Skip to content
Fundamental Analysis

Cheap or broken: a procedure for a low multiple

A screen has handed you a company at six times earnings and below book value. Five checks, in order, that separate a mispricing from a correct discount — and the sentence you have to be able to write at the end.

Fundamental AnalysisAdvanced14 min read
Browse Fundamental Analysis(169)

A screen returns a company at six times earnings, 0.7 times book, with a dividend yield above what a fixed deposit pays. It has been listed for thirty years, it is profitable, the promoter holding is stable and there is no obvious scandal attached to the name. A great many other people have run a screen close enough to produce the same list this month, and most of them reached the conclusion you are about to reach, which is that the market has missed something. Occasionally that is true. The purpose of this lesson is to make you earn the conclusion rather than reach it, and the work is a fixed procedure rather than an instinct.

Think of it like this
The flat priced below the lane

A flat in a good building is listed 20% below what the neighbours have been getting. There are exactly two kinds of explanation and they look identical in the advertisement. Either the seller is moving abroad next month and wants it done — in which case you have found a genuine discount — or the title has a dispute, the completion certificate never came, the society has a levy pending, or the flat is above a transformer. A buyer who assumes the first without checking for the second is not being optimistic; they are declining to do the only part of the transaction that matters.

In the market

A low multiple is a listing price below the lane. The market is not usually asleep. It is usually discounting something specific, and your job is to find out what and then decide whether the discount is too large. That is a different exercise from noticing that the number is low.

The five checks, in order

Each one can end the investigation, which is the point of the order
  1. 1
    1. Is the E real?

    Normalise before you do anything else. Strip one-off gains and losses after tax. Check whether the profit is cyclical-peak profit — for a commodity or a capacity-cycle business, the lowest multiple of the decade usually coincides with the best year of the decade, and the multiple is low precisely because the market expects the earnings to fall. Then set five years of net profit beside five years of operating cash flow. If profit has consistently arrived without cash, the E is an opinion supported by accounting policy, and no multiple applied to it means anything.

  2. 2
    2. Is the E about to be different?

    A low multiple is often a correct forecast rather than a mistake. Look for structural decline: volumes falling while the industry’s volumes rise, the substitute taking the increment, realisations that never recover after each downturn, capital expenditure that produces no additional revenue. Telling a cycle from a genuine decline is worked through at length earlier in this track; here it is only a gate, and the question is whether the stream is stable enough for any multiple applied to it to mean anything. If the market is discounting a stream that is genuinely shrinking, the multiple is not cheap — it is arithmetic, and the shrinking business is a separate valuation problem with its own method.

  3. 3
    3. Is the B real, and separable?

    For a case resting on book value, ask what the book is made of. Land carried at decades-old cost is usually understated, and it is real. Goodwill from acquisitions, capital work in progress that has not moved for four years, receivables from related parties and inventory in a business with falling volumes are none of those things. Liquidation value is a floor only where the assets could actually be sold separately and where somebody with the power to sell them would ever want to.

  4. 4
    4. Can the cash ever reach you?

    A company can be genuinely cheap and still be a bad holding, because the value is unreachable. Cash sitting in a subsidiary you own a minority of, a promoter with no history of paying dividends and no interest in starting, cash earmarked against contingent liabilities or a disputed demand, a holding company whose only asset is a stake it will never sell. The earnings yield on the screen assumes the earnings belong to you. Check the last five years of dividends, buybacks and related-party flows before believing it.

  5. 5
    5. Is this a solvency question rather than a valuation one?

    Look at debt maturing in the next two years against cash generated in the last two, and at interest cover through the worst year of the cycle rather than the last one. Equity in a heavily indebted company is closer to an option than to a share of a business: it is worth a great deal if the refinancing happens and very little if it does not, and a low multiple on that is not cheapness, it is the price of a binary outcome.

