The theoretical purpose of the stock market is to raise money for new companies. But if that's what it's for, it's doing a bad job: publicly-traded stocks return more capital than they raise, as the annual volume of buybacks in the US exceeds the total amount of IPO and secondary share issuance. So, in a sense, this theory is backwards: the stock market exists as a convenient vehicle for corporations to return capital to investors.

If markets are primarily a channel that turns savings into productive investment, they're bad at it. Other channels do roughly what they're supposed to do, albeit sometimes in indirect ways. For example, banks do intermediate between short-term depositors and long-term borrowers, but when they do that, it's by way of a third party: central banks are implicitly helping them achieve the maturity transformation whereby their liabilities, which include demand deposits that can vanish overnight, fund their assets, which can include loans and mortgages with a term of years or decades. The banks are still intermediating, and are still mostly providing long-term capital—if a local bank makes a small business loan, that loan typically funds some kind of investment, or at least helps the borrower roll over their loan so they don't have to liquidate an investment.

But there's another use case for equity markets: they provide the private benefit of liquidity, i.e. individual investors benefit from the fact that they can quickly turn stocks into cash and turn that cash into shares of other stocks. And they provide a public benefit by providing information. If you read an ominous headline about a sudden resurgence in inflation, for example, one thing you might do is look at the shares of companies that have more inflation exposure on the revenue line than the cost line, to see if they're acting as if there's runaway inflation. (You might also look at treasury bonds, or gold; the two-year treasury, for example, ends up being a very useful proxy for how investors are reacting to near-term inflation and growth expectations, since two years is a pretty good guess for the length of a typical rate hiking/cutting cycle.)

That pricing function is useful in another way: you can think of the price/book value ratio by industry as a rough guide to how much actual value gets created per dollar expended in different fields. This will be a very, very imprecise statistic, since many things that are economically an investment will show up as an expense on the P&L—when a company refunds a customer after a bad experience, the cost of doing that is subtracted from book value, while the present value of that person's continued patronage shows up in market cap. You can also look at a company's market cap compared to its net paid-in capital as another heuristic, though that ignores the time value of money. (This can be a good tool generally because sometimes there are surprisingly valuable companies that achieved almost all of that value from the judicious use of internal funds.)

Companies are slowly shifting to more stock-based compensation. This has been popular in the tech world for a long time—Fairchild Semiconductor was on the wrong side of an options trade, since they gave Fairchild Instrument an option to buy their entire business at a fixed price as part of their funding, but the Fairchildren went on to use widely-distributed stock options as a tool for aligning incentives. In other industries, it's gotten more common over time, and private equity is doing this, too ($, WSJ).

Once that happens, the value of prices as a signal goes way up, because the content of that signal is not just "this is a compelling business," but "working in this business will make you richer than the alternatives." That kind of career sorting is extremely valuable, because these businesses tend to create the infrastructure for the next generation of similar companies—topic-specific search products like Yelp, LinkedIn, and TripAdvisor all benefited from general-purpose search traffic, and consumer brands grew faster when social media provided an efficient way to scale their marketing.

This can eventually go too far (markets usually, eventually, do). But "too far" is a relative term; the tech market of early 2000 was wildly overextended, but would have been even more so if the tech stock market of 1995-1999 hadn't prompted so many people, and so much capital, to build things online. That's ultimately the function public markets serve: they're a helpful vehicle for allocating savings, they provide simple ways to return the value of that investment to shareholders over time, but the real thing they do is provide a signal to talent about where to work.

Read More in The Diff

The Diff often covers these dynamics, at varying levels of precision. Some examples:

• We looked at Blackstone granting equity to rank-and-file employees at firms they acquire ($).

• Startup compensation takes advantage of the wide bid-ask spread in private markets ($)—if there's a wide range of possible valuations for a company, paying with equity is partly a way to sell stock to people who put the highest valuation on it. (You can interpret this in a sociopathic way, by exploiting people, but in practice, it ends up being a nice motivator—as a founder, you're working for shareholders, and those shareholders include people who took a pay cut because they believed in you.)

• We've also looked at the history of equity compensation. Born as a tax dodge, it grew up into something else.

• And we've explored reflexivity in tech returns. Money chases talent, and talent chases money.

Join the discussion!