Asymmetric Information: Adverse Selection, Moral Hazard & the Lemons Problem

Across most economies, marketable securities are not the primary source of external business finance — banks are. Every mortgage requires collateral. Most loan contracts include restrictive covenants. The concept that explains all of this is asymmetric information — the single most important idea for understanding why financial institutions exist and why financial markets are structured the way they are.

What Is Asymmetric Information in Finance?

Asymmetric information occurs when one party to a financial transaction has more or better information than the other. A borrower knows more about the true risk and expected return of an investment project than the lender. A company’s managers know more about the firm’s financial health than outside investors.

Adverse selection is a pre-transaction problem. Because lenders cannot perfectly distinguish good borrowers from bad ones, the riskiest borrowers are the most eager to seek loans — and the most likely to be “selected.” This is analogous to the insurance market, where individuals who know they are high-risk are the most eager to buy coverage.

Moral hazard is a post-transaction problem. Once a borrower receives funds, they may engage in riskier activities than the lender anticipated — because the borrower captures the upside while the lender bears the downside. This change in behavior after the transaction is what makes moral hazard so difficult to control.

Economists broadly recognize several responses to adverse selection, including screening (where the uninformed party gathers information, such as credit checks) and signaling (where the informed party voluntarily reveals quality through costly actions). In Mishkin’s analysis of financial structure, the core solutions are private information production, government regulation, financial intermediation, and collateral — each addressing the lemons problem from a different angle.

The Lemons Problem: How Adverse Selection Works

The most famous illustration of adverse selection is George Akerlof’s 1970 “Market for Lemons” model — work that later earned him the 2001 Nobel Prize in Economics. The model demonstrates how information asymmetry can cause markets to unravel.

Akerlof’s Used Car Market

Consider a used car market where sellers know the true quality of their cars but buyers cannot tell good cars (“peaches”) from bad ones (“lemons”). Because buyers cannot observe quality, they are only willing to pay a price reflecting the average quality of cars in the market. But sellers of good cars know their vehicles are worth more than this average price — so they withdraw from the market. With good cars gone, only lemons remain, and buyers adjust their willingness to pay downward. Depending on the distribution of values and reservation prices, the result can range from partial market unraveling — where the market still functions but only for low-quality goods — to complete market collapse.

Application to Financial Markets

The same logic applies to firms raising capital through security issuance (direct finance). When investors cannot distinguish high-quality firms from low-quality ones, they offer an average price for new securities. Managers of genuinely strong firms recognize their securities are undervalued at this price and refuse to issue — they seek alternative financing instead (typically bank loans, where the bank produces private information). Only weaker firms, whose securities are overvalued at the average price, are willing to sell. Rational investors anticipate this and become reluctant to purchase new securities at all.

This is precisely why marketable securities are not the primary source of external business finance. Across most countries and time periods, businesses rely far more heavily on bank loans than on stock or bond issuance for external financing. The lemons problem in security markets is a central explanation for this pattern.

Information asymmetry also manifests in initial public offerings (IPOs), where companies going public for the first time face significant information gaps between insiders and outside investors. IPO underpricing — where shares are offered below their true market value — is one application of how information asymmetry affects primary market pricing, though multiple factors beyond the classic Akerlof mechanism contribute to this phenomenon.

Adverse Selection Example: Calculating the Lemons Discount

To see how adverse selection drives market unraveling, consider a numerical example using Akerlof’s framework. The key distinction is between buyer values (what the car is worth to a buyer who knows its quality) and seller reservation prices (the minimum a seller will accept).

In financial markets, this same dynamic explains why high-quality firms avoid issuing public securities (the “lemons discount” makes their shares too cheap) and instead turn to banks that can privately evaluate their creditworthiness.

Moral Hazard and the Principal-Agent Problem

While adverse selection occurs before a transaction, moral hazard arises after funds have been transferred. The borrower or manager may change their behavior in ways that harm the lender or investor — because the informed party captures gains while the uninformed party bears losses.

The Principal-Agent Problem in Equity

Consider a firm where managers (agents) own only 10% of the company while outside shareholders (principals) own the remaining 90%. If a manager spends $100,000 on a private office renovation, the manager bears only $10,000 of the cost (their 10% ownership share) while capturing the full benefit. Outside shareholders bear $90,000 of the cost with no benefit. This misalignment of incentives — where agents do not bear the full consequences of their decisions — is the principal-agent problem.

Moral Hazard in Debt Markets

Debt contracts face their own form of moral hazard. After receiving a loan, a borrower has an incentive to shift into riskier projects than originally proposed. If the risky project succeeds, the borrower keeps the excess profits above the fixed loan payment. If it fails, the lender absorbs much of the loss. This asymmetric payoff structure encourages excessive risk-taking — known as the asset substitution problem.

Real-World Examples

The 2008 mortgage crisis illustrates originator moral hazard: mortgage