Is Amazon cheap?

Good morning. Today is Fed day, which means that tomorrow’s letter will probably be all about rates and inflation. So today I have avoided those topics, and look again at Amazon, which I think remains the most interesting company in the world — in part because it is so hard to assess from the outside. I’m very keen to hear readers’ thoughts on what Amazon’s future looks like and what the company is worth. Email them to me: robert.armstrong@ft.com.

Is Amazon cheap?

Amazon reports tomorrow, and some observers are a little worried that the results might disappoint.

Amazon’s cloud computing business, Amazon Web Services, likely kept roaring ahead in the June quarter. But the retail operation has been reeling a bit. Now that the coronavirus pandemic-driven online shopping boom has blown over, it has been left with excess capacity, falling margins and little growth. Retail sales fell slightly in the last quarter, when third-party sales on Amazon’s site are excluded, and the unit has run at a loss for the last three quarters. High fuel prices will have increased shipping expenses this quarter, too. The result could be a “choppy” quarter, according to Morgan Stanley analyst Brian Nowak. That’s a euphemism, I think.

But if there was ever a company where it was dumb to worry about a single quarterly result, it’s Amazon. This is a company that has remade two industries in the last 20 years, all while showing a lordly disdain for short- term profit. If you like the big story at Amazon, a terrible quarter is nothing but a buying opportunity.

So is Amazon cheap, in the big picture? It’s certainly cheaper than it was. Its market capitalisation peaked at $1.9 trillion about a year ago. Now the company is worth $1.2tn. That’s a lot less.  

It feels odd to even ask if a company trading at 50 times next year’s earnings estimate is undervalued. But because Amazon (and, crucially, its investor base) care so little about near-term profit, multiples of earnings or cash flow don’t tell you much.

So what do you do? The relatively easy part is valuing AWS as an independent business. It has churned out operating profit of $21bn over the past 12 months. Tax that at 21 per cent, and that comes to net income of $16.5bn. A software business growing at nearly 40 per cent in an expanding industry could be valued, conservatively, at 45 times earnings (that’s around where Salesforce trades, and Salesforce is not growing as fast). So AWS is worth $740bn.

Or rather $740bn-ish. We’re talking big, round numbers here, which are the only kind of numbers that are worth talking about with a business and an industry changing as fast as Amazon’s. If you find yourself cutting it fine you’re just kidding yourself. I’m confident that this is a conservative valuation, and if you don’t like my estimate, Barron’s had a nice feature last weekend that went through several other analysts’ sum-of-parts models.

Now we turn to the hard part: the retail side of the business. The market cap of Amazon as a whole (which is an acceptable measure of value to use because the company has no net debt) minus the AWS value just arrived at comes to $425bn or so. So, our question becomes, is Amazon’s retail business worth more or less than $425bn?

Just to give a sense of how big that number is, Walmart, including its debt, is worth about $390bn. Walmart’s sales this year will run to about $600bn, growing in the low single digits. Its operating margin runs to 4 or 5 per cent. Amazon’s retail business is a third smaller and has much tighter margins; it used to grow much faster than Walmart, but didn’t last quarter.

Now, it used to be possible to use the Walmart comparison to come up with an estimate for the value of the retail side of the business. You’d estimate what level Amazon sales would reach in five or 10 years, pencil in a Walmart-like margin structure (on the assumption that Amazon would eventually invest less and earn more), put a growth retail multiple on the resulting net income number, and voila.

But if this sort of back-of-a-napkin hijinks ever made sense, it does not any more, for two reasons.

The first reason is that the retail side of the business is now a very heterogeneous group of businesses. While we know what they earn collectively (previously, a little; right now, nothing) we have no idea what the potential earnings power of the individual businesses are. Amazon’s online advertising businesses (it is mostly ads that appear alongside Amazon product searches) had $33bn in sales in the last year. How much of that is profit? I don’t know, and I don’t think anyone else outside of Amazon does, either. It’s a unique business: you can’t just assume Google’s margins, or whatever. A similar story applies to Amazon’s growing physical stores business, and its various video, audio and ebook units. All this stuff is in a black box, profit wise.

The second reason it is hard to model Amazon’s retail business is that, in the wake of the pandemic boom, it is unclear how fast its core operations will grow in the next few years. Right now, with the tough comparisons to the pandemic, it is not growing at all. That won’t go on forever, but will we see double-digit growth again? The really tricky thing is, if the sales growth rate is now permanently lower, one starts to wonder about how competitive the online retail industry will soon become, and therefore how much Amazon can expand its retail margins in the glorious future Amazon investors have been dreaming of for two and a half decades. Are you confident that Amazon can aspire to Walmart-like mid-single-digit operating margins in its online retail operation? I’m not. At its core, Amazon’s diversified online retail empire might not be very attractive, economically.

At a gut level, I feel that Amazon’s retail operation, with its reach, diversification and superior infrastructure, has to be at least as valuable as Walmart, and is probably worth significantly more. If that’s true, Amazon is probably pretty cheap right now. But the key word there is “feel”. The more I think about Amazon’s online retail business, the less I understand it.

One good read   

Good luck to this poor guy.