Give a sponsor one budget and one quarter, and offer two paths.
Path A is the one almost everyone takes. Rebuild the financial model. Add the scenario engine, the sensitivity tabs, the monthly build, the working-capital detail nobody asked for. Three months later the model is genuinely better — more careful, more granular, more defensible on its own terms.
Path B costs less and is duller. Commission an independent engineer to certify the technical numbers already in the model. Get the auditor to issue on the prior year. Convert the two largest customer understandings into signed contracts, conditions intact.
At the end of the quarter, Path A’s sponsor has a superb document that a credit committee will treat exactly as it treated the old one. Path B’s sponsor has moved the same underlying business up two rungs.
Neither sponsor changed the business. One of them changed who was speaking for it.
The remedy that cannot work
The market treats investment readiness as a quality problem. Something in the pack is not good enough, so the remedy is always more work on the pack: better model, deeper market study, tighter deck. Sponsors spend real money and real quarters this way, and the improvement is real — it simply is not the improvement that counts.
The grading system capital allocators actually run does not read quality. It reads provenance. It asks, of every number that matters, a single question: who is standing behind this, and what happens to them if it is wrong? A document you produced yourself answers that question with your own name, which is the answer the reader already had before opening it.
This is not a Nigerian quirk or a lender’s fussiness. It is written into the disclosure architecture of the world’s capital markets, in three places that have nothing to do with each other.
The Evidence Ladder has four rungs, and only one way to climb
Every claim in a capital raise sits on one of four rungs.
| Rung | What sits on it | Who is answerable |
|---|---|---|
| 1 | Audited, certified, regulator-issued audit opinions; licences and permits; reserve or resource reports; independent engineer certification | A named third party with something to lose |
| 2 | Contracted, signed, externally verified executed offtake, EPC, O&M, lease, concession; LCs and guarantees; verified title and counterparty credit | A counterparty, bound |
| 3 | Management-prepared management accounts; the financial model; internal projections; LOIs, MoUs, term sheets in negotiation | You |
| 4 | Asserted “agreed in principle”; “in discussions with”; verbal undertakings; market-position claims | Nobody |
The only move that changes a rung is a change in who is answerable for the statement.
The author sets the rung, not the polish
Three disclosure regimes, built by different people in different decades for different markets, encode the same rule.
In mining, the JORC Code requires that a public report on exploration results, mineral resources or ore reserves “must be based on, and fairly reflect, the information and supporting documentation prepared by a Competent Person,” whose name the company must disclose, and whose “prior written consent as to the form and context in which it appears” the report must carry. The Code then defines who qualifies: a professional member of a recognised organisation with “a minimum of five years relevant experience in the style of mineralisation or type of deposit under consideration.”
Canada’s NI 43-101 does the same job in near-identical terms. “A technical report must be prepared by or under the supervision of one or more qualified persons,” and the issuer must file a certificate “dated, signed, and if the signatory has a seal, sealed, of each qualified person responsible for preparing or supervising the preparation.”
Read those two side by side. Neither says the report must be accurate, thorough, or well-presented. Both say a named, qualified, personally consenting individual must stand behind it. The regulator is not grading the document. It is identifying the person who can be pursued.
Accounting reaches the same place by a different road. IFRS 13 sorts valuation inputs into three levels and states the ordering principle plainly: Level 1 inputs are “quoted prices (unadjusted) in active markets for identical assets or liabilities”; Level 2 are inputs “observable for the asset or liability, either directly or indirectly”; and Level 3 inputs are “unobservable inputs for the asset or liability” — which, the standard says, an entity develops “using the best information available in the circumstances, which might include the entity’s own data.”
The entity’s own data sits at the bottom. Not because it is wrong, and not because the entity is suspected of anything. Because nobody outside the entity can check it.
That is the whole mechanism, and it is why Path A cannot work. A management model is rung three however sophisticated it becomes. An independent engineer’s report on the same numbers is rung one. The numbers did not move. The author did.
