Felix Salmon

No dividend, no worries

Felix Salmon
Nov 30, 2011 01:52 UTC

Karl Smith made a funny point in response to my post about Apple’s falling p/e ratio: since Apple’s not returning any money to shareholders in the form of dividends or buybacks, he says, shareholders aren’t getting any return on their investment.

Unfortunately, Matt Yglesias didn’t seem to get the joke:

The crux of the matter, as I see it, is Apple’s ever-growing cash horde which went from $70 billion in liquid assets at the end of Q2 to $82 billion in liquid assets at the end of Q3. The company is earning huge profits, which is great, but since it seems determined to neither return those profits to shareholders nor to re-invest them in expanded operations it’s hard to see how investors aren’t going to discount the value of the enterprise.

This is trivially wrong. If Apple’s cash pile is growing, that will increase its p/e ratio, rather than decrease it. On April 20, Apple reported Q2 earnings of $6.40 per share, or $20.98 over the previous 12 months. It closed the following day at $350.70, which corresponds to a p/e of 16.7 on a TTM basis. On July 19, Apple reported Q3 earnings of $7.79 per share, or $25.26 over the previous 12 months. It closed the following day at $386.90, which is a p/e of 15.3 on a TTM basis. The earnings were up, the price was up, but the p/e ratio was down.

Now Apple has roughly 1 billion shares outstanding, so let’s say that its “cash horde” went from $70 per share to $82 per share over the course of the third quarter. That’s more than 20% of the share price, right there. Take the cash away, and the p/e ratio falls to just 12. Even if you value the cash horde at just 50 cents on the dollar, the p/e ratio still falls, to 13.7 from 15.3.

It’s possible that shareholders would like to receive the cash as a dividend payment — although if and when that ever happens, they will have to pay income tax on it. They might even value Apple more highly if they can see themselves getting a modest income from their Apple stock without having to sell any shares. But we’re talking very marginal effects here: there’s no real sense in which turning a dollar of cash into a dollar of dividend payment increases the value of a company. Indeed, once the dividend is paid, the stock price will go down, since it no longer reflects the value of that cash.

I suppose it’s theoretically possible that investors are valuing Apple’s cash at zero, on the grounds that they’re never going to see any of it. But even if they are valuing the cash at zero, that doesn’t change Apple’s p/e ratio, which is still falling. What makes no sense is Yglesias’s idea that Apple with zero cash would somehow be worth more than the same company with $82 billion in the bank.

Smith’s point is a bit more subtle, and is probably best expressed in terms of the theoretical idea that a company’s share price should equal the net present value of its future dividends. If it never pays a dividend, and will never pay a dividend (or get bought), then the value of the company is zero.

I’ve been critical of Berkshire Hathaway’s no-dividend policy, but largely because the company’s shares are so ludicrously expensive that you can’t raise cash by selling just a few of them. Anybody who started with a decent Apple shareholding and then rebalanced annually to keep Apple a certain percentage of their total portfolio would indeed have received very healthy cash dividends, over the years, from the proceeds of all the shares they sold. And meanwhile, Apple’s shareholders get to hold on to all of the company’s earnings tax-free. (In fact, insofar as those earnings are kept overseas, they’re saving on tax twice: first when Apple repatriates the money and pays corporate income tax on it, and secondly when they pay personal income tax on their dividend income.)

It’s very easy, of course, to run a discounted cashflow model on Apple: such things model earnings, not dividends. And although there are some mutual funds which only invest in stocks which pay a dividend, I don’t think their absence from Apple’s shareholder base explains any part of its low p/e ratio.

And in any case, the joke behind Smith’s post is just that even if the lack of a dividend can explain a depressed p/e ratio, it can’t explain a falling p/e ratio. No one expected Apple to pay any dividends two years ago, when the stock was trading on a p/e of 32. Why should they suddenly care about such things now?


