Moral Hazard in Lending: What Borrowers Do After Approval
Moral Hazard in Lending: What Borrowers Do After Approval
Give an owner $40,000 of debt on a $100,000 project and they choose the safe plan. Give them $80,000 and they rationally choose a riskier one worth $11,500 less, because if it fails the lender absorbs most of the loss. Nothing about the owner changed. That is moral hazard: not dishonesty, but incentives the contract created. The tools that contain it are costly, the concept is often overstated, and lenders have it too.
What you'll learn
The problem that starts after approval
Adverse selection is a problem of who applies: the lender cannot tell good borrowers from bad ones. Moral hazard is a problem of what borrowers do after the money arrives: the lender cannot observe or control how it is used, and the borrower's incentives change once someone else is carrying part of the risk.
Jensen and Meckling set out the core mechanism in 1976. A borrower protected by limited liability keeps the upside of a project but shares its downside with the lender. The more of the project is financed with debt, the more of the downside the lender bears, and the more attractive risk becomes to the borrower. Holmström's work on moral hazard and observability formalized the consequence: when effort and choices cannot be observed, contracts must be designed around what can be.
Risk shifting, worked
A business owner has two ways to deploy $100,000. The safe plan succeeds with probability 95% and returns $130,000. The risky plan succeeds with probability 70% and returns $160,000. Either plan returns nothing if it fails. The safe plan is worth more: $123,500 in expectation against $112,000.
| Financing | Owed to lender at 10% | Owner's expected payoff, safe | Owner's expected payoff, risky | Owner chooses |
|---|---|---|---|---|
| $40,000 debt | $44,000 | $81,700 | $81,200 | Safe |
| $80,000 debt | $88,000 | $39,900 | $50,400 | Risky |
This is the point that makes moral hazard different from fraud. The owner in the second row is not dishonest. Given limited liability and $88,000 owed, the risky plan genuinely is better for them, because if it fails the lender absorbs most of the loss. Per our signaling analysis, adverse selection is about hidden types; moral hazard is about hidden actions taken by ordinary people responding to the incentives the contract created.
Where the switch point sits
The owner is indifferent between the plans when the expected payoff after debt is equal. With success probabilities pS and pR and payoffs RS and RR, that happens when the amount owed, D, satisfies pS(RS − D) = pR(RR − D), which gives:
D* = (pSRS − pRRR) / (pS − pR)
Here that is ($123,500 − $112,000) / 0.25 = $46,000. Three features of the formula matter for underwriting.
- The numerator is the value the safe plan adds. When the safe plan is only slightly better, the switch point is low and even modest leverage tips the choice.
- The denominator is the gap in success probability. When the risky plan is only a little less likely to succeed, the switch point rises; when it is much riskier, the switch point falls and the lender is exposed sooner.
- The relevant quantity is the amount owed, not the loan. Interest raises D, so a higher rate pushes a borrower toward the switch point. This is the moral hazard counterpart to the Stiglitz and Weiss argument about adverse selection: past some point, raising the rate makes the borrower riskier.
The third point is the one lenders feel without naming. Pricing a marginal borrower higher to cover risk can create the risk it is pricing for, which is part of why, per our loan pricing analysis, lenders often decline rather than charge more.
The tools lenders use, and what each one targets
| Tool | How it reduces moral hazard |
|---|---|
| Equity requirement or down payment | Keeps the borrower's own money first in line to be lost, below the switch point |
| Covenants | Smith and Warner showed bond covenants restrict exactly the actions that transfer value from lenders. Per our covenant analysis, the same logic governs small business loans |
| Collateral | Raises the borrower's loss on failure, which makes the safe plan relatively more attractive |
| Personal guarantees | Per our personal guarantee analysis, removes limited liability, so the owner bears the downside directly |
| Monitoring | Holmström and Tirole showed intermediaries add value by monitoring borrowers who could not otherwise credibly commit to behaving |
| Short maturities and lines that must renew | Give the lender repeated chances to stop lending if behavior changes |
Every tool in the table has a cost to the borrower, and most fall hardest on borrowers who would have behaved well anyway. A personal guarantee prevents risk shifting and also exposes a careful owner's home to a business failure that was no one's fault. Per our guarantee analysis, those costs accumulate in ways nobody prices.
Moral hazard in consumer credit
Consumer moral hazard is smaller in scale and harder to see, and much of what is called moral hazard in consumer lending is not.
- Strategic default. A borrower who can pay and chooses not to is the clearest case. It exists, and it is far rarer than lenders often assume. Per our borrower option analysis, the threat of a damaged credit file is what keeps it rare.
