The Screening Cost of Liquidity

This paper argues that principals optimally force counterparties to borrow at above-market rates to use the form of finance as a screening device, where preserving exposure to outside financing separates types despite the costs, a mechanism that explains the coexistence of early payments and contingent compensation in various financial markets.

The Gist

Imagine you are a wealthy, smart business owner (let's call you The Principal) who wants to hire a team of workers (the Counterparties) to build something valuable. You have plenty of cheap money in your pocket, but your workers are broke and need cash right now to buy materials.

Usually, logic suggests you should just hand over the full amount of cash upfront. "Here's the money, go build it," you'd say. That way, your workers don't have to borrow expensive money from a loan shark, and you get the project done.

But in the real world, big companies and investors (like Venture Capitalists) rarely do this. Instead, they give a small chunk of cash upfront and promise the rest later, but only if the worker hits certain goals.

Why don't they just pay everything upfront?

This paper argues that the way you pay someone isn't just about money; it's a lie detector test.

The Core Idea: The "Pay-As-You-Go" Lie Detector

Think of your workers as having two types:

1. The Stars: They are talented, efficient, and will build a great product.
2. The Stragglers: They are lazy or inefficient and will likely fail.

You don't know who is who before you hire them. This is the problem.

• If you pay 100% upfront: Everyone gets the same check. The Stars and the Stragglers both get paid the same. You can't tell them apart. The Stragglers might take your money and run, or do a bad job. You lose money.
• If you pay 0% upfront and 100% later (based on results): The Stars are confident. They know they can build a great product, so they are willing to wait for their money. The Stragglers, however, are scared. They know they might fail, so they won't accept a deal where they get paid only if they succeed. They might walk away.

The Sweet Spot (The Optimal Contract):
The paper says the smart move is a mix. You give a little cash upfront (to help them buy materials) but keep a big chunk of the payment tied to future performance.

• Why? Because the Stragglers are afraid of the "performance" part. They know they can't deliver, so they will reject the deal. The Stars, knowing they are good, accept the deal.
• The Catch: By not paying them fully upfront, you force them to borrow money from a loan shark (outside lenders) to cover the gap. This is expensive for them. But you want them to pay that cost! It's the price they pay to prove they are a Star.

The Analogy:
Imagine you are hiring a chef.

• Full Pre-payment: You give them $1,000 to buy ingredients. Anyone can take it, even a bad chef.
• Performance Pay: You say, "I'll pay you $1,000 after you cook a perfect meal." A bad chef won't take the risk.
• The Mix: You give them $200 now (so they can buy the basics) but say, "I'll pay the rest $800 only if the meal is perfect."
  • The Bad Chef says, "I can't guarantee a perfect meal, and I can't afford to borrow money to buy the rest of the ingredients. I quit."
  • The Good Chef says, "I'm great. I'll take the $200, borrow the rest, and cook a masterpiece."

The "costly" part (borrowing money) is actually a feature, not a bug. It filters out the bad workers.

The Big Surprise: "Helping" Can Hurt

The paper has a second, even more surprising twist.

Imagine you are a government or a big bank. You see that your suppliers are struggling with high loan interest rates. You decide to subsidize them. You say, "Hey, I'll lower the interest rates for everyone so they can borrow money cheaply!"

You think this is great. Everyone is happier, right?

Wrong.

If you make borrowing cheap for everyone, the "filter" breaks.

1. Because borrowing is cheap, the Bad Chefs can now afford to borrow the money they need to try the "Performance Pay" deal.
2. They take the deal, fail, and you end up paying them (or dealing with the mess).
3. Because the Bad Chefs are now in the pool, you have to change the deal to protect yourself. You have to lower the "performance" requirement or pay more upfront to keep the Good Chefs happy.
4. Result: You end up paying more to the Bad Chefs and getting less value from the Good Chefs.

The "Contagion" Effect:
If you have two suppliers who work together (like a car manufacturer and a tire maker), making credit cheap for the tire maker can actually hurt the car manufacturer.

• The tire maker gets cheap credit, so they accept riskier deals.
• This changes the "selection" of who gets hired.
• Suddenly, the car manufacturer is forced to hire worse tire makers because the "filter" is broken.
• Conclusion: A subsidy that helps everyone can actually make the whole business less valuable. It's like lowering the bar for a sports team; suddenly, you have to play with weaker players, and the team loses.

The "Sufficient Statistics" (The Simple Rule)

The author says you don't need to be a genius mathematician to figure this out. You only need to know two numbers:

1. How expensive is it for your partner to borrow money? (If it's very expensive, you should give them more cash upfront).
2. How easy is it to measure their success? (If it's very easy to tell if they did a good job, you can rely more on "performance pay" and give less upfront cash).

