Decision Engineering™ · The Decision Architect

DecisionEngineeringTM

The discipline of engineering what deciding actually means.
Issue #004 · Deepak Aggarwal
Previously in this series
Issue #001 — Layer 1 · Purpose showed what happens when institutions deploy systems without encoding their Purpose into what those systems actually do. Coutts. NHS triage. The gap between stated intent and operational reality accumulates silently — until it doesn't.
Issue #002 — Layer 2 · Strategy examined what happens when Strategy drifts from the Purpose it was designed to serve. Credit Suisse. NHS England. Eight years of drift. Nobody asked whether the two were still compatible.
Issue #003 — Layer 3 · Intent examined what happens when a mandate is issued without a boundary. Wells Fargo. Kaiser Permanente. The mandate was real. The outcome was not intended. The gap between them was never closed.
This issue examines Layer 4 — Rules. Where Intent gets codified into formal policy. And where the rules that are written often govern a world that no longer exists.
The institutions referenced in this issue are cited on the basis of publicly documented regulatory findings, official investigations, and other publishd reports. All analysis is educational. Nothing here constitutes legal, regulatory, financial, or investment advice.
00 · This Issue
The rulebook was current. The world had moved on.

Every institution has rules. Compliance frameworks. Risk appetite statements. Operating procedures. Regulatory limits. The rules exist because Intent — however clearly stated — cannot govern every decision made at the execution layer. Rules are the mechanism that codifies Intent into something repeatable, auditable, and enforceable.

The Layer 4 failure is the most counterintuitive in the Decision Integrity Chain™. It is not the failure to have rules. It is not the failure to follow them. The Layer 4 failure is what happens when the rules are followed precisely — and the outcome is catastrophic anyway. Because the rules were written for a world that has since changed. And nobody asked whether the rulebook still governed the institution it was supposed to protect.

Wirecard's auditors followed their engagement rules. BaFin followed its classification rules. Orpea followed its care rules — every staffing ratio, every nutrition budget, every medication threshold. All compliant. All followed. All governing an institution that had already become something the rules were never designed to contain.

The institution heard the rules say yes. What the rules could no longer say was no.

01 · Signals
Two institutions. Two sectors. The rules held. The institutions didn't.
Wirecard AG
Financial Services · Germany
2002 — 2020

Wirecard's Purpose was straightforward: process payments, connect merchants to consumers, operate the infrastructure that made digital commerce function. Its Strategy was to scale that infrastructure globally — acquiring payment licences across jurisdictions, expanding into Asia, positioning as a full-service financial technology institution rather than a processor. By 2018, Wirecard sat on the DAX 30. Its market capitalisation exceeded Deutsche Bank's.

Its founding regulatory classification was as a technology company — not a bank or financial institution. That classification determined who supervised it, at what frequency, and at what intervention threshold. BaFin followed its rules. EY followed its audit engagement rules. Both sets of rules were accurate in 2002. Neither had been updated for the institution Wirecard had become.

€1.9 billion — absent from balance sheet
€24 billion — peak market capitalisation
DAX 30 member — until collapse June 2020
EY — signed off accounts for 10 years
Insolvency filed — within days of disclosure
BaFin — no bank-level supervision applied
Where the rules broke — Wirecard
What the rules assumed (2002)
A payment technology company processes transactions on behalf of banks. It does not hold material client funds. Bank-level supervision is not required. Standard technology audit scope is sufficient.
What the institution became (2018)
Wirecard held billions in third-party merchant trust accounts across multiple jurisdictions. It operated payment licences in Asia. Its balance sheet complexity matched — and exceeded — a mid-tier European bank.
The gap
BaFin classification never updated. EY engagement scope never expanded. €1.9 billion accumulated in accounts the rules did not require anyone to verify directly.

In June 2020, Wirecard announced that €1.9 billion — a quarter of its balance sheet — probably did not exist. The trustee accounts in the Philippines could not be confirmed. The institution filed for insolvency within days. Every annual report had been signed. Every classification rule had been followed. The rules simply governed a Wirecard that had ceased to exist years earlier.

Wirecard's Purpose was real. Its Strategy was documented. Its Intent was codified in rules written for a 2002 payment sector. The institution scaled past every assumption those rules depended on. Nobody broke a rule. The rules broke the institution.
Orpea Group
Healthcare · France / Europe
2012 — 2022

Orpea's Purpose was the care of elderly residents across its network of private nursing homes. Its Strategy was to scale that network aggressively — by 2021, Orpea operated over 1,100 facilities across 23 countries, making it the largest private nursing home operator in Europe. Its Intent was to deliver care that met the standards its licences required.

France's Agence Régionale de Santé — the ARS — set the rules: minimum staffing ratios, nutrition budget floors, medication management thresholds, inspection frequencies. Those rules were written for a private care sector of modest scale, locally inspected, where residents were a knowable population. They were not written for a publicly listed institution managing 1,100 facilities, answerable to shareholders, with procurement centralised across jurisdictions.

1,100+ facilities — 23 countries
Largest private care operator in Europe
€3.8 billion — 2021 revenue
ARS — compliant throughout inspections
French Senate investigation — 2022
Share price — fell 60% within weeks of publication
Where the rules broke — Orpea
What the rules assumed (written for)
A modest-scale, locally-inspected nursing home operator. Staffing ratios and nutrition minimums set for facilities where inspectors knew residents individually. No centralised procurement pressure.
What the institution became
A publicly-listed institution with centralised procurement across 23 countries, shareholder margin pressure, and care delivery standardised at industrial scale. Same rules applied unchanged.
The gap
Residents rationed on incontinence pads. Food budgets cut to improve margin. Medication optimised for procurement cost. All within the rules. The rules had become the architecture of the harm.

