AMZN - Amazon.com Inc.

[PatientCheetah]

15.5% if terminal multiple is 15x

12.6% if it is 20x

10.3% if it is 25x

 

solver is pretty nifty

Amazon_DCF.xlsx

[JAllen]

One thing to keep in mind about AMZN in addition to the sales growth rate is the gross profit growth rate, which is consistently greater than 30% (33% last q/q).  Would you rather have your sales or gross profit growing faster?  This is because gross margins have increased 600 basis points over the last four years and continue to do so with AMZN's increasing digital sales.

 

 

Something else that is instructive for us is looking at historical capex margins (say 2002-2009), and how amazingly low they were, prior to 2010.  You can then also look at the immense ramp in capex since then and ask yourself how much of the current capex AND expenses that are purely designed to give cash/benefits to AMZN's customers instead of paying 39.6% to the federal government are actually required to maintain the business. 

 

 

Another instructive thing is to look at AMZN's North American operating margins that are pressured since they expense more than half of their technology expenses in the North America. It makes perfect sense to stuff your expenses in the top corporate tax jurisdiction.

 

 

Also look for additional discretionary accounting and customer expense items, such as 'accelerated stock compensation' - much faster than the actual vesting schedule.

 

 

Would love to engage more about AMZN's financials with those who've taken time to understand their financials.  Last time I asked only one person responded that they had read their annuals.

[Laxputs]

Basically I think Amazon is a ground breaking company and I would like to own them but I suck at math and am wondering what kind of growth rate and multiple contraction result in at least 10% return on my investment.

 

If the above math is correct that they need 16% growth for 10 years, I am interested.

 

I really like this write up by Tarasoff:

 

http://www.scribd.com/doc/98208572/ValueXVail-2012-Josh-Tarasoff

 

[Laxputs]

Read through the 10-K and 10-Q. Trying to understand Amazon better. Some thoughts and questions.

 

1. Operating expenses were ~18% before 2010 and now around 25%, and WMT is below 18% If AMZN should have lower costs in these areas, can we deduce that much of the Advertising/Marketing, and Techonology/Content is in fact spent on growth? And does the massive spend on fulfillment centers as compared to prior to 2010 support that?

 

2. I don't understand their capitalization costs associated with internal use software/website development. They capitalized 581m for 2013 and amortized 451m for 2013 that was there from previous years. Does it make sense to amortize the capitalized asset if it is still productive? This is another non cash expense that takes away from a more accurate approximation of owner earnings.

 

3. If we apply similar cap ex spending of .5% of revenue that WMT spends on marketing, we get drastically lower marketing cap ex needs, and can deduce the rest is for growth.

 

4. Is investments in PPE including internal use software and website development (3.4M) on the cash flow, completely separate from Marketing and Technology and Content Expenses (9.7m) on the earnings statement?

 

5.What is a reasonable way to deduce how much of the Technology and Content spend is for maintaining their earning power and what percent is for growing it?

 

6. Is AMZN revenue comparable to WMT? AMZN records 40% on a net basis and 60% on a gross basis. WMT records 100% on gross basis. Therefore approximating net profit using WMT margins may not be accurate.

 

7. It can be said AMZN earnings are likely very far from accurate if we use comparable WMT maintenance cap ex spending, and make other add-back adjustments for amortization. A range from 3.8b to 4.8b on 2013 seems reasonable depending on how much cap ex we associate to maintenance versus growth.

 

8. Amazon showed consistent FCF margins of 6-10% prior to the massive fulfillment center build, 2010. Their high scalability should mean that those margins increase.

 

9. There are 20 outstanding lawsuits. Any concerns there?

 

10. Gross margins are increasing because of service sales increasing (AWS, advertising,  3P sales--less shipping costs and and general expenses, higher margins). Is that sustainable?

 

A couple of statements I like that I've read:

-Amazon has 8x the ROIC of the average company. 6x the earnings growth. And trades at 2.5x (normalized earnings) of the average company's multiple.

-If amazon charged 63 cents per shipment they would double their profits. Untapped pricing power. Who would care about paying 63 cents for shipping?

-There were 46 fulfillment centers in 2013, 23 were built in the last 3 years. That is enormous growth cap ex spending.

