Price-to-Earnings — What you're really paying for when you buy a stock
Worked example: Microsoft — 2013–2019
In January 2000, the S&P 500 traded at 32 times trailing earnings. By October 2002, it had fallen 49%. Corporate earnings — the E in P/E — declined only 18% over that period. The rest of the damage came from multiple compression: investors who had been willing to pay 32 times a dollar of earnings stopped paying 32 times and reverted toward 15 times. The business fundamentals changed modestly. The price investors were willing to pay for those fundamentals changed dramatically. Understanding that distinction — earnings change versus multiple change — is the foundation of every serious valuation framework.
The concept in 60 seconds
The price-to-earnings ratio is the price of one share divided by the earnings per share that share represents. At its simplest, it answers the question: how many dollars are investors paying for each dollar the company earns each year?
A P/E of 20 means the market is paying $20 for every $1 of annual earnings. A P/E of 10 means it is paying $10. The ratio shows up in every financial database, every earnings report, and every analyst note — which is precisely why it gets misused so often.
There are two common versions. Trailing P/E uses the last 12 months of reported earnings. Forward P/E uses analyst estimates for the next 12 months. Neither is inherently better. Trailing earnings are known facts; forward earnings are educated guesses. Both are useful in context, and both can mislead.
The inverse of P/E is the earnings yield: earnings divided by price, expressed as a percentage. A P/E of 20 equals an earnings yield of 5%. A P/E of 10 equals a yield of 10%. The earnings yield is useful because it can be directly compared to bond yields — a key input in the decision framework below.
P/E ratios vary widely by sector and business model. Capital-light compounders with predictable growth (software, consumer staples) typically trade at 25–40 times earnings. Cyclicals and commodity businesses often trade at 8–12 times — and even that can overstate the true multiple when earnings are at a cyclical peak.
Mental model
Think of P/E as the market's patience score — a measure of how long investors are willing to wait for their money back through earnings alone.
A P/E of 10 says: at current earnings, you recover your purchase price in 10 years (ignoring growth and dividends). A P/E of 30 says: you are willing to wait 30 years at current earnings — or you expect earnings to grow fast enough that the wait is much shorter in practice.
Every P/E embeds a growth assumption. A stock trading at 30 times earnings is not "expensive" in any absolute sense. It is only expensive if the growth required to justify 30 times is unlikely to materialize. A stock at 10 times earnings is not "cheap" if the earnings are about to collapse.
This reframes the investor's job. The question is not "is this P/E high or low?" The question is: what growth rate is the market pricing in at this multiple, and do I believe that rate is achievable? A P/E is not a verdict — it is a question.
One more anchor: compare the earnings yield to the 10-year Treasury yield. When a 10-year bond yields 5% and a stock's earnings yield is 4%, the stock offers less income certainty than a government bond — with more risk. When the 10-year yields 2% and the earnings yield is 6%, equities offer a meaningful risk premium. The gap between these two numbers — the equity risk premium — is one of the most powerful inputs in any valuation decision.
Worked example: Microsoft, 2013–2019
In 2013, Microsoft was widely dismissed as a legacy technology company — tied to Windows, declining PC sales, and a failed mobile strategy. Its shares traded at roughly $33, earnings per share were approximately $2.76, and the market assigned a P/E of 12 times. A multiple you would expect for a slow-growing industrial business, not one of the most profitable software companies in history.
Over the next six years, something changed — but it was not the underlying business so much as the market's perception of it. Satya Nadella refocused Microsoft entirely around cloud computing. Azure grew from a minor product line into the second-largest cloud platform in the world. Office 365 converted millions of one-time license customers into recurring subscription revenue. Earnings grew, but the multiple grew faster.
By 2019, Microsoft shares were trading at $136. Earnings per share had reached $5.06 — an 83% increase from 2013. The P/E had expanded from 12 times to 27 times — a 125% increase. A minority of the total share price gain came from earnings growth. The majority came from the market deciding it would pay more than twice as many dollars per dollar of earnings.
The lesson is not that multiple expansion is good or bad. When a business transitions from low-quality earnings (lumpy, cyclical, dependent on one product) to high-quality earnings (recurring, predictable, platform-driven), the market re-prices those earnings upward. Identifying that transition before the market does — or recognizing when the market has already priced it in — is where P/E analysis becomes genuinely useful.
| Year | Share price | EPS (TTM) | P/E | What the market was pricing |
|---|---|---|---|---|
| 2013 | $33 | $2.76 | 12x | Legacy PC software, flat growth |
| 2015 | $47 | $1.48 | 32x | Cloud transition recognized; GAAP depressed by restructuring |
| 2017 | $73 | $3.25 | 22x | Azure scaling; recurring revenue model accepted |
| 2019 | $136 | $5.06 | 27x | Cloud compounder; durable, high-margin recurring growth |
Historical pattern
1999–2002: The dot-com compression. The S&P 500 peaked at 32 times earnings in March 2000 — nearly double the long-run average of 15–17 times. The embedded assumption was a prolonged period of exceptional corporate earnings growth driven by the internet economy. When that growth failed to materialize at the expected pace, the multiple contracted to 15 times even as earnings themselves declined only modestly. The S&P fell 49%. Most of that decline was a P/E story, not an earnings story.
2008–2009: The value opportunity. By March 2009, the S&P 500 traded at 11 times depressed earnings. Normalized earnings — stripping out the financial crisis write-downs — implied a multiple closer to 8 times on through-cycle earnings. Investors who bought at that multiple earned exceptional returns over the following decade, not because the economy recovered spectacularly, but because 8–11 times through-cycle earnings for high-quality businesses was genuinely cheap.
