Microsoft Stock: Wonderful Business. Awful Price (NASDAQ:MSFT)

Two types of growth investors

There are two types of growth investors in my experience. First, growth investors who believe P/E does not matter for growth stocks. Second, investors who believe a good growth stock, even a behemoth like Microsoft Corporation (NASDAQ:MSFT) with a $3+ trillion market cap, can perpetually grow at a rate faster than discount rates.

Arguing with the first type of investors becomes largely a matter of investment style and won’t be the focus of this article. Arguments with the second type of investors, however, can be made on a more technical basis. And it is the goal of this article to analyze if MSFT’s can perpetually outgrowth discount rates.

As seen from the chart below, analyst estimates for MSFT’s EPS growth in the next 10 years imply its earnings can compound at a CAGR of ~11% in the next 10 years (from an EPS of $11.7 in 2024 to about $32.74 in 10 years). Whenever the growth rates exceed discount rates (typically assumed to be between 9% to 10%), a simple discounted dividend model (or discounted cash flow model) would show the fair price of the stock is infinite. This is what the current consensus model seems to imply. In the remainder of this article, my goal is to scrutinize and challenge these assumptions.

A great business becomes greater thanks to Azure and CoPilot

Before I start challenging anything, let me first make it abundantly clear that there is no denying that MSFT is a great business. It features a well-balanced combination of mature cash cows and new growth frontiers. For example, its operation system and office software suite, with the subscription model, have essentially achieved an annuity status.

At the same time, the company has established leading positions in several key new frontiers and poised itself well for future growth. The top two new frontiers in my mind are its cloud and the integration of AI. Thanks to its Azure and other cloud services, its Intelligent Cloud segment has rapidly become a significant revenue stream to the tech giant (see the chart below). As of Q4 2023, Azure was the second largest cloud infrastructure service provider, capturing a market share of 24%, far leading the third place (Google Cloud with an 11% market share) and narrowing its gap to AWS.

On the AI front, MSFT is also at the forefront of the wave. Access to Office Copilot has recently expanded with the introduction of Copilot Pro, a subscription ($20 per month) service that provides access to AI features inside Microsoft apps like Word and Excel. I happily paid for the subscription myself. I can easily see why I won’t be able to live without it anymore (just like Windows and Office) and how this would add another annuity-like income stream to MSFT.

From here on, I see endless possibilities that the use of AI can spread fast and widely with almost all other Microsoft applications, including Dynamics 365, LinkedIn, and Windows. The company’s Security Copilot, a generative AI product designed to thwart cyberattacks, could help strengthen security across the Internet. Moreover, AI features could go beyond software and help to reignite MSFT’s PC sales too. As an example, a new Copilot key has been added to the Windows PC keyboard, creating a way to seamlessly use Copilot with our computing hardware.

Awful price

Now let’s look at valuation. As hinted above already, I will rely on the discounted dividend model (“DDM”) in this article. If you’ve used the model (or any other similar model such as the discounted cash flow model), then you know why these models are so popular and powerful. The worth of a company is the sum of the present value of its future dividends (or cash flow, net profit, etc.). It is always self-evident.

But you would also know that beneath the simplicity on the surface, all the trouble hides in two numbers: the growth rates and the discount rate you assume. Any answer can come out if you change these two numbers in a range wide enough. In particular, as mentioned earlier, if you assume a growth rate that exceeds the discount rates, then any price is a good price to buy the business.

Enough general philosophy and let’s examine MSFT specifically. The table below shows my analysis of MSFT’s average cost of capital using the so-called WACC, weight average cost of capital model. This link provides a detailed account of the WACC model and the gist is:

WACC represents a company’s average after-tax cost of capital from all sources, including common stock, preferred stock, bonds, and other forms of debt. According to this model, MSFT’s WACC has been averaging around 9.3% in the past 10 years. Since the WACC is the minimal hurdle a company has to overcome to justify the use of capital, it is widely used as an approximation for the company’s discount rate.