A claim takes its best support. A package takes its worst.
Here is the distinction that saves sponsors from over-correcting, and it is the one most people get backwards.
Grade a single claim and the strongest artefact governs. If your production figure is supported by both a management schedule and an independent engineer’s certificate, the claim is graded on the certificate. The weaker document does not drag it down; it simply becomes redundant.
Grade a composite — a valuation, a coverage ratio, a whole investment case — and the rule inverts. IFRS 13 states it exactly: a measurement “is categorised in its entirety in the same level of the fair value hierarchy as the lowest level input that is significant to the entire measurement.” One unobservable input that matters puts the whole measurement in Level 3, regardless of how much Level 1 data surrounds it.
Both rules are correct because they grade different objects, and the practical consequence is precise. Adding a second management document to support a claim you have already evidenced buys nothing. Leaving one load-bearing assumption on rung three or four caps the entire case, no matter how strong everything around it is.
Which means the useful question is never “is our pack strong?” It is “which of our claims is both weak and load-bearing?” The list is usually short.
Two corollaries follow, and both routinely surprise sponsors.
Audited accounts do not evidence what they omit. A set of audited financials is rung one for the line items it covers and rung three or four for everything it does not. “We’re audited” is a narrower answer than sponsors think it is, and it does not reach a segment margin the accounts never disaggregated, or a volume the accounts never reported.
Evidence expires. A licence past its term, an expired letter of credit, audited accounts materially out of date — these are not the rung they used to be. A data room assembled two years ago and untouched has quietly slid down the ladder while nobody was watching.
Every weak claim has a named price
The reason this framing is worth adopting is not diagnostic elegance. It is that it converts an open-ended anxiety into a list.
For every claim that is both weak and load-bearing, four things can be written down: the artefact that would lift it, who has to sign that artefact, how long they take, and what they charge. Not “improve the technical case” but “an independent engineer’s certification of the production profile, signed by a firm the lender will accept, at their quoted lead time.” Not “strengthen the commercial evidence” but “convert the two largest offtake understandings into executed contracts, conditions intact.”
Readiness stops being a journey and becomes a shopping list with a cost column. Most sponsors have never seen theirs written down, which is why they experience readiness as a fog rather than as a procurement exercise. It is a procurement exercise.
Written down, the list also exposes something uncomfortable but useful: the cheapest items usually move the most. A signature that takes three weeks frequently outperforms a modelling programme that takes three months, because the signature crosses a rung and the modelling programme does not.
“This is bureaucracy. My business is real.”
It is the right objection, and half of it is correct.
The ladder does not measure whether your business is real. It measures whether a stranger can verify it without taking your word. Those are different things, and plenty of excellent Nigerian companies sit low on the ladder while running perfectly good operations. Nobody is calling you a liar; the system is not built around suspicion of you specifically. It is built around the fact that an allocator deploying other people’s money cannot act on truth it has no way to check, and must price the gap you leave.
The second half of the objection has more force: some rung changes are genuinely expensive. A full audit where none existed, a certification programme, a title perfection exercise — these are not trivial, and a sponsor with limited capital cannot buy every upgrade.
Nor should they. That is why the load-bearing test matters. Rank your claims by whether they are actually constraining the outcome, buy the constraining ones, and leave the rest at rung three with an explanation. A pack that is honest about what it has not evidenced reads far better than one that quietly hopes nobody checks — and the reader, who does this for a living, checks.
Back to the fork
Path A’s sponsor spent a quarter making a rung-three document into a better rung-three document. Path B’s sponsor spent less, produced nothing anyone would call impressive, and moved the case two rungs.
You cannot upgrade a document by improving it. You upgrade it by changing who is answerable for it.
So before you commission another study, take the three claims your entire case rests on, and beside each one write the name of the person or firm who would have to put their signature on it. Then find out what they charge. That list is your readiness plan, and it is almost certainly shorter and cheaper than the one you are currently working through.