Google … Or Auto Parts?
If you’re going to buy a tech stock, Google (GOOG) is hard to beat.
It’s not terribly expensive, at around 25x earnings. It has incredible prospects: its cars are driving themselves around San Francisco. It dominates the high-margin business of online advertising with its famous Google search engine. It has an enormous user base with its ubiquitous Gmail platform. And then there’s YouTube, which seems to be quietly winning the streaming wars.
Google is one of the most powerful tech companies in the world, and probably the biggest winner of the internet revolution that began in the late 1990s. But… is it a good business?
Its earnings just came out, so we have lots of timely and relevant data to consider.
When I judge the quality of any business, I’m primarily looking for return on equity (“ROE”): I want to see how much money the businesses managers can produce with their assets, because – assuming share count remains the same – that’s the rate at which the business is likely to compound over time. There are different ways to generate ROE (profit margin, asset-turns, and leverage). So, my next question is, how profitable is the business… and, then, finally, how much growth is there.
(These questions all assume that the company’s balance sheet is stable, that its products or services are widely embraced, etc.)
Last year, Google produced an enormous amount of revenue – up 14% over last year to $350 billion! That makes it one of the largest businesses in the history of capitalism. And, the company’s accountants claim (more about this below) that the firm ended the year with after-tax profits of $100 billion. Again, that makes it one of the most profitable businesses in the history of capitalism.
A few other figures we need to know. Google has total assets of $450 billion and total debts of $125 billion, giving it a net asset value of $325 billion.
Thus, we can see this is a pretty darn good business: $100 billion in after-tax profit on $325 billion in equity shows the firm is compounding its equity at more than 30% a year. That would make it one of the greatest businesses in America.
So… then why would the stock have fallen so much today – down 7% as of 2 pm ET? Seems like this is a great business, that’s growing pretty fast…
Well, the trouble with technology is that it’s always changing. Among Google’s listed expenses are $49 billion (!) on research and development (R&D). The company also has to pay a lot of really smart people to keep creating great software and services. Selling, general & administrative costs are also close to $50 billion a year. And that’s not including the $22 billion that Google spent giving employees shares last year.
When you add up all of this overhead, you discover that Google’s net income margin is only 28% a year.
While that’s not bad… that’s not the kind of profit margin you’d expect at a software company. And that’s because provisioning web services is pretty expensive: cost of revenue at Google last year was $146 billion.
So… maybe Google is only a good business, not a great business.
And there are a few other problems too.
The biggest problem is that the depreciation rate on tech investments keeps accelerating. Moore’s Law (which says that the speed of computer chips doubles every two years) used to make new computers obsolete every four to five years. But, Nvidia’s parallel-processing revolution has blown those growth rates to pieces.
As a result, Google (and all of the other big tech firms) are having to spend a lot more, year after year, just to keep up. That’s because most people using the internet don’t really care if they’re using Google search… or Gmail… or YouTube. There’s no moat around web properties, because all of these things are free. You’ll use whichever web service offers you the best technology. And that means that in order to keep its audience – and to keep selling advertising – Google must continue to spend heavily to make sure its web services are the best.
Last year, it spent $52 billion on these major capital investments. As you can see, that’s a major portion of the money it made.
Simple question for you: If a business has to spend billions and billions every year on major equipment upgrades, then shouldn’t those expenses be included as a line item on the income statement?
Google doesn’t count these as expenses because its accountants view these costs as investments. The only trouble is, as I mentioned, tech investments have been depreciating faster and faster every year.
If these tech upgrades were treated as expenses (and I’d argue they should be) then Google’s business would look a lot different. It would have made less than half as much money last year – only $48 billion on sales of $350 billion (rather than $100 billion), for a net income margin of only 13%.
That’s not a good business at all. Judging the business on an after-capital investments basis would leave you with a paltry 14% ROE, which is barely better than average.
And then there’s this troubling fact: While Google expenses the stock options it gives to employees ($22 billion is charged against income), it actually spends way more than it expenses to prevent its share count from rising year after year.