 

What's a fair multiple for a company with greater than 25% growth rate for the foreseeable future. Massive international expansion. That has very little competition. Has some of the most talented management. Has exceptional scalability. Has a virtuous cycle of lower prices, attracting more customers, offering more SKUs, attracting more customers, lowering prices, etc.; a strong moat. Has more deflationary costs than traditional competitors. Is growing a loyal customer base/unparallelled focus on customer satisfaction. Will have untapped pricing power when a focus on accelerated expansion slows. What multiple of approximate normalized current earnings provides a margin of safety if that multiple drops to 15x in 10 years?

 

Thanks.

[Palantir]

^Good post!

[Grenville]

I didn't realize the Diapers.com/Quidsi founders had left Amazon. The other founders of acquired companies are still there.

 

Diapers.com co-founder quietly working on new startup called Jet

http://tech.fortune.cnn.com/2014/05/13/diapers-com-founder-startup/

 

Amazon’s Quidsi Sees Founders Depart to Work on New Projects

http://www.bloomberg.com/news/2013-07-08/amazon-s-quidsi-sees-founders-depart-to-work-on-new-projects.html

[JAllen]

 

 

Here are some thoughts I wrote in an email a while back.  I hope it spurs more AMZN discussion.

 

 

"

• Amazon's margin history is the most misunderstood aspect of the company, in my opinion, and suggests that the company trades at an EV/FCF multiple of 20-25.  For instance, AMZN generated a 12% FCF margin in 2009, just prior to the massive, greater-than-sales-growth-rate infrastructure expansion the company embarked on (2010-2012 57% square footage growth against 20% average square footage growth 2002-2009)
  

• Comparing AMZN's historical operating expense margin to Walmart's is instructive.  AMZN's opex margin was mid-teens prior to its massive spending ramp in 2010 compared to ~19% now, compared with Walmart's stable 18% opex margin.  I believe that it's likely Amazon's long-term opex margins will be less than Walmart because AMZN doesn't have expensive retail real estate, and this is corroborated by AMZN's prior-to-2010 opex margin history - mid-teens opex margins noted above.
  

• AMZN's gross margins have expanded materially to 28% for 2013 compared with ~22% prior to 2011.  When combined with mid-teens opex margins noted above, there's the possibility AMZN may have long term operating margin/FCF margins (which aren't perfect substitutes over short periods of time due to various non-cash expenses like stock comp. etc. but should approximate one another ex-taxes and interest) into the double-digits, which I doubt the market has considered (although Skzutak mentions the potential double-digit margins in a CC).
  

Generally the market seems to believe AMZN has no pricing power and is making zero profit.  But we haven't seen anyone analyze the company's historical margins thoroughly and considered where they might go.  Even using AMZN's current gross margin - 28% - and WMT's opex margin of  ~18% gets us to 10% operating margins for AMZN (Walmart's operating profit margin are very reliably 6% so I think of 6% as a base for AMZN long-term).  I personally believe that AMZN's North American operations are running at least 8% underlying operating margin now, excluding massive non-revenue generating expenses like digital video licensing (~$1B/year); crazy things like Amazon Fresh and hardware development with non-trivial expenses (e.g. hundreds of Silicon Valley employees and dozens of Fresh trucks and employees).  The 8% North American operating margin is also supported by the fact that the N.A. operating margin is already 4% despite the fact that AMZN books all stock-comp in N.A., does so using an accelerated method which results in >1% of N.A. revenue being expensed prematurely (in 10-K), and almost surely has most R&D type expenses booked here (which makes perfect sense to do in your highest corporate tax jurisdiction, doesn't it?). 

 

 

Another thing that AMZN does well is to invest using expenses, not always capex, which as you know are immediately tax-deductible and help maximize long-term cash flows by minimizing taxes (remarkably similar to John Malone's modus operandi actually), the only thing Bezos cares about.  So the market thinks AMZN's underlying operations aren't profitable at all when in reality the profit is being obscured by the company's massive growth expenses, much of which are growth expenses.  A great example is the significant sum AMZN is now spending on HBO shows.