Another point is that the book value of Apple is increasing as they hold on to retained earnings. Assets, after all, do have value. Especially cash. If they are able to continue growing revenues without reinvestment of capital, why not keep the asset as cash? In the future if Apple finds a project they estimate will warrant an investment of capital for lucrative future returns in a more friendly business climate, they will have the capital on hand to do so. Why invest the money now in an unfriendly business climate with a low expected return? Obviously, Apple sees what a lot of other businesses see now, regardless of political rhetoric. There is not a lot of confidence that in the future, there will be a market for the public to adopt new innovations in a stagnant economy. If the risks of the cash investment losing value didn’t outweigh the probable expected return on the reinvestment, they would be reinvesting. If all it took to raise a stock price was to pay dividends, every company would be paying out everything they could in dividends. Plus the tax implications already pointed out.

Posted by Anonymous | Report as abusive

The Fed’s 1-cent-a-share dividend cap

Felix Salmon
Mar 23, 2011 20:19 UTC

Antony Currie wants a bit more clarity on why the Federal Reserve seems to be happy with Bank of America paying a dividend of 1 cent per quarter, but unhappy with a dividend of 5 cents per quarter. “It’s not clear,” he writes, “whether the Fed is worried about BofA’s core earnings falling short or about potential losses being higher than the bank projected, to pick a couple of possible concerns.”

My feeling is that the Fed’s logic wasn’t nearly that granular. There are basically three states that a bank can be in: it can pay no dividends at all; it can pay the minimum possible token dividend of 1 cent per quarter; or it can pay out a full and generous dividend. The Fed, in the wake of its latest stress tests, has determined that both Citigroup and Bank of America belong in the middle group. They can pay one cent a share, but no more — and in Citi’s case, they can only pay that much after doing a 1-for-10 reverse stock split and bringing the share price up above $40.

Paying a dividend is important from a symbolic perspective. It signals that the bank isn’t worried about insolvency, and it forces the bank to pay all dividends on preferred shares in full. If banks are paying dividends, that helps shore up confidence in the banking sector. If banks aren’t paying dividends, that keeps investors worried about what problems might lurk under the surface.

At the same time, the Fed has every interest in forcing banks to keep anything over one cent per share as precious capital, rather than sending it out to shareholders. The shareholders can’t do much with the money — not in this environment — while the system as a whole is clearly more robust the greater the amount of capital that banks manage to build up.

When Shira Ovide, then, characterizes the Fed’s stance towards BofA as “No Dividend for You,” she’s going too far. The Fed’s happy with BofA paying a dividend — indeed, astonishingly, BofA has been paying a dividend all along, with no objections from its top regulator. It just doesn’t want BofA paying out a dividend which is linked to profits. Not yet. For the time being, just like Citigroup, it’s being kept at that flat 1-cent-per-share level.

Update: I’ve heard a lot about mutual funds which are only allowed to invest in stocks which pay a dividend — sometimes with an explicit Berkshire Hathaway exception. Insofar as such funds exist, and I’m unclear on how prevalent they are, it makes a lot of sense to pay a de minimis dividend of one cent.


“I can say with 100% confidence that BofA’s deposits are money good.”

I very specifically was talking about corporate revolvers, not FDIC insured deposits. Corporate revolvers are not FDIC insured because 1) they are liabilities of the corporation not assets and 2) they are well over the FDIC limit. Corporate revolvers are an interesting product because, somewhat like swaps, the credit risk goes both ways in the transaction. The banks rely on the corporates to repay their borrowings under the revolvers and the corporates count on the banks to be able to fund the undrawn balance of the revolver whenever the corporation needs liquid funds. Some borrowers had to find new lenders to replace Lehman when it wasn’t able to fund its revolver commitments.

“They have repaid their goverment funding with interest and now they should be allowed to resume their dividend payout at a low rate.”

Shouldn’t they be earning money to pay out a dividend? If the dividend is higher than earnings (it is since earnings are negative) then leverage goes up (assuming steady asset levels). This is easy math here. Also, lets not pretend that BAC’s GSE settlement wasn’t a transfer of value from the Govt to Bank of America.

“All banks, every one depend on a an explicit govermental guarentee via the FDIC which can borrow directly from the U.S. Treasury.”

Which is exactly why regulators Govt regulators get to approve or deny dividend policy.

“If you want to curb the growth of the evil mega-banks and promote the growth of Credit unions and Mutual savings banks (like mine)”

I work for a competing “evil mega-bank” so I can assure you that was not the motivation of my post.

Posted by TurtleBay | Report as abusive