Graham's Margin of Safety, Rebuilt for a Trading Account

Open risk $2,000. Cash $12,000. Those two lines sit at the top of a hypothetical trading plan for a $40,000 account, above the watchlist and the setups, and they do more to keep the account alive than any entry signal on the page below them.

They are a margin of safety. Benjamin Graham used that phrase for buying a security far enough below an estimate of its value that errors in the estimate would still leave room. He was writing about investors valuing businesses. A trader’s version of the error is different in kind, since it is about price and timing, and the buffer has to be measured in a different place.

What the buffer protects against

Graham’s investor is wrong about value. The trader is wrong about the next few days of price, which happens constantly, and occasionally wrong in a way no stop can contain: a gap on earnings, a halt, several positions falling on the same morning for the same reason.

So the trader’s margin of safety is a distance. It runs from the worst realistic loss to what the account can absorb. It gets built in layers: a stop placed beyond normal noise, a cap on total open risk, and buying power left unused.

A stop the market can breathe past

A stop set inside the stock’s ordinary daily swing gets hit by the swing, which turns a sound idea into a string of small losses. Placing it beyond normal noise, at the price where the idea is actually wrong, costs you shares, because a wider stop at a fixed dollar risk buys fewer of them. Take a hypothetical stock that routinely swings $2 in a day. A stop $1.50 from the entry would let you buy $400 / $1.50 = 266 shares, rounded down, and the ordinary swing would probably take it out before the idea had been tested at all; a stop $4 away buys $400 / $4 = 100 shares and survives the noise. That cost is the first premium you pay. It is paid in size. The lesson on placing a stop where the idea is wrong covers how to find that price.

A cap on everything open

Risk per trade is only half the sum. The other half is how many trades are open at once.

The cap is the part people skip. Five positions at $400 of risk each is $2,000 exposed if every stop is hit, and a sixth trade, however good it looks, has to wait until one of the five closes or its stop moves to breakeven. Adding up that total is simple, and the answer on adding up the risk on all your open trades walks through it, including positions that tend to move together. The number of positions follows from the cap, and the answer on how many swing trades to have open at once looks at the same limit from the other side, starting with the positions and working back to the risk.

Buying power you leave alone

The 30% cash floor is the last layer. It means that margin, if the account has it, stays mostly unused, and that there is money on hand if a position has to be defended, a call has to be met or a better opportunity appears after a bad week. It also means the account never runs so close to its borrowing limit that the broker’s house rules start making decisions. Cash looks idle on a good day. On a bad one, it holds still.

Testing it on a bad morning

Does it hold when everything goes wrong at once? Assume the worst realistic case for this plan: all five positions open past their stops, each by half the stop distance again.

A 7.5% drawdown in one morning hurts. It needs a gain of about 8.1% to repair. It is also survivable, which is the point, and the account still has its $12,000 of cash, its rules and its trader’s nerve roughly intact. Without the cap, the same morning with ten positions open at full risk would cost twice as much, and a margined account could be looking at a call before the opening bell had finished ringing.

The objection: buffers cost returns

They do. In a strong, steady market a trader running six or eight positions at full risk, with the cash deployed and some margin in use, will earn more than the buffered account for as long as the market cooperates. Every layer above is a drag on returns in good conditions.

That cost is the premium on staying in business. An account that is still trading after the bad morning gets to compound; one that lost a third of itself has to earn 50% just to get back, and the buffer is priced far below that repair bill. See the margin of safety entry for the original investing sense of the idea.

Where the buffer runs out

The plan above is sized for gaps of half a stop distance. A stock halted on news and reopened far lower can gap several stop distances at once, and one position that opens at three times its planned loss costs $1,200 on its own, which the cap does not see coming. Correlation shrinks the buffer too. Five positions that move together behave like one large one. The call holds for an account spread across unrelated positions, with stops in place and a cap the trader actually obeys. Concentrate it and the margin of safety is thinner than the plan says.