 1%. 306 paid between 1% and 5%. And only 29 paid over 5%, which would be considered a good return on capital for a normal investment. Rounding the numbers to the closest integer, we can generate the following chart for FY2021:
And its not just struggling companies that aren’t paying dividends. If we restrict our selection to companies that made at least $1B over the past year, the numbers get better but still surprisingly low given that the S&P500 overall generated 11% compounded annual returns over the past 30 years.
We might also suppose that 2021 is an exception COVID year and dividends were better in the past but the trends for dividends are consistently down over the past decades:
So where does the value of all these stocks come from, given that a lot of them generate mediocre dividends back to their shareholders? The naive answer would be “Ponzi scheme”: people buy the stock hoping that the next person buys it from them, until eventually the entire scheme collapses when the world runs out of suckers. However reality is more complicated once we look at all other sources of shareholder value other than dividends.
Berkshire Hathaway famously never pays dividends despite being one of the most profitable companies in the world. However it does provide value back to shareholders in the form of stock buybacks, which are triggered by Warren Buffet whenever he believes the stock is undervalued on the open market. This year Berkshire spent at least $31 billion on buybacks, providing an effective dividend rate of about 5%. There’s a lot of criticisms around buybacks as naively one might think they’re only done to manipulate the market and prop up stock value. However they’re simply a different mechanism for returning value to shareholders, which also has a distinct advantage of tax optimization. With regular dividends investors have to pay income tax every year, while with buybacks investors will only pay tax if they decide to sell their shares back to the corporation.
Mergers and acquisitions
Many successful companies will eventually end up acquired by another corporation or merging with them. In both scenarios the holders of stock will be compensated for their share by receiving stocks or cash from the newly formed larger venture. For example, Slack was recently bought by Salesforce for $27b, creating direct value for its shareholders. Canadian mobile operator Shaw merged with Rogers this year, generating $26b of value. Investors might have wait for a long time for this to happen, but when it does happen it tends to generate excellent returns.
While any small time stock holder is unlikely to benefit from the power of voting during annual shareholder meetings, this property of stocks does represents a share of their intrinsic value. A paper by Hauser and Lauterbach estimated in 2003 that 20% of stock value in Israel can be attributed to voting rights. Lease, McConnell and Mikkelson estimated in 1983 that 5.4% of stock value in the US is attributed to voting rights, based on a comparison between stocks that provide voting power and those that do not for the same corporation. Voting could be used to generate intrinsic value by encouraging the corporation to hire better personnel or issue buybacks/dividends to start generating direct value for shareholders.
In 2014 Warren Buffet included a paragraph in Berkshire’s annual shareholder letter about eventually starting to pay dividends when market conditions are right. Similarly new companies growing at a fast pace usually choose not to pay dividends while in the growth phase. Investors who stick with the company’s stock long enough will thus end up seeing a return on their investment. For example, Varta paid dividends for the first time this year, 4 years after going public.
It feels strange to bring this up in a post about stock value but as with all investments there’s always the risk of a company going belly up, collapsing the enterprise. In this case shareholders will receive whatever money is left after paying out other creditors. This value is roughly correlated to the “intrinsic value” of the stock, which can be roughly obtained by calculating the value of the company’s assets minus the company’s liabilities. If all other avenues of obtaining value, this remains as a final way for shareholders to extract value from their investment.
Prediction of the future
At this point one may ask: what about companies like Tesla where the stock price shoots up sky high despite the fact that the company might take decades to payback its value in dividends, isn’t that a Ponzi Scheme? However its important to remember that the stock market is effectively a prediction market for future performance of the underlying assets, over many decades to come. This works despite the fact that the median time for holding a stock is currently around 5 to 8 months. Even those who are buying the stock to hold it for at most a year are effectively making a prediction about the full shareholder return of the company for many decades to come. This can be shown using the induction method:
If you buy a stock right now (lets call it year 1) and plan to sell it in a year, you’re effectively predicting that another market participant will buy it from you for an equal or higher price in one year. Alternatively, you might be predicting that the stock will pay dividends, issue a buyback or be acquired by another company.
When year 2 comes along, you will only be able to sell the stock for a higher price if someone else predicts that the stock will go up in price or generate intrinsic value by year 3.
When year 3 comes along, the person who bought the stock in year 2 will have to find someone who makes a positive prediction for year 4.
When year 4 comes along, someone will need to find a person to make a positive prediction for year 5.
We thus can say that anyone buying a stock in year 1 is making a very long term prediction that the company in question will pay back its current value to shareholders over the long horizon.
So why is the price of Tesla so high? There’s obviously at least some component of irrationality but at the end of the day its still a decades long bet on the success of Elon Musk and his company, even if no one from the present day is planning to hold it long enough to see speculative value being converted into direct profits. Below is one version of what the future of Tesla could look like from the perspective of someone buying a stock today. The details will of course differ from person to person - some perhaps hope that Tesla will take over every industry on the planet while others will merely hope for stable returns. But on the aggregate the price indicates a belief in future value being received back by shareholders.
Comparison with pump-and-dump schemes
But wait - don’t Ponzi Schemes work exactly like that, with people buying and selling shares with the hope that someone else will buy it from them? Aren’t they making a decades long bet on their version of stocks? The difference is that Ponzi Scheme organizers rely on opacity to successfully run their scam. Bernie Madoff had an entire secretive floor rented out for running his operations, where no clients were allowed access. He falsified everything to hide his tracks, so at the end of the day his customers were making bets based on a collection of falsehoods. In comparison, Tesla is a fairly transparent company with quarterly reports, real world products and tens of thousands of corporate and factory employees who could potentially blow the whistle if something was amiss. Its investors might be wrong about just how valuable an electric car is or how much of a monopoly Tesla would have in the future, but at least they can personally see and drive those cars before making a final decision.
Okay, but what about highly dubious investments where almost everyone knows that the underlying value is zero and people are simply playing a game of who will be the last bagholder, such as penny stock rallies and short-lived cryptocoins? Aren’t they making iterative predictions which imply that the asset will stay profitable for decades? The answer is that while they do engage in iterative reasoning, their timelines are significantly compressed. Your average stock investor is making a prediction for 5-12 months ahead at a time, while your average memecoin prediction is for a week ahead at most. Iterative predictions gain uncertainty with each step, so the shorter the median bet is, the shorter is the implied longevity of the asset.