Worked example
One screen output, triaged
Four companies from the same screen — illustrative, and each of a familiar type
Company A — 5× earnings, commodity producerRecord realisations for three years; the ten-year average profit is a third of this year’s. The multiple is low because the earnings are at a peakFails check 1
Company B — 7× earnings, established brandVolumes down for five consecutive years while the category grew. A shrinking stream correctly discountedFails check 2
Company C — 0.6× book, asset-heavyTwo-thirds of the book is goodwill from an acquisition and capital work in progress that has not moved in four yearsFails check 3
Company D — 6× earnings, cash-richHalf the market capitalisation is cash, no dividend in eleven years, and a growing loan to a promoter-controlled entityFails check 4
Companies remainingWhich is the ordinary outcome of a screen, and is information rather than a disappointmentNone
What the exercise costTwo failed inside twenty minutes on a normalisation and a volume series; the other two needed about an hour eachAround two and a half hours in total
What it producedWhich is a checklist you re-use on every screen you ever run afterwardsFour described failure modes
Four cheap companies, four different reasons, none of them a mystery and none of them requiring information the public record did not hold. This is the ordinary result. A screen is a search tool, not a conclusion, and the value of running one is that it delivers candidates fast enough that you can afford to reject nearly all of them. The investor who finds one genuine mispricing a year and rejects two hundred is doing this correctly.
Where the two look identical, and where they separate
Cheap — the discount is larger than the problem
  • Earnings that convert to cash, through a full cycle rather than in the good years
  • Volumes flat or growing in line with the industry, so the stream is not shrinking
  • A book made of assets with an identifiable outside buyer
  • A record of the cash actually reaching shareholders — dividends paid, buybacks done, no growing related-party balances
  • A specific, nameable reason for the discount: a bad year, a dull sector, no analyst coverage, a governance fix already made
Broken — the discount is the market being right
  • Profit that has never turned into operating cash, or that is at a cyclical peak
  • Volumes falling while the industry grows, with the substitute taking every incremental sale
  • A book dominated by goodwill, stalled projects and receivables from connected parties
  • Cash that exists but has never moved, in a structure that gives you no way to make it move
  • No specific reason for the discount, only the observation that it is cheaper than it used to be
◆ Checkpoint

Module checkpoint: when growth ends

5 questions. Answers are revealed once you submit all of them.

1.A company reports 18% revenue growth. Volumes are up 3%, realisation per unit is up 8% against an input cost per unit up 9%, and an acquisition made mid-year contributed the balance. How much of the 18% describes the business you owned last year performing better?

2.A business has added ₹450 crore of revenue in each of the last five years, taking it from ₹1,800 crore to ₹4,050 crore. What has actually changed?

3.Why does reading the weakest listed company in an industry give an earlier signal than reading the leader?

4.A company’s earnings rise 12% while its multiple falls from 50 to 25. A colleague concludes that the share price must therefore have fallen by more than 50%. What is wrong with that?

5.A screen shows a company at six times earnings with half its market capitalisation in cash, no dividend in eleven years, and a growing loan to a promoter-controlled entity. What is the correct reading?

0 of 5 answered
Simple bhasha mein
Gali se sasta flat

Achhi building mein flat padosiyon se 20% sasta laga hua hai. Ya toh seller ko agle mahine bahar jaana hai — sacchi bargain. Ya title pe jhagda hai, OC nahi aayi, society ka levy pending hai. Ad dono mein ek jaisa dikhta hai. 6 P/E wali company bhi wahi flat hai: pehle poocho kamai asli hai kya, ghat toh nahi rahi, book mein rakha kya hai, aur cash aapke haath tak pahunchega bhi ya nahi.

What to remember
  • Normalise the earnings before applying any multiple; a cyclical peak produces the lowest multiple of the decade.
  • A low multiple is often a correct forecast of a shrinking stream rather than a mistake about a stable one.
  • Ask what the book is made of, and whether anything in it could actually be sold.
  • Value you cannot reach is not value — check five years of dividends, buybacks and related-party flows.
  • Write one sentence naming what the market believes and what you think it has wrong; without it you have a number, not a thesis.

Common questions

Short, direct answers to what people ask about this topic.

value trap meaning
A value trap is a share that looks cheap on its multiple or its book value but whose discount turns out to be correct — because the earnings are about to fall, are not backed by cash, or can never reach a minority shareholder. The market is rarely asleep on a low multiple; it is usually discounting something specific. The work is finding out what, and then judging whether the discount is larger than the problem deserves.
profit adjusted for one-off items and the stage of the cycle is called
Normalised earnings. Normalising means stripping one-off gains and losses after tax, asking whether the figure is cyclical-peak profit, and setting five years of net profit beside five years of operating cash flow. If profit has consistently arrived without cash behind it, the earnings are an opinion supported by accounting policy and no multiple applied to them means anything.
how do I check whether a low PE stock is genuinely cheap
Work through five checks in order: is the reported profit real once normalised, is it about to be structurally different, is the book value real and separable, can the cash ever reach a minority shareholder, and is this a solvency question rather than a valuation one. The order matters because most low multiples fail at the first two, which are the quickest to perform, so the slow balance-sheet work goes only to ideas that have already survived twenty minutes.
why is a cyclical company cheapest at the top of the cycle
Because the lowest multiple of the decade usually sits on the best earnings of the decade — the denominator is peak profit, and the market is already discounting the fall that follows. For a commodity or capacity-cycle business a very low price-to-earnings ratio at the top of a cycle is a forecast rather than a bargain. That is why normalising earnings across a full cycle comes before applying any multiple at all.
is a stock trading below book value a bargain
Not until you know what the book is made of. Land carried at decades-old cost is usually understated and genuinely real; goodwill from acquisitions, capital work in progress that has not moved in four years, receivables from related parties and inventory in a business with falling volumes are not. Liquidation value is a floor only where the assets could actually be sold separately and where somebody with the power to sell them would ever want to.