Benjamin Graham’s most mechanical rule was also his most neglected: buy businesses whose entire market price sits below the value of their short-term assets after subtracting every liability they owe. The name describes the subtraction twice over – current assets less total liabilities, hence net-net – and the appeal is brutal simplicity. If the arithmetic holds, you’re paying less than the liquidatable working capital for a claim on the whole enterprise, receiving the fixed assets, the brand and future earnings without charge. It sounds too easy because, applied carelessly, it usually is.
Getting the definition right first
Take current assets – cash, receivables, inventory – and subtract all liabilities, short-term and long-term, including debt, taxes owed and provisions. Divide by shares outstanding. That per-share figure approximates what would remain from short-term items if operations stopped today, before liquidation costs and before anything sold realises less than book. Graham recommended acting only at a substantial discount to that number, conventionally around two-thirds, precisely because the inputs are optimistic by construction: receivables age badly, inventory becomes obsolete, provisions understate what turns up later – the balance sheet explains why book values mislead. Older drafts of this page conflated the approach with a famous insurance holding acquired by Graham’s partnership in the late 1940s. That purchase wasn’t a net-net trade; it was concentrated, held for decades, and depended on management being exceptional – the opposite of a mechanical screen applied across many obscure names. Confusing the two has taught generations of beginners the wrong lesson about what makes cheapness pay off.
Run the arithmetic with invented round numbers so you can check it: suppose current assets of $40 million against total liabilities of $25 million leaves $15 million of net working capital across one million shares, or roughly fifteen dollars per share. Two-thirds of that suggests acting beneath about ten – provided the composition survives interrogation, which is where nearly every failure occurs.
What kills net-net investors
Burn. A business consuming several million a year in operating cash will have consumed its balance sheet within a few periods regardless of how attractive today’s snapshot looks – cash conversion exposes the difference between profit and survival. Cheapness measured against a shrinking denominator decays fast, which means the strategy’s real skill is recognising decay early rather than spotting discounts – most apparent bargains are simply businesses heading somewhere else at a slower speed than expected.
Realisation quality. Book inventory assumes sale at ordinary margins; distressed inventory clears far below it. Book receivables assume collection; aged books collect partially. So experienced practitioners apply punitive assumptions – inventory at perhaps half of stated value, receivables discounted according to how overdue they run, deferred tax assets treated as worthless until proven otherwise – and see whether the discount survives such treatment. A candidate that only looks cheap at full book was never cheap.
Minority-shareholder reality. Owning a claim on assets does not compel anyone to distribute them. Entrenched families, cross-holdings, related-party dealings and comfortable managements have presided over decade-long cheapness in plenty of markets while cash migrated toward affiliates or sat earning nothing. Governance isn’t decoration here; it decides whether the discount ever closes, and in some jurisdictions structural reasons exist for discounts persisting indefinitely.
Dilution. Companies trading far below asset value rarely raise equity on good terms – they raise it because they need money, issuing claims on those very assets at prices punishing existing holders. Monitor share counts across years before committing; a rising count converts your bargain into somebody else’s rescue financing. This connects directly to Graham’s wider screening framework, whose conservatism on debt and dividends exists partly to filter out exactly these cases.
Practically the approach demands breadth and patience rather than conviction. Any individual position can stay absurdly cheap indefinitely, be delisted, or be voted into irrelevance, so disciplined application means many small holdings with written exit triggers – price reaching the conservative asset estimate, evidence of capital returning to owners, persistent cash destruction prompting withdrawal – and time horizons measured in years rather than weeks. Academic work examining many markets over long stretches found extreme cheapness of this kind historically preceded strong average returns, most consistently in smaller companies in certain Asian markets, though the pattern weakens once everyone knows it, transaction costs bite harder in illiquid names, and averages conceal individual outcomes brutally.
Sequence worth following: filter broadly using free data, then eliminate candidates failing governance or cash-consumption tests before doing any valuation work, since the interesting question is never “how cheap” but “why hasn’t this closed, and what would close it.” Read the filings themselves for related-party transactions, going-concern language, auditor behaviour and share issuance; read notes for pension deficits and tax contingencies excluded from the naive calculation. Then size positions so a total write-off changes nothing meaningful – the honest admission that this genre contains many winners offset by many slow bleeders, discoverable only afterwards.
The concept remains valuable even if you never buy such a stock, because it teaches the load-bearing lesson of valuation: price and worth are separate quantities, markets can be wrong for years, and asset-rich cheapness is evidence worth investigating rather than a conclusion to act on – protection comes from that gap, not from the label “cheap”.
This is educational explanation of a historical investing technique, not financial advice or a recommendation to buy or sell anything; all figures above are invented round numbers supplied so the arithmetic can be checked independently, no company is named, liquidation values are estimates that frequently prove generous, and investing involves risk – including strategies that look arithmetically safe on paper.