- Forbearance and relief. The fear is that offering help encourages borrowers to claim hardship. Per our forbearance analysis, the evidence for that effect is weaker than the evidence that lenders over-collect.
- Insurance-like products. Per our credit as insurance analysis, a borrower who knows a payment can be deferred may plan less carefully, which is a genuine but modest effect.
The common error is labelling hardship as moral hazard. A borrower who stops paying after losing a job has not changed behavior because of the loan contract; the circumstances changed. Treating that borrower as a strategic defaulter misreads the cause and, per our correlated failure analysis, usually arrives at exactly the moment the borrower can least respond.
The lender has moral hazard too
Incentives cut both ways. A lender that will not bear the losses on the loans it makes has the same problem as the over-leveraged owner.
- Originate-to-distribute. Keys, Mukherjee, Seru and Vig found evidence that mortgages more easily securitized were screened less carefully, consistent with lenders relaxing effort when they did not keep the risk.
- Volume-based compensation. Per our intermediary analysis, an originator paid on volume rather than performance has the borrower's incentive in reverse.
- Deferred recognition. Per our loss recognition analysis, a lender close to its capital limit has reason to extend loans rather than recognize them, which is risk shifting toward its own creditors.
Risk retention requirements, clawbacks and performance-based pay are the lender-side equivalents of equity requirements and guarantees. They keep the decision-maker exposed to the outcome of the decision.
The case that moral hazard is overstated
The concept is frequently invoked to justify withholding help that would cost little and prevent much. Arguments against forbearance, debt relief and insurance often cite moral hazard without measuring it, and in several of the settings where it has been measured carefully, the behavioral response was modest relative to the benefit. The concept is sound; its use as a reflex is not. A lender or policymaker that cites moral hazard should be able to say which hidden action it expects, how large the incentive to take it is, and why the proposed safeguard is cheaper than the harm it prevents.
Testable implications
- Among otherwise similar business borrowers, default rates should rise more steeply with leverage than observable risk alone predicts, reflecting choices made after funding.
- Adding personal guarantees should reduce default more among highly leveraged borrowers than among lightly leveraged ones, since the switch point matters only for the former.
- Loans originated for sale should underperform otherwise similar loans held by the originator, with the gap concentrated in characteristics that are costly to verify.
- Relief programs should show smaller take-up by borrowers able to pay than critics predict, where access requires documentation.
The second implication is the one with direct practical use. Per our small business underwriting analysis, lenders typically require guarantees uniformly; the theory says their value is concentrated where leverage puts borrowers near the switch point, and that is where requiring them is easiest to justify.
Frequently asked questions
The tendency of borrowers to take more risk after funding because the lender bears part of the downside, distinct from who applies.
Above a debt level the risky plan pays the owner more because the lender absorbs failure. Here that switch is at $46,000 owed.
Equity requirements, collateral, personal guarantees, covenants, monitoring and short maturities, each with costs to good borrowers.
Often. Hardship is frequently mislabeled as moral hazard, and the concept is cited more often than it is measured.
Key takeaways
- Moral hazard concerns hidden actions after funding; adverse selection concerns hidden types at application.
- Limited liability lets borrowers keep upside and share downside, so leverage makes risk more attractive.
- At $80,000 of debt the owner prefers a plan worth $11,500 less; the switch happens above $46,000 owed.
- Equity, collateral, guarantees, covenants, monitoring and short maturities each target the incentive.
- Each safeguard costs borrowers who would have behaved well, and guarantees cost the most.
- In consumer credit, hardship is often mislabeled as moral hazard.
- Lenders face moral hazard too: originate-to-distribute, volume pay and deferred loss recognition.
- Citing moral hazard should come with a measured incentive and a safeguard cheaper than the harm.
References
- Jensen, M. C., & Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, 3(4), 305–360.
- Holmström, B. (1979). Moral Hazard and Observability. Bell Journal of Economics, 10(1), 74–91.
- Smith, C. W., & Warner, J. B. (1979). On Financial Contracting: An Analysis of Bond Covenants. Journal of Financial Economics, 7(2), 117–161.
- Holmström, B., & Tirole, J. (1997). Financial Intermediation, Loanable Funds, and the Real Sector. Quarterly Journal of Economics, 112(3), 663–691.
- Keys, B. J., Mukherjee, T., Seru, A., & Vig, V. (2010). Did Securitization Lead to Lax Screening? Evidence from Subprime Loans. Quarterly Journal of Economics, 125(1), 307–362.
This analysis is educational and does not constitute financial or legal advice. All figures are stylized illustrations using assumed project payoffs, probabilities, leverage, and interest rates, and do not describe any particular borrower, lender or loan. The cited studies are summarized for context; consult the original works for their models, data and conclusions.