Harbor Coffee, our fictional coffee business, trades at $66 a share. Its forward P/E is about 20.37. Suppose next year's earnings forecast falls 20% while the price stays put. The multiple jumps to about 25.46.
The share costs exactly the same. Each dollar of expected profit now costs more.
The change is in the forecast, the number beneath the price. Before comparing forward P/Es, you need to know whose estimate it is and which year it covers.
The earnings have not happened yet
Keep the P/E calculation, but replace reported earnings per share with expected EPS for a named period. For Harbor, that is next fiscal year, FY4, following its latest reported year, FY3.
$66 ÷ $3.24 = about 20.37 times expected FY4 earnings.
Sell-side analysts, researchers at brokerages and investment banks, build earnings forecasts. A data provider combines the forecasts that meet its rules into a consensus earnings estimate.
The provider may use the mean (average) or median (middle of the ordered estimates). A very high forecast can pull up the mean without moving the median. Consensus combines opinions; it does not require agreement.
The calendar matters too:
- Next fiscal year covers the company's next full financial reporting year.
- A next-twelve-month earnings estimate, or NTM estimate, covers a rolling year ahead. Halfway through a fiscal year, that window spans parts of two years.
Fidelity's glossary describes a forward P/E field using next-fiscal-year median estimates. Other fields use rolling forecasts. Two screens can show different forward P/Es without either doing the division wrong.
Keep the price fixed
Harbor's teaching figures use the $66 FY3 year-end price and GAAP diluted EPS throughout. The last row cuts the FY4 forecast by 20%; price stays fixed to isolate the effect, not predict a market reaction.
| Earnings period | EPS | Price | P/E |
|---|---|---|---|
| FY3 · reported | $3.00 | $66 | 22.00 |
| FY4 · forecast | $3.24 | $66 | 20.37 |
| FY4 · 20% cut | $2.592 | $66 | 25.46 |
Trailing P/E is $66 ÷ $3.00 = 22. After the cut, $3.24 × 0.8 = $2.592, and $66 ÷ $2.592 gives a forward P/E of about 25.46.
A 20% cut leaves 80% of the forecast. Dividing by 0.8 multiplies the ratio by 1.25: a 25% increase. The same price now buys a smaller expected profit.
Moving left along the curve means shrinking the forecast. It gets steeper as earnings approach zero: the same dollar cut does more damage to a smaller denominator.
High or low compared with what?
Harbor's original forward P/E is below its trailing P/E because $3.24 is 8% above $3.00. Its apparent discount comes entirely from expected growth. After the cut, forecast EPS falls below last year's $3.00, and forward P/E rises above the trailing 22.
That reading works because the earnings definitions and annual periods line up. Compare GAAP trailing EPS with adjusted earnings in the forecast, and the gap may come from excluded expenses instead of growth.
For peers in the same industry, match the forecast horizon, earnings basis and quote date. Then ask whether growth and risk explain the gap. There is no universal “under 15 means cheap” line.
Forecast dispersion is the spread among analysts' estimates for the same period. A wide spread means the consensus hides disagreement. A narrow spread can hide a shared mistake. Agreement is different from accuracy.
Read the label before the ratio
Five details tell you what you are dividing by:
- As-of date. Check when the price and consensus were last updated.
- Period. Find the fiscal year-end date, or confirm that the field uses NTM. FY1 often means the earliest fiscal year with unreported results, which can already be underway.
- Earnings basis. Match GAAP with GAAP, or adjusted with the same adjustments. Keep both EPS figures on a diluted basis.
- Contributors. Check how many analysts are included and how far apart their high and low estimates sit.
- Revisions. Compare forecasts for the same earnings period at different dates.
An earnings estimate revision changes the forecast for the same period. A rollover changes which year the screen shows. FY1 is a moving label; Harbor's FY3 and FY4 name fixed years in our example. A switch to a later year's forecast can change P/E without any analyst changing their mind.
A March 2023 NBER working paper found both optimistic and conservative bias in annual earnings forecasts for a sample of U.S.-listed stocks over 1986–2016. Optimism was more prevalent. Treat that as a reason to test the forecast's assumptions, not to dismiss every analyst.
Management's own outlook is earnings guidance. An EPS forecast estimates company profit; analyst ratings and price targets express a view of the stock. They answer different questions.
When the forecast stops being useful
Near zero, even a small dollar change in forecast EPS can cause a huge swing in P/E. At zero, you cannot divide; below zero, the ratio cannot usefully rank stocks from cheap to expensive. A blank forward P/E can mean missing forecasts, not just expected losses.
A rebound forecast after a weak year can make the ratio look appealing before lasting profitability is clear. The unusual-year warning still applies: a forecast does not turn a temporary recovery into dependable profit.
Harbor's answer is 25.46 times expected earnings after the cut. The original $3.24 consensus and our $2.592 stress case both cover FY4 GAAP diluted earnings. Before trusting another displayed multiple, verify its earnings period and basis just as carefully.
The next question is whether the earnings growth on offer justifies the price. The PEG ratio brings growth into that comparison.
In short
- Forward P/E prices a forecast, so name its earnings period and definition.
- Consensus combines analysts' estimates; it is no promise of future profit.
- At an unchanged price, a 20% forecast cut raises forward P/E by 25%.
- Compare matching forecasts and check revisions before calling a multiple low.