Downside at Microsoft

Quality of Earnings, Peter Lynch and Warren Buffet.

“We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten. Don't let yourself be lulled into inaction.”

—Bill Gates

---

Business is HARD

Microsoft may be one of the most successful companies in the world, but it missed out on the race for mobile, search and social — and it’s only 2024!

Consider that Microsoft was everywhere even before Apple launched the iPhone, before Google launched its eponymous search engine and before Zuck launched Facebook (and then bought Instagram and Whatsapp). And yet it still lost.

Microsoft launched MSN Messenger 25 years ago and most of us were on it at some point. We all had a hotmail address way before gmail ever came along. Steve Ballmer famously disregarded the launch of the iPhone in 2007 in this viral video — but do you know anyone that owns a Windows phone today? I thought so. No one uses MSN Messenger and barely anyone uses a hotmail address nowadays.

And that’s not because Microsoft abandoned those businesses willingly, in fact they doubled down in an attempt to take on Google and ultimately win.

So much for that winner-take-all theory…

Winning in Cloud

In the past 10 years, Microsoft Sales are up 2.4X, EBIT margins expanded by 10 whole points and EPS is up 4.2X. But the stock price is up 12X…!!!

Half of that 12X was since November 2022 — AKA the Chat GPT trade. More on that here. Besides their other smaller businesses, Microsoft is simply riding the Cloud.

Since y/e 2021 (MSFT’s year end is June 30th), they grew “Microsoft Cloud” revenues by c.$20bln a year ending at $111bln y/e 2023. Total revenues for 2023 were $212bln but Q2 ‘24 closed at $62bln — giving us annualised 2024 sales of $248bln (+17% growth y/y).

Microsoft isn’t the only player in Cloud obviously — Amazon’s AWS leads the pack, Microsoft comes next and Google is third in line. But all three of them are fighting for Cloud dominance with an all-in mentality. The big 3 Cloud players comprise 2/3 of the Cloud market, and the fight is still on.

The Peter Lynch Ratio

In his 1989 book, “One up on Wall Street”, Peter Lynch famously wrote:

"The P/E ratio of any company that's fairly priced will equal its growth rate"

For 2023/2022 Microsoft grew Revenues by 7% and EBIT by 6%. Microsoft is currently selling for $3.1trln and even assuming $90bln in net income for 2024 — that’s a P/E of 34X.

In the past decade revenues grew with an annualised rate of 9.3% — making Microsoft shares overvalued by 3.65X, according to Peter Lynch and his PEG formula.

This is of course an oversimplification of company valuation and nuances always apply, but the point is why pay 34X earnings for a company that’s growing at a ~10% clip?

Because of AI? Ok, your call…

Quality of Earnings

In the past decade Microsoft has been transforming under the hood — but sadly not in a good way. In 2014 the company generated 97 cents of FCF for every $1 in EBIT, but in 2023 FCF conversion dropped to 67 cents for every EBIT dollar.

The reason? CAPEX.

Microsoft simply became a more capital-intensive company. In 2014, the company generated $8.7 in sales for every dollar of net PPE (property, plant and equipment) — in 2023 that ratio crashed to $2.4!

Gone are the days where you built one piece of software and sold it to the whole world with practically zero friction. Now, Microsoft is big in the Cloud and that takes serious Capex. In the past 3.5 years net PPE doubled from 53bln at y/e 2021 to 110bln at half-year 2024.

But that’s not all…

Effective from fiscal year 2023, Microsoft changed its depreciation accounting for both server and network equipment, increasing their estimated useful lives from 4 to 6 years. This added $3.7bln of operating income in 2023.

I don’t know what the useful life is for this equipment — but what I do know is this equipment is expensive to buy and costly to run.

Considering the competitive climate for Hyperscalers and the amount of money they need to spend to keep up with developments — they could use some depreciation accounting headwinds to make the books look good! If things don’t work out they could always throw an impairment charge in the mix.

Standing on their Tiptoes & Second-Order Problems

But it’s not just depreciation that we should consider when thinking about this business. We know there’s a very strong trend towards AI solutions from corporates, on top of the trend with cloud solutions.

Microsoft’s Copilot uses AI (from Open AI, I wrote about their relationship here.) in an attempt to “boost productivity, unlock creativity and help you better understand information..” — hold that thought!

No company wants to be left behind — so they spend whatever is needed to execute their AI strategy. This is a boon for Microsoft, but is it a boon for them?

Warren Buffet’s second-order thoughts on investing-just-to-keep-up:

But the promised benefits from these textile investments were illusory. Many of our competitors, both domestic and foreign, were stepping up to the same kind of expenditures and, once enough companies did so, their reduced costs became the baseline for reduced prices industrywide.  Viewed individually, each company’s capital investment decision appeared cost-effective and rational; viewed collectively, the decisions neutralized each other and were irrational (just as happens when each person watching a parade decides he can see a little better if he stands on tiptoes).  After each round of investment, all the players had more money in the game and returns remained anemic.

Corporates are already overstretched and squeezed from all sides, not everyone has it as easy as Microsoft or Apple! We have already established that AI models are in the path of becoming commoditised, and that competition in the Cloud is fierce.

After the honeymoon-phase of broad AI adoption is over, companies (i.e. clients of Microsoft) will once again start to look for cost efficiency and bargains. And if a service doesn’t deliver, they will look to cut it.

Rounding it all up

The question now remains, how much return on capital can Microsoft achieve on all that Capex spend?

Will AI and Cloud solutions become commoditised with time, and if so, how will margins change for Hyperscalers going forward? How much are Hyperscalers growing profitably Vs cannibalising each other?

And considering these unknowns through the prism of Microsoft’s current valuation — is there sufficient margin of safety here?

Sincerely,

Philo