It's perhaps the greatest debate raging in the market right now. In a market trading at 13-15 times earnings, where even the Apple (AAPL) behemoth commands just 11 times forward 2012 earnings (and that's not excluding the cash it holds), Amazon (AMZN) stands tall, trading at 3000 times TTM earnings or 143 times forward 2013 earnings.
This type of discrepancy can only happen because the market expects Amazon to suddenly spring a huge gusher of profits. Some of this is warranted, because if we look into Amazon's P&L, we'll notice a huge TTM $1.86 billion charge for depreciation and this happens on just $5.7 billion in fixed assets. This is a very fast, 3-year depreciation schedule, betraying the nature of the assets being depreciated. They are probably mostly AWS technological assets such as servers, which depreciate fast. This depreciation is surely not for fulfillment centers, as those sit outside the balance sheet (in operating leases). Were Amazon to slow down AWS investment, this depreciation charge would start running off after 2 years or so. If Amazon were to cut depreciation in half, $0.93 billion would flow to the bottom line, at a 30% tax rate and 460 million shares, such would mean a positive EPS impact of $1.41 per share.
But obviously, Amazon is not trading as if it expects $1.41 or $1.50 in earnings per share. What Amazon shareholders also expect is for the profit nirvana to extend further - for the former glory to be re-attained, for Amazon to at the very least match and exceed Wal-Mart's (WMT) profitability while keeping its own faster growth. Were Amazon to match Wal-Mart's net profitability, now standing at around 3.5%, and its estimated 2012 revenues of $62 billion, it would produce a net profit of $2.17 billion. This would give Amazon an EPS of $4.72, and thus a P/E of 53.6, high but no longer outlandish.
There's just one problem, and also one clue
Thing is, Wal-Mart's business is not exactly like Amazon's. In Wal-Mart, the customers drive to the fulfillment centers, pick and pack their purchases, pay and leave - handling the delivery themselves. As for Amazon, when a customer orders something, he fully expects Amazon to eat the cost of picking, packing and worse still, most times, delivering. And Amazon, when making this delivery, has few choices. It can use United Parcel Service (UPS), FedEx (FDX) and at times USPS, but that's about it.
Basically, for part of its value chain where Amazon has to eat the costs, Amazon must outsource a service that's provided by an inflexible oligopoly. This is where we can get some type of clue as to what happens in a similar situation. Because, oddly enough, there's just another completely unrelated industry where a very favorable situation is paired with exactly the same kind of bottleneck.
Meet the PRB, Powder River Basin. This region, in southeast Montana and northeast Wyoming, supplies 40% of the U.S. coal needs. And it does so cheaply, very cheaply, because you can actually just dig it out of the ground with no underground mines and all. Its cost can actually run lower than $10 per ton, where other more traditional mines in the Appalachia hit $40-$60 for slightly higher quality coal. Were there no other effects, the PRB would bury every other coal mine in the country (in thermal coal).
Yet, this is not necessarily the case. This happens because PRB coal, much like Amazon's products, has to be delivered to the final customers through an outsourced delivery method which is for all intents naturally structured as a monopoly/oligopoly. That would be the rails, including Burlington Northern Santa Fe Corporation which Warren Buffett's Berkshire Hathaway (BRK.B) took over during 2010, and Union Pacific Corporation (UNP).
What happened, then? Coal from the PRB could be had for $10 per ton, but then transporting it would run another $30 per ton. In the end, this is what happened to the profits of Union Pacific Corporation:
Now compare those to the profits of, say, Arch Coal (ACI), which has most of its coal coming from the PRB - and for a second, forget the fact that ACI had a coal bubble in there due to Chinese demand:
UNP has seen mostly increasing earnings over time, whereas ACI has seen volatile, and ultimately plunging, earnings. Obviously also affected by things like the shale boom and the economic crisis, but still it remains obvious that UNP's economics were highly favored in this relationship. As long as the volume flows, UNP wins. Not so with ACI.
Amazon's situation is similar
Now fast forward to Amazon's relationship with FedEx and UPS. Below we see how FedEx and UPS did over time. FedEx and UPS did take a severe hit during the 2008 economic crisis since they were still very vulnerable to wide fluctuations in shipping volume, but quickly reestablished profitability.
And now Amazon … here's your online, technological, coal digger:
So what's the most probable answer for how much money Amazon will end up making? The answer is "not much". Other than what it stands to show as profits if it slows down the AWS investment, there's not a whole lot of reasons to believe Amazon will profit that much from the e-commerce trend. On the other hand, it's highly likely that UPS and FedEx will continue to see gains from such a trend, also evident in how both companies keep on raising tariffs even when crude is rather stable.
As we go forward, online commerce stands a good chance of seeing most of its economic returns being absorbed by the oligopoly standing downriver from it. Amazon will be like the coal producers of the PRB, even if they stand on an incredible opportunity to service the market cheaper, all or most of the economic benefits will flow to the rails. In this case "the rails" are UPS and FedEx.
Obviously this is not what the market presently discounts. The market presently discounts that Amazon will rise from the profit ashes as if it were a phoenix, able to out-profit even the mighty Wal-Mart. The market prices this as if it is a near-certainty. Due to what I explained in this and other articles, however, this is far from a certainty and might even be quite unlikely to happen. The end result will be a severe drop in Amazon stock.