Wednesday, October 29, 2014

"If you build it (revenues), they (profits) will come": Amazon's Field of Dreams!

I have a long standing fascination with Amazon from its inception as a dot-com poster child in the late 1990s to its current standing as online retailer to the world. I have always liked the company's willingness to challenge established rules on how business should be done and admired Jeff Bezos for being to willing to leap into places where others only tip toe. As an investor, though, I have found the company to be cheap at times in the last 15 years and expensive at others, and the most recent earnings report led me to revisit it, partly to examine whether the market's negative reaction to the most recent earnings report was appropriate and partly because I may learn something.

A short history of Amazon
For those are twenty five or younger, it is hard to imagine a world without online retailing, in general, and Amazon, in specific, but it was just over 20 years ago (in July 1994), that Amazon was founded by Jeff Bezos in his garage, continuing the long tradition of garage-founded companies in the United States. The company caught the dot-com wave of the late 1990s and was listed on the NASDAQ in 1997. Initially focused on book retailing, the company remained small in operating numbers, relative to other retail giants, and generated only $1.6 billion in revenues in 1999, while reporting an operating loss of almost $600 million. Its market capitalization, though, rocketed up (with the rest of the dot-com sector), hitting $ 35 billion in early 2000. In fact, it was one of the companies that I used as a prop for a book I had on valuing young, technology companies. At the risk of gravely embarrassing myself, this was my valuation of Amazon in January 2000, close to its peak:

My valuation of Amazon in January 2000 (The Dark Side of Valuation)

It is never flattering to the ego to compare actual to forecasted numbers, especially for young growth companies but it is a process that has never bothered me, because it comes with the territory. I compare my forecasted revenues & operating income for Amazon (from my January 2000 valuation) to the actual revenues & operating income for the company (from 2000 to 2013) in the table below.
Comparison of my forecasts in 2000 to actual numbers

I will cheerfully confess that I did not have the foresight to predict the behemoth that Amazon would become in retailing and the tentacles that it put into other businesses (including media and cloud data) but my forecasted revenues were higher than the actual numbers every year through 2010. Since 2010, though, the company has blown the lid of my revenue forecasts but that outperformance has come at a price. I may have been pessimistic in my assessments of Amazon's capacity to scale up its revenues, but I was also overly optimistic in assuming that it would find a pathway to strong profitability. After mounting a steady improvement in margins in the first half of the last decade, the company seems to have relapsed in the last few years. 

A Field of Dreams company
A couple of years ago, James Stewart wrote an article in the New York Times, using Amazon to draw a contrast between short-term markets and long-term managers. The discussion about whether markets are short term and if so, why, is one well worth having, but I took issue with Mr. Stewart on his use of Amazon as an example of short term markets. In fact, I would argue that markets have been extraordinarily forgiving of Amazon's long loss-making history and have given Mr. Bezos breaks that very few companies have received through time. If anything, they have been too "long term" in their thinking, not too "short term".

In keeping with my obsession with popular culture, the movie that comes to mind whenever Amazon reports yet another earnings report, with strong revenue growth and decreasing profits, is the Field of Dreams, with this scene, in particular, playing out.

As I see it, Jeff Bezos has built the ultimate field of dreams company (and I don't mean that in a dismissive way), where he has sold investors on the notion that if he builds revenues up, the profits will come. The losses at Amazon are thus a deliberate consequence of the way the company approaches business, selling products and services below cost and with lots of hype, with the intent of inserting itself in peoples' lives so completely that they will be unable to abandon it in the future. To provide a simple illustration of this process, consider one of Amazon's most successful services, Amazon Prime, to which I am a subscriber. At $99/year, it is a bargain, since the shipping costs I save vastly exceed the cost of the service. While that may reflect my family's profligate spending habits, there is some evidence in Amazon's own financials that the cost of providing this service significantly exceeds the revenues that they collect from it. In the figure below, I compare shipping revenues and costs reported by Amazon each year for the last few years.

From Amazon financial filings

Not only has Amazon lost billions on shipping each year, but its losses have become larger over time.  In fact, you can take many of Amazon's recent innovations (including the Kindle) and put them to the profitability test and will find them falling short. I am sure that Amazon's cheerleaders will argue that both the Prime and Kindle create synergistic benefits to Amazon, but that argument would have more resonance, if the company made money in the aggregate.

