Why a Point Forecast Quietly Lies to the Board
The problem with a single-path budget is not that the number is wrong — every forecast is wrong — but that it discards the shape of the distribution it was drawn from. When the FP&A team hands the board a revenue line of $84.2 million, that figure is the expected value, or worse, the modal value, of a distribution that might run from $61 million to $103 million. The board sees the point and never the spread. It then makes commitments — headcount, a lease, a debt covenant, a dividend — that are safe at the mean and ruinous in the left tail. The forecast did not merely fail to inform the decision; it actively concealed the variable the decision most needed, which is dispersion, not central tendency.
This is the deeper objection to point forecasting in board finance. A board's job is not to predict the future accurately. It is to commit capital and constraints in a way that survives the range of futures that can actually occur. A point estimate answers the wrong question precisely. A scenario set answers the right question approximately, and approximate answers to the right question dominate.
What Banding Actually Buys You
Scenario planning replaces the single line with a small number of internally coherent worlds — typically a downside, a base, and an upside, each built from a consistent set of driver assumptions rather than from arbitrary haircuts. The mechanism that makes this valuable is not the breadth of the band; it is the coherence within each branch. A real scenario forces you to state that if revenue lands at $61 million, then it did so because a named driver moved — a key account churned, a price war compressed gross margin, a launch slipped two quarters — and that same driver also moves working capital, hiring, and cash conversion in the same direction. The branches couple the line items the way reality couples them. A sensitivity table that flexes revenue while holding cost structure fixed produces numbers that can never happen.
From those coupled branches the board extracts the only outputs that matter for an irreversible commitment:
- Breakpoints — the revenue or margin level at which a covenant trips, the runway crosses a financing deadline, or a contractual minimum bites.
- Decision triggers — the leading indicator whose reading tells you which branch you are on early enough to act, converting a forecast into a monitoring discipline.
- Robustness — which commitments survive every branch (make them now) versus which survive only the base case (stage them, option them, or defer them).
The Trade-Off and the Failure Mode
Banding is not free, and its costs are where it goes wrong in practice. The first failure mode is theater: three columns labeled bull, base, and bear, where the bull is the base plus ten percent and the bear is the base minus ten percent. This is a point forecast wearing a costume. It widens nothing the board did not already fear and narrows nothing the board needs to decide. The second failure mode is the opposite — scenarios so numerous and so finely conditioned that the board cannot hold them, and the exercise collapses back into whichever single path the CEO prefers. The discipline lives in a narrow band: enough branches to bracket the consequential outcomes, few enough to govern. Three or four genuinely distinct worlds, each tied to a different dominant driver, is almost always the right count.
There is also a real cost in accountability. A point forecast can be scored — you hit it or you missed it — and that crispness is why management clings to it and why boards tolerate it. A band cannot be missed in the same way, which tempts management to treat the whole range as pre-excused. The corrective is to score the driver calls, not the aggregate: did the named assumptions in the base branch hold, and when a downside trigger fired, did the pre-agreed action follow?
The Decision Implication
For the board, the practical shift is to stop asking management for the number and start asking for the band, the breakpoints, and the triggers. Approve the base case for resourcing, but condition every irreversible commitment on its survival across the downside branch, and attach to each downside trigger a pre-authorized response so that the board is not convening in a crisis to discover it has no playbook. The point forecast remains useful as a communication device and a coordination anchor — it tells the organization what to plan around. It must never again be the artifact on which the board bets the balance sheet. Confidence is not the same as information, and the board's fiduciary duty runs to the latter.