P/E, forward P/E and PEG
Lesson 9 · about 10 min
A multiple is a price divided by something the business produces. It lets you compare a $40 stock with a $400 stock, or a company with 100 million shares with one with 2 billion. The earnings multiples are the most quoted, the most useful for a quick read, and the most abused. This lesson covers what they say, what they assume, and where each one breaks.
Trailing P/E
Price ÷ earnings per share for the last twelve months (often written TTM, trailing twelve months).
ACME: $40 ÷ $2.10 = 19.0
The intuition: at 19 times, you are paying $19 for every $1 of last year's earnings. If earnings never grew and were all paid out, it would take 19 years to get your money back. The inverse, 1 ÷ 19 = 5.3%, is the earnings yield, which is comparable to a bond yield.
What it assumes: that last year's earnings are representative. For a stable business, fine. For a cyclical business at the peak of its cycle, last year's earnings are the best they will be, and the trailing P/E looks deceptively low (module 1, lesson 3).
Forward P/E
Price ÷ consensus estimated EPS for the next twelve months (or next fiscal year).
ACME: consensus next-year EPS $2.40, so $40 ÷ $2.40 = 16.7
The forward P/E is what most professionals actually use, because a stock is a claim on future earnings, not past ones. Its weakness is obvious: the denominator is a forecast, and forecasts are wrong. Analysts as a group tend to be too optimistic at the top of a cycle and too pessimistic at the bottom, so forward multiples flatter cyclical stocks at the wrong moment.
For a trader the forward P/E has a specific use: it is the multiple the market is paying for the number the company is about to be judged against. If ACME reports and guides to $2.60 instead of $2.40, the stock at $40 is suddenly at 15.4 times, and if the market decides 16.7 was the "right" multiple, the price should be $43.40. That is one way to estimate a post-earnings move: new estimate × old multiple.
| Scenario after earnings | New forward EPS | At the old 16.7 multiple | Implied move |
|---|---|---|---|
| Guidance raised | $2.60 | $43.40 | +8.5% |
| Guidance maintained | $2.40 | $40.00 | 0% |
| Guidance cut | $2.15 | $35.90 | −10.3% |
In practice the multiple moves too (a raise often earns a higher multiple; a cut a lower one), so the real moves are wider. But the arithmetic gives you a floor for what to expect.
Key idea: Trailing P/E prices the past; forward P/E prices the forecast. When guidance changes, hold the multiple constant and re-multiply to get a first estimate of where the stock "should" go.
PEG
P/E ÷ expected annual earnings growth rate (in percent).
ACME: forward P/E 16.7, expected growth 14%, so PEG = 16.7 ÷ 14 = 1.19
The PEG tries to answer the objection that a high P/E is fine if growth is high. A rule of thumb says PEG around 1.0 is "fair", below is cheap, above is expensive. Like every rule of thumb it is a starting point, not a verdict.
Where PEG breaks:
- Low growth. A company growing 2% at a P/E of 12 has a PEG of 6. That does not make a stable utility "expensive"; it makes the ratio meaningless at low growth rates.
- Which growth? Next year's? A five-year average? Analysts' long-term estimate, which is often a guess dressed as a number? Two sources can give PEGs of 0.9 and 1.8 for the same stock.
- Growth has quality. 14% growth funded by acquisitions and debt is not the same as 14% organic growth with buybacks on top.
Use PEG to compare companies in the same sector with similar growth rates. Do not use it across sectors or for companies growing under about 5%.
Reading a P/E in context
A P/E of 19 means nothing alone. Three comparisons make it mean something.
Versus its own history. If ACME has traded between 14 and 24 times over the last five years with a median of 18, then 19 is ordinary. If it has ranged 25 to 35, then 19 is either a bargain or a sign the market thinks growth is over.
Versus its sector. If industrial peers trade at 22 times, ACME at 19 is a slight discount. If they trade at 15, it is a premium, and the question is what ACME does better to deserve it.
Versus the market. A broad index at 20 times makes 19 average. An index at 14 makes it expensive.
| Comparison | ACME | Reference | Read |
|---|---|---|---|
| Own 5-year median | 19.0 | 18.0 | Roughly in line |
| Sector median | 19.0 | 22.0 | Modest discount |
| Broad index | 19.0 | 20.0 | Roughly in line |
A stock at a discount to peers with growth in line with peers is one worth understanding; either the market is missing something or there is a reason (a legal issue, a customer concentration, an ageing product line). Read the risk factors.
Where earnings multiples do not work
- Negative earnings. A company losing money has no meaningful P/E. Use price-to-sales or EV/revenue (next lesson).
- Tiny earnings. A company earning $0.02 a share at $10 has a P/E of 500. The number is technically correct and useless; a penny of extra profit halves it.
- One-off items. A large gain on the sale of a division makes trailing EPS jump and the P/E collapse for exactly one year.
- Banks and insurers. Their "earnings" depend heavily on provisions and reserves; price-to-book (next lesson) is the usual multiple.
Try it: For any company, find trailing P/E, forward P/E and consensus growth. Compute PEG. Then compute what the price "should" be if next year's EPS came in 10% above consensus at the same forward multiple. Compare that with the implied move in the options market before earnings (module 4); they are often surprisingly close.
Recap
- Trailing P/E = price ÷ last twelve months' EPS; forward P/E = price ÷ next year's consensus EPS.
- Forward P/E is the multiple the market pays for the number the company will be judged against; re-multiply a new guidance figure by the old multiple for a first estimate of the move.
- PEG = P/E ÷ growth rate; useful within a sector, meaningless at low growth.
- A P/E means something only against the stock's own history, its sector and the market.
- Earnings multiples fail with negative or tiny earnings, one-off items, and financial companies.