Dividends are a good case for the classic bell curve meme:

There are two historical factors that made dividends more important in the past. First, if you go back far enough, the default policy for most companies was to set a dividend that represented their long-term estimate of all of the company's profits. That seems perverse today: didn't they want to grow? But that made sense at the time: more companies had limited room to grow, either because they were railroads with high capital requirements or because they were manufacturers who, before the early 20th century, could only expand one factory at a time. These chunky increments of expansion meant that retaining earnings didn't really make sense, except when the company's capital needs changed. One year’s earnings couldn’t be reinvested at the same return as what the company’s existing business got, and if they wanted to expand, they’d have to raise a big slug of capital.¹ So from the very early investors' perspective, a stock looked a lot like a very risky bond that could (and would) suspend payment when the issuer was unprofitable.²

Retaining earnings for growth was a norm that was slowly established in the early twentieth century, at which point dividends started to matter—not because of constraints on companies' growth—but because of constraints on their ability to return capital through any means other than a dividend. That changed in 1982, when the SEC's changed some of its rules, and suddenly it became clear that buying back stock on the open market didn't constitute illegal market manipulation. (Before that, companies did buy back stock, but they did it all at once through tender offers at a premium to the prevailing price rather than through an ongoing process.)

But companies didn't all instantly switch to returning capital by way of buybacks instead of dividends, although there's an argument that they should have. Let’s look at how these options differ—in one case a stock has a 4% dividend yield, and in one case a company buys back 4% of its outstanding shares each year:

1. The dividend recipient pays taxes on the dividend (0%, 15%, or 20%, depending on the recipient's annual income). When a company buys back stock, it pays a 1% tax on the net buyback (in other words, for companies buying back stock to offset shares they issue through options, the tax does not apply). Shareholders who want to create a "synthetic dividend" by selling shares will pay capital gains taxes, but only on the appreciation of the stock, not on the absolute amount sold, so a buyback is more tax-efficient than a dividend for the shareholder unless the average shareholder is up 100x or more on their investment.

2. Buybacks increase the value of the company's shares when they're undervalued, and decrease it when they're overvalued. Dividends arguably do the opposite: when the stock is cheap, the industry is probably out of favor and retained earnings will have a higher expected value. When the stock is expensive, continuing to invest in the business probably produces poor returns, and those high valuations attract competitors who will bring down the value of the firm's assets.

One thing dividends offer that buybacks don't is a credible signal of where management thinks profits will trend over time. Most companies will pay a lower dividend than their earning power, so they can maintain it, which means that cutting the dividend is a signal that management has overestimated long-term earning power.³ Some companies highlight their impressive track records of maintaining or increasing dividends; AT&T was well-regarded for keeping its dividend at a constant $9/share throughout the Great Depression, and Hormel has raised dividends for 57 years in a row. That isn't a promise that they'll raise the dividend for the fifty-eighth year in a row in a few weeks, but it certainly sets that expectation.

Ultimately, a dividend just sets a publicly visible minimum goal for a company's average annual free cash flow. And if you dig through a proxy statement, you'll often find plenty of metrics that management is compensated for, and can use those to get a sense for what they're incentivized to do. Most investors don't do this, or may worry (correctly!) that the incentives will get adjusted based on what's achievable. A dividend is easier to check and harder to revise. So it's ultimately a way to keep management focused on fundamentals.

Read More in The Diff

We’ve covered capital retention policies and their valuation impacts a few times in The Diff.

• Electrocloud elaborates on the argument that electrification created the concept of a growth stock.

• In The Equity Risk Premium at 0% Interest ($) and Should We Expect Valuations to Mean-Revert Over Time? we look at dividends as an input into valuation models.

• Should Meta Be Run for Cash Flow? ($) looks at this at the level of a single company, with a focus on whether they should keep investing in their business or not. (Disclosure: writing that piece eventually convinced me to buy some Meta stock; still long.)

1. Banks and insurers had different norms, and often would retain more of their earnings. In the case of banks, though, limitations on branch networks meant that their capital requirements were based on how much business a specific location could do, not the total set of opportunities in a region or country. Insurers were similarly constrained.

2. This effect even extended to share prices. In the early 20th century, stocks usually had an issue price of $100 a share, and eyeballing the price was actually informative, much the same way that if you look at two recent de-SPACs, one of which trades at $2 and one of which trades at $20, you can instantly tell which one has done better.

3. Some companies will deliberately pay an unsustainable dividend, and fund it by issuing stock. This is extremely bad behavior: the stock price reflects investors naively buying for yield, and those investors are partly buying shares from the company itself. Over long periods, this achieves the same return as just paying a lower dividend, except that a) investors get taxed, and b) bankers get fees. Doing this repeatedly should be considered a violation of fiduciary duty, especially because part of the result is a continuous wealth transfer from unsophisticated buyers to professional short-sellers.

Help us learn more about you!

We're building a media kit and need some anonymous aggregated subscriber data. If you work in finance and have time for a 3-question survey, that would be very helpful:

Join the discussion!