Three quantified risks, one reserve, and a risk that has happened

Calculation

Risk	P	Impact	EMV
R1	0.25	£180,000	£45,000
R2	0.15	£90,000	£13,500
R3	0.60	£70,000	£42,000
Total			£100,500

The reserve of £100,000 matches the EMV sum almost exactly - which is precisely the trap. The sum of EMVs is not a contingency reserve.

Why the EMV sum is the wrong basis

1. It is an expected value, not a confidence level. It is the average outcome across many simulated futures. In a single project you do not get the average; you get one outcome. A reserve equal to the EMV sum gives roughly a 50% chance of overrunning, which is exactly what the simulation says (P50 = £95,000).
2. The sponsor's appetite is 20%, i.e. P80. The simulation says P80 = £210,000. The reserve is therefore £110,000 short of the level the sponsor has said they require.
3. Correlation is ignored. R1 and R3 both drive rework through the same small technical team. If both land in the same quarter, the combined impact exceeds the sum and the schedule damage is compounded - the simulation will capture this only if the correlation was modelled.
4. It ignores the shape of impact. EMV treats a 25% chance of £180,000 as equivalent to a certain £45,000. For a fixed regulatory date, an eight-week delay is not a linear cost.

Now the position has changed. R3 has occurred. It is no longer a risk; it is an issue with a known £70,000 liability. Leaving it inside the reserve means:

• residual reserve = £100,000 − £70,000 = £30,000, against residual risks R1 and R2 with a combined EMV of £58,500 and a P80 far above £30,000;
• the reserve looks untouched while being 70% consumed.

Recommendation

1. Transfer R3 out of the risk register and into the cost baseline as an approved issue, funded from the reserve through change control - this needs sponsor authorisation, not PM discretion.
2. Re-run the simulation on R1 and R2 only, including correlation, to get a fresh P80.
3. Request that the reserve be topped up to that new P80, and state plainly that the current £100,000 does not meet the sponsor's own stated appetite.
4. Note the governance point: contingency is held for risks that may occur; management reserve is held for unknown unknowns. R3 was a known risk and should have been funded before it happened.

Straight EMV comparison

• Do nothing: 0.25 × £180,000 = £45,000
• Mitigate: £18,000 + (0.10 × £180,000) = £18,000 + £18,000 = £36,000
• Net EMV benefit: £9,000

On EMV alone the case is positive but thin - a £9,000 saving for £18,000 of committed cash. Spend it only if the qualitative factors support it, and here they do.

Why the EMV understates the case

1. The impact is non-linear. £180,000 is £60,000 of rework plus eight weeks of £15,000 damages. An eight-week delay to a fixed regulatory date may not be recoverable at all - the damages figure is the contract's estimate of the loss, not the loss. Missing a regulatory date can stop the business process entirely.
2. Tail risk. This is the single largest risk on the register and it sits outside the remaining reserve. Reducing probability from 25% to 10% halves the frequency of the worst outcome, which is exactly what a reserve too small to absorb a £180,000 hit needs.
3. It is the only lever on R1. There is no second source and no workaround; the mitigation is the only action available short of accepting the risk in full.

Check before signing

• Is the cost recoverable? If the contract requires the supplier to achieve certification at their own cost, the £18,000 may be a supplier obligation or recoverable from them. Read the contract before spending project money.
• Does it actually work? Ask the supplier for evidence that the support contract moves the probability that far. A 10% residual is a claim, not a fact.
• Fee structure. Make part of the £18,000 contingent on certification being achieved, so the supplier shares the risk rather than simply being paid to try.
• Authority. £18,000 from a £30,000 residual reserve needs sponsor approval under most delegated limits; check the governance before committing.

Recommendation: proceed, subject to the contract check and sponsor approval, and record the expected £9,000 EMV benefit and the qualitative justification in the risk register entry rather than on a calculation alone.

Resist it. The amount is not the issue; the control failure is.

Why informal absorption is wrong

1. It conceals the risk history. R3 was a registered risk with 60% probability and a £70,000 impact. If it disappears into a forecast without trace, the register shows a reserve of £100,000 when the true residual is £30,000. Every subsequent risk decision is then made on false data.
2. It breaks the linkage between the register and the budget. Contingency exists so that when a known risk materialises the money is available, visibly and quickly. Absorbing it in forecast quietly converts contingency into an unexplained overspend.
3. It removes the decision from the person who owns it. Drawdown of contingency is a sponsor decision under the delegated authority structure. A forecast adjustment moves that decision to whoever prepares the spreadsheet.
4. It destroys the audit trail for the next gate. The board will ask why the budget moved; 'it was absorbed' is not an answer that survives a stage-gate review or an audit.

What to do instead

• Raise a formal issue record with the confirmed root cause (data profiling results, date, owner).
• Raise a change request to draw £70,000 from the contingency reserve into the cost baseline, with the impact on the remaining reserve stated explicitly: £100,000 → £30,000 against residual risk EMV of £58,500.
• Take it to the sponsor, since the reserve is sponsor-approved, and present it alongside the request to top up.
• Keep the forensic point visible: the risk was known and quantified for nine months. The lesson is about the adequacy of the original reserve, and it will be lost if the money is absorbed quietly.
• Update the risk register (close R3, re-score R1 and R2 in the light of the reduced reserve) and the cost forecast in the same cycle, so the numbers reconcile.

Concede the legitimate point: the finance manager is right that the money must appear in the forecast quickly. Speed and informality are not the same requirement - a change request can be raised and approved within the week, and the forecast updated in the same cycle. Do it fast, but do it on the record.