Valuing Amazon
At this stage, the value of Amazon rests on how much you trust the vision that Jeff Bezos has for the company and whether you believe in his capacity to fulfill that vision. In fact, the value of Amazon will be largely determined by your assumptions about revenue growth and operating margins. To provide perspective, let’s start by looking at where Amazon falls in the competitive spectrum by looking at the retail sector as a whole. In the table below, I list the ten largest retailers in the US and globally, in terms of revenues:
Largest Retail Companies: Trailing 12 month data (on 10/29/14)
Note that while Amazon makes the top ten lists in terms of revenues both in the US and globally, it lags in terms of profitability with paper-thin operating margins.  To get a measure of profitability in the retail sector, I estimated operating margins (converting leases to debt) for all retail firms and report the distribution in the graph below (for both the conventional pre-tax operating margin, which is operating income as a percent of sales, and a lease-adjusted operating income, where leases are converted to debt).

Since many of the firms in this sample are small, with revenues of a billion or less, I looked at the pre-tax operating margin for firms in different revenue classes and the results are not surprising, with margins decreasing as revenues increase.
Source: S&P Capital IQ, Trailing 12 month data (October 2014)
The median pre-tax operating margin for a US retailer with at least $1 billion in sales is  7.67% and the 75th percentile is 11.99%, but the median operating margin for US retailers with more than $10 billion in sales drops to 5.14% and the 75th percentile is 10.17% (and the 25th percentile is only 2.85%). It is true that Amazon also draws revenues from its media and cloud computing businesses and that the margins are higher at least in the media business. Using a (revenue) weighted average of the operating margins across the businesses (with the weights based on Amazon's mix of media and retail) yields values of 4.35%, 7.38% and 12.84% for the 25th percentile, median and the 75th percentile.

If you assume that Amazon will continue its steep revenue growth into the future (and is able to grow revenues to about $250 billion by 2024) and that its operating margin will converge on the weighted median operating margin for the retail and media sectors (7.38%), the value of equity that you obtain is about $81 billion (or $175/share). You can download the spreadsheet that contains the Amazon valuation. If you are bullish on Amazon, at its current stock price, you have to be either be expecting even higher revenues (than $250 billion) in 2024 than or much higher steady state margins (than 7.38%), with the best-case scenario being one where Amazon continues growing revenues significantly, driving its competitors into bankruptcy, and then uses its market power to charge higher prices and generate high profit margins. Thus, assuming a 12.84% operating margin (the weighted average of the 75th percentiles), in conjunction with the revenues forecast in the base case, would yield a value per share of $345/share, higher than the current stock price of $295.

Rather than play scenario games, I chose to vary revenue growth and operating margins to see the combinations that deliver values (shaded in yellow) that exceed the current stock price ($295) in the table below:

I draw three lessons from this table. The first is that there are pathways that Amazon can follow that deliver values greater than $292 but they are narrow and require a combination of high revenue growth and high operating margins, and some of these combinations may expose the company to anti-trust action down the road. The second is that the variable that makes the bigger difference is the operating margin, not revenue growth. In fact, if the margin stays at 2.5%, higher revenue growth causes value to decline as the cost of increasing revenues (acquisitions and reinvestment) exceed the benefits. The breakeven operating margin at which growth even starts to create value is about 4%. and if the operating margin stays at 7.5% or lower, you cannot get above the current stock price, even with Walmart-like revenues. The third is that the potential for explosive returns is low, given the current stock price. While there are combinations of revenue/margin that deliver values well above $295, they seem improbable, requiring Amazon to have revenues like Walmart and margins like Lululemon.

Bottom line
In a world of cookie-cutter CEOs, uninspired and uninspiring, eager to please analysts (rather than investors) and playing the me-too game (You can buy back stock, me too! You can do acquisitions, me-too!), Jeff Bezos offers a refreshing contrast. He has a vision for Amazon, has communicated it to markets with passion and has acted consistently with that vision, and has been rewarded by markets with a high market value for his company, even in the absence of profitability. However, the peril with charismatic CEOs is that the strength and single-mindedness that make them so successful can become weaknesses, if they start believing the hype. As a consumer, I am delighted that I get Amazon Prime for $99 a year, that the Kindle costs a lot less than an iPad and that I can (though I don’t plan to) pick up the Amazon Fire for nothing, but as an investor, this is not a winning game.  Mr. Bezos has delivered on half of his field of dreams vision by building up the revenues for Amazon, but the other (and more difficult) half of the vision requires that the “profits” arrive. Much as I would like to believe in miracles, it will take far more work to make Amazon profitable than it will to make Shoeless Joe Jackson show up in a cornfield in Iowa!

  1. Amazon 10K from 2013
  2. Amazon 10Q from September 2014
  3. Amazon Valuation (Late October 2014)
  4. Global retail companies (revenues & margins)
  5. Global media companies (revenues & margins)


Jeff Chmielewski said...

I've been pretty negative on $AMZN since late 2013 - my biggest issue is the cash flow quality (and where it is being invested).

Neg CCC and stock based comp are big numbers. And then look where investment is going - AWS (cloud). And who are most of AWS's customers? Start-ups.

Not too much of a leap of faith to argue that Amazon is subsidizing cap-ex for start-ups with vendor float... scary proposition if growth ever slows.

