Bolton–Scharfstein Deep-Pocket Predation Model

The Bolton–Scharfstein deep-pocket predation model is a foundational synthesis of corporate finance, contract theory, and industrial organization in business and economics theory, developed by Patrick Bolton and David S. Scharfstein in their 1990 American Economic Review article, "A Theory of Predation Based on Agency Problems in Financial Contracting." For decades, industrial organization economists debated the deep-pocket (or long-purse) hypothesis—the claim that a financially liquid incumbent firm can use predatory pricing or aggressive market expansion to inflict short-run operating losses on a wealth-constrained entrant, exhausting the entrant's external financing and forcing its exit so the incumbent can subsequently recoup monopoly profits. Whereas the Chicago School critique (McGee, 1958; Telser, 1966) argued that deep-pocket predation is irrational if capital markets are efficient because outside investors will always continue funding a positive-net-present-value entrant regardless of past predatory losses, Bolton and Scharfstein proved that endogenous financial constraints arise from optimal dynamic contracting under asymmetric information and managerial cash-flow diversion. To deter a wealthless entrepreneur from secretly diverting or misreporting first-period profits, an outside investor must make second-period refinancing contingent on realized Period-1 repayments (β_H > β_L). However, this termination threat creates a strategic vulnerability in product-market competition: by preying in Period 1 to depress the probability that the entrant realizes high cash flows, a deep-pocketed incumbent lowers the probability that the entrant's financier renews funding for Period 2, rationalizing predatory behavior in subgame-perfect equilibrium and generating a fundamental trade-off between mitigating financial agency costs and deterring product-market predation.

Classical Deep-Pocket Predation vs. Telser's Capital-Market Critique

In traditional antitrust law and early industrial organization (such as the Standard Oil and American Tobacco cases), courts and economists viewed predatory pricing through the lens of asymmetric financial reserves: a dominant incumbent with large internal cash reserves ("deep pockets" or a "long purse") cuts prices below cost during a price war until a smaller rival with limited liquid capital ("shallow pockets") runs out of cash and exits the industry, after which the incumbent raises prices to the monopoly level.

John S. McGee (1958) and Lester G. Telser (1966) mounted a theoretical challenge to this narrative. Telser pointed out that an entrant's lack of internal cash does not explain why the entrant must exit if its post-predation continuation value is positive:

1. Forward-Looking Capital Markets: Suppose that once a price war ends, duopoly competition yields positive continuation profit π̄₂ > I (the per-period fixed capital requirement). If the entrant suffers a temporary loss in Period 1 due to predation, a rational outside creditor evaluating Period-2 refinancing looks only at future profitability (π̄₂ − I > 0), treating Period-1 losses as a sunk cost.
2. Backward Induction Breakdown: If outside creditors always refinance a viable entrant in Period 2 regardless of Period-1 performance, Period-1 predation fails to induce exit in Period 2, so the incumbent will never prey in Period 1.

Earlier attempts to rescue the long-purse story assumed exogenous credit rationing or incomplete information about cost or demand (Kreps and Wilson, 1982; Milgrom and Roberts, 1982; Fudenberg and Tirole, 1986), leaving unanswered why a rational bank would write a financing contract that commits to terminating a known viable firm simply because it suffered low cash flows in Period 1.

Bolton and Scharfstein's Agency-Contracting Resolution

Bolton and Scharfstein (1990) answered Telser's critique by replacing exogenous capital-market imperfections with an explicit dynamic financial contract under non-verifiable cash flows (parallel to the Hart–Moore incomplete-contracts debt theory):

• When realized operating profits are privately observed by the firm's manager and can be diverted for private consumption, a static one-period loan contract fails: the manager claims that profits were low (π₁ = π_L) and defaults.
• In a two-period relationship (t = 1, 2), the investor controls the right to withhold Period-2 capital I and liquidate the firm at the end of Period 1. By promising a higher probability of Period-2 refinancing (β_H) when the manager repays a high amount R_H in Period 1, and a lower refinancing probability (β_L < β_H) when the manager repays only R_L, the investor induces the manager to turn over Period-1 cash flows so as to preserve the manager's continuation rent in Period 2.
• Thus, the bank commits to terminating the entrant after low Period-1 cash flows (β_L < β_H) not because the bank believes the entrant is inefficient in Period 2, but because ex-post termination after low repayments is required to incentivize truthful cash-flow repayment in Period 1.