If you know these two things, you can figure out the perfect mix of "Cash Now" vs. "Cash Later" without needing to know exactly how talented your partner is.

Real-World Examples

• Trade Credit: When a big store (like Walmart) buys from a small supplier, they often pay a little bit early but hold back a chunk until the goods are verified. This isn't just about cash flow; it's a way to make sure the supplier is serious and capable.
• Venture Capital: Investors don't give a startup all the money at once. They give a little, then check milestones (did you build the app? did you get users?). This forces the "bad" startups to drop out early because they can't handle the pressure of the "performance" check.
• Internal Company Budgets: A big company might give a division some budget but tie the rest to performance. If they gave the division all the money upfront, the division might slack off. Tying it to results keeps them honest.

Summary

1. Money is a Filter: The way you structure a payment (cash now vs. cash later) is a tool to separate the "good" partners from the "bad" ones.
2. Don't Pay Everything Upfront: Even if you have cheap money, you should leave some of the bill for your partner to pay. This "pain" of borrowing is what makes the bad partners quit.
3. Beware of Free Money: If you make borrowing too easy for everyone, you might accidentally let the "bad" partners into the game, which can lower the value of your entire business.

In short: Sometimes, making things harder for your partner (by not paying them fully upfront) is actually the best way to ensure you get a great partner.

Technical Summary

1. Problem Statement

The paper addresses a central puzzle in financial contracting: Why do principals with access to cheap capital (e.g., large manufacturers, venture capitalists, corporate headquarters) not simply pre-finance their constrained counterparties (suppliers, startups, divisions) to eliminate expensive external borrowing costs?

Standard intuition suggests that a principal with lower cost of capital should fully advance the working capital requirement ($K$) to the counterparty, thereby eliminating the counterparty's need for costly outside finance and capturing the full surplus. However, empirical evidence shows that mixed payment structures—combining early advances with deferred, performance-contingent transfers—are ubiquitous.

The paper investigates the optimal financial contract between a principal and a counterparty where finance serves two distinct, competing functions:

1. Liquidity Provision: Reducing the counterparty's need for costly external borrowing.
2. Screening: Separating high-quality counterparties from low-quality ones based on private information (adverse selection).

2. Methodology and Model Setup

The Environment:

• Agents: A risk-neutral Principal ($P$) and a risk-neutral Counterparty ($C$).
• Private Information: The counterparty has a privately known type $\theta \in [\underline{\theta}, \bar{\theta}]$ (representing quality, cost, or ability), drawn from a distribution $F$.
• Timing:
  • $t=0$: Principal offers a contract $(a, T(\cdot))$. $a$ is an advance payment; $T(x)$ is a transfer contingent on a verifiable signal $x$.
  • $t=0$ (Post-contract): Counterparty borrows the residual working capital $\ell = K - a$ from outside lenders at a cost $\Phi(\ell; R)$, where $R$ indexes the tightness of external credit.
  • $t=2$: Production occurs, signal $x$ is realized, and $T(x)$ is paid.
• Key Frictions:
  1. Costly External Finance: $\Phi_\ell > 0$ and $\Phi_{\ell\ell} \geq 0$.
  2. Non-Pledgeability (Assumption [NP]): The contingent transfer $T(x)$ cannot be pledged to outside lenders at $t=0$. Only the advance $a$ relaxes the borrowing constraint. This is the critical friction distinguishing the model from standard screening problems.
  3. Limited Liability: $T(x) \geq 0$.
  4. Informative Signal: $x$ satisfies the Monotone Likelihood Ratio Property (MLRP), allowing $T(x)$ to screen types.

The Mechanism:
The principal designs a direct mechanism $(a(\theta), b_0(\theta), b_1(\theta))$ where the transfer is affine: $T(x) = b_0 + b_1 x$.

• The advance ($a$) provides liquidity but is uninformative (flat payoff across types).
• The contingent slope ($b_1$) provides screening power (steep payoff across types) but forces the counterparty to borrow externally, incurring financing costs.

