1. Home
  2. Use cases
  3. Capital committee

Capital committee · One pooled plan

If every list met on one table, what would we fund?

The committee places every candidate from the four decisions on one envelope and one risk appetite. It shows the pooled plan beside today's separate plans on the same money, the shock that would break it, and the thresholds that would change the answer.

Output: the pooled plan beside the separate plans; the shock that would break it and its estimated return period; the thresholds that would change the recommendation; a decision record with votes, dissent and tripwires.

Who it is for

For the people who own the whole envelope

CFOs, capital and investment committees, boards and the strategy teams that prepare for them. The committee does not take decisions away from budget holders; it shows what their lists look like when they share one envelope and one set of risks.

  • CFO
  • Capital committee
  • Investment committee
  • Board
  • Strategy and planning

Questions it answers

  1. What is the best plan across all four decisions, on the same money?
  2. How much better is it than funding each list separately, and in the bad year?
  3. Which shock would break the plan, and how often should we expect one?
  4. Which threshold, if crossed, should reopen the decision?
  5. What did we decide, who dissented, and what run did we rely on?

What Capibud produces

A committee pack built on one set of scenarios

Each candidate is charged a hurdle by risk class, so a venture ticket and a mandatory safety line are compared fairly. Risk-class premia are a labelled policy setting you can edit.

01

Pooled vs separate

Value after hurdles, bad-year value and capital, for the pooled plan and for today's mode plans on the same money.

02

The moves

The handful of changes that turn today's plans into the pooled one: what to release, what to fund and what each is worth.

03

Risk appetite

Choose how much of the bad year to accept, or require the pooled plan to be no worse than today's in the worst 5% of scenarios.

04

Reverse stress

The smallest joint factor move that breaks the plan, its plausibility as a return period and its nearest historical analogue.

05

Thresholds

How far each factor must move, in standard deviations and native units, before the pooled plan stops beating the alternative.

06

Decision record

Votes, dissent and tripwires, the alternatives on the table and the run relied on. The committee can choose any feasible alternative.

Worked example

A refiner's four decisions on one envelope

Illustrative example — figures invented; not a client or a Capibud demo

The example follows a coastal refiner, invented for illustration, through all four decisions. Pooling 22 candidates (new units, venture tickets, recovery moves and capex groups) on 3,000 shared scenarios gives more value after hurdles and a better bad year than the separate plans, with less capital.

The gain comes from a few moves: release one venture commitment, fund extra crews on two units in construction, and fund the best unfunded reliability capex. The plan's value after hurdles reaches zero only under a joint move of about three standard deviations, led by petrochemical and refining margins.

Read the full example in the coastal refiner case.

Pooled plan vs separate plans Illustrative example · USD
MeasureSeparate plansPooled planChange
Value after hurdles$812M$865M+$53M
Bad-year valueP10: nine scenarios in ten do better$497M$540M+$43M
Capital usedBaseline$45M less−$45M

Illustrative example — figures invented for illustration; not a client or a Capibud demo.

Hurdle premium over WACC, by risk class Editable policy setting
Mandatory · recoverystatutory, HSE, protects a committed programme+0 pts
Sustainingreliability and integrity+1 pt
Brownfield · greenfielddebottleneck · new product or market+2 to +4
Venturenew technology+6 pts

Illustrative settings over an 11.25% cost of capital, in line with survey evidence that firms set hurdles several points above their cost of capital.

How it works

From four lists to one recorded decision

1GatherEvery candidate from the four decisions arrives with its spend profile, value distribution and risk class, on the same scenarios.
2ChargeEach candidate carries a hurdle by risk class over your cost of capital; synergies and exclusions carry across decisions.
3SolveThe whole book is solved for the highest value after hurdles within the envelope and the risk appetite you choose.
4StressHistorical stresses, reverse stress and thresholds are shown beside the recommendation, with alternatives on the same page.
5RecordThe committee votes. Dissent and tripwires are recorded with the run, and the brain watches the tripwires from then on.

What it connects to

The committee is where the four decisions meet

It reads from every decision, hands tripwires to the brain and writes every vote to the record.

Decisions

Four inputs

Options, tickets, recovery moves and capex lines, on shared factors.

All use cases
Product

The committee view

Envelope, appetite, frontier, stress, thresholds and packs in one place.

The capital committee
Brain

Tripwires watched

Thresholds that reopen the decision if live data crosses them.

The brain
Record

Votes and dissent

Kept with the alternatives and the run, replayable to an identical result.

The record

Honest limits

What the committee does not do

Pooling makes trade-offs visible. It does not make them for you.

  • Hurdles are policy, not physics. Risk-class premia drive which candidates win. They are labelled settings for your committee to own and edit.
  • Gains depend on overlap. Where decisions share little money or risk, the pooled plan will look much like today's. The demo gain is modest by design.
  • Return periods are model-based. Reverse-stress plausibility comes from the factor model's distribution, not from history alone.
  • Nothing moves money. "Apply" changes a plan inside Capibud and nothing else. Capibud does not move funds or place orders.

Put your four lists on one table.

A Decision Sprint can pool one portfolio's decisions on your own files in six weeks.

Talk to us →