You can’t trust EBIT (AMZN/ASC)

The other day I was discussing with one of my colleagues the differences between tech valuations in the UK and in the US. There’s a good reason the real tech giants in the world are US companies, as far as I’m concerned, and you see a reflection of general investor sentiment in that respect when you look at the pricing of the stocks and the attitudes of management teams.

The simple fact is that US public equity investors seem to have more of a stomach for ongoing losses in exchange for top-line growth. Is this a good thing? Clearly, as a value investor, you wouldn’t expect me to say anything other than the usual spiel about racy tech stocks and their absurd valuations. One thing I do buy, though, is that the distinction between operating expenses (opex) and capital expenditures (capex) isn’t quite as solid as line as many investors seem to think.

Amazon is the perfect example of this. Amazon ‘makes very little money’, but it’s valued by the markets at $140bn. It has ‘only’ invested $10bn in actual tangible assets in the last 10 years, and during that whole 10 year period the company only made cumulative pre-tax operating profit of $7.5bn. What’s going on? Clearly you aren’t paying for tangible assets  – though the infrastructure in place is formidable and not to be sniffed at as naively as I have here – and clearly you aren’t paying for earnings.

A simplistic analysis like the one I’ve just displayed misses one clear facet; in Amazon’s case, opex is capex. They are spending money at the operating level, but it is in effect a long-term investment. Consider the following example of a traditional capital expenditure, on a new piece of factory equipment:

Widgets Plc. invests £1m in the latest German widget press

The widget press gives them a sizable edge over the competition, by enabling them to produce widgets at a lower unit cost

The company is hence exchanging cash now for increased future profitability

Clearly, this is a run-of-the-mill procedure, and simple from an accounting point of view. The asset goes on the balance sheet, the cash comes out of the cash-flow statement, and the P&L is unaffected.

We all have no problem designating this as capex. This is short-term investment in exchange for a long-term gain. But Amazon’s whole business model is based on something which is arguably very similar:

Amazon has the opportunity to sell products at significantly higher markups than it does, but this would hamstring its growth

Instead, they choose to keep margins extremely low (does anyone think on a run-rate basis Amazon will actually make a 1% operating margin?), continuing to grow the top line and creating scale efficiencies, which enables them to stock more products at a lower unit cost – their ‘virtuous cycle’.

… The company is hence exchanging cash now for increased future profitability

And, again, this is a run-of-the-mill procedure, and simple from an accounting point of view. Your report weak margins on your P&L, nothing goes on the balance sheet, but – if your sums are good – you’ve made a value accretive decision.

There is one group of companies where this conflict – the treatment of investments for future growth – is particularly interesting, and that’s in technology businesses. Technology businesses get a bit of flexibility in how they treat their R&D expenditures; some are expensed, and run through the P&L (reducing profits) and others are capitalised (stuck on the balance sheet, with no effect on current P&L) and then amortised over a fixed time period. There are all sorts of accounting rules that regulate that treatment, I should add, but when you look at the landscape it becomes obvious there’s more than a touch of flexibility. Some companies expense it all. Some companies capitalise over two thirds.

This sort of distinction is another example of how accounting statements – while decent representations of the business – can’t and won’t ever be perfect. You might reasonably argue that a company reporting £5m in operating profit after expensing £10m in R&D is more valuable than a company reporting £5m in operating profit while bleeding their customers and not ‘investing’ in lowering prices and improving efficiency or product ranges, an accusation that’s been levelled at Tesco.

This whole discussion might also bear some relevance for anyone casting a cursory glance, as I have, at ASOS post their share-price cliff. For what it’s worth, I don’t think Asos is doing what Amazon is doing – they talk about investing in reducing prices and aggressive expansion, but their margins aren’t that bad, and their prices never seemed that cheap to me. I’m probably just a cheapskate, though.


The above shows the breakdown of Asos’s operating costs in their last annual report. The company’s narrative is that many of these costs are to support and encourage future growth – and they’re hence more like investments than running costs. The difference isn’t trivial, either; if you think that the company could have run this year’s operations on last year’s cost base – implicitly saying that their investment is all for future growth that hasn’t come yet, and that the company is operating at well below full capacity – you basically double their operating profit figure.

That’s not a small difference. Add 5 percentage points onto their gross margin (currently ~50%) and you again double operating profit from its base level – perhaps a possibility if you think that Asos has been passing on the cost savings of its growing bargaining power, and could at some point in the future decide to instead price more competitively.

Think my assumptions are complete bunk? They probably are, but to have an opinion on the stock – long or short – you should have an opinion on exactly what Asos’s run-rate cost structure really is, and whether they’ll be able to mature into the sort of margins we might intuitively think a big, efficient e-tailer should be able to achieve.

2 Replies to “You can’t trust EBIT (AMZN/ASC)”

  1. Investing Sidekick

    At the risk of sounding like a fool, I do think there is some truth to the usual tech investor mantra that even though some businesses look expensive in terms of earnings, some have the unique position of being a ‘gateway’ on the internet, and that in itself is very valuable. I mean look at Google, they often looked expensive in terms of earnings at the beginning, but they managed to monetise being the ‘gateway’ of the internet for most users.

    I think Amazon is similar, for people wanting to buy something online, most will go directly to Amazon first, then maybe do a search on Google shopping to check if there’s anything cheaper. Amazon have monetised this via the Amazon marketplace, question is can they do more? Or will their new ‘investments’ come good one day?

    A lot of these tech companies are very difficult to value, if not impossible, who knows what kind of profits these investments will make in 10 years? But it’s clear that there are only a handful of winners in the end, back the wrong one and you’re in for some heavy losses.

    • Lewis

      All the relevant questions – I guess the only real point of debate is the last bit! They are difficult to value, but someone has to do it. That someone usually isn’t me – I think the returns to my time are better spent on the inefficient small-cap end of the market and not global retailing behemoths.

      Aswath Damodaran, for instance, has a great blog where he does exactly this:

      And it’s instructive to see how his valuations develop as time goes on. In the early stages – as in his valuation of GoPro – he makes very broad-brush assumptions. By their nature, these are often proved wrong. But as more data comes in, he updates his model and tightens his incredibly wide valuation band.

      I’m sure some people far more intelligent than me make great money trading or investing in these names… but I can’t shake the knowledge that, on average, you make considerably better money investing in un-sexy value plays and businesses with proven track records instead of pre-profit or concept stocks.

Leave a Reply

Your email address will not be published. Required fields are marked *