2013–2019: The Microsoft re-rating. As described above, Microsoft's P/E expanded from 12 times to 27 times as the market recognized a fundamental shift in earnings quality. This pattern — a re-rating driven by business model transformation rather than simple earnings growth — appears repeatedly in the histories of great long-term compounders.
2020–2022: The rate cycle compression. By late 2021, the S&P 500 traded at 37 times earnings. The earnings yield was approximately 2.7% — barely above the 10-year Treasury yield of 1.5% at the time, offering almost no equity risk premium. When the Federal Reserve began raising rates aggressively in 2022, the 10-year yield rose from 1.5% to over 4.0%. Suddenly a 2.7% earnings yield looked deeply unattractive against a risk-free 4%. The market compressed its P/E from 37 times to approximately 18 times — a 51% compression — even as S&P 500 earnings held up relatively well. The 2022 bear market was almost entirely a multiple compression event.
Decision framework
Step 1 — Establish the earnings baseline. Before using any P/E ratio, ask: what are you dividing by? Trailing GAAP earnings can be distorted by one-time items, restructuring charges, or cyclical peaks and troughs. Normalize earnings by using a 3–5 year average, or by converting to owner earnings (operating cash flow minus maintenance capex). The multiple is only as reliable as the E you use.
Step 2 — Calculate the earnings yield and compare it to the 10-year Treasury. Invert the P/E to get the earnings yield. Compare it to the current 10-year Treasury rate. The difference is the equity risk premium you are receiving for taking on business risk over a risk-free bond. A premium below 1–2% is historically low and suggests limited margin of safety. A premium above 4–5% suggests the market is being excessively pessimistic.
Step 3 — Identify the growth assumption embedded in the multiple. Use a simple rule of thumb: a P/E of 15 implies modest growth in line with GDP; a P/E of 25 implies moderate growth of 8–10% per year; a P/E of 40 implies strong growth of 15%+ sustained for many years. Ask whether the business is realistically capable of delivering the growth the market is pricing. If the embedded assumption requires best-case outcomes for a decade, the multiple offers little margin of safety.
Step 4 — Assess earnings quality. Not all earnings are equal. Recurring subscription revenue is more valuable than lumpy project revenue. Cash earnings are more valuable than heavily adjusted GAAP earnings. A business earning $5 per share in highly predictable, capital-light cash earnings deserves a higher multiple than one earning $5 per share in cyclical, capital-intensive earnings that may not repeat. Check the conversion from net income to free cash flow — divergence is a warning sign.
Step 5 — Cross-reference with sector context and own history. Compare the current P/E to the company's own 5–10 year history and to direct competitors. A company trading at a 20% discount to its own historical average, with no fundamental deterioration, is making a different argument than one trading at a 50% premium. The company's own history is often the most relevant benchmark.
Common mistakes
Trusting forward P/E without adjusting for analyst optimism. Sell-side earnings estimates are systematically biased upward. On average, forward earnings estimates exceed actual reported earnings by 5–10%. A stock that looks attractive at 18 times forward earnings may be trading at 22 times what it will actually earn. Always stress-test the E before accepting the P/E at face value.
Comparing P/E ratios across different sectors. A retailer at 12 times and a software company at 30 times are not telling you the same thing about relative value. Industries have fundamentally different capital requirements, growth rates, and earnings volatility. Using the same P/E threshold across sectors produces systematic errors — either labeling capital-light compounders as expensive when they are not, or missing cyclical traps hiding behind temporarily depressed multiples.
Ignoring the difference between GAAP earnings and cash earnings. A company that reports $5 per share in net income while generating only $2 per share in free cash flow is not a business worth 15 times its stated earnings. The P/E ratio uses accounting earnings, not cash. Always check how much of the stated earnings converts to actual cash in the bank. Heavy stock-based compensation, aggressive revenue recognition, or large non-cash charges can make the GAAP P/E mislead significantly in both directions.
Treating a low P/E as automatically cheap. The value trap is the most common mistake made by beginning value investors. A stock trading at 8 times earnings looks cheap — until you examine why. It may be cheap because earnings are at a cyclical peak and will fall. It may be cheap because the business is in structural decline. It may be cheap because the capital structure is dangerous. Low P/E is a starting point for investigation, not a conclusion. The cheapest-looking stocks in any market screen are usually cheap for a reason.
How VI Stack uses this
P/E is one of five valuation lenses applied in Gate 3 — The Forensics. It is never used alone.
In the Gate 3 valuation model, P/E is calculated on three bases: trailing GAAP, normalized (3-year average), and owner earnings (operating cash flow minus maintenance capex). The three figures tell different stories, and the divergences are often more informative than any single number.
The earnings yield is always compared to the current 10-year Treasury yield, with the equity risk premium tracked over the full research period. A business that passes Gates 1 and 2 — genuinely understandable, genuinely high quality — needs to offer a meaningful risk premium over risk-free alternatives to earn a position.
In Gate 4 — The Pitch — the P/E multiple used in the target price is justified explicitly: what earnings are assumed, what multiple is assumed, and why both are defensible. References to "industry average" multiples without further support are not accepted. The multiple used must be justified by the quality of the earnings, the durability of the growth, and the current interest rate environment.
What's next
val-02 — EV/EBITDA explores the enterprise value multiple — a valuation tool that captures the full capital structure of a business, not just the equity. It is the preferred metric for comparing businesses with different debt levels and is widely used in private equity and M&A. Understanding when EV/EBITDA is more informative than P/E — and when it is not — sharpens the valuation toolkit considerably.
vistack.io