Last week the fight over new capital requirements for large banks took a strange turn. Michael Hsu, acting director of the Office of the Comptroller of the Currency, tried to connect the controversy to banks’ dividend and share buyback policies.

His statements were bizarre because they needlessly politicized the issue–capital distribution policies have next to nothing to do with capital requirements.

In the Financial Times, Hsu implied that if the new federal capital requirements raise banks’ borrowing costs, it doesn’t really matter because banks can simply cut their dividends and buybacks. According to Hsu, “The banking industry has said the new rules are going to hurt all kinds of folks in the real economy. I have encouraged them—provide analysis on what your share buybacks and dividend policies are going to be under these different scenarios. Because there’s a choice to be made with capital.”

This statement is strange for several reasons. First, whether companies choose to pay dividends or repurchase shares with residual profits does not hurt people in the so-called real economy. Second, while the cost of banks’ capital ultimately affects their profit and, therefore, how much they have left to distribute to their shareholders, switching their payout method (or reducing their payout) does not lower their cost of capital.

To be fair, perhaps Hsu is referring to some other “choice to be made with capital.” But the first sentence of the article says “A top US regulator has pushed back at banks’ claims that stricter capital rules will increase borrowing costs, saying the lenders could always cut dividends and buybacks instead.”

Additionally, near the end of the article, Hsu states that “…if banks are saying it’s really hurting the real economy, but they’re providing increases in their share buybacks, I would suggest that there’s some questions for the bank about what the priorities are with regards to that comment.”

Hsu is clearly promoting the idea that if a company increases its share buybacks it hurts the real economy, and that’s a great indication of how political he’s making this issue.

Companies create value by investing in people and capital goods so that they can sell products to consumers. When successful, they can return cash to the owners. Failing to sufficiently invest in either capital goods or people over the long term defeats the very purpose of forming a company.

This idea is so obvious it’s not worth debating, but critics such as Senator Elizabeth Warren (D‑MA) have been attacking share buybacks for years. These critics insist that companies repurchase shares to “pump up” their share price rather than investing in their business. That might seem like a great short-term strategy for corporate managers, but it isn’t.

First, managers who do things at the expense of taking positive investment opportunities earn a reputation worthy of never being hired to run another company. Furthermore, a share repurchase is an exchange of cash for shares–it uses part of the company’s value (cash) to buy back another part of the company’s value (equity). The result might be a boost in earnings per share, but not a boost in value.

For all these reasons, Hsu’s comments are just a distraction from the debate over whether large U.S. banks need more stringent capital rules. It is far from clear that they do. Yet, the new requirements will impose higher costs on these banks and possibly put them at a disadvantage to foreign-based banks. Any real gain in safety and soundness is also murky.

In fact, multiple federal banking regulators–Fed chair Jerome Powell, Fed governor Michelle Bowman, FDIC vice chair Travis Hill, and FDIC board member Jonathan McKernan, just to name a few–have voiced concerns with the proposed rules. Aside from their concerns, the 316-page proposal devotes less than five pages to explaining the “impact and economic analysis” of the new requirements, and that discussion is so general it’s virtually pointless.

Whether the rule is truly arbitrary is almost sure to be litigated, but this whole exercise exposes a bigger problem: The regulatory capital framework itself is highly flawed.

Under this regime, virtually no weight is given to people’s ability to manage their own safety. Instead, a handful of government officials dictate how carefully banks must treat all kinds of financial assets and activities.

Supposedly, this regime is an improvement over a simpler flat capital requirement. But the first version of Basel failed because regulators got the risk weights wrong on (among other items) mortgage-backed securities. And while the second version was being finalized, the 2008 financial crisis hit, so regulators started working on Basel III and never fully implemented Basel II. Some folks portray the new rules as the final touches on Basel III, but the new proposal is sort of Basel IV-ish.

Regardless of the version, this approach will always fail because nobody has the knowledge to correctly assign risk weights to thousands of assets and activities. Nonetheless, federal regulators have an enormous amount of discretion to set banks’ capital rules as if they do have the knowledge.

The best way to get closer to the “right” bank capital levels is to put them to the test of the market. Even then, there is no reason to think that bank managers will always get the amounts correct, but that’s not really the point. The point is that banks want to stay in business, so they would have to convince people they’re not too risky. (Yes, government backing, even if implicit, makes this virtually impossible, but that’s a problem for another column.)

Nobody–not even many of the folks in the banking industry–wants to hear this, but the whole issue comes down to what is best: operating in a free enterprise system or in one where government officials decide, at a very detailed level, how people get to use their money.

Tragically, the United States has moved further in the direction of the latter approach for decades. But it doesn’t work, just like micromanaging the broader economy doesn’t work.

Bankers have little incentive to say it, but Congress should ditch the current approach and reduce regulators’ discretion. For a very high cost, the current regulatory regime provides a false sense of security and little else.