That brings me to our newest position, which will no doubt make some question our credentials as value investors: Amazon (AMZN).
Consensus forward earnings for Amazon are a little over a dollar. At the median forward P/E multiple, Amazon would be priced in the low $20s. So, even though the stock fell $124 from its January high of $408 to a May low of $284, its P/E ratio remained in nosebleed territory. But we have never believed the P/E ratio was the be-all and end-all for valuation. Amazon is a retailer – a very efficient retailer. When we compare stocks in the same industry, we often compare their market caps to their sales rather than their earnings. Since 2001, Amazon has generally traded at a cap-to-sales ratio of two to four times that of the average bricks-and-mortar retailer. Having fallen to just under two recently, one might say that, as an advantaged retailer, Amazon looks somewhat attractive.
But that metric misses an important change in Amazon’s business. Third-party sales (sales on amazon.com where the seller is not Amazon) have grown more rapidly than Amazon’s direct business. And on those transactions, accounting rules credit only Amazon's commission as revenue. So if you buy a $100 item on amazon.com from a third party, Amazon is only allowed to show about $13 of revenue, nearly all of which is gross profit. For third-party sales, Amazon is effectively functioning as the mall owner, collecting a percentage of sales as rent. Amazon earns less gross profit on that sale than an average retailer would, but it is also a much lower risk endeavor. For that reason, we think a dollar of third-party sales should be worth about the same as a dollar that Amazon sells directly.
It gets interesting when we adjust our cap-to-sales ratio comparison to include estimated gross third-party sales. Instead of selling at twice the ratio to sales of the average bricks–and-mortar retailer, Amazon is selling at only 80%. So, relative to gross sales, Amazon's stock would have to increase 25% to be priced consistent with the very companies whose survival Amazon is threatening. On that metric, Amazon has never been cheaper.
Should Amazon sell at a discount on sales? The answer largely rests on what Amazon could earn if it wasn’t investing so heavily for future growth. For most asset heavy businesses, growth investment is primarily on the balance sheet, and is slowly expensed on the income statement as depreciation throughout its useful life. In an asset–lite business like Amazon, however, most growth spending gets directly expensed to the income statement, creating a much larger immediate reduction in income. We believe that if Amazon sharply curtailed its growth spending so that it only grew at the rate other retailers grow, it could produce similar operating margins. But we don't want them to do that. We believe that management is maximizing value by investing heavily for super-normal organic growth. So, yes, Amazon is a rapidly growing business. But at this price, we believe it is also a value stock.