The PEG multiple's weaknesses — when a low PEG fools you

Everyone loves financial ratios that feel exact. PEG is the most tempting of them all: a single number claimed to capture both what you pay (P/E) and what you get (the growth). The rule of thumb sounds simple — PEG below 1 is cheap, above 1 is expensive. But the more exact a financial ratio feels, the more important it is to ask what the number actually hides. According to classic value investing practice, a valuation multiple is a conclusion packaged as a number — and PEG packages two conclusions in a single division.

PEG in its simplest form

PEG = P/E ÷ expected earnings growth, expressed in percentage points.

• Company A: P/E 18, expected growth 12 % → PEG 1,5
• Company B: P/E 24, expected growth 30 % → PEG 0,8

On paper B is cheaper: you do pay more per krona of earnings, but the growth is more than twice as fast. That far, the mathematics is friendly. The problem begins in the denominator — and there almost all the weaknesses live.

The worked example: the sensitivity in the denominator

Take a concrete example in Swedish kronor. Company X has a stock with earnings per share of 10 kr, and the course of the share price stands at 220 kr. P/E becomes 22. The consensus counts on 20 % earnings growth per year ahead. PEG = 22 ÷ 20 = 1,1. Decent, not cheap.

Now the same company, but with a more cautious growth assumption of 10 % instead of 20 %. PEG = 22 ÷ 10 = 2,2. Note what happened: the growth was halved — and PEG doubled. The mathematics of division amplifies every uncertainty in the denominator. A financial ratio that doubles when a single input changes by a factor of two is not a decision basis. It is an amplifier of your own ignorance about the future.

And it is worse than that. Growth is not a fact — it is an analyst's guess, often aggregated into a consensus that historically tends to be optimistic at peaks and pessimistic at bottoms. The same mechanism that makes PEG low when optimism is highest is exactly the mechanism that makes PEG most dangerous then.

Weakness 1: G is a guess — and growth mean-reverts

According to classic value investing practice, high growth is more exposed to regression toward the mean than low. The causes are structural: competition, commoditization, customer concentration, the mathematics of size. A company growing 30 % per year doubles in barely three years — and it is considerably harder to double a large business than a small one.

PEG completely ignores this. The denominator is treated as eternal, when in practice it is an assumption of perhaps one to three years. Always ask yourself: where does G come from? Trailing (last year's growth) is at least a fact — but a backward-looking one, full of one-off items and base effects. Forward (consensus) is a forecast. Do not mix randomly: a PEG built on trailing growth and one built on forward growth are two different numbers that happen to share a name.

Weakness 2: cyclical peaks give the lowest PEG — precisely when it is most dangerous

Here is PEG's biggest and most Swedish trap. The Stockholm exchange is dominated by cyclical businesses: forest, steel, engineering, cars, construction. For a cyclical company, the result is a function of the business cycle position, not of a stable organic growth engine.

Imagine a Swedish engineering company in the middle of an upturn. After a recession bottom, the result has grown 35 % in a year — not because the company has gotten better, but because it is recovering from a deep recession. P/E has been pressed to 8 because E is at a record high. PEG = 8 ÷ 35 = 0,23. Screamingly cheap by the rule of thumb — and in practice often a cycle peak. The next year demand falls, E halves, P/E doubles and the growth turns negative, which makes PEG meaningless (a negative denominator gives a negative PEG — the number stops communicating altogether).

Invert the picture: at the bottom, when the cyclical company is actually cheap, PEG shows negative or extremely high. The valuation multiple says cheap when it is expensive and expensive when it is cheap. That is not a random error — it is systematic, and it arises every cycle.

In the AKM1 model, this is precisely what V12 (revenue stability) exists to capture: how stable are the revenues through the business cycle? A company with a high V12 score — stable, diversified, recurring revenue — is the only kind of company where PEG at all begins to be a meaningful number.

