Why the institutions that “suddenly” fail were dead long before — and solvent the whole time. The Great Homecoming research programme · June 2026
Every organisation runs on two clocks. The solvency clock counts the stores: cash, credit, brand, legitimacy, infrastructure. The coherence clock counts whether the thing still works as one purposeful system: whether bad news travels, whether stated purpose and actual behaviour still match, whether the bonds that hold people to the work are being renewed or merely consumed. The clocks can diverge — and when they do, the solvency clock always reads later. Stored surplus keeps a hollow organisation looking alive: the bigger the store, the longer the masquerade, which is why empires take centuries to fall and startups take quarters. Financial-distress models read the solvency clock superbly; almost nothing reads the coherence clock. The gap between the clocks — not either clock alone — is the diagnostic, and it is observable years before the balance sheet notices.
When a large institution fails, the story is always told the same way: nobody saw it coming; the numbers looked fine until the end. Both halves of that sentence are usually true — and that is precisely the problem. The numbers did look fine, because the numbers were reading the wrong clock.
Ask instead when the organisation last functioned as a coherent thing — when bad news last travelled upward intact, when the stated mission last described actual behaviour, when people last stayed for the work rather than the exit package — and the date moves back by years, sometimes a decade. The death certificate and the death did not happen in the same period. What filled the gap was money.
Stored surplus — capital, brand equity, political legitimacy, accumulated infrastructure, the loyalty of customers who haven’t re-decided lately — has a property that makes diagnosis treacherous: it keeps paying out after the engine that earned it has stopped.
An organisation that has lost its coherence does not stop. It coasts. Reserves cover the losses; reputation covers the reserves; the habits of counterparties cover the reputation. From outside — and usually from inside too — everything reads as health. We call the coasting distance the runway, and we state its law as a testable claim: time from coherence-breach to overt failure scales with the size of the stored surplus. That single claim explains a pattern the collapse literature treats as separate puzzles — why imperial declines run for centuries while a startup’s takes two quarters, and why the richest institutions produce the most spectacular “sudden” failures: the deeper the store, the longer the mask holds, the bigger the gap between the clocks when it breaks.
And the claim can die: if runway length turns out to be uncorrelated with stored surplus across cases, the mechanism is wrong. That is a falsifier we accept and have built into our test designs.
The runway mechanism has a darker corollary: surplus does not merely mask the rot — it can deepen it. Resources buy the silence of those who would have objected. They paper over the frictions that would have forced a correction. They let leadership defer every reckoning one more quarter, and each deferral makes the eventual one larger. In systems language: the buffer that absorbs the error signal also destroys the information in it.
This is why “but we’re profitable” is not the rebuttal it sounds like. Profitability is a reading of the solvency clock. It says the stores are still paying out. It says nothing about whether the engine still runs — and in the masked regime it is exactly the healthy-looking reading you would expect.
The standard instruments of failure prediction — Altman-style scores and their descendants — are good at what they do. What they do is read the solvency clock: ratios of the balance sheet, which is an inventory of the stores. They detect the end of the runway. They cannot, even in principle, detect the start of it — the coherence breach that happened while every ratio was green — because the breach doesn’t live in the balance sheet. Insolvency is a lagging proxy for a death that occurred on the other clock.
This is not a criticism of those models; it is a statement of the open lane beside them. (Whether our instrument adds measurable predictive lift over those baselines is an empirical question we are testing, not asserting — out-of-sample, with the scoring rules fixed in advance.)
Nothing exotic — the variables are observable today, in any organisation, without waiting for a crisis:
whether error signals transit — does bad news arrive upward intact, slowed, softened, or not at all; whether the say-do gap is opening — the distance between declared purpose and observed allocation of money, attention and promotion; whether bonds are renewed or consumed — are people, partners and customers re-choosing the institution, or staying by inertia and exit cost; and where effort evaporates — units of work that produce dashboard movement but no outcome movement.
None of these requires inside access to read approximately, and all of them sharpen with it. Read together, they time-stamp the coherence clock.
Neither clock alone tells you much: rich-and-coherent is simply healthy; poor-and-coherent is early or unlucky; poor-and-incoherent is visibly dying and needs no instrument. The reading that matters is solvency fine, coherence falling — the masked quadrant. The width of that gap is one of the most decision-relevant numbers we know how to produce for an organisation, for two reasons. It is the earliest available warning: it appears at the breach, not at the end of the runway. And it predicts the shape of the end: the wider the gap when it breaks, the more sudden and total the failure looks from outside — by the time the solvency clock shows distress, there is typically no almost-crisis phase left, because the years that would have held one were spent on the runway, masked.
For a board, an owner, a funder, three questions follow directly. Which clock is our reporting actually reading — and is anything in the institution reading the other one? If our solvency readings are strong, what would tell us whether they are strength or runway? And when did we last verify that bad news transits — not by asking (everyone says yes) but by tracing what happened to the last three pieces of it?
These questions are answerable by measurement. A reading of the coherence clock, taken now and repeated on a cadence, with thresholds agreed in advance, converts “we feel solid” into a statement that can be checked — and converts the masked quadrant from an ambush into a warning. That is what we build, and we make its claims the same way this article does: stated in advance, falsifiable, and scored whichever way they land.
The Great Homecoming is an independent research programme on why systems cohere or fragment. Companion piece: “Why Systems Don’t Die From Shocks.” The instrument described is under live forward test; we describe it that way until its register adjudicates. Contact: Wim Van Laere.