Salesforce and the power of cash

Good morning. The theme on Wall Street yesterday was slowing growth. The softer economic reports discussed in the last letter, plus a weak manufacturing ISM survey, seem to have given everyone the idea that maybe owning more bonds is not such a bad idea. The yield on the 10-year Treasury fell 11 basis points yesterday, which is a lot. Unhedged is not panicking yet. If you are, email me: robert.armstrong@ft.com. 

Why hasn’t Salesforce been more magnificent?

The following analysis only exists because of a dumb coincidence: that perennially slow-growing HP (computers, printers, ink) and perennially fast-growing Salesforce (enterprise software-as-a-service) reported on the same day last week. The stock in the former rose a lot, and the latter fell a lot, as discussed then. And it wasn’t just that one day. Depending on the starting date you pick, the tortoise (HP) has kept pace with the hare (Salesforce) for quite a few years now. 

This got me thinking about Salesforce. Why haven’t its shares done better, given that it has grown very rapidly during the market’s love affair with fast-growing technology companies?

Picking a somewhat random date, I had a look at performance at a bunch of tech companies since the end of 2016 — not just Salesforce and HP, but several of the Magnificent Seven world-beaters and another slow grower, Oracle. Here they are, ranked in terms of total return performance: 

Looking from left to right, the striking thing is that Salesforce shares have performed like slow-growing Oracle and HP, despite the company having by far the strongest net income and earnings per share growth in the group, and trailing only Meta in revenue and free cash flow per share growth. 

The primary reason for this is obvious: at the end of 2016, Salesforce was a wildly expensive stock, as reflected in a price/earnings ratio of 146 (!). With a price like that, a 37 per cent annual EPS growth rate is a disappointment — though it should be noted that Salesforce generates more cash than earnings, so that its free cash flow valuation was a bit less insane (or at least looked that way). The secondary reason is equally plain: the company’s weak outlook for the next quarter of this year suggest the company could be, all of a sudden, leaving its high-growth era. This goes back to a principle that Unhedged holds dear: long-term corporate growth rates are very hard to predict. 

But there is more to the story than a swivel-eyed valuation and a growth shock. Over the past seven fiscal years, Salesforce has generated $34bn in free cash flow (defined as operating cash flow less capital expenditures). But over that same period, the company has paid out $15bn in stock-based compensation. Regular readers will know that Unhedged regards stock comp as a cash expense that has been deferred into the future. When all those stock options begin to vest, stock compensation must be paid for in cash, in the form of the stock buybacks required to keep the share count from rising and earnings per share from falling. This is exactly what has happened to Salesforce. In the past two fiscal years, Salesforce has bought back $12bn in shares just in order to keep its share count flat. 

So what we might call Salesforce’s true free cash flow — free cash flow adding back stock compensation expense — is just $19bn over seven years. But then we come to the acquisitions: $22bn of them in seven years. In terms of cash flow that could be returned to shareholders, then, Salesforce has burned $3.4bn in seven years. One could look past the acquisitions if the company was still growing like wildfire. But it isn’t. 

Two lessons. First, stock compensation expense really matters, even for companies that are growing rapidly. When it converts to a cash cost, redubbed as stock buybacks, markets notice. Second, the tech stocks that have done so well in recent years, such as Google and Microsoft, have done so not just because of rising profits and high margins. They generate oodles of distributable cash. In that sense, this market has been attuned to value, not just growth.

Valuations and liquidity

In the past week or so a fair amount of this newsletter has been devoted to various theories of why US stock valuation has been so high in the past 30 years. One theory that many people seem to favour is the fact that interest rates have been low, or at least falling, for much of that period.

The economist Andrew Smithers emailed me to object to this idea. In his view the crucial thing is not rates but unfunded government stimulus. “The prolonged overvaluation of the US stock market is surely the natural response of the private sector to having had a massive cash injection,” he wrote. 

The fact that the stimulus is not funded — that is, the central bank, rather than the public, has bought most of the bonds that fund the stimulus — is important because otherwise bond yields would rise, offsetting the stimulative impact of the spending. The cash handed to the private sector by the stimulus, if it is not consumed or otherwise invested, finds its way to the stock market.

The cash that is forced into the system must show up as bank reserves. And the relationship between those rising reserves and a rising stock market has been close since 2020, as this chart provided by Smithers shows: 

I’m sympathetic to this view. My only objection is that valuations have been notably higher since the 1990s, well before the Fed started expanding its balance sheet and bank reserves started to climb.

One good read

Everyone needs a fixer.

FT Unhedged podcast

Can’t get enough of Unhedged? Listen to our new podcast, for a 15-minute dive into the latest markets news and financial headlines, twice a week. Catch up on past editions of the newsletter here.