Weakness 3: PEG ignores the balance sheet and the quality of growth

Two companies with an identical PEG 0,9 can be completely different businesses:

• The first is capital light: grows without tying up capital, net debt near zero, gross margin (V07) stable for ten years
• The second grows just as fast but with large liabilities, capitalized development costs and an EBITDA margin (V08) that is pressed every quarter

P/E — and thereby PEG — sees no difference. P/E builds on earnings after net financial items, but the capital structure determines how much of the future growth actually ends up with you as a shareholder and how much goes to creditors. Therefore every PEG exercise should be complemented with a capital-structure-neutral measure such as EV/EBITDA (V06 in AKM1) and a look at net debt and the debt-to-equity ratio (V10).

In addition, PEG says nothing about what the growth costs. Growth bought when the return is higher than the cost of capital creates value; growth bought expensively destroys it — even when delivered in full. That is the whole difference between growing valuable and growing poorer, and it is a question PEG is structurally blind to.

When is P/E alone enough?

Despite all the above, there are situations where the ordinary P/E does the job better than PEG:

• When you compare companies in the same sector with similar business models and capital structures — then the growth differences are often already priced, and P/E ranks them honestly
• When the company is mature with a predictable result: stable customer base, high revenue stability (V12), little acquisition expansion — then the difference between trailing and forward is small and P/E becomes robust
• When you compare the company's current P/E with its own 5-year historical valuation multiple, rather than with a growth forecast — history is at least a fact
• When the growth is roughly the same for all candidates — then the division just scales all the P/E:s down by the same number and adds no information at all

The disciplined order is: first quality (V07, V08, V09, V12), then the balance sheet (V10), then the valuation. A valuation multiple — any one — comes last. Whoever starts with the multiple and hunts for motives afterwards is not doing analysis; that is rhetoric.

PEG's place in the ecosystem — and in the time horizons

In AKM1, PEG has no variable of its own, and that is deliberate: the valuation (V04 P/S, V05 P/B, V06 EV/EBITDA) and the growth (V01, V02) are kept apart, so that you always see what you are paying for what. PEG can still be used as a complement — a third eye — when the conditions above are met.

The AK1TS perspective is at least as important: a growth forecast for next year is a short-term tool. The same company's growth on the medium and long term can be a fraction of the consensus forecast — competition catches up, markets saturate. Before you act on a PEG, ask: does the assumption hold on all the time horizons I intend to own the company? If PEG only works on the shortest horizon, you do not have a valuation argument — you have a momentum argument in disguise.

Next step

• Test the sensitivity yourself in the calculator: enter three companies in the same sector and calculate PEG twice per company — once with consensus growth, once with your own, lower forecast. See how the ranking changes.
• Look up V12 (revenue stability) for the same companies in the syllabus before you trust any growth narrows. The course V09: ROE shows how you simultaneously check that the growth is value-creating.
• Run a complete deep analysis with scenarios instead of a single growth forecast — then the denominator is forced to compete with a base case, not just a wish.

FAQ

What is the PEG ratio and how is it calculated?

PEG är P/E-talet delat med väntad vinsttillväxt, uttryckt i procentenheter. Ett bolag med P/E 22 och väntad tillväxt på 20 procent får PEG = 22 ÷ 20 = 1,1 — tumregeln säger att tal under 1 betraktas som låga och över 1 som höga.

Why can a low PEG be misleading for cyclical companies?

I konjunkturtoppar är vinsten rekordhög, vilket pressar både P/E och PEG samtidigt. Ett verkstadsbolag med P/E 8 och 35 procents tillväxt ur en recessionsbotten får PEG 0,23 — lågt enligt tumregeln men ofta en cykeltopp, eftersom både vinsten och tillväxten vänder nedåt nästa år.

When is P/E enough on its own?

När du jämför mogna bolag i samma sektor med likartade affärsmodeller, eller när du ställer bolagets nuvarande P/E mot dess egen 5-åriga historiska multipel. Om tillväxtantagandena är ungefär desamma för alla kandidater tillför PEG-divisionen ingen ytterligare information.

This is educational financial analysis, not investment advice.