So, in this case, if you believe MSFT can grow faster than 9.3% perpetually, then it makes perfect to buy it at any price. For those of us who do not, our next resort is a staged discounted model. In this model, we assume the company can grow at a faster rate for a number of years, and then at a slower rate (the so-called terminal growth rate) thereafter. It is an approach that introduces a new problem in order to solve an old problem – the new problem is that now we have to estimate a terminal growth rate.

My approach for estimating the terminal growth rates involves the return on capital employed (“ROCE”) and the reinvestment rate (“RR”). The method has been detailed in my other articles, so I will just quote the end results here:

The ROCE for MSFT is around 66%, in line with the average of other tech giants as seen in the chart below. MSFT’s reinvestment rate is about 5% on average, also in line with the average for other major tech firms. As such, I think there are reasonable numbers to use that can represent long-term trends. With these inputs, MSFT’s perpetual growth rate would be ~3.30% (66% ROCE x 5% reinvestment rate = 3.30%). Note this number is the real growth rate without inflation. To obtain a notional growth rate, one would need to add an inflation escalator.

After this point, we are just left with some simple number crunching. I will use its FWD dividend of $3.0 per share in the two-staged DDM calculations.

Given the company’s consistent dividend growth for almost 20 years, it is reasonable to assume the dividend represents its true owner’s earnings in the long term. I won’t bore you with details (which are shown in the table below), and will just quote you the key assumptions and results here:

1. I assume” A) MSFT can grow at 11% CAGR for the next 10 years as consensus projected; and B) its dividends grow at the same rate too. Under these assumptions, MSFT’s dividend would grow to $8.52 per share in 10 years. The present value of the dividends a shareholder collects in this 10-year period adds up to $32.69 per share in the first stage of the DDM. It’s also interesting to note that MSFT’s total market cap would exceed $10 trillion then under these assumptions at today’s P/E of ~35x.
2. Stage 2 of the DDM then assumes MSFT’s earnings and dividends to grow at a terminal rate of 3.3% afterward. I also assumed an inflation escalator of 2.5% (which is the long-term average), so the notional growth rate would be 5.8%. With these inputs, the present value of the second stage of the DDM turns out to be $243.

Thus, the fair value of the stock is $276 under these assumptions. The current price of $418 represents an overvaluation of more than 50%.

Other risks and final thoughts

As a most popular stock on the Seeking Alpha platform, I won’t repeat the risks generic to MSFT (competition, data privacy concerns, competition, et al). Instead, I will focus on a few risks that are more particular to MSFT and my approach.

First, I do not want to leave you with the impression that the WACC model gave the “correct” discount rates. One of the key concepts in the model involves the cost of equity, which is an elusive concept in my mind. I totally resonate with Charlie Munger’s following comments on this concept:

“Obviously, considerations of costs are important in business, and we always had that kind of thinking. Of course, capital isn’t free, and you can easily figure out the cost of loans. But the theorists had to make a measure for what equity costs, and there they just went bonkers.”

That is why Warren Buffett promotes the use of risk-free rates as discount rates for businesses that he understands well (like Coca-Cola). If you believe you understand MSFT’s long-term prospects well enough to apply the risk-free rates as the discount rates, then MSFT’s fair price would be much higher. I want to emphasize that I do not intend this to be a mockery or even a challenge. My only intention is to point out a genuine technical limitation in my analysis.

Second, in terms of downside risks, MSFT relies heavily on sales of legacy software products like Windows and Office. I view this as a concentration risk, especially under the current heightened geopolitical tensions. Certain regions or countries might consider the application of these products as a national security issue (rightfully so in the digital age) and restrict their sales.

To conclude, Microsoft is an undeniably strong company with a differentiating business model, consistent growth, and a dominant market position in several segments. However, the current stock price reflects this optimism to such a degree that the valuation risks are too high – even under some quite aggressive growth assumptions in my view. As such, I see it as a good example of a truly wonderful business at an awful price under current conditions.