Stock Buybacks are a problem
(Dan here….written last year, still relevent)
By Steve Roth (originally published at Evonomics)
Bernie Sanders and Chuck Schumer’s New York Times op-ed, “Limit Corporate Stock Buybacks,” has thrown internet gasoline on the buyback debate. The left is waving the flag, and the right is trying to tear it down.
The core Sanders/Schumer argument: buybacks extract money from firms, money that could be used to pay workers more, and fund productive investment (including worker training and upskilling).
The counterargument: how are buybacks any different from dividend distributions that way? Both transfer cash from firms to households. We don’t hear people complaining about dividend distributions stealing money from workers and investment.
That counterargument is absolutely right, even while it’s completely wrong. Because both sides miss the overwhelming effect of stock buybacks (vs dividends). Buybacks are a massive tax dodge for shareholders.
Imagine Megacorp wants to transfer a billion dollars to its shareholders (notably including the huge shareholders in its C suite and on its corporate board). Whether they distribute dividends or buy back shares, either way Megacorp has a billion dollars less on its balance sheet. Its book value drops by $1B.
But what happens on the household, shareholder side? With a dividend distribution it’s simple; households get $1B in taxable dividend income. With a buyback, households that sell shares also receive $1B in cash, but they give up their shares, which obviously have value.
That’s where the (perfectly legal) tax avoidance lies — perhaps best explained by example:
Suppose the average shareholder’s shares were purchased for $20 each. That’s the shareholders’ tax basis. If Megacorp pays $25 a share (for 40M shares), the shareholders who sell have cap gains of $5 a share — $200M in taxable income — versus $1B if the same cash is paid out via dividends.
Dividends and long-term cap gains in the U.S. are currently taxed using the same rates and brackets: 15% if your income is above $38K, 20% if it’s above $425K. If Megacorp chooses a $1B stock buyback, our imagined shareholders pay $40M in taxes (at the 20% rate), versus $200M in taxes on a dividend distribution.
Neither of these has any effect on corporate taxes, by the way. C-corporation profits (as opposed to S corporations and other “pass-throughs”) are taxed at the corporate level, whether they get distributed or not, and no matter which distribution method is used.
There are other important problems with buybacks, mainly having to do with boards’ and C suites’ greater discretion over the timing and amounts of buybacks, and the potential for self-dealing price manipulation. (Contra Schumer and Sanders’ odd bank-shot approach to this problem, we could just repeal Rule 10B-18.)
But the right is right: Buybacks are no more pernicious than dividends in “stealing money” from firms, that could otherwise be used for worker pay and productive investment.
And they’re dead wrong that there’s no difference between the two. Buybacks steal money not so much from corporate wages, but much more obviously and explicitly: from taxes that contribute to our common public purse.
2019 February 4
“Neither of these has any effect on corporate taxes, by the way. C-corporation profits (as opposed to S corporations and other “pass-throughs”) are taxed at the corporate level, whether they get distributed or not, and no matter which distribution method is used.”
Thanks for noting this. Now if capital gains were taxed as they accrued at the same rate as dividends, then this tax dodge would disappear.
Oh come now. With a dividend, the shareholder receives $25 in cash and keeps a share they could sell tomorrow for no gain at $20. So you might as well just benchmark they sell their share and end up $20 cash ahead of a strict share sale at $25 bucks. So you describe a less beneficial transaction as some tax avoidance scheme? The dividend recipient pays more because $45 is more than $25.
Eric, your numbers don’t match what was given.
Company buys some shares – remaining shares increase in value. Most shares aren’t sold, have a capital gain that (maybe) gets taxed in the future. Shares that are sold get taxed on capital gain. So most of the distribution doesn’t get taxed. As a shareholder, it’s the same as a dividend if you sell, better if you hold. If the buyback is evenly distributed over all shareholders, they all win, as most remains untaxed.
The assumption behind the tax avoidance critique is that share buybacks result in a persistent capital gain. Unfortunately, the analysis I found on this was relatively sparse but tended to argue against its efficacy.
They can push up EPS to some degree, but that effect can easily be drowned out by other factor and the stock price is only loosely coupled to the current EPS and more so to a company’s future outlook.
Dividends have the benefit to the economy of introducing actual capital which can be selectively redeployed. The share repurchases pull money out of the company and put it into the market at a purely notional value. Typically since repurchases are done by healthy companies this would more than likely be a wash with capital largely swamped by other factors. The individuals selling their shares would have been able to find buyers at roughly the same price, so there is no net return of capital.
My intuition is that the repurchases result in a waste of resources for companies and a decrease in deployable capital to the economy as a whole.
All of that being said, I don’t think the share repurchases should be restricted, companies should be able to purchase their own and other shares. It is simply bad strategy. One option might be to allow companies to repurchase the shares but not to reabsorb them. The shares would simply be listed as assets like they would if they held shares in other companies. This would maintain the capital stock and not allow the companies to use the approach to manipulate their EPS statistics.
The ripple effects through the economy are arguable however the addition to many CEO’s incone is demonstrable?