Introduction
.
The median P/E (excluding negative earnings companies) of the top 20% of US stocks greater than or equal to $200 million market cap is 69 compared with the bottom 20% which is at 12. This 58 P/E point spread is higher than the historic average of 50 since 1994.
.
What does a high P/E imply? Super rates of growth? The implied growth rate of the high P/E stocks is 15.0% and only 0.8% for the low P/E stocks.[i] Thus, the implied growth spread is 14.3% (15.0% – 0.8%), compared to an average of 14.1% since 1994. Since 1994, the realized spread in 5-year forward growth between the top and bottom P/E companies is 16.5%, thus it may surprise many people that the current implied growth spread may be reasonable.
.
Amazon trades at a 205 P/E, while Walmart is at only 18X (based on 2016 EPS). This means that AMZN’s P/E is 11X Walmart’s. Is this reasonable? With some simplifying assumptions described later, to justify AMZN’s $1003.00 value (June 6, 2017), its earnings would have to grow about 61% in 2017 and fade steadily to 7% by year 10 (33% average growth for the next 10 years). Compared to Walmart, Amazon would have to grow earnings by 50% per year for over 11 years before its discounted earnings stream would be equal to a like sized investment in Walmart that grows at 5%. This could happen, but it is obviously a rosy scenario.
.
Is Amazon’s P/E worth 11X more than Walmart’s?
.
Amazon earned $4.90 a share in 2016, up from $1.25 a year earlier and -$0.52 in 2014. EPS improved as sales jumped 25%, 16%, and 20% in 2016, 2015, and 2014, respectively. The stock’s 205 P/E is dependent on both continued high rates of sales growth, and increasing margins which drives up EPS to even more impressive rates of growth. Although, its price has risen nicely with sales without EPS growth or even positive EPS in the past (figure 2). AMZN’s P/S multiple is at one of its highest in 17 years (figure 3); it appears that investors have become even more optimistic over the last couple years. Amazon’s P/S multiple has risen with the overall market’s. Figure 4 shows that the P/S ratio for the S&P 500 has generally risen/fallen with profit margin; therefore, we can extrapolate that AMZN’s rising P/S predicts improving profitability for the company.
.
To evaluate whether the stock is worth a 205 P/E, it is useful to compare it to Walmart. While AMZN has transformed into more than just a retailer – $12 billion of its $134 billion in sales is from Amazon Web Services which generates 74% of its operating profits, has an operating margin over 25%, and has rapidly rising sales (sales were less than $5 billion in 2014) – a comparison versus Walmart is at least a good place to start.
.
Amazon’s North America retail sales of consumer products and subscriptions is finally profitable on an operating income basis; operating income has grown from $360 million in 2014 to $2.4 billion in 2016 (operating margin of 3.0%) as sales rose from $50.8 billion to $79.8 billion. International retail sales of consumer products and subscriptions is still unprofitable (operating income was a loss of $1.3 billion) even though sales were $44.0 billion in 2016. In comparison, Walmart had operating income of $22.8 billion (4.7% margin) on sales of $485.8 billion.
.
.
Long-term, sales must translate to earnings (and cash flow) to become valuable to investors. Amazon is one of the few companies in America that has essentially been given 20-year free pass to exist (and the stock to rise) without making earnings. However, it does generate positive free cash flow (defined as net income plus depreciation – capital expenditures and acquisitions) as depreciation eats up a huge share of potential earnings and is greater than capital expenditures; depreciation was a total of $19.1 billion from 2014-16 while acquisitions and capital expenditures totaled $18.1 billion. The firm made $2.4 billion in net income in 2016 during a year when depreciation was $8.1 billion (capital expenditures and acquisitions were $6.9 billion), so depreciation is a huge drag on earnings. Still, even with a positive free cash flow, the free cash flow yield ($3.6 billion free cash flow/$479.4 billion market cap) is only 0.8% (the S&P 500 is at around 4-5%), Even with this long free-ride, we must assume that investors will eventually need to see the AMZN produce stronger earnings and cash flow, and not just from Amazon Web Services. Let’s take a deeper dive to determine what is necessary to justify its 11X PE premium vs Walmart.
.
.
Figure 5 provides some comparison information for Amazon and Walmart. On June 6, AMZN closed at $1,003. Let’s assume that one share was purchased, so the total investment is $1,003. That same $1,003 investment could have purchased 12.71 shares of Walmart at $78.93. Amazon made $4.90 per share while WMT made $4.34 in EPS in 2016. Since a $1,003 investment in Walmart gains one 12.71 shares, this means that earnings to the investor is $55.15. This compares to an investment in Amazon that only generates $4.90 ($50.25 behind WMT). However, investors are paying up for Amazon’s potential growth in earnings, and may be more sanguine on Walmart’s outlook.
.