 

Regarding size of Walmart now compared with AMZN's eventual size.  I've generally thought of AMZN's current addressable U.S. retail sales at ~$4T, which is less than $5T, primarily because it excludes fuel sales but includes groceries now (pretty sure retail + groceries is ~$4T).  But  yes, I believe AMZN will eventually be larger than WMT as a percentage of the U.S. economy (WMT is now ~10-11% of U.S. retail sales) because AMZN is now in four large businesses and sells 100X+ (WMT Supercenters are ~150,000 SKUs compared with Prime at 15M+) the number of products WMT sells and is not limited in the sectors it can participate in because it doesn't rely on the physical retail infrastructure Walmart relies upon (think AMZN art + cars with no physical capital required).

 

 

Regarding raising prices: aside from the proposed Prime price increase, I've seen a couple of potential price increases anecdotally.  I ordered something a few weeks ago on a Thursday evening.  They offered me one-day shipping for Saturday  - for $8.99, instead of the usual $3.99 - but for almost 48 hours later, not one day shipping really.  The free shipping date was the following Tuesday - a full 3+ shipping days after my order.  So not sure what that's about but it could be a shortage of that particular product (computer cord), a large shipping price increase, or some combination thereof.

 

 

Bezos has mentioned the need to "check in and make sure we can still make a profit" (this is not verbatim and I can't remember where he said this) a few times.  They may be doing this now, but I'm not sure.  But the Prime price increase suggests they believe they have pricing power with Prime, which I imagine they do.

 

 

Anyways, I'm happy to discuss AMZN more.  It's our second largest position at the moment and I hope to hold it for a decade or two.  I believe it will be the largest U.S. company by revenues and market cap in 10+ years or so.

 

 

One last thing I just remembered: the market focuses on sales growth.  This is wrong for AMZN because their gross profit is growing much faster - 34% in 2013 - than sales - 24%. This is because of the third-party retail commission growth (that should have much higher than normal retail gross margins of ~24%) and the shift to digital sales.  It's a little hard to explain, but the gist is that if AMZN were selling all of the third-party products themselves and not selling more digital products, sales growth would be closer to gross profit growth instead of sales growing at about two-thirds the rate of gross profit growth.  So AMZN in effect is generating the gross profit from sales growth equal to gross profit growth because it's capturing the profit with lower revenue, higher margin commissions, if that makes sense.  So I think of AMZN being a 30+% grower instead of 23%, which is a totally different story from a long-term shareholder return perspective.

 

 

I agree about your ROIC type calculations, although I've been using FCF for AMZN ( OCF - CAPEX and normalizing that for historic times when they were expanding less) instead of EBITDA.    This is primarily because OCF fully deducts taxes and interest (there aren't much of either though) and you've got a conservative pre-capex FCF number there.  Also, Bezos focuses on the cash flow statement - the operating cash flows (it's listed first in the reports; can either of you recall another company doing that?).  You can pick your FCF or whatever margin.  I think of it as high single digits now for OCF, which is less than what they were doing before they ramped spending in 2010 as I noted above and is only 2% of revenues more than what the income statement shows - 7.4% OCF margin for 2013.  I use 2% for the capex margin, because their capex as a percent of revenues was a very steady 1-2% until 2010 when they massively ramped spending (see above chart again).  So I get mid to high single-digit FCF margin. As far as invested capital for the ROIC denominator, AMZN's tangible invested capital is about $10B (leaving nearly all of the cash as invested which is probably conservative).  If you look at OCF/tangible capital it's quite attractive at ~55%. I realize this doesn't include any capex, I'm just using numbers that we know for certain to ballpark FCF ROIC numbers: OCF and balance sheet numbers."

 

 

Maintenance capex is of course the wild card here, and I don't know what % of revenues it will average.  I would love to read others' arguments why it will be substantially greater than the ~2-3% it averaged for seven years prior to the massive spending ramp in 2010.  If we use 3% as maintenance capex margin and operating margin of ~10% we get to 7% FCF margin which is entirely possible and would result in an approximate 50% FCF/ROIC

[Liberty]

Excellent post, JAllen. Thanks! The Malone comparison when it comes to avoiding taxes is very apt, IMO. Better to spend that money to invest in growth than to pay it in taxes, even if it makes GAAP earnings look horrible.

[Palantir]

JAllen - what adjustments are you making to OCF to get the true "economic" OCF? I assume your argument is that economic FCF is far higher than what the statement shows, so just wondering where to look to find adjustments.