Last year, Google spent $62 billion buying back shares.
And, again, most of these costs didn’t count against its earnings. That seems odd, doesn’t it? Google must continue to invest in both building out new data centers and it must continue to buy back the shares it gives to employees. (Over the last decade, Google has lowered share count by 11%.)
These are real costs of Google’s business… but they don’t count against its earnings.
If they did, Google wouldn’t have made $100 billion last year… it would have lost $14 billion.
And now Google says it will have to spend $75 billion on still more capital investments this year. That, combined with ongoing big stock-options expenses could make it impossible, once again, to actually make any money at all this year.
I’d be willing to bet that there are not many investors who realize any of these facts.
To give you some further perspective on how important it is for investors to understand all of the financial statements (not only the income statement – which, as you can see, leaves out many of the actual costs of running the company), let’s take a look at a different business that I’d wager most of you don’t think will outperform Google in the years ahead: AutoZone (AZO).
AutoZone is the very definition of a boring business that shouldn’t be worth very much. It’s not growing fast (annual revenue growth is around 5%). It doesn’t have high margins – it’s a physical retailer. On annual revenue of $18.5 billion, it earns a paltry 14% ($2.6 billion) in after-tax net income.
This is, I’m sure many would argue, a pretty lousy business. Or is it?
As I mentioned earlier, the real test of a company’s quality is how fast it can compound its equity. You can think of ROE as being like the annual return of a mutual fund. If a company is compounding its equity at high rates (over 20%), then it will almost certainly outperform the stock market over time.
There are three ways that companies can increase ROE. 1) They can increase their profit margins – that’s the most obvious way. 2) They can use debt to add leverage to their business model, which is also a pretty obvious way of increasing returns on equity (because adding debt reduces net assets). Or 3), the hardest way: by increasing the inventory turnover ratio. Selling multiples of your company’s assets each year can dramatically increase the amount of earnings being generated from the asset base, increasing total ROE.
And while AutoZone can’t do much about its profit margins (retailing auto parts is extremely price competitive), it has become a master at using its balance sheet and turning its inventory.
AutoZone keeps only a sliver of equity on its balance sheet – in fact, its net equity is less than zero. Beyond the stores it owns ($14 billion worth), AutoZone funds virtually all of its operations with vendor financing ($7.5 billion) and leases ($3 billion). It is as asset-light as possible, so to calculate ROE (we can’t divide into a negative number), we have to approximate a net asset value by subtracting long-term debt ($9 billion) from total assets, which leaves us with $8 billion.
As you can see, with $18.5 billion each year in sales, it’s “turning” a multiple of its net assets every year, driving ROE higher. Likewise, by financing so much of its operations with vendor financing and debt, it uses a lot of leverage. The combination leads to a return on net assets of approximately 25% a year.
That’s a great business.
Plus, unlike Google, AutoZone doesn’t have to hire the world’s smartest (and most expensive) employees. And it doesn’t have to give them a bunch of stock. So, year after year, because AutoZone spends virtually all of its earnings buying back its own shares, the share count rapidly falls. Since 1998, Autozone has retired 89% of its shares. Over the last decade, the share count has fallen from 30 million to 16.6 million, a decline of 44%.
Will Google maintain its supremacy in the world of high-tech web services? I bet it will. But I also believe that’s going to cost a lot more than most investors expect. I doubt that Google will work out very well for investors over the next five years as the AI bubble booms and busts.
In the meantime, I’m virtually certain that AutoZone will continue to compound at over 20% a year in a simple business.
So… which stock should you own?
Get Porter in your inbox… Every Monday, Wednesday, and Friday, Porter Stansberry will deliver his Porter & Co. Daily Journal directly to your inbox. He puts his 25+ years of investment knowledge into every punchy, fresh, and insightful issue… that’s free, with no strings attached. Everything is uncensored, and nothing is off limits. To get the Daily Journal, click here.