Graham's net-current-asset method values a company at current assets minus total liabilities and assigns zero to every fixed asset. That was a defensible blanket rule in 1934 and it is a poor one now, because it is simultaneously too harsh on land and too kind on receivables. The modern discipline is the same idea run line by line with defensible haircuts.
Six principles do the work.
Value each asset at what it fetches in a forced sale within a stated period, not at book and not at a valuer's optimistic estimate. The period is part of the assumption and should be written down.
Age the receivables. A ledger's oldest bucket is not worth 90% of a fresh one; a receivable over 180 days old in an industrial business is usually a dispute rather than a delay.
Split the inventory by stage. Raw material has a resale market. Finished goods have a discount market. Work in progress usually has neither, because a half-machined casting is worth its scrap value to everyone except the company that was going to finish it.
Advertisement
Take land at market value, not book, which in India is frequently the largest single adjustment in either direction, and take plant at a heavy discount to net book value, because specialised equipment in a distressed sale is worth what a competitor will pay to add capacity, which in a downturn is very little.
Write off goodwill, capitalised development costs and deferred tax assets entirely. A deferred tax asset is a claim on future profits that the liquidation assumption has just abolished.
Then subtract the things that are not on the balance sheet: the cost of running the business while it is wound down, the severance owed to employees, and the professional and process fees, which are typically a percentage of realisations and are large.
Only after that do you subtract the liabilities, all of them, at full value, including the statutory dues and employee claims that rank ahead of unsecured lenders.
How do you test a liquidation value?
The worked example runs this for an Indian rolling mill and produces ₹126.15 a share against a market price of ₹105. Then it does the thing that makes the number usable, re-running it at the pessimistic and optimistic ends of the two haircuts that matter, land and plant, to get ₹48.34 and ₹191.69.
A liquidation value is a band. Anyone who quotes it as a point estimate has not tested it.
The private-market cross-check is the other half. Take normalised operating profit, apply the multiple that unrelated buyers have actually paid for similar assets in recent transactions, subtract net debt, and divide. It answers a different question from the liquidation value, and where the two are far apart the difference is the value of the business as an operating whole rather than as a pile of assets. In the worked example the private-market value is ₹316.67 against a liquidation value of ₹126.15, and the going-concern discounted value is ₹82.04, which is the lowest of the three. That ordering is the interesting fact: the current owners are running the assets at a return below the cost of capital, the estate is worth more than the operation, and an industry buyer would pay more than either because they would run it differently. Every one of those three sentences is a different investment thesis, and the three numbers are how you tell which one you are actually holding.