Among the annexes of a file placed before an investment committee, one commonly finds a three-column table labeled downside, base, and upside. Read across the columns, the input line items prove identical and only the values differ — three gradations of the same demand growth rate, three levels of the same input cost, three variants of the same commissioning schedule. The committee reads this table as scenario analysis and approves the file as one whose alternative futures have been examined. What the table actually produces, however, is not three futures but a margin of error drawn around a single one; none of the three columns would require the project to be configured differently from the way it is configured today.
The same pattern appears in the deliberation that precedes the decision. In meetings where a plant investment, a capacity expansion, or a long-dated supply commitment is negotiated, several paths remain live in the early portion of the discussion, yet by the later portion those paths have quietly collapsed into one. The collapse is not enacted by a decision; it occurs because a question goes unasked. By the close of the meeting, what is being debated is no longer which future to prepare for, but at what pace and through which financing structure the selected future will be built.
The mechanism beneath this narrowing — scenario-planning failure, meaning commitment to a single path without constructing multiple internally coherent futures — is not a form of carelessness. Institutional decision processes are engineered to operate around one agreed narrative: budgets are built on a single revenue curve, incentive systems are tied to a single target set, credit models rest on a single cash flow projection, and the organizational chart reflects the accountability split implied by that one narrative. Keeping several futures simultaneously alive imposes an additional cost at every layer of this architecture, which means the collapse into one narrative is rewarded by the system's own operating logic.
A second layer of the mechanism is that constructing scenarios is cognitively demanding work. Building a coherent alternative future requires more than flexing a variable; it requires tracing what else must have changed in the world where that variable moved. In a future where input prices are structurally elevated, competitors' capacity decisions, customers' substitution behavior, the regulator's appetite for price intervention, and the financing market's risk tolerance are all different as well, and following that chain without dropping its end is considerably more laborious than appending a column to a table. The return on that labor remains invisible in the near term, because a well-constructed alternative scenario only reveals its value once that scenario begins to materialize.
The third layer concerns a misdefinition of what scenario work is institutionally for. Scenarios are constructed not to know which future will occur, but to know in advance which future demands which decision. Where that distinction is not drawn, scenario work degenerates into a forecasting contest, and in a forecasting contest the column that appears most plausible wins while the remainder is archived. The plausibility of the winning column typically derives not from its accuracy but from its offering the least resistance to the plan the organization already holds.
The institutional cost of this tendency does not appear as a single line in the income statement; it accumulates in the degree of irreversibility written into contract documents. An organization committed to one path will typically sign supply agreements carrying minimum offtake obligations, EPC contracts calibrated to a single capacity envelope, liquidated damages tied to a single commissioning date, and covenant packages drafted around a single cash flow profile. Each of these documents is defensible standing alone; taken together, they raise the cost of migrating to an alternative future to a magnitude comparable with the project's own equity.
A second cost category surfaces in working capital and in the organization itself. An inventory policy sized to one scenario, a field team hired against one scenario, and a supplier pool assembled for one scenario must all be resized once conditions shift, and the cost of resizing is rarely proportionate, since downward adjustments generate severance, contract termination, and the loss of institutional memory. On the diligence table this cost tends to leave its trace in the staff turnover ratio of the preceding three years and in the frequency of supplier substitution, quietly compressing the multiple the buy side is prepared to apply.
The third cost is that commitment itself halts the production of information. Once a single path is selected, internal reporting is calibrated to measure progress along that path; signals pointing toward an alternative future — permitting timelines accelerating unexpectedly in a particular region, payment terms shortening on a particular input line, order sizes contracting within a particular customer segment — never enter the organization's decision space because no report carries a line for them. This is not an oversight but a design consequence of the measurement system: what is not measured is not managed, and a single-scenario organization measures only its own scenario.
What neutralizes this tendency is not a broader imagination on the part of the decision-maker but the architecture of the decision itself, and that architecture separates into four components. The first is an irreversibility inventory: a single table recording, for every commitment attached to the project, the date through which and the price at which it can still be unwound, presented to the committee as a standalone document ahead of FID. The second is trigger definition: for each alternative scenario, the three earliest observations that would indicate its materialization, written down in advance together with the source from which each will be measured. The third is an option budget: the price of holding the alternative path open — an additional land option, qualification of a second supplier, phased rather than single-block capacity design — carried as an explicit budget line. The fourth is distributed scenario ownership: each scenario assigned a named owner inside the organization, whose mandate is defined as keeping the scenario live rather than advocating for it.
Across the capital-intensive projects it manages, BEIREK operates this architecture through decision-record discipline. The record opens at the moment of proposal rather than the moment of approval; when an investment option enters the agenda, the alternative paths live on that day, the distinct contract structure each would require, and the date on which each closes are entered into the record, and the record is not closed for the life of the project. The irreversibility inventory is maintained as a separate document from the term sheet stage onward, updated by every executed document, so that the committee can see on a single page how much commitment the organization has assumed and which doors remain open at any given moment.
The second line of intervention is to tie the review rhythm to thresholds rather than to the calendar. A monthly progress meeting measures advancement along the selected path and is, in that form, blind to alternative scenarios; projects therefore carry an additional agenda item that opens automatically whenever any of the predefined trigger observations occurs. The subject of that session is not whether the project should be halted, but which decisions remain reversible and through what date; its output is not a recommendation but an updated inventory. For the sponsor, this mechanism means early warning; for the senior lender, the prevention of covenant breach; for the developer, knowing the duration of a negotiating position.
The quality of an investment is measured not by whether the selected future proves correct, but by how much the organization can recover when it does not. That measure is fixed on the day the file is approved and no subsequent effort widens it; the operative question, accordingly, is not which scenario will materialize but how many of the documents signed today constitute a price already paid for the scenario that will not.