The PEG ratio is an attempt to fix the most obvious flaw in the earnings multiple – that eighteen times profit means different things depending on what those profits are doing next. The construction is simple: take the trailing or forward multiple and divide by a growth rate expressed as a plain number rather than a percentage, so twenty-five times earnings growing twenty-five becomes one, while the same multiple against eight becomes roughly three. Popularised by Peter Lynch as shorthand for paying a fair price for growth rather than any price, its appeal is obvious: it makes expensive-looking stocks comparable. Its failure modes are equally obvious once you look at what the division quietly assumes.
Two companies, identical multiples
Both trade at twenty-five times earnings. One is expected to grow profit about 25% a year; the other about 8%.
| Â | Multiple | Assumed growth | Ratio |
|---|---|---|---|
| Company A | 25Ă— | 25 | about 1.0 |
| Company B | 25Ă— | 8 | about 3.1 |
Read on the folklore reading, A is reasonably priced and B looks dear. Nothing in that table should make you confident, because every input came from a forecast and none of the qualities that decide outcomes appear anywhere in it. If A’s growth arrives through borrowed money, temporary pricing, or a competitor’s disaster, the flattering ratio described a year rather than a business. Growth rates are not properties of companies; they’re claims about futures that were drafted by someone.
Where the ratio misleads hardest
It depends entirely on which growth number you fed it. Next-year estimate, five-year consensus, the last three years’ actuals, management’s own target – each produces a different answer for the same share price, sometimes by a factor of two. Any screen advertising “low PEG names” has made that choice silently on your behalf, usually selecting whatever produced the prettiest list. Ask which horizon, which source, whose model.
Base effects flatter the fastest growers. A company rebounding from a collapse can post spectacular percentage growth precisely because the comparison period was awful, producing a tempting ratio right when the underlying franchise is weakest. Percentages hide their own denominators, and this ratio divides by one of them.
Cyclicals invert it. Commodity and shipping producers at the top of a cycle combine strong current profit with analysts extrapolating recent momentum, generating cheap-looking readings just before normalisation. The earnings multiple’s trap repeats, amplified by optimism in the denominator.
Quality is invisible to it. Two businesses growing 20% can differ completely in how much capital that growth costs, how defensible margins are, and whether the growth requires borrowing. Returns actually earned on reinvested capital do most of the explanatory work – decomposition shows why – and neither version appears in the division. A business growing quickly by destroying value per unit of growth scores beautifully.
Dividends, buybacks, leverage and payout policy are absent. A flat-profit company returning cash generously can reward holders while scoring appallingly here; a leveraged non-payer can score brilliantly. Comparisons across industries inherit all of these differences plus accounting variation, which is why ranking by this ratio across sectors mostly ranks business models.
Negative or near-zero growth makes it meaningless. With a zero or negative denominator the ratio collapses into nonsense, and screens frequently exclude such names entirely – quietly deleting entire categories of investment from whichever portfolio you later regret not owning.
Use the instinct, skip the number
What survives scrutiny isn’t the ratio; it’s the habit behind it – never judge a multiple without a view on the future it presupposes. Turn it into questions with documentary answers. Over what horizon does this growth run, and who wrote that figure? How much of it comes from demand versus prices versus costs versus fewer shares outstanding? What has to be true operationally for it to arrive, and what evidence supports each piece? What does the growth cost in capital – compare reinvestment requirements against profit growth, since growth bought with debt at unfavourable terms converts into mediocre returns even when the headline rate impresses – cash generation exposes the gap. Sensitivity beats point estimates: halve the growth assumption, raise the required return slightly, and see whether the case survives. Cases that survive only their own best scenario were never cases.
If you keep the arithmetic anyway, treat sub-one readings as a prompt for reading rather than a purchase signal, stay within a peer group where cost structures resemble each other, name the horizon and source out loud – and be able to explain, in ordinary sentences, why this particular company’s growth is likely and sustainable. That explanation is the actual analysis; the ratio is a way of noticing you haven’t written one yet, which valuation work covers properly. Nobody who beat markets did it by sorting a column, and nobody who lost badly avoided it either – the people who got hurt had stopped asking what sat beneath the number.
This is educational explanation of a valuation shortcut, not investment advice or a recommendation of any kind. Every multiple and percentage above is invented round-number arithmetic used to show relationships, no company is named because forecasts and ratios expire the moment they publish, forecasts are frequently wrong, and any holding can lose value including one screening as cheaply as this one can make a stock appear.