If Amazon’s earnings grow at 50% (the expectations are for 55% for 2017-19) and Walmart’s grow at only 5%, then it takes seven years for Amazon’s earnings for the investor to catch up with Walmart’s (see the second and third columns in figure 6). Seven years is a long time to wait to for earnings catch up, but then again, Amazon’s shareholders have waited about 20 years before the firm has shown it can earn consistent profits. Albeit, this is largely due to Amazon’s huge technology investments (technology and content cost $16.1 billion in 2016, up from $9.3 billion in 2014, and is growing faster than sales and was expensed on the income statement) and depreciation. The level of earnings in any one year is not too important for value (investors seem to care way too much about short-term earnings, but what is important for intrinsic value is the discounted value of all earnings (or cash flow)). Amazon investors must realize this or they would not pay up so much for AMZN’s current earnings. Still, eventually one should expect investors to require that the firm’s investments pay off.
.
.
Of course, earnings in year seven is not worth as much as earnings today since one cannot spend or invest today earnings to be received in seven years! Discounted earnings (or more importantly discounted cash flow) are what we should care about as investors. Assuming Amazon is more risky than Walmart (during the 2017 Berkshire Hathaway Annual Meeting, Charlie Munger implied that what AMZN has accomplished was very difficult and there were a lot of easier investments out there), let’s discount Amazon’s earnings at 12% and Walmart at 8%. It takes eight years for discounted earnings of AMZN to surpass WMT’s (see 4th and 5th columns of figure 6). Still, discounted earnings in one year is not too relevant; we should care about the cumulative discounted earnings (or cash flow) power of a company. The last two columns of figure 6 show that it takes over 11 years for the cumulative earnings of Amazon to catch up with Walmart. 50% growth for Amazon could last that long, but this is quite unlikely for the average or even above average firm. Figure 15 in The Expectations Clock: A Model for Cycles and Sentiment shows that above-average sales growth of the top quintile of stocks only lasts about seven years, and the starting point of growth for these stocks is only in the low teens.
.
The question now becomes, “Can AMZN grow at 50% for the next 11-12 years?” The answer is yes. Let’s take a look.
.
The first panel in figure 7 shows sales and earnings for AMZN and WMT, and Walmart’s share of US total retail sales and food services. Walmart has about double the net profit margin of Amazon, and WMT’s sales in the US are about 5.6% of total retail and food sales in the US.
.
If we assume (see middle panel) that AMZN’s sales rise to $500 billion in year 7 and net profit margin grows to 5%, then this equates to 20.4% sales growth and 40% net income growth (below the 50% threshold discussed above). 20.4% sales growth is slightly less than the past three years (the five-year, 10-year, and since 1997 sales per share growth rates are 22%, 27%, and 39%, respectively), but one would expect sales growth to slow over time. However, the 5% margin assumption is only about half of the S&P 500’s. Perhaps if we are a little bit more optimistic…
.
.
See the bottom panel in figure 6. If sales rise to $1 trillion in year 12, or 18.1% per year, then Amazon would have 12.8% of US market share (the current share is about 1.5%) of retail and food services sales by that time. US E-commerce sales was 8.5% of retail sales in the first quarter of 2017 (source: US Census Bureau), up from 2.7% in 1Q 2016 (12 years ago). The pace of adoption and change in the world appears to be accelerating, and Amazon is a leader/a disruptor, so perhaps this 12.8% market share is achievable. Also, 36% of Amazon’s current retail sales are generated internationally; if 1/3 of its retail business in year 12 is outside the US, then the US market share assumption falls from 12.8% to just 8%-9%. Perhaps this is optimistic but reasonable? Finally, Amazon Web Services is 9% of sales and is growing rapidy (up 163% in two years). Maybe this business, which is highly profitable, or other new businesses could add significantly to future sales.
.
Finally, if we increase the profit margin assumption to 6%, net income growth jumps to 58.7% over the next 12 years. 6% margin is slightly more than double Walmart and S&P 500 food and staples retailers, about the same as S&P 500 specialty retailers, and about a percent more than S&P 500 distributors. Thus, it appears that Amazon really could generate 50%+ earnings growth over the next 12 years and be worth a P/E of over 11X the P/E of Walmart.[ii]
.
Much of its capital expenditures and R&D must be to fund growth. If AMZN stopped going for “broke” (i.e., growth) and R&D fell 75% from 2016 levels and capital expenditures dropped 50%, it may interest you to know that EPS would rise to $20.70, the P/E would fall to 48.4, free cash flow would rise from $3.6 billion to $14.6 billion (a free cash flow yield of 3.1%), and the net margin would jump to 7.4%. These numbers are still expensive versus the market, but maybe if growth slowed the firm would focus more on improving margins further. If they can get to 10% (around peak net margins for the S&P 500), then earnings are lifted to $28.10, P/E would be 35.7, free cash flow would rise to $18.2 billion (a free cash flow yield of 3.8%). These numbers are still expensive versus the S&P 500, but far more reasonable and maybe only 25% (FCF yield) to 45% (P/E) too high if a market multiple is appropriate. When faced with a drop like this, maybe the firm and investors prefer to go for growth? Another way of interpreting this is that growth opportunities – above the market rate of growth – are 25% to 45% of the stock’s value.
