A downside is not a probability

A deterministic downside answers what the specified schedule does under a specified change. It does not answer how likely that change is. The distinction is central to honest financing analysis. A severe case can expose a weakness, identify a funding need or test an intervention without having a measured probability. Calling the case probable requires a separate evidence basis. The committee should retain separate columns for the event description, plausibility reasoning, numerical assumptions and any supported likelihood. This prevents a transparent stress from acquiring empirical authority through a familiar label.

A base case also has an epistemic status. It can represent an expected operating path, a planning convention, a lender's selected funding basis or a sponsor's preferred proposal. These uses should not be interchangeable. A case chosen for financing prudence may deliberately differ from an unbiased expectation. A case used for operating planning may contain stretch objectives. The author should state what the case is intended to do and what evidence supports its material inputs. A label such as realistic does not define a forecast method or establish that adverse conditions have been adequately considered.

Scenario plausibility comes from mechanism coherence and relevant evidence, not merely from numerical severity. A delayed ramp can be plausible if customer or operating readiness depends on unresolved milestones. A sustained margin pressure can be plausible if required input costs lack pass-through. A combined case can be plausible if a common event affects several parts of the system. The analyst should explain that chain. Combining the worst independent value of every line can create a severe hypothetical schedule while offering little insight into an event that could actually occur. Its usefulness should be described accordingly.

Empirical likelihood requires an appropriate population, measurement and inference. Historical outcomes from different assets, contractual structures or jurisdictions may not describe this project. A small or selected sample can be especially weak for rare failures. The committee should ask what is counted, how comparable it is and what uncertainty remains in the estimate. This paper supplies no empirical failure distribution. Its frozen studies are deliberately authored illustrative schedules. Their calculated outputs can be reproduced exactly, while the future likelihood of the corresponding project events remains outside the evidence claimed here.

Sensitivity is useful for identifying decision dependence. Moving a discount assumption while holding cash constant can show whether valuation is sensitive to that chosen input. Changing an operating cash schedule can show whether coverage or value has little margin. Neither exercise automatically yields a probability. The analyst should distinguish an input range chosen for exploration from a range justified by evidence. A committee can use the exploration to prioritise diligence even when the probability is unknown. The conclusion is that the decision depends on the unresolved input, not that the explored value has been measured as likely.

The public app separates a rate sweep from its mechanical operating downside. The rate sweep holds the cash schedule constant and changes the valuation rate. The operating downside applies a specified rule to future cash and supplied cash available while retaining service inputs. These tests answer different questions. The first examines valuation dependence on a preference or rate assumption; the second examines a cash-pattern change. A result from either should retain its method label. Describing a lower valuation caused by a higher rate as an operating loss would confuse the event and the analytical perimeter.

The combined-pressure fixture changes the initial outflow, early ramp, mature operating cash and supplied service together. It is a deliberate stress study, not a claim that those changes share a measured correlation. The inputs are published so the reader can challenge whether the combination is useful for a particular decision. The case demonstrates how several adverse assumptions can interact in an otherwise simple cash system. It does not establish an actual construction overrun, customer delay or loan term. Its role is to identify what a committee would need to fund or redesign if a similarly shaped path were considered credible.

Correlation should not be inferred from scenario grouping. Putting several changes in the same fixture states that they are being tested jointly. It does not estimate statistical dependence among them. A probability model requires distributions, dependence assumptions and evidence or a transparent basis for expert judgment. A simulation can propagate those assumptions but cannot create the missing basis. The committee should know whether a combined result is an intentionally severe test, a coherent causal event or a probability-weighted estimate. These categories can all be useful, but they justify different conclusions.

Reverse stress can help when the committee wants to understand the failure boundary. Rather than choosing a broad downside and observing the result, the team can identify which conditions would cause the specified obligation or decision test to fail, then investigate whether those conditions are credible. This is a decision method, not a feature calculated by the public app. It should retain the accepted cash definition and response mechanism. A failure boundary can guide evidence collection and monitoring even when no reliable probability exists. It should not be described as a measured tolerance without a validated calculation and input basis.

Scenario ownership improves accountability. The operator may be best placed to describe an outage path; commercial teams can explain collection or demand mechanisms; treasury specialists can define rate and currency conventions; technical advisers can examine completion and replacement. The financial analyst assembles a coherent cash path but should not silently replace specialist judgment with arbitrary multipliers. The committee should identify who accepted each mechanism and where disagreements remain. This makes a scenario reviewable and prevents the authority of a polished model from hiding the absence of responsible input owners.

Decision language should match the evidence. A result can support the statement that a supplied schedule produces a specified valuation or coverage value. It can support the inference that a dated cash gap needs a funded response within that illustration. It cannot by itself support a claim that a real project will fail, that financing will be available or that a protection is legally effective. The committee package should state the strongest supported conclusion and the conditions under which it would change. This restraint makes the analysis more useful because the reader can distinguish calculation from judgment.

Scenario review should preserve adverse evidence instead of negotiating it away. A sponsor may reasonably challenge assumptions that are inconsistent with the actual service or contract. The response should be to correct the mechanism and show the revised result. A challenge based solely on discomfort with the output should not justify replacing an input. The record should state the reason for a change, the supporting evidence and the decision effect. This allows a committee to understand whether a downside became less severe because knowledge improved or because the proposed tolerance was reduced without explanation.

Monitoring should convert scenarios into observable conditions where possible. A delayed-ramp scenario can identify readiness milestones; a collection scenario can identify invoice acceptance and overdue cash; a cost scenario can identify quantities, prices and required work. The scenario assumptions should not remain detached from operations after approval. When a leading condition moves, the team can revisit the relevant cash path before the service date becomes threatened. This preserves the original analytical purpose: identify an event, understand its cash consequence and make a decision while response options remain available.

The practical protocol is to publish the case purpose, changed inputs, held constants, mechanism, evidence status, decision consequence and prohibited inferences. A reviewer can then reproduce the output and assess whether the assumed path is relevant. No single case label should carry all of that meaning. The committee should accept uncertainty explicitly where it cannot be measured, choose a funded response where warranted and refuse unsupported claims of probability. A downside gains authority through traceable assumptions and coherent reasoning, rather than through an invented likelihood.