Earlier this week, Microsoft issued long-term debt for the first time in its history, selling $3.75 billion of 5-, 10-, and 30-year bonds. From a corporate finance perspective, I guess this makes sense, since it got to lock in historically low borrowing rates. Treasuries are low, and Microsoft paid only about one percentage point more than the U.S. government, which makes sense since it does have over $20 billion in cash, no other long-term debt, and – let’s not forget – a virtual monopoly on computer operating systems and basic desktop software. (In some ways, Microsoft looks like a safer place to lend mone than the U.S. Treasury, except for the ability of the latter to print its own money.) But it’s still a little sad.
First, we have an (optional) Corporate Finance for Beginners interlude:
There has been a ton of research on the optimal “capital structure” for companies, meaning the ratio of debt to equity. Conceptually, companies raise money from two sources: debt (selling bonds) and equity (selling common shares). Debt, as we know, has to be paid back; equity doesn’t. In the 1950s, Franco Modigliani and Merton Miller proved that, under certain (quite large) assumptions, the capital structure of a firm doesn’t matter. In the real world, however, there are a number of reasons why it should.
The simplest question is probably why any company would borrow money that it doesn’t need. After all, Microsoft has over $20 billion in cash, and it continues to make money. Taking on unnecessary debt isn’t something that I would do in my personal finances. But the idea is that companies are different: all of their money comes from and eventually belongs to someone else – for convenience, let’s call them bondholders and shareholders – so it’s just a question of which source is better. For one thing, a company could simply raise money by selling bonds, and then turn around and give the money to its shareholders. So from one perspective, you should ask the shareholders if they would rather have more cash in their pockets, or own a company that had less debt and was therefore worth more.
One major consideration is taxes. (The simple form of the M&M theorem assumes no taxes.) The interest on bonds is tax-deductible for the company, while dividend payments to shareholders are not; they are paid out of after-tax income. So in that sense debt financing is cheaper than equity financing.
But the broader issue has to do with the cost of capital. The cost of debt is pretty simple – it’s the interest rate you have to pay on the bonds. But the conceptual point is that equity has a cost as well – it’s the rate of return that shareholders expect to get by buying common stock. The higher that required rate of return, the less they will pay for your shares, and the more expensive it is to raise money by selling shares (because you need to sell more shares to get the same cash in return). Even if you (like Microsoft) have enough cash on hand that don’t need to sell shares, you could be using your cash to buy back shares on the market; by not doing so, you are implicitly “selling shares” and paying the cost of equity. That’s the concept, at least; trying to estimate your own cost of capital can be very difficult, and trying to estimate the cost of capital for a marginal transaction is pretty close to impossible.
Now, back to Redmond.
A company like Microsoft – with no debt, lots of cash, and a highly profitable business – has an extremely low cost of debt. If Microsoft took on a huge amount of debt, that cost of debt would go up, because investors would see that debt as riskier; the more debt you have, the higher the risk that you will default. But given where they are on the curve, the textbook answer is that they could afford to take on some debt.
But what is Microsoft going to do with that extra cash? There seem to be two main theories: (1) buy back more shares than it is already buying back and (2) buy companies.
I’ve never fully understood why some people think that buying back shares is always a good thing for shareholders. If your market value is $100, and you use up $10 in cash to buy back shares, then it’s true that there are 10% fewer shareholders that the value has to be divided between; but it’s also true that your company is worth exactly 10% less. In practice, share buybacks can boost share prices because they act as positive signals: if the company thinks its stock is underpriced, then maybe it is. But I don’t see how it can work as the long-term strategy that some companies, like IBM, think it is.
As for buying companies, Microsoft already had enough cash to buy any company that could actually have benefited by being bought by Microsoft. Sure, Yahoo! and SAP are out there, but they’ve already done all the innovation they’re ever going to do; from this point these companies are just playing a game of adding their earnings together and trying to minimize the number of shares to divide them by.
In short, issuing debt looks like just the latest step on Microsoft’s way to being a company that uses financial engineering to boost its share price rather than inventing new products. Now I know that Microsoft has thousands of very smart and ambitious employees, so the fact that it has become a sinkhole where talent goes in and nothing new comes out is sad. The simplest explanation is probably that Microsoft is not too big to fail (although maybe it is – what would happen to our economy if nobody were around to fix security holes in Windows and IE??!!), but simply too big to manage. In addition, software has a tendency to get more and more unwieldy and difficult to modify as it gets bigger and older, and Windows is one of the biggest and oldest programs around.
So maybe it’s a smart move. But it isn’t anything for Bill Gates to be proud of.
Update: Microsoft credit default swaps are trading at 37.5 basis points – 2 basis points higher than U.S. Treasuries, and lower than just about every other sovereign government except France, Germany, Finland, and Norway.
By James Kwak