Bias Is Optional; Capture Is Structural

Rational managers can produce system harm when system welfare is outside the payoff space

The paper argues that the 2008 crisis does not require a story about irrational executives. Lehman’s leaders optimized the incentives they faced while the stability of the financial system was absent from the formal decision calculus.

• Lehman filed for bankruptcy with $639 billion in assets.
• The crisis destroyed about 11 trillion in household wealth, caused 8.8 million U.S. job losses, and required 700 billion in taxpayer-funded bailouts.
• The named managers were Richard Fuld, Christopher O'Meara, Ian Lowitt, and Erin Callan.

The paper accepts bias research but rejects it as the cause of 2008

Malmendier and Tate, Ben-David, Graham, Harvey, Kahneman, and Lovallo identify real cognitive mechanisms. The paper’s claim is narrower: those mechanisms did not cause the shadow-banking Hollow Win.

• CEO overconfidence, CFO miscalibration, inside-view bias, and temporal myopia are treated as real but incomplete explanations.
• The firm-level behavioral literature keeps the firm as the dependent variable.
• System welfare appears, at most, as a control variable rather than as a payoff dimension.

The bilateral game G = A, B leaves the system C out of the choice set

Regulated firms and counterparties bargain inside a four-outcome game. The system is affected by the outcome but is not a strategic participant in the payoff function.

• Party A is the regulated firm; Party B is the direct counterparty; C is the regulatory framework, public fisc, social contract, infrastructure, or environment.
• The Missing System Theorem says W is excluded by construction in G = A, B .
• Only the transformed game G1 = A, B, C makes the Win-Win-Win outcome (1,1,1) visible.

Pre-2008 finance fits the paper’s Hollow Win cell: C=0, A=1, B=1

Banks and counterparties gained while the financial system became more fragile. The paper names this outcome as the canonical signature of Missing System Theorem failure.

• Banks booked record profits, so A=1.
• Institutional investors, pension funds, and insurers received high yields on AAA-rated mortgage-backed securities, so B=1.
• The system degraded through use above 30:1, opaque counterparty risk, and dependence on rising housing prices, so C=0.

When φ exceeds 0.3, higher rationality moves managers toward Hollow Win

The Capture-Without-Bias Theorem states that with system welfare excluded from the payoff space, fully informed expected-utility maximizers converge toward Hollow Win as r approaches 1.

• Capture index φ is the proportion of regulatory oversight endogenous to industry influence.
• Manager rationality r=1 means perfect expected-utility maximization under full information.
• The threshold in the theorem is φ > 0.3.

Capture lowers effective compliance cost and raises β_W

The proof uses a firm payoff that contains revenue and compliance cost, but not system welfare. As capture rises, the firm expands revenue until system costs are shifted outside the private calculus.

• Firm payoff: U_i = Π_i - (1-φ)·K(Π_i).
• First-order condition: K'(Π_i) = 1/(1-φ).
• The derived marginal system welfare loss is β_W = φ/(1-φ).

Debiasing raises harm when it raises r without changing G

The paper’s Debiasing Paradox follows from Axiom 8: interventions that improve managerial optimization do not alter the payoff function. They only make managers solve the system-excluding problem better.

• Corollary 1 applies in industries with φ > 0.5.
• A debiasing intervention that increases r by Δr increases β_W by about [φ/(1-φ)²]·Δr.
• The mechanism is more efficient extraction from C, not worse information.

Disclosure changes information, not incentives, unless it changes payoffs

2 says system-welfare disclosure has zero effect on equilibrium outcomes in a captured bilateral game unless disclosure triggers liability, regulatory action, cost-of-capital effects, or another payoff change.

• Disclosure changes the information set from I = ∅ to I' = W_i(Π_i) for all i .
• The manager still maximizes U_i = Π_i - (1-φ)·K(Π_i).
• The paper’s ESG corollary applies this logic to disclosure mandates without enforcement.

Regulators and firms both prefer W unmeasured when φ is high

The Conflictoring Theorem formalizes why formally adversarial parties can align around information suppression. The regulator avoids reputational damage from acknowledged system loss; the firm avoids system-cost internalization.

• Regulator payoff: U_R = B_R - L_R(φ) - D_R(W).
• Firm payoff: U_A = Π_A - (1-φ)·K(Π_A).
• For φ > 0.3, both payoffs rise with φ and fall when acknowledged W increases.

Decision Accounting works only when Field 17 becomes enforceable

The proposed transformation R changes the payoff function by making SYSTEM_WELFARE a legal decision variable rather than a reportable fact.

• New firm payoff: U' = Π_i - (1-φ)·K(Π_i) - λ·W_i(Π_i).
• 4 requires mandatory Field 17, regulator independence with φ < 0.3, and automatic enforcement.
• In the linear case, λ > φ/(1-φ) is sufficient to preserve C.

Sweden’s post-1992 banking model is the paper’s main empirical reform case

The paper uses the Nordic banking model to show that lowering capture and enforcing system-welfare accounting can reduce β_W without reducing shareholder returns.

• Sweden, 1992-2023: φ fell from 0.8 to 0.2.
• β_W fell from 22.4 to 1.8.
• The paper also reports the comparative range φ = 0.15-0.25 and β_W = 1.8 for Sweden, versus U.S. β_W = 8.3-12.7 with φ = 0.8-0.9.

The theory can fail if game structure is not doing the work

The paper presents the Capture-Without-Bias Theorem as empirically testable, not as a metaphor for bad management.

• A controlled or natural experiment where debiasing reduces system harm in an industry with φ > 0.5 would refute the Debiasing Paradox.
• The SAPM corpus reports 49-52 confirmed domains with zero counterexamples.
• Across 61 domains, the corpus reports β_W values from 0.76 to approximately 51 and a Reform Dividend of roughly $73.8T per year.