Technology, Cash-Flow States, and Moral Hazard

Consider a wealthless firm (Firm B, the entrant) operating over two periods, t = 1, 2, with zero discount rate and risk-neutral parties:

• Investment and Stochastic Profits: At the start of each period t ∈ {1, 2}, the firm requires a fixed outlay I > 0 raised from a competitive investor. Operating profit π_t takes one of two values: low profit π_t = π_L with probability 1 − θ, or high profit π_t = π_H > π_L with probability θ ∈ (0, 1), satisfying π_L < I < π̄(θ) ≡ (1 − θ)π_L + θπ_H.
• Non-Verifiability of Cash Flows (Diversion Moral Hazard): Realized operating profit π_t ∈ {π_L, π_H} is privately observed by the manager and cannot be verified by a court. Because π₂ in the final period (t = 2) is also unverifiable and there is no Period 3 with which to threaten termination, the manager in Period 2 never repays more than the verifiable floor π_L. Conditional on funding the firm in Period 2, the investor collects π_L in Period 2 (incurring a net Period-2 loss of I − π_L > 0), while the manager captures an expected Period-2 continuation rent of:

Incentive Compatibility and Individual Rationality Constraints

By the Revelation Principle, a two-period financial contract specifies, as a function of the manager's Period-1 profit report π̂₁ ∈ {π_L, π_H}, a Period-1 cash transfer R_i (i ∈ {L, H}) and a refinancing probability β_i ∈ [0, 1] (i ∈ {L, H}) that the investor supplies Period-2 capital I and collects R₂ = π_L at the end of Period 2. The contract must satisfy three constraints:

1. Period-1 Cash-Flow Feasibility: R_L ≤ π_L and R_H ≤ π_H.
2. Period-1 Truth-Telling / Incentive Compatibility (IC₁): When Period-1 profit is high (π₁ = π_H), the manager must prefer reporting π̂₁ = π_H (paying R_H and receiving continuation probability β_H) over claiming π̂₁ = π_L (paying R_L, pocketing π_H − R_L, and receiving continuation probability β_L):

The maximum extra cash R_H − R_L that the investor can extract in a good Period-1 state is bounded above by the refinancing spread (β_H − β_L) multiplied by the manager's Period-2 continuation rent θ₂Δπ. If the investor promised unconditional refinancing (β_H = β_L = 1, as assumed by Telser), then R_H ≤ R_L ≤ π_L < I, so the investor would lose money in both periods.

1. Investor's Ex-Ante Break-Even Constraint (IR_I): Across both periods, the investor's expected net payoff Π_I must cover the initial outlay I:

Setting R_L = π_L, R_H = π_L + (β_H − β_L)θ₂Δπ, and β_H^* = 1 while minimizing inefficient liquidation (β_L^* < 1) shows that optimal contracting under agency costs requires β_L < β_H = 1, creating an endogenous shallow pocket whenever Period-1 profit is low (π₁ = π_L).

Now embed this financial contracting game into a duopoly product market where Firm B (the wealth-constrained entrant financed by an outside investor) competes against Firm A (a deep-pocketed incumbent with internal cash reserves).

Effect of Incumbent Predation on Cash-Flow Probabilities

• Period 1 (Predation Decision): In Period 1, the incumbent Firm A chooses between normal accommodation (a₁ = N) and predation (a₁ = P):
  • Under normal competition (a₁ = N), the entrant Firm B achieves high Period-1 profit (π₁ = π_H) with probability θ₁ = θ_N > 0.
  • Under predation (a₁ = P), the incumbent incurs a direct Period-1 profit sacrifice (cost of preying) c > 0, which depresses the entrant's probability of realizing high Period-1 profit from θ_N down to θ_P < θ_N (lowering it by Δθ ≡ θ_N − θ_P > 0).
• Period 2 (Monopoly vs. Duopoly Payoffs):
  • If the entrant's financier refinances Firm B for Period 2, the two firms compete as duopolists in Period 2: the incumbent earns Period-2 duopoly profit Π_A^d, and the entrant's probability of high profit in Period 2 is θ₂ = θ_N.
  • If the entrant's financier terminates Firm B at the end of Period 1, the incumbent becomes an unchallenged monopolist in Period 2 and earns monopoly profit Π_A^m > Π_A^d, capturing the monopoly prize ΔΠ_A ≡ Π_A^m − Π_A^d > 0.

The Incumbent's Incentive to Prey

Under a contract with refinancing probabilities (β_L, β_H), the probability that the entrant is terminated at the end of Period 1 is Pr(Exit | a₁) = 1 − β_L − θ(a₁)(β_H − β_L). Preying in Period 1 (a₁ = P instead of N) increases the probability that the entrant is liquidated by:

Comparing the expected Period-2 monopoly gain from inducing exit against the Period-1 predatory cost c yields the Incumbent's Predation Condition:

Conversely, to deter predation (a₁ = N), the entrant's financial contract must satisfy the No-Predation Constraint:

Observable Contracts and Shallow vs. Deep Pockets

When the entrant's financial contract (β_L, β_H) is observable by the incumbent before Period-1 competition begins, the entrant and its investor face a strategic commitment vs. agency cost trade-off:

1. Steep Termination Sensitivity (β_H − β_L = 1 — "Shallow Pocket" Contract): Setting β_H = 1 and β_L = 0 maximizes the manager's incentive to pay out cash in Period 1 (R_H − R_L = θ₂Δπ), relaxing the investor's break-even constraint IR_I. However, this contract invites predation whenever ΔθΔΠ_A > c.
2. Flat Termination Sensitivity (β_H − β_L = δ̄ < 1 — Synthetic "Deep Pocket" Contract): Raising the refinancing probability after low cash flows to β_L = 1 − δ̄ > 0 reduces the termination spread β_H − β_L to δ̄ = c / [ΔθΔΠ_A], insulating the entrant from predatory exit. However, raising β_L shrinks the maximum repayment R_H in good states (R_H − R_L ≤ δ̄θ₂Δπ) and forces the investor to inject β_L(I − π_L) in low-profit states. Substituting β_H = 1 and β_L = 1 − δ̄ into IR_I shows that a predation-deterring contract is feasible if and only if:

When ΔΠ_A is large or c is small (δ̄ small), Π_I^deter < 0: the investor cannot afford to give the entrant a predation-proof contract because the resulting managerial agency problem prevents breaking even.

Unobservable Contracts, Renegotiation, and Empirical Evidence

• Unobservable Financial Contracts: If the incumbent Firm A cannot observe the clauses (β_L, β_H) of the loan contract, the entrant cannot use β_L = 1 − δ̄ as a strategic commitment device to alter a₁, leaving the entrant exposed to predation whenever Δθ(β_H^* − β_L^*)ΔΠ_A > c.
• Renegotiation-Proofness of Termination: Once π₁ = π_L is realized and the entrant pays R_L = π_L, the manager has zero remaining cash to compensate the investor at the end of Period 1, and in Period 2 the investor can never collect more than π_L < I because Period-2 cash flows are also unverifiable. Thus, liquidation after a genuine low-cash-flow realization is renegotiation-proof.
• Empirical Validation: Judith Chevalier (1995) documented that when supermarket chains undertook leveraged buyouts (LBOs), rival non-LBO chains with deep pockets entered and cut prices to prey on the leveraged stores, while Gompers (1995) and Kaplan and Strömberg (2003) verified that staged venture capital financing uses performance-contingent continuation threats (β_H > β_L) to mitigate agency costs.

How does the Bolton–Scharfstein model defeat Telser's (1966) Chicago School argument that banks will always refinance a viable firm after a price war?

Telser assumed that if a project has positive net present value in Period 2 (π̄₂ > I), the bank can capture enough of that Period-2 surplus to justify investing I in Period 2. Bolton and Scharfstein showed that when cash flows are non-verifiable (subject to managerial diversion), the bank actually loses money on the Period-2 investment viewed in isolation (collecting only π_L < I in Period 2 while the manager captures the Period-2 rent θ₂Δπ). The bank only agrees to fund Period 2 after a high Period-1 repayment (β_H = 1) because that ex-ante promise induced the manager to hand over R_H > I in Period 1. When Period-1 cash flows are low (π₁ = π_L), terminating the firm (β_L = 0) saves the bank from incurring the Period-2 loss I − π_L > 0 and preserves Period-1 truth-telling incentives.

Why is termination after a low Period-1 cash flow renegotiation-proof if the project has positive social NPV in Period 2?

Although continuing in Period 2 generates positive total social surplus π̄₂ − I > 0, that surplus is not transferable from the entrepreneur to the bank at the end of Period 1. Because Period-2 cash flows above π_L cannot be verified by a court, the entrepreneur cannot write a credible promise to share Period-2 profits above π_L with the bank; and because the entrepreneur has zero cash left at the end of Period 1 after suffering a predatory loss (π₁ = π_L and R_L = π_L), the entrepreneur cannot pay an upfront cash side-payment to cover the bank's Period-2 loss I − π_L.

When should an entrant sign a "shallow-pocket" vs. a "deep-pocket" financing contract?

If the incumbent's cost of preying c is small relative to its monopoly gain ΔΠ_A, a shallow-pocket contract (β_H = 1, β_L = 0) invites predation. When the entrant's expected profitability θ_N allows the investor to break even under a flatter termination rule (β_H − β_L ≤ δ̄), the entrant and investor commit to a deep-pocket credit line (β_L = 1 − δ̄ > 0) to deter predation.

1. Bolton, P., & Scharfstein, D. S. (1990). "A Theory of Predation Based on Agency Problems in Financial Contracting." American Economic Review, 80(1), 93–106. https://www.jstor.org/stable/2006736
2. Telser, L. G. (1966). "Cutthroat Competition and the Long Purse." Journal of Law and Economics, 9, 259–277. https://doi.org/10.1086/466627
3. McGee, J. S. (1958). "Predatory Price Cutting: The Standard Oil (N.J.) Case." Journal of Law and Economics, 1, 137–169. https://doi.org/10.1086/466547
4. Chevalier, J. A. (1995). "Capital Structure and Product-Market Competition: Empirical Evidence from the Supermarket Industry." American Economic Review, 85(3), 415–435. https://www.jstor.org/stable/2118181
5. Fudenberg, D., & Tirole, J. (1986). "A 'Signal-Jamming' Theory of Predation." RAND Journal of Economics, 17(3), 366–376. https://doi.org/10.2307/2555717