.
.
Is Amazon Fairly Valued?
.
Just because Amazon may (the key word is may as it relies on one continuing to be optimistic) be worth the 11X premium to Walmart, this does not mean it is fairly valued. To determine if its valuation is fair, we can compare the growth rate implied in its valuation to forecasted growth. Based on the analysis reviewed below and in figure 8, EPS growth must be about 33% over the next 10 years for the stock to be fairly valued.
.
The first step to value Amazon is to determine the growth rate that Amazon’s growth fades to after year 10. I assume Amazon’s growth fades to the market rate of growth after 10 years. To compute the rate of growth for the market, I solved for an implied growth rate that results in a value of the S&P 500 equal to today’s (6/6/2017) price using the Gordon Growth Model where price today equals [earnings today * payout ratio today * (1 + growth rate)] / (required rate of return – growth rate). The required rate of return is assumed to be 10% (about the long-term return of the market), and the payout ratio is FCF as a percent of earnings (about 96%) based on 2016 results. This results in a long-term growth rate of about 5%, which is about 1% above the current rate of nominal GDP (historically, S&P 500 growth has been slightly above the rate of GDP growth). A similar method can be used to determine the implied growth rate for Amazon. If Amazon grew at 9.4% forever, its $1,003 price is justified. This assumes an 11.6% required rate of return and its current payout ratio of 409.2% would last forever. FactSet defines FCF as funds from operations (operation cash flow excluding changes in operating assets and liabilities) plus changes in working capital less capital expenditures. Of course, that payout ratio cannot stay at this rate forever (e.g., working capital added $3.9 billion to free cash flow in 2016, net unearned revenue was nearly $2 billion – unearned revenue is primarily associated with payments for Amazon Prime memberships and AWS services). If it was 96%, then the value of the stock would be $236. These growth rates – 9.4% and 5.0% – justify P/E multiples of 204.7 and 20.4.
.
Second, one must determine Amazon’s first stage growth rates. I made a simplifying assumption that growth would decline linearly to 7% by year 10 from a higher rate in year 1. I assume a 11.6% required rate of return in the first stage, but just 10% in the second stage when the growth rate falls to the market’s (5%). I also assume that the firm’s current payout ratio declines linearly to a more reasonable payout rate of 96% (the market’s) by year 10 as the company matures. All that is left to do is solve for the first-year growth that equates the value with the current price of $1,003, and doing so yields a rate of 60.6%. This is above expected growth for this year (36.1%), but the next two years’ projected growth rates in the valuation model (54.6% and 48.7%) are lower than consensus estimates (70.2% and 62.3%).
.
Overall, Amazon must grow EPS at a 32.7% growth rate over the next 10 years to justify its the current valuation. Is it reasonable? Maybe. At least, it appears achievable based on the prior analysis. If it can grow at 50% for 10 years as discussed earlier, then the stock can more than double to $2,550.
.
However, just because a firm has high EPS growth does not mean it outperforms. While firms with the highest five-year historical growth had great past 5-year returns (column 8 of figure 9), they do not outperform by much during the next five years (column 7). Perhaps future growth slows? Perhaps their P/Es were high at the start of the period of above-average growth (the median P/E was 29.9 five years before the start of the growth period for the highest growth quintile) and already priced in the high rates of growth. Valuation matters. While the future growth of high P/E stocks is quite high (column 4), high P/E stocks underperform low P/E stocks over time (column 3). They must have underperformed because the P/E fade (or decline) over time for the high P/E stocks relative to low P/E stocks more than offsets their relatively higher growth rates (remember, P/E * EPS = price, so (1 + percent change in P/E) * (1 + earnings growth) – 1 = price return).
.
.
To conclude, it appears possible for Amazon to deliver very high rates of earnings growth. However, we really should say it better deliver high rates of growth (more than 30% per year for a decade) or the P/E fade as growth slows (assuming it slows) could overwhelm the benefit of growth and the stock could languish over the next decade.
.
[i] Note that this comparison ignores the fact that free cash flow to earnings and not earnings is what is important to valuation. It also ignores earnings post the period when discounted earnings is the same for both companies, or alternatively, it also makes the assumption that growth and risk for Amazon fade to be the same as Walmart after the period when cumulative discounted earnings of the two companies converge.
[ii] This assumes above or below average growth lasts 20 years, the discount rate is 11.6% (above the market of 10.0%), FCFE is 90% of earnings (the S&P 500’s free cash flow is at about 95% of income), and growth post year 20 growth is 6% (about the average rate of the market over time).