3. Key Contributions

1. The Liquidity-Screening Tradeoff: The paper establishes that the form of finance is an endogenous screening device. Because contingent transfers are non-pledgeable, they cannot serve as collateral. Therefore, increasing the advance (liquidity) reduces the room for contingent transfers (screening). The principal optimally leaves the counterparty partially exposed to outside finance to preserve the information content of the contract.
2. Screening Contagion: In a multi-counterparty setting with complementarity, the paper identifies a novel "screening contagion" effect. Cheaper financing for one relationship can reduce the principal's total value. This occurs because cheaper finance shifts a contract toward more contingency, which expands the set of implemented types for the complementary relationship, thereby increasing information rents paid to marginal types.
3. Sufficient Statistics: The optimal contract is characterized by two sufficient statistics: the marginal external financing cost and the marginal screening return. The principal does not need to know the full type distribution to determine the optimal advance share.
4. Policy Implications: The paper demonstrates that uniform subsidies that cheapen finance for all counterparties can reduce the value of every relationship in a firm's portfolio due to informational externalities.

4. Main Results

A. Optimal Bilateral Contract

• Interior Solution: The optimal advance $a^*$ is strictly interior ($0 < a^* < K$). The principal deliberately retains outside-finance exposure ($\ell^* > 0$) even though she has cheaper funds.
• First-Order Condition: The optimum equates the marginal liquidity benefit of an additional dollar of advance with the marginal information cost of the resulting flattening of the contract:
  $$\underbrace{\Phi_\ell(K - a^*; R)}_{\text{Marginal Liquidity Benefit}} = \underbrace{\frac{\partial}{\partial a} E\left[ b_1^* \mu'(\theta) \frac{1-F(\theta)}{f(\theta)} \mid q=1 \right]}_{\text{Marginal Information Cost}}$$
• Dominance of Mixed Contracts: A mixed contract (combining advance and contingency) strictly dominates both pure advance (full pre-financing) and pure contingent contracts.
  • Pure Advance: Eliminates financing costs but destroys screening (pooling).
  • Pure Contingent: Maximizes screening but imposes maximal financing costs.
  • Mixed: Balances the two, exploiting the comparative advantage of each instrument.

B. Comparative Statics

• Credit Tightness ($R$): As outside finance becomes more expensive (higher $R$), the optimal advance share increases. The principal substitutes toward liquidity to save on financing costs, accepting lower screening power.
• Signal Informativeness: As the signal becomes more informative, the optimal contingent slope increases, and the advance share decreases.
• Welfare Impact: Tighter financial conditions degrade the quality of screening (flatter contracts), widening the gap between the second-best and first-best allocations.

C. Screening Contagion (Two-Counterparty Extension)

• Mechanism: When relationships are complementary ($\delta > 0$), the implementation of one relationship lowers the threshold for the other.
• The Contagion Effect: If outside finance becomes cheaper for Counterparty $j$, the principal optimally increases the contingency in $j$'s contract. This lowers $j$'s selection threshold, increasing the probability of implementing $j$. Due to complementarity, this lowers the threshold for Counterparty $i$, forcing the principal to implement lower-quality types in $i$. These new marginal types have high information rents.
• Result: If the increase in information rents for $i$ exceeds the direct financing savings for $j$, the principal's total value decreases.
• Policy Paradox: A uniform subsidy reducing $R$ for all counterparties can reduce the value of every relationship simultaneously.
• Organizational Scope: The principal may optimally drop a complementary relationship (single-sourcing) to avoid the screening spillover, even if the relationships are technologically complementary.

5. Quantitative Implications

Using a uniform-quadratic parameterization calibrated to trade credit and venture capital data:

• Advance Shares: The optimal advance ranges from 17% to 54% of working capital, consistent with empirical observations in trade credit (Petersen & Rajan, 1997) and VC staging (Kaplan & Strömberg, 2003).
• Value Loss: The screening-contagion effect can reduce principal value by up to 15% at moderate levels of complementarity.
• Hump-Shaped Value: Principal value can be non-monotonic in financing tightness; moderate tightening can increase value by reducing information rents more than it increases financing costs.

6. Significance and Conclusion

This paper fundamentally reinterprets the role of financial instruments in contracting. It argues that the "form" of finance (advance vs. contingent) is not just a financing decision but a critical screening device.

• Theoretical Contribution: It bridges the gap between adverse selection models (Myerson, 1981) and financial contracting models with costly external finance (Holmström & Tirole, 1997). It introduces non-pledgeability as the specific friction that creates a tradeoff between liquidity and information.
• Practical Significance:
  • Explains why firms do not fully pre-finance suppliers despite having cheap capital.
  • Provides a rationale for "staging" in venture capital and mixed payment terms in supply chains.
  • Offers a cautionary note for policymakers: broad-based credit subsidies may inadvertently degrade screening mechanisms, leading to value destruction in firms with complementary supplier/division networks.

The paper concludes that the optimal financial structure is determined by a simple rule: advance funds until the counterparty's marginal external borrowing rate equals the principal's marginal return from screening. This rule holds regardless of the specific type distribution, relying only on local sufficient statistics.