Sizing Contingency and Reserves: A Field Checklist You Can Use This Week
Almost every troubled project I have reviewed had a contingency line. The trouble was how it got there: a round percentage applied to the whole budget, set once at approval, then drawn down quietly until it was gone two-thirds of the way through. With material prices jumping and lead times still unpredictable, a guessed-at number is more dangerous than usual. Contingency should be sized from your actual risks, kept separate from management reserve, and governed with rules everyone agreed to before the pressure started.
Here is a checklist you can run against your project this week. It does not require a fancy model — just discipline and an honest risk register.
Get the definitions straight first
Contingency reserve covers identified risks — the ‘known-unknowns’ on your risk register. It sits inside the project budget and the project manager controls it.
Management reserve covers the truly unforeseen — ‘unknown-unknowns.’ It sits above the project budget, in the sponsor’s or steering committee’s control, and is released only by formal change.
Padding hidden inside individual estimates is neither. It hides risk, destroys your ability to see where money is going, and should be stripped out before you size anything.
The sizing checklist
Start from the risk register, not a percentage. List your identified risks with a probability and a cost impact. The expected value (probability × impact) summed across the register is your evidence-based floor for contingency. A flat 10% is a guess; this is a calculation you can defend.
Account for correlation and tail risk. Risks rarely fire independently — a supplier delay and a price spike often arrive together. If your project cannot survive several risks landing at once, size above the simple expected value, or run a quick Monte Carlo to see the spread.
Separate the two reserves on paper. Show contingency and management reserve as distinct lines. Blending them lets the project quietly spend the sponsor’s buffer, and removes the early warning a depleting contingency is supposed to give.
Tie each draw to a risk that fired. Contingency is released against a specific realized risk, with a note in the register. ‘We were over, so we used contingency’ is not a draw — it is a cost overrun wearing a disguise.
Re-baseline the reserve as risks close. When a risk passes without occurring, its share of contingency should be reviewed, not silently spent elsewhere. Reserves should shrink as uncertainty resolves, not drift to zero by attrition.
Set the burn-down expectation up front. Plot how contingency should deplete over the schedule. If you are at 30% of the timeline and 70% of contingency is gone, that is a signal to surface now — not a surprise at the next gate.
Agree the governance before you need it. Write down who approves a draw, the threshold above which it escalates, and how management reserve is released. Decide this in calm conditions; you will not negotiate it well mid-crisis.
A note for the current market
When inputs are this volatile, resist the temptation to inflate one giant contingency to cover everything. A single fat number invites the team to treat it as slack and dissolves accountability. Instead, size tightly from your risks, hold a clearly separate management reserve for genuine shocks, and revisit both at every gate. Visible, governed reserves keep the conversation about risk where it belongs — in the open, before the money is spent.
Setting up contingency and reserve practices that hold up under scrutiny — and stand up at the project gate — is something XNM's program & project delivery advisory helps clients get right from the start.