Thoughts on and why despite the insane run the stock has had I think there's more to go. There are a fair few assumptions here so my numbers are going to be wrong but hopefully sufficiently close to right that the conclusions still stand.
1/ I think you have to look at the business as Azure / everything else. Not because there's some SoTP angle or anything but because the businesses are at such different stages in their life, agglomerating them together misses too much.
2/ Let's start with everything else. Last year did $143b of revs o/w ~ $20b was Azure, leaving ~ $123b for the rest. My guess is that Azure was modestly profitable, let's say a 5% EBIT margin or $1b of EBIT, last year. That leaves the rest doing $52b of EBIT / a 42% margin.
3/ That passes the smell test for me with at ~ 43% EBIT margins (2021e) and at ~ 46% margins (also 2021e).
/4 The "rest" of has been growing revenues at a ~ 10% CAGR and EBIT at a ~ 20% CAGR for the last few years. The mix-shift driven EBIT margin expansion won't recur but consensus estimates of ~ 8% revenue growth and ~ 11% EBIT growth out to 2025 don't strike me as crazy.
/5 That leaves us with FY25 EBIT of $89b, NOPAT of $74b and at 30x NOPAT (seems OK for a monopoly generating low double-digit capital-light growth) a "rest" of valuation of $2.2trn.
/6 Then moving to Azure. Gartner sizes the cloud TAM ex SaaS at $204b in 2022. If you keep growing that at a low-double-digit clip you'll be at $500b by 2030. I think Azure can take 35% of the market, let's call it 45% AWS, 35% Azure, 10% GCP, 10% everyone else.
7/ I don't see why they can't make 35% EBIT margins at scale if AWS has been printing 30% for the last three years while growing at 40% on average. The RoIC is harder given the shifting mix of IaaS / PaaS / SaaS. Let's say 25% for the sake of argument.
8/ That would imply an invested capital base, in 10 years, of ~ $205b. If you extrapolate TTM Azure revenue of ~ $20b vs my estimate of cumulative Azure-specific CapEx in the region of $25b and some underutilization today that improves over time, that would suggest ~ $195b.
9/ So we've got $175b of revenue, $61b of EBIT, $51b of NOPAT. If that's also worth 30x NOPAT, that's $1.5trn of value. Then deduct the ~ $170b of CapEx they'll have to spend to achieve this vision. That's $1.3trn. Discounted back to 2025 at 8%, that's $920b.
10/ So we've got $2.2trn for ex Azure in 2025 + $920b for Azure for a total of $3.1trn. They've got $65b of net cash today and should generate ~ $300b of cumulative NOPAT between now and then getting you to a total of $3.5trn in 2015.
11/ That's a share price of $462, +114% from Friday's close and a ~ 23% IRR over 4 years (since these are forward multiples on FY25 numbers). Pretty compelling if you ask me. I'd love to hear constructive feedback on the above or any key points I've missed / math I've got wrong.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
TCI has almost a quarter of their 13F book in the Class 1 rails (namely & ). Competitively-advantaged, irreplaceable, well-run companies? Sure. But I just can't see how the math works out such that these are compelling investments at today's prices.
1/ Let's look at (because that's the one I have an up-to-date model for and because they're all pretty darn similar). In the last five years (2014 - 2019) CSX carloads shrunk at a ~ 2% CAGR, revenue/carload grew ~ 1% so total revenue shrunk at ~ 1%.
2/ Margins exploded, with EBITDA margins going from 38% in FY14 to 52% in FY19 while CapEx invensity shrunk with CapEx/Sales going from 19% to 14% over the same period. EBITDA - CapEx margins therefore more than doubled from ~ 18% to ~ 39%.