Anonymous said...

Is it time to talk about a stock that is under-value? For past couple weeks/months, you have not highlighted any stock that is currently under-valued (whether or not you have long positions) expect Yhoo (indirect investment in YHoo JP & Alibaba). Time to talk about undervalued stock ... would love to see if you post any of those postings .... as student, reader, and coachee, we love hearing about you every time u post something.

Anonymous said...

I was wondering how you justified a terminal grwoth rate of 6% - was that ther expectation of long terminal nominal growth c. 2000/2001?

Anonymous said...

Almost all of your value is derived from assumptions what will happen 10 years from now (DCF). This is probably useless in a dynammic and disruptive envrionment such as AMZN playing field where we have pronounced non-linearities.

Aswath Damodaran said...

On the comment that most of your value is derived from the terminal value, you are wrong. While the terminal value might mathematically be a huge proportion of the current value (and it should for any young growth company), it is very sensitive to what you assume about revenue growth and margins in the next 10 years. And what exactly do you propose to do about the world being dynamic and disruptive? I tried by allowing my drivers of value to not only change but in nonlinear ways.

On the Amazon valuation in 2000, note that the ten-year bond rate was at 6.5%, thus including a much higher inflation rate than real growth than you have today. I could not use that growth rate today.

UniverseofRisks said...

Amazon could (after killing its competition and becoming indispensable to people) just charge an annual membership fees (like cost-co) and gain profitability while still running on razor thin margins. It would also be very good for their cash-flow. Just saying...

Anonymous said...

Regarding non-linearities: Your factors are based on a linear development of the business model (capital intensity constant, margin developments in context with brick and mortar retailers and so forth). AMZN already started disrupting the cloud and media. What's next? Margin's could rise exponentially once the product mix accelerates towards AWS. LULU type profit margin's of 15% might be the future while capital intensity goes down boosting cash flows. I think DCF has too many shortfalls in a such a transitioning business model.

Rohit said...

Does accounting for 3rd party rev and 1st party revenue in Amazons total revenue make a difference? Some high profile value investors have suggested that the 1p/3p split in revenues is 60/40 and that it is moving in the direction of 3p revenues. the 3rd party revenues seem to be accounted at 100% gross margin as AMZN seems to include only their margin ($13 or so) instead of $100 merchandise value in their revenue accounting.

If we adjust revenues fort his split, current revenues are about 50% higher and then they revenues actually become comparable to other retailers.

I would assume that would have significant impact on your valuation.

Aswath Damodaran said...

Accounting for third-party revenue does push up revenues (though I am not sure that you want to do that). By that accounting, Alibaba has larger revenues than Amazon.
That does not solve your basic problem, which has little to do with revenues and everything to do with profits. Until Amazon starts making serious money, it is difficult to get to today's price.

Aswath Damodaran said...

Nonlinear anonymous,
I am afraid that you are mistake the inputs into DCF as DCF. There is nothing that stops you from making your inputs move in nonlinear ways in a conventional DCF, if you are willing to model the nonlinearities.
On a different count, I would be interested to see what exactly you propose as your alternative to DCF. If it is real options, i would love to see your real options valuation of Amazon. You do realize that a DCF value is a key input into real options. If it is something else, I would like to see what it is.
Trust me. I am aware of DCF's limitations more than most anti-DCF people are and I am also aware that most of the critiques of DCF are based either on a surface-level understanding of DCF or a profound misunderstanding of what it is.

Non-linear Anonymous said...

Agreed, I was referring to your model assumptions.

Anonymous said...

Interesting that Jeff has been negative on AMZN since late 2013, especially looking in the rear-view mirror to see that it is near AMNZ's all-time high.

It seems AMZN has had the same business model for many years... interesting that cash flows all of a sudden became

ThinkTank said...

I get the feeling that Bezos doesn't care much about the stock price of his company. Not that you want your CEO to be a slave to the market or anything, but he really does sort of listen to the beat of his own drum. For investors, this is very bad news. For the consumer, it's completely awesome.

What would happen if Amazon's stock price were to plummet by 50%? Would anything fundamentally change about the company? They wouldn't be able to make acquisitions as easily, I suppose, but they haven't been exactly going gangbusters on that front, either (

The fact is, investors have been giving a interest-free loan to Amazon for years and years, and they have built such a huge head start in so many ways -- AWS, fulfillment, media streaming -- that I just don't see them suffering all that much if investors start pulling out. Hordes will still buy everything from this company for years to come.

Anonymous said...

Love the post.
I just wish you included that monte carlo graph of the range of values like you did with Facebook. If I remember right it looked like FB would be a good buy at $17- and it was!
I think AMZN is a good buy in the $90-$120 range.

Sebastian said...