In January 2022, investigative journalist Victor Castanet published Les Fossoyeurs — The Gravediggers. Confirmed by a French Senate investigation and parliamentary hearings, the account documented what rule-compliance at scale looked like inside Orpea: residents rationed on incontinence pads, food budgets cut to improve margin, medication management optimised for procurement cost rather than clinical need. Not by accident. By system.

The Senate investigation did not find that Orpea broke the rules. It found that the rules were insufficient for the institution that had been operating under them. The ARS inspection regime had been designed for a different scale, a different ownership model, and a different era. The rules had not moved. Orpea had.

Orpea's Purpose was care. Its Strategy was scale. Its Intent was codified in rules written for a sector that bore no resemblance to the institution following them. The rules permitted everything that followed. That is the Layer 4 failure — not misconduct within the rules, but rules that had become the architecture of the harm.
02 · Diagnosis
Rule-following is not the same as right-following.

Both failures share the same structural signature. Different sectors. Different jurisdictions. Different scales. The same underlying collapse.

In both cases, rules existed. In both cases, those rules were followed. In both cases, the institutions operating under them were audited, inspected, and found compliant. And in both cases, the outcome — €1.9 billion missing from a DAX company's balance sheet; systematic harm to elderly residents across a thousand facilities — was the direct product of institutions operating precisely within the rules they had been given.

The diagnosis is not that the rules were broken. The diagnosis is that the rules had become obsolete — and the governance architecture contained no mechanism to notice.

Rules are written at a point in time, for an institution of a certain scale, operating in a certain environment. Wirecard's classification rules were written for a 2002 payment technology sector. Orpea's care rules were written for a French nursing home sector of modest, locally inspectable scale. Both institutions changed fundamentally — in size, in ownership structure, in the nature of their operations, in the risks they carried. The rules did not change with them. Nobody was required to ask whether they should.

This is the Layer 4 failure in its clearest form. It is not visible at the moment the rules are written — they are accurate then. It is not visible during compliance reviews — the institution is compliant. It only becomes visible when the gap between the world the rules govern and the world the institution actually inhabits becomes large enough to produce an outcome that no rulebook would have sanctioned as an explicit choice.

By the time that gap is visible, it has already done its damage.

A rulebook that is never reviewed is not a governance framework. It is a record of what the institution once decided to prevent.

The deeper problem is structural. Institutions treat rule-compliance as the end of the governance obligation. Once the checklist is green, the question is considered closed. But rule-compliance is not evidence that the rules are adequate — it is only evidence that the institution has met whatever standard was set, at whatever point in time it was set, under whatever assumptions were current then.

In the DIC™, Layer 4 is not a static layer. Rules must be treated as perishable. They carry an expiry condition — not a calendar date, but a set of assumptions about the institution, its scale, its environment, and its risk profile. When those assumptions change materially, the rules that depended on them have already failed, whether or not anyone has noticed.

What makes this the most operationally invisible failure in the chain is that the very mechanism designed to catch it — compliance review — cannot see it. A compliance review asks whether the rules were followed. It has no mandate to ask whether the rules still govern the right institution. That question requires a different review entirely.

03 · Engineering Note
What Layer 4 governance actually requires.
Three mechanisms every institution needs at Layer 4
01
Rule Expiry Conditions
Every rule was written for a specific type of institution. When the institution changes materially — in size, structure, or risk — the rule needs a trigger for review. Not a calendar date. A condition: what would have to be true for this rule to stop working?
Wirecard
A balance sheet threshold above which a technology company classification triggers mandatory supervisory review. Wirecard crossed it years before collapse. No trigger existed.
Orpea
An operator size above which minimum care ratios trigger mandatory reassessment. Orpea exceeded it by 2015. No reassessment was required because no threshold had been set.
02
Minimum vs Adequate — Not the Same Thing
Rules set floors. Institutions optimise to floors. A floor designed for a small institution, applied at scale, permits outcomes it was never designed to allow. Every minimum rule needs a stated assumption about the institution type that makes it adequate.
Wirecard
Audit scope adequate for a payment processor. Inadequate for a €24bn institution holding cross-jurisdictional trust accounts. The minimum never scaled. The institution did.
Orpea
Staffing and nutrition minimums adequate for 50 locally-inspected facilities. At 1,100 facilities under centralised procurement, the same minimums produced systematic harm.
03
Institution-Rule Alignment Review
Compliance review asks: were the rules followed? This review asks a different question: does the institution being governed still resemble the institution the rules were written for? This is a board-level question. Not a compliance function.
Wirecard
Asked formally in 2017: is this institution still appropriately classified as a technology company? That question, placed on a board agenda, changes the supervisory outcome. It was never asked.
Orpea
Asked formally in 2016: do rules designed for locally-inspected care homes remain adequate for a 700-facility publicly-listed operator? That question changes the regulatory outcome. It was never asked.
The Question for your institution
Your institution has rules. When did it last ask whether those rules still fit the institution it has become?

Not: are we compliant? That question is for the compliance team.
This one is for the board.
Decision Integrity Chain™ · Layer 4 of 8
L1Purpose
L2Strategy
L3Intent
L4Rules
L5Judgment
L6Decision
L7Outcome
L8Feedback
Rules
The codification of Intent into formal constraints — compliance frameworks, risk appetite statements, operating procedures, regulatory limits. Rules are legitimate when they govern the institution that actually exists. They become a failure mode when they govern the institution that once existed — and the world has moved on without them.
Issue #005 moves to Layer 5 — Judgment. Where the rules run out and human discretion begins. And where the gap between what the institution permits and what it actually decides becomes visible for the first time.
← Issue #003 · Intent Layer 4 of 8 Issue #005 · Coming