The Calculation
The formula is direct:
P/E Ratio = Share Price ÷ Earnings Per Share (EPS)
If a stock trades at $50 and reported $2.50 of earnings per share over the trailing twelve months, the P/E is 20. The aggregate equivalent is that market capitalisation equals twenty times the most recent annual net income attributable to common shareholders. The two formulations describe the same relationship. The per-share version is conventional in published research; the aggregate version is more informative when the share count is changing materially through buybacks or new issuance, since EPS movement under those conditions can be partly mechanical rather than economic.
Trailing P/E and Forward P/E
Two variants are standard. The trailing P/E (TTM) uses reported earnings over the past twelve months and is the default value displayed on most data services. Its advantage is that it rests on audited or reviewed figures rather than projections. The forward P/E uses sell-side consensus estimates for the next twelve months and reflects what the market is pricing the company to earn going forward.
Neither variant is inherently superior. The trailing version is generally more reliable for stable businesses with predictable earnings. The forward version is the more informative anchor for companies whose earnings power is changing rapidly, where last year’s reported figure no longer represents the prospective run-rate. The analyst should be aware that consensus forward estimates are systematically optimistic in aggregate, particularly several quarters out, and should adjust accordingly.
What Constitutes a Reasonable P/E
There is no universal threshold. A P/E of 15 may be expensive for a regulated utility growing earnings at low single digits and inexpensive for a software business compounding earnings at 25% annually. The ratio acquires meaning only against a reference: the company’s own multiple history, the multiples of its sector peers, the prevailing market multiple, and the trajectory of the underlying earnings stream.
| P/E Range | Typical Interpretation |
|---|---|
| Below 10 | Cheap on the surface. The analyst’s task is to identify why — cyclical trough, structural decline, distressed balance sheet, or genuine mispricing. |
| 10–15 | Below the long-run market average. Common in mature, slow-growth sectors such as utilities, regional banks, and industrial conglomerates. |
| 15–25 | The historical range for the broad market. Most established companies with moderate growth trade in this band. |
| 25–50 | A growth premium is being paid. The investor is implicitly forecasting earnings expansion sufficient to justify the price. |
| Above 50 | High-growth, speculative, or both. The multiple requires explicit justification, not a general appeal to growth. |
| Negative / N/A | The company has no earnings. The P/E is not meaningful and revenue- or asset-based multiples should be used instead. |
Sector Context
The P/E is most useful as a relative measure within a single sector. Technology companies have generally traded at higher multiples (roughly 25 to 40 in normal conditions) because incremental capital is reinvested into growth at high marginal returns. Utilities trade lower (roughly 10 to 18) because their growth ceiling is regulated and the bulk of earnings is distributed as dividends. A direct comparison of multiples between the two sectors is generally uninformative absent independent corroboration of the underlying economics.
The S&P 500’s long-run average P/E is approximately 15 to 17, with substantial dispersion around that figure. When the index multiple rises materially above the long-run range, equities are collectively expensive in historical terms; when it falls below 12, they are collectively cheap. Mean reversion in aggregate market multiples is one of the better-documented regularities in the empirical literature, though the timing of any specific reversion has been notoriously difficult to forecast in advance.
The PEG Ratio: P/E Adjusted for Growth
The PEG ratio, popularised by Peter Lynch in One Up on Wall Street, addresses the most cited weakness of the simple P/E: it does not incorporate the rate at which earnings are growing. The PEG divides the P/E by the expected earnings growth rate, expressed as a percentage:
PEG = P/E Ratio ÷ Annual EPS Growth Rate
A P/E of 30 paired with 40% expected earnings growth produces a PEG of approximately 0.75; the same P/E paired with 5% growth produces a PEG of 6. As a rough convention, a PEG below 1.0 is treated as potentially undervalued relative to growth, a PEG near 1.0 as fairly valued, and a PEG above 2.0 warrants explicit scrutiny of the underlying growth assumption. The convention is rough because it is sensitive to the growth horizon used, the reliability of the estimate, and the durability of the projected rate beyond the explicit window. A PEG of 0.5 calculated against an unsustainable one-year growth burst is not a bargain.
Where the P/E Misleads
Cyclical companies. Cyclical sectors — auto manufacturers, steel, mining, semiconductors at certain points in the capital-equipment cycle — produce their lowest P/E ratios at the top of the earnings cycle, when trailing earnings are temporarily inflated, and their highest P/E ratios at the trough, when trailing earnings are temporarily depressed. Buying a cyclical purely on a low headline multiple is a recurring error. The cyclically adjusted multiple, normalised across a full cycle, is the more appropriate reference figure.
Non-recurring items. A reported earnings figure inflated by a one-time gain — asset sale, litigation recovery, tax benefit — produces an artificially low P/E; conversely, a large impairment or restructuring charge produces an artificially high or negative one. The analyst should reconcile reported earnings to a normalised figure before drawing conclusions. The footnotes to the 10-K are the primary source for this work, and there is no acceptable shortcut.
Negative or thin earnings. A company posting losses has no meaningful P/E. For unprofitable businesses, price-to-sales, price-to-book, or enterprise value to revenue are the appropriate substitutes. Forcing a P/E onto a company in early-stage growth or in mid-turnaround tends to produce numbers that look precise and convey little.
Share buybacks. Repurchases reduce the share count and thus mechanically increase EPS. A company whose net income is flat but whose EPS is rising 6% annually because of buybacks will display a declining P/E without any underlying improvement in the operating business (one sees this most clearly when the aggregate net income line is plotted alongside the per-share line over a five-year window — the gap between the two curves is often the entire story). The analyst should generally treat aggregate net income trends and per-share trends as separate questions, since they answer different ones.
Accounting differences. US GAAP earnings, IFRS earnings, and various company-defined non-GAAP measures (adjusted earnings, cash earnings, distributable earnings) can produce materially different EPS figures for the same enterprise in the same period. Cross-jurisdictional and cross-company comparisons require reconciliation. The work is tedious and there is no acceptable shortcut for it either.
Applying the Ratio
In practice the P/E is most useful as a filter rather than a decision rule. The procedure is approximately as follows. Begin with the sector average and the company’s position within the peer distribution. Calculate the PEG to determine whether the multiple is reasonable given the growth profile. Examine the company’s own historical P/E range over a full economic cycle — a stock currently at 25× that has historically traded between 12× and 18× is expensive by its own standards regardless of where the sector trades. Investigate the trajectory of the multiple over the past several years to gauge whether sentiment is expanding or contracting. Reconcile reported EPS to a normalised figure, removing the one-time items.
The case of Wells Fargo in 1990 is instructive. The stock traded near $20 at a single-digit headline multiple, calculated on earnings that were temporarily depressed by elevated credit-cycle provisions. On a normalised figure that backed out the cyclical provision, the multiple looked considerably higher; on the reported figure, considerably lower. The normalised figure was the operative one for the underlying business, and the reported figure was the operative one for an investor willing to underwrite the cycle. Both were correct as far as they went; neither, in isolation, settled the question.
The P/E does not, on its own, identify investments. It identifies candidates for further work. A serious valuation conclusion requires the broader apparatus of fundamental analysis and due diligence — an examination of the balance sheet, the cash flow statement, competitive position, capital allocation history, and the durability of the earnings stream itself. The technical literature is extensive; Aswath Damodaran’s published treatments of relative valuation provide a thorough reference for readers who want to extend the framework beyond what an encyclopedia entry can reasonably cover.
See also: Fundamental Analysis · Peter Lynch · What Are Dividends? · Value Investing · What Is Market Capitalization?