Great post Prof. Damodaran, I really appreciate your valuations. However there is one aspect of Amazon I think your're missing, at least in your blog, and that is their negative net working capital. HBR wrote about this some days ago - At Amazon, It’s All About Cash Flow

Have you accounted for the fact that at Amazon operating free cash flows can actually be bigger than operating earnings for ever? Is your "Reinvestment" number lower because of this?

Aswath Damodaran said...

It is embedded already in the current invested capital and in the current reinvestment. I also think that this is a distraction rather than central to value. You may be add a couple of billion to your value because your working capital is a source of cash flow rather than a use, but not a hundred billion. Here is a simple rule. If the HBR talks about valuation and investing, you can safely assume that it is either wrong or misdirected.

Aswath Damodaran said...

There was a comment on Amazon's reinvestment that I tried to publish but it seems to have disappeared. The question was why I set it equal to the average for the retail sector, since Amazon is so different from other retail firms. I agree and it is not that Amazon invests less than the rest of the sector. It just invests differently: in acquisitions, product development (R&D) and related costs. I considered going back in time and collecting information on all of these but then I noticed that raising the sales to capital ratio for Amazon from what I used (3.68) to even an astronomically high number (10, for instance) has very little effect on value (with the value per share going up to $214). Lowering it significantly can be devastating, though. It reinforces my point that at Amazon, the key concern is profit margins and how to raise them.

ANON said...

In you model, I don't see you adding depreciation and amortization leading to negative FCF for the next 5 years. In reality, AMZN showed positive FCF for the last 10 years and is (according to bloomberg) projected to generate positive FCF over the coming years.

Aswath Damodaran said...

Reinvestment = Cap Ex - Depreciation - Change in Working capital. Depreciation is, in effect, being added back. And if your consolation prize is that the cash flow is positive, it is only because Bloomberg adds back stock-based compensation and assumes that the supplier-credit based working capital model will continue in perpetuity...................

mavericks said...

Prof Damodaran,

I valued Amazon in March this year, projecting continued revenue growth 18.65% for the next 5 years before its starts to converge towards t- bond by year 10. Interestingly, with this assumption Amazon reaches 250 billion mileston by year 8 but the biggest riddle was the operating margins. Based on those numbers , my estimate for Amazon was $62/ share and I shared with my readers how operating margins are the biggest value driver now as opposed to revenue growth. You can read my post at this link.

Aswath Damodaran said...

It looks like we agree on more than we disagree. I do think that $62 is a little too low for the value/share. I will do a Monte Carlo simulation in the next week or so and add it to this post.

Anonymous said...

I noticed that terminal-year reinvestment is calculated based on EBIT (1-t)/ROIC * terminal growth rate. Would you explain your rational behind this?

kayser said...

I am surprised that there was only a cursory mention about Amazon's cloud computing business. Enterprise IT spending is a USD1trn market. Data center and Infrastructure services are the largest segment over there.

Amazon is disrupting that business. Though they don't disclose it separately, they are likely to have sales of USD2bn. That business should be highly profitable; a 20%+ EBITDA margin is not unlikely.

Investor said...

Aswath sir,

I began valuation by reading your little book on valuation in may 2014,,than now moving your original book on valuation & reading blog regular,,
sir if we capitalize R&D expense than ROIC & ROE all parameter looks interesting,,as per last chapter of little valuation,,should that not apply to amazon??

Beppaun said...

Dear prof., have you read a great analysis made by Benedict Evans, partner @ Andreessen Horowitz?

Here the link:

Anonymous said...

Professor! How do you account for negative Working Capital in FCFF? Thank you.

Jason shwartz said...

Building revenue really can be something that is difficult. What really is interesting is that there are so many companies that are able to have that art down to a tee. Something that would be great to have for small companies are some kind of background in big companies that have figured out that skill. Thank you for sharing this information.

TomB said...

I am a year late in reading this post, so I'd be surprised if you answer. I like your blog and would appreciate an opinion. A little thought experiment:

Amazon has no earnings but generates cash - according to the cash flow statement - through the following: working capital, depreciation/amortization, and share-based payments. Basically, all of these three cash generators stay alive only if revenue grows.

Working capital because it is negative, depreciation/amortization because new assets generate new sales but need no immediate replacement, and share-based payments because shareholders surely care more about revenues than about earnings.

Is there a possibility that amazon could be growing sales to stay liquid? Instead of stating it the other way around: that they are producing CFs although they have no earnings?

I'd like to hear your opinion about this question, because I don't get satisfactory answers to it. Anyway you look at it: the price - it is 668 today, only a year after your post - is insane. Which doesn't mean that it couldn't go to 1000 anyway...

And by the way: growing sales is not that difficult when you're selling without profit.

(No, I am not a short seller. Just wondering about the hype.)