Benjamin Graham's defensive policy in The Intelligent Investor sets a range rather than a target: never less than 25 percent in common stocks and never more than 75 percent, with the balance in high-grade bonds. Most readers remember the ceiling and forget the floor, which reverses the intent, because the floor is the harder of the two to hold and does more work.
The ceiling exists to stop a bull market from doing your asset allocation for you. Leave a 50/50 book alone through a long rise and equities become the majority of it by drift, so the exposure you carry into the eventual fall was chosen by the market rather than by you.
The floor exists to stop a bear market from ending your participation. An investor at 25 percent equity in a crash is unhappy. An investor at zero has converted a drawdown into a permanent loss, because getting back in requires a decision that nobody makes at the bottom. The floor is a commitment device against your own future self, and the fact that it will feel wrong at the moment it binds is the reason it has to be written in advance.
Advertisement
How often does the 25 to 75 band actually bind?
Now the honest finding about how often the band actually binds, which Worked Example 4 computes in full. Start at 50/50 and run five strong equity years, a geometric mean of 28.57 percent against bonds at 7.00 percent. After five years the equity weight is 71.47 percent. The ceiling never binds, and the arithmetic says it would take 5.98 years of that run to reach 75 percent from a 50/50 start. On the other side, it takes 3.78 consecutive years of a 20 percent equity decline against 7 percent bonds to drive a 50/50 book down to the floor.
So a 25 to 75 band is a ruin guard, not a rebalancing rule. It binds after a decade-long one-way market, which is precisely when abandoning a policy feels most reasonable. If you want the discipline to bite in ordinary conditions you need a tighter collar inside it, and the collar is what generates trades.
What does a 45 to 55 collar cost and save?
Worked Example 4 runs a 45 to 55 collar on the same path, rebalancing to 50/50 on breach. Three trades result, all of them sales of equity into strength, and the collar costs 314,776 rupees of terminal wealth over the five bull years, 6.84 percent. Then it runs both books through a three-year crash. The uncollared book draws down 32.21 percent; the collared book draws down 21.38 percent and ends 320,981 rupees ahead. That pair of results is the whole trade-off, quantified: a collar is an insurance premium paid in bull markets and collected in bear ones, and whether it pays over any particular path is unknowable in advance.
Which is why the collar is chosen for its behavioural properties rather than its expected return. It sells into strength and buys into weakness automatically, at moments when doing so by judgement is nearly impossible, and the drawdown it prevents is the drawdown that ends investing careers.