When to Sell a Net-Net Stock: Graham's Exit Rules

Graham set a profit objective and a two-to-three-year limit

Graham described a 50–100% profit objective and a maximum holding period of two to three years.

Benjamin Graham gave his selling policy in a 1976 interview with the Financial Analysts Journal, "A Conversation with Benjamin Graham", September–October 1976. An investor, he said, should set "a reasonable profit objective on each purchase--say 50 to 100 per cent--and a maximum holding period for this objective to be realized--say, two to three years. Purchases not realizing the gain objective at the end of the holding period should be sold out at the market."

In the same interview Graham described the net-net approach itself: buying common stocks below their working-capital value, giving no weight to plant and other fixed assets and deducting all liabilities in full. Over a 30-odd year period at Graham-Newman, he said, "we must have earned an average of some 20 per cent per year from this source." He also counted more than 300 such issues in the Standard & Poor's Stock Guide in January 1976, about 10% of the list, which is a reminder that the opportunity set expands and contracts with the market.

Two things in that rule matter more than the exact numbers: the objective is set before the purchase, and the clock runs whether or not the market cooperates. This profit-and-time policy is distinct from this guide's NCAV-based exit framework: the interview did not prescribe selling exactly at NCAV per share.

Studies tested one-year and five-year holding periods

Oppenheimer tested one-year portfolios, while Xiao and Arnold examined five-year holding periods. Each test's holding period is part of the context for its results:

• Oppenheimer (1986) bought US stocks at or below two-thirds of NCAV per share each December 31 and held them for a year, repeating from 1970 to 1983: a mean return of 29.4% a year against 11.5% for the NYSE-AMEX index, or about 28.2% compounded.
• Montier (2008) tested a global basket from 1985 to 2007 and found more than 35% a year against 17% for an equally weighted universe.
• Xiao and Arnold (2008) measured five-year holding periods on the London Stock Exchange and still found market-adjusted returns of up to 19.7% a year.

These are historical portfolio tests, not proof of one best holding period or a guarantee of future returns. A yearly review or rebalance is not the same as a rule requiring every holding to be sold after one year. For the full list of studies and the caveats on their headline numbers, see the net-net strategy's historical returns.

Sell when the discount to NCAV per share closes

Sell when the share price reaches the latest NCAV per share and the discount closes. Recalculate NCAV per share from the latest reported balance sheet, not from the figure that was true when you bought. P/NCAV = share price ÷ NCAV per share, equivalent to market capitalization divided by total NCAV. At P/NCAV = 1, the discount to net current assets is zero, although the site's screen includes P/NCAV ≤ 1.

Buying at two-thirds of unchanged NCAV per share and selling at that value produces a 50% price gain before costs. This arithmetic illustrates the lower end of Graham's quoted profit range; it does not explain its historical origin. Profits may increase NCAV per share, but do not necessarily do so: recalculate it from the latest balance sheet rather than assuming the target has risen.

Sell when balance-sheet deterioration invalidates the thesis

Exit when balance-sheet deterioration invalidates the investment case, even if the price target has not been reached. Watch for changes that undermine the asset backing:

• Cash burn. Operating losses consume the current assets that are the whole investment case. A year of burn can remove the discount without the share price moving at all.
• Write-downs. Receivables that go uncollected and inventory marked down destroy NCAV directly. If a company's assets keep being written down, value them the strict way instead: net-net working capital counts receivables at 75% and inventory at 50%.
• New liabilities. They reduce NCAV when not matched by an increase in current assets. Borrowing retained as cash increases assets and liabilities equally and alone does not reduce NCAV; spending that cash or adding liabilities without corresponding current assets does.
• Dilution. Shares issued below NCAV per share hand part of your discount to the new holders — the same trap covered in the 10-point net-net checklist.

Sell when the agreed holding deadline expires

Sell if the pre-agreed maximum holding deadline expires without reaching the profit target. Graham's quoted example was two to three years. A statistically cheap company can stay cheap for years, so set the deadline before buying rather than extending it under pressure. A review or rebalance cadence is not automatically a mandatory holding deadline.

The point of the deadline is that it removes the decision from the moment it is hardest to make. Without one, a dead position quietly becomes a permanent one. Graham's phrase "sold out at the market" means disposing of the position at prevailing prices, not an instruction to submit an unprotected market order.

Record the target, deadline and early-exit triggers before buying

Before buying, record the target price (latest NCAV per share, or purchase price × (1 + the chosen profit target)), the date the holding period ends, and the balance-sheet triggers that invalidate the investment case and end it early. A rule written before you own the stock is easier to apply consistently than one invented after a price decline; it does not put a floor under losses.

Diversification reduces dependence on any one exit but cannot eliminate losses. Graham held these as a group rather than as individual bets, and recommended at least 30 names — see what a net-net stock is and why he diversified. Even a diversified basket can suffer substantial losses.

Allow for spreads, fees and thin trading when selling

Allow for bid-ask spreads, fees and thin trading when planning an exit. Net-nets are mostly micro-caps: a position that took days to build can take days to unwind while moving the price against you. Commissions, currency conversion and applicable taxes can reduce realized returns. More frequent trading can increase these costs; a yearly review alone does not imply that every holding must be replaced.

None of that argues for abandoning the rule. It argues for sizing positions you can actually exit, using limit orders, and accepting partial fills over several sessions.

Frequently asked questions