[JAllen]

JAllen - what adjustments are you making to OCF to get the true "economic" OCF? I assume your argument is that economic FCF is far higher than what the statement shows, so just wondering where to look to find adjustments.

 

 

For now, I'm using actual OCF as a floor that no one can argue with on its own (except for what I discuss below).  AMZN does have growth expenses as I mentioned, and these do reduce OCF as well as income, so your intimation that OCF is probably understated is correct, IMO.  OCF is just a number that is better than GAAP operating income to work with because op. income is being intentionally minimized.

 

 

There are some major expenses OCF excludes in addition to not deducting anything for maintenance capex (which I do deduct for FCF in my discussion above as perhaps 3% of revenues), the biggest one being stock-comp. expense (which is itself overstated - see below), so I think OCF is a very rough proxy for current operating margins after subtracting growth expenses.  We would need to adjust OCF downward for stock-comp, by something like .4-.5% of revenues (actual stock-comp. expense is ~2% of revenues but AMZN's vesting schedule is not 50% the first year of grants which is how they expense stock-comp[1]; it is 5%-15%-40%-40%) but adjust it upwards for all of the cash growth expenses and benefits that are thrown at customers but aren't charged for like the Prime benefits, Amazon Fresh, digital video (this is a huge tax-deductible expense, probably something like 1-1.5% of revenues now), other general expansion and whatever else they are spending money on that we have no clue about. 

 

 

So the moral of the story is that much of AMZN's growth spending is not capex but are tax-deductible expenses.  It makes perfect sense to target growth expenses if you are truly concerned about long-term FCF like Bezos is. 

 

 

Would you rather give 39% of your income to the Federal Govt. and get zero benefit or throw your OCF at customers to make them happier, more loyal, build trust, gain a larger percentage of their annual retail spending and drive your taxable income to essentially zero?  If you're not concerned about current years EPS, which Bezos is so clearly not, this is exactly what you would do.  And then Buffett would say you're essentially the best American CEO.

 

 

[1] We utilize the accelerated method, rather than the straight-line method, for recognizing compensation expense. For example, over 50% of the compensation cost related to an award vesting ratably over four years is expensed in the first year. If forfeited early in the life of an award, the compensation expense adjustment is much greater under an accelerated method than under a straight-line method. (pg. 19 in 10-K)

[PatientCheetah]

Excellent analysis, it took me a few years to finally understand the AMZN story and I blame myself for reading too many bearish thesis. The steady increase in price and my own personal dependency on AMZN should've tipped me off a long time ago.

[dwy000]

Great analysis JAllen.  Good summary and it's tough to argue with any of the math.

 

At the end of the day though I just struggle with the concept that Amazon has always been an investment I justify by starting with "when they finally start to...." or "when they stop doing.....".  The sales growth has unquestionably been fantastic but I always come back to the fact that free cash flow over the past 5 years has been steadily declining: 

    2009  - $2920

    2010  - $2516

    2011  - $2092

    2012  - $ 395

    2013  - $2031

All while sales have grown from $24bn to $74bn (and other than 2012 the share count has continued to increase).

 

There are a ton of adjustments you can make that suggest the FCF potential is enormous but it always seems to be just that, potential.  The higher capex to build out fulfillment centers has hurt the past few years but I can't see that stopping any time soon if they want to keep growing the top line and taking delivery down to same or next day.  It will also increase the ongoing fulfillment opex as they operate all these new and larger facilities.  If you assume growth slows they lose the cash flow benefit of the operating cycle (working capital) as well as the direct cash generated by unearned revenue growth ($400M of FCF in 2013).

 

Would love to get your view on a couple of items:

- what's your thoughts on the impact of them having to start charging state sales tax in the next year or two?  I haven't done the homework but it would seem that for the states this has been implemented they took a pretty sharp hit to revenues (probably one time and then resumed growing);

- the price declines in AWS services have been significant over the past year.  Microsoft, Google and IBM have all dropped prices hugely (over 40% I think) in just the past year and have stated they will match any price.  This is high margin, sticky business for AMZN.

- they don't really buy back shares but they certainly expense them (your point on timing is a good one but over a longer term this will even out).  The share count hasn't gotten out of hand over the past few years but it certainly isn't helping from a FCF/share perspective.

 

In buying this stock I worry very much about the reaction you saw after the 1Q earnings and outlook.  People seem to be pricing it off of the potential FCF it can earn in the future.  But at the FCF multiples it trades at today there's a big risk of the Microsoft situation - i.e. the stock is dead money for 10 years while the cash flow and earnings grow into the stock price as the multiple comes down.

 

 

 

 

 

[dwy000]

sorry for droning on but I forgot to add a couple of other points:

 

- The gross margins are strong and growing but I really think it makes sense to adjust those for the technology and content spend which is also growing rapidly, given they are including AWS in the sales number but there aren't many COGS expenses for this as it is largely a fixed cost business and we don't know what support costs for AWS are in COGS vs tech & content.

 

- on a similar vein, again, they don't break it out but a good portion of the capex growth over the past 2 years has to be related to AWS.  And that's a business that will require a lot of additional capex (both facilities and hardware) to support growth.  The hardware would likely need to be replaced every 3-5 years.

 

- Google shopping now offers same day delivery here (San Fran) on items purchased from major stores.  It's free for the first 6 months.  Who knows how they'll price it long term or if it takes off (I think it will die a quiet death at the next downturn much like all those delivery businesses in the internet bubble) but it's a deep pocketed competitor aiming directly at the Amazon core.

[Laxputs]

I'm curious on the investment argument: how do you price in the future growth given your hurdle rate?. I never do DCF analysis, I usually buy companies priced at very attractive earnings yield, with growth not necessary for the investment. So supposing this is 25x normalized earnings, what kind of minimum return are you expecting when accounting for growth? I'm hoping a few examples and the math involved can be offered.

 

It's interesting that multiple contraction is not really an issue with this stock, if the math of your assumptions is roughly correct. It means the company is currently trading at the multiple some anitcipate it will contract to.

 

Also, JAllen, what is your largest holding?

 

Thanks.

[JAllen]

Great analysis JAllen.  Good summary and it's tough to argue with any of the math.

 

At the end of the day though I just struggle with the concept that Amazon has always been an investment I justify by starting with "when they finally start to...." or "when they stop doing.....".  The sales growth has unquestionably been fantastic but I always come back to the fact that free cash flow over the past 5 years has been steadily declining: 

    2009  - $2920

    2010  - $2516

    2011  - $2092

    2012  - $ 395

    2013  - $2031

All while sales have grown from $24bn to $74bn (and other than 2012 the share count has continued to increase).

 

AMZN is for me and Bezos a very long-term stockholding, so I'm really not overly concerned about what the company is earning now.  If you don't view AMZN as a long-term hold and focus on what it will earn in five to ten years, it just doesn't work, because it's not cheap on a current FCF basis.  That point is inarguable. Since the vast majority of us will be alive in 10-20 years, why not have a holding period perspective that long?  There's the great hockey quote that applies in droves here: skate where the puck is going, not where it is.  So I believe that when the massive spending ramp (again from 20% growth to 57% growth - this is huge!) cools back down to sales or gross margin growth, we will see growth in operating income and free cash flow, but I have no idea when that will happen.  If one has a VERY long-term perspective, you hope this doesn't happen for a very long time, as that would mean AMZN has run out of places to invest its capital.  This won't even begin to be a problem for AMZN for a number of years though, at least ten I'm guessing.

 

 

There are a ton of adjustments you can make that suggest the FCF potential is enormous but it always seems to be just that, potential.  The higher capex to build out fulfillment centers has hurt the past few years but I can't see that stopping any time soon if they want to keep growing the top line and taking delivery down to same or next day.  It will also increase the ongoing fulfillment opex as they operate all these new and larger facilities.  If you assume growth slows they lose the cash flow benefit of the operating cycle (working capital) as well as the direct cash generated by unearned revenue growth ($400M of FCF in 2013).

 

Again, this comes down to a difference in time-horizon.  I'm viewing AMZN as a semi-permanent shareholder, not as 1-2 year shareholders.

 

Regarding the operating cycle, I expect that to continue due to Prime + Prime Fresh and AWS 1 and 3 year reserved instances (that are quite attractively priced) revenues.  I don't expect growth to slow anytime soon because of the virtuous cycle caused by having the best selection, prices, and service/convenience.  AMZN has by far the most momentum in very important industries that are the future of how humans do business and consume.  This should not be overlooked.  It can be easy to miss the forest for a tree.  AMZN having great momentum, first mover advantage, scale and the best management in very large businesses while still being tiny in the grand scheme of things is the forest (AMZN's North American revenue is still just 1% of addressable sales).  Being concerned with how much FCF the company is currently generating is a tree.

Would love to get your view on a couple of items:

- what's your thoughts on the impact of them having to start charging state sales tax in the next year or two?  I haven't done the homework but it would seem that for the states this has been implemented they took a pretty sharp hit to revenues (probably one time and then resumed growing);

 

The sales tax cost advantage was just one attraction from a consumer's perspective.  AMZN started charging sales tax in CA, which is almost surely its highest revenue state at least a couple of years ago.  When they do start charging in a new state, they can lower shipping expenses building warehouses closer to population centers.  For SF, packages used to come from Reno and Las Vegas, 4.5 and many hours away respectively.  Reno was over the Sierra Mountains, which I'm sure was risky and sometimes treacherous due to snow for four months of the year.  Now there are two warehouses 1.5 hours away.  That cuts down on costs so AMZN can lower prices.  The sales tax advantage was always an impermanent advantage and everyone knew it was going away.  This is an anecdote, but I probably spend 2X with AMZN compared with when they started charging sales tax.  It was a bummer the first time I saw I was going to pay sales tax, but that didn't once cause me to even consider driving to Wal-mart or walking to Best Buy! People love AMZN for lower prices that they still have, convenience, selection, reviews, great customer service (I regularly complain for things like massive price drops right after I purchase things and they'll give me money back no questions asked - this is the right way to build trust and loyalty).  Don't forget that now the field is level from a sales tax perspective, but since AMZN has lower op ex. costs, because they're more software-focused, automated and don't operate retail real estate, they can still charge lower prices.  Also, driving to stores is a huge pain and WMT.  I'm 30+ minutes from a WMT (SF).  That's another mark against WMT relative to AMZN - driving to a store, finding what you're looking for, and waiting in line sucks!

 

I will respond more tomorrow.  It's late here now.  I'm really glad there's more discussion about AMZN now!

[rkbabang]

I agree with you about the sales tax advantage.  I think people made too much of the sales tax advantage and its demise.  I live in a state (New Hampshire) where Amazon has never had a sales tax advantage and my purchases from Amazon have increased every year.

 

[saltybit]

Just a personal anecdote, but I know quite a few people who worked at AMZN over the years. While the reviews of working there are mixed, they unilaterally say that they are more likely to keep being an AMZN customer after working there than before.

It's rare to say that about something after seeing "how the sausage is made".

[dwy000]

JAllen - I agree with your 5-10 year time horizon on any stock.  I guess the issue was one where 5-10 years ago people we're saying the exact same thing for amazon (ie just wait until they start throwing off cash flow) and there's no indication that they won't be in the same position 5-10 years from now.  I'd like to see at least some ability or willingness to prove the cash flow machine can work and is not just a "some day" potential.

 

The other concern on the same point is that with it trading at such ridiculously high FCF multiples (actual FCF not potential), the stock price will be relatively stagnant as the multiple comes down and real FCF goes up.  Even fit they can start to churn out cash it's kind of already reflected in the stock price.

 

Don't get me wrong, I love the business and the moat.  This will be around for a long time and remain dominant.  But I'm not bought into it as a value stock story.

[Laxputs]

JAllen - I agree with your 5-10 year time horizon on any stock.  I guess the issue was one where 5-10 years ago people we're saying the exact same thing for amazon (ie just wait until they start throwing off cash flow) and there's no indication that they won't be in the same position 5-10 years from now.  I'd like to see at least some ability or willingness to prove the cash flow machine can work and is not just a "some day" potential.

 

The other concern on the same point is that with it trading at such ridiculously high FCF multiples (actual FCF not potential), the stock price will be relatively stagnant as the multiple comes down and real FCF goes up.  Even fit they can start to churn out cash it's kind of already reflected in the stock price.

 

Don't get me wrong, I love the business and the moat.  This will be around for a long time and remain dominant.  But I'm not bought into it as a value stock story.

 

People that bought it 5-10 years ago probably aren't complaining.

 

I think you are missing the point. If our math is close to correct, the company is already trading at 20-30x owners earnings.  The multiple won't come down, it will just get readjusted.  And if our very reasonable assumptions are correct then the company is doing exactly what an owner of the business would hope it would do: reduce taxes as much as possible while investing cash flows at high ROIC. This is what benefits shareholders; not making the bottom line look good. Think like an owner.

 

I'm hoping somebody can go over the math used in calculating a rough return assuming it trades at 25x earnings, and using different growth rates over a 5 and 10 year holding.

[dwy000]

Lax,

 

I'm just back-of-the-enveloping this but my thinking is along the lines of this:

 

- assume within 5 years they can be generating $10bn in FCF (real cash, not adjusted.  Cash flow from operations minus actual capex).  That's 5x what they are generating today.

- assume that FCF is without any share dilution;

- put say, a 20x multiple on that cash flow;  ignore debt (that's factored into the FCF);

- that translates to a market cap of $200bn;

 

With market cap today at $140bn that translates into about a 40-45% total return over 5 years.  And that's with some pretty aggressive and uncertain assumptions that are inconsistent with what they've done in the past.

 

I'm happy to be wrong on this.  How are you doing the math that makes the stock price today look like a good value investment?

 

[dwy000]

Curious on your point that they are reinvesting cash flows at high ROIC.  Despite revenue tripling over the past 5 years, FCF has declined.  Where are you seeing the high ROIC?

[PatientCheetah]

Many brick and mortar stores are struggling right now. I think Barnes and Nobles is the only national book stores left. Some of their problems undoubtedly has to due with AMZN. AMZN is investing in margins right now. It is not inconceivable that when enough brick and mortar stores go under, like any other monopolists, AMZN will probably rise prices.

[JAllen]

Curious on your point that they are reinvesting cash flows at high ROIC.  Despite revenue tripling over the past 5 years, FCF has declined.  Where are you seeing the high ROIC?

 

 

It helps to estimate growth and maintenance capex and subtract maintenance capex from OCF.  Also look at tangible invested capital - it's ~$10B.  So you can look at OCF/tangible capital, compare AMZN's to others and see that it's quite high.  I wrote about AMZN's ROIC in a previous post above and how we can estimate it and I'll write more here.

 

 

Current FCF, or OCF minus current year capex, doesn't account for maintenance versus growth capex at all.  And we all know that Amazon is investing for the future. It helped us to look at AMZN's pre-2010 capex margins, which were very reliably 2-2.5%.  If you assume that maintenance capex for AMZN is now ~$2B, OCF of $6B gets us to $4B of FCF on ~$10B in tangible capital which is is a nice return on tangible capital (Buffett singled out his 60% tangible-return companies in an 80s annual letter).  Of course I don't need to tell anyone that 40% ROTCs are attractive.

 

 

Call us crazy but we actually believe that AMZN's actual ROTC is greater than 50% and that's its true current FCF is almost 10% of sales, somewhere in the range of $7B-$9B now. 

 

 

Here is some of the evidence we have for this:

• current gross margin minus historical operating expense margin would equal 13% operating margins or $9B of operating income
  

Gross margins cannot be fibbed or manipulated, but operating expenses are inflated by massive investing and spending, e.g. paying HBO, AWS expansion, China competition etc.  AMZN's gross margins are continually increasing, along with its opex.  Since WMT's operating expense margin is 18%, we can conclude that AMZN's will be a few points lower because AMZN doesn't operate any retail real estate over time. A pure warehouse infrastructure is cheaper than warehouses + retail to deliver same products.

 

• AMZN was generating FCF margins well into the very high-single digits from 2002-2009
  

Prior to the massive spending ramp which increased operating expenses and capex, AMZN generated FCF margins in the 7-12% range.  This linked chart deducted all capex as maintenance capex, despite the fact that the company was growing rapidly.  This means that the capex margin is overstated in the above chart and that FCF would have been higher if you knew how much was growth capex was.  2009 had a large working capital benefit, but FCF would still have been almost 10% of sales.

[dwy000]

JAllen,

 

The ROTC is certainly phenomenal.  No argument.  But the stock is trading at 1400% of Tangible Capital which takes the return on YOUR capital invested very low.  Again, my argument is not the quality of the company but the price of it.

 

It's nitpicking but I'd take the other side on a couple of your assumptions.  First, the concept of maintenance vs growth capex and looking back to 2010 as a basis for maintenance.  Unless Amazon decides to stop growing they are going to need to invest in growth capex and those $ comes straight out of free cash flow.  Also, Amazon has never broken out it's capex spend between maintenance and growth or even that related to AWS vs the retail business.  It's a dangerous assumption (from an investing perspective) to assume that 2010 capex levels can be used as a proxy for maintenance.  For example, they opened 12 new fulfillment centers in 2012 that cost $1.4bn (within capex) - but they still spent $2.4bn beyond that.  And the number was $3.4bn in 2013 and I don't believe they opened another 12 new centers.  Either way, unless you assume growth stops, it's inconsistent to assume growth capex stops.

 

The second point I'd nitpick with is the opex and that it is inflated due to one time growth initiatives.  It's true that current gross margins less historic opex equate to 13% but that's comparing apples to oranges.  I guess I will believe that it is truly inflated when I see that they are able to bring it back down to historic levels without impacting either sales growth or gross margins.  I think you also need to take into account that gross margin includes growing AWS which is a low COGS, high opex business - so that will inflate gross margins but require an ongoing increased level of opex.

 

Final point - even if using the $4bn of FCF that you indicated (using $2bn of maintenance capex and $6bn of Op CF), and increasing that by 250% to $10bn in 5 years (and they've indicated no sign of stopping gowth capex) at 20x FCF it's still only a 45% gain over 5 years.  And that's assuming they buy back stock equal to everything issued over the next 5 years.  How are you working the math on the upside for the stock.

 

Great discussion even if we don't see eye to eye on it.  I appreciate having my assumptions challenged.

[JAllen]

- the price declines in AWS services have been significant over the past year.  Microsoft, Google and IBM have all dropped prices hugely (over 40% I think) in just the past year and have stated they will match any price.  This is high margin, sticky business for AMZN.

 

If someone asked me what bothered me most about AMZN recently it would be cloud competition (the second would be lack of apparent traction in China).  A 38% price decline is never welcome.  The flip side of the coin is that we know AWS almost surely had greater than 38% gross margins, which is awesome, and that AMZN is currently the cloud-computing leader.  This is hugely valuable.  We like AWS because the costs decrease rapidly with time (AWS has decreased price by about 20% per year).  AMZN will almost surely be one of the leaders in cloud computing, along with Google and a few others over the years to come.  Just like retailing, there is room for more than one competitor to generate strong returns. 

 

AWS is the segment we would most appreciate hearing from others about.  We are still taking the time to fully understand this segment, its economics and potential.

 

 

- they don't really buy back shares but they certainly expense them (your point on timing is a good one but over a longer term this will even out).  The share count hasn't gotten out of hand over the past few years but it certainly isn't helping from a FCF/share perspective.

 

The share count grows pretty steadily at about 1%.  This is pretty reasonable for a company that is still growing rapidly in our view.  Also, when employees leave, their unvested stock-compensation is forfeited which results in the previous stock-comp. expense being more overstated.

 

Regarding buying back shares: if you look back, they have bought back shares, twice I think, and both times the stock is up multiple times in the years following the repurchases.  I did some calculations a while back and the CAGR from buying back the stock they did was something like 40% to date.  Bezos is very much concerned about FCF per share.  I'd bet that they will aggressively buy back stock in the future but don't have any idea how they determine at what point to do so.  It would be interesting to perhaps see a price/sales chart showing what the price/sales was when they did repurchase.

 

In buying this stock I worry very much about the reaction you saw after the 1Q earnings and outlook.  People seem to be pricing it off of the potential FCF it can earn in the future.  But at the FCF multiples it trades at today there's a big risk of the Microsoft situation - i.e. the stock is dead money for 10 years while the cash flow and earnings grow into the stock price as the multiple comes down.

 

 

This all depends on what FCF multiple you believe it's trading at today and the long-term growth rates AMZN will experience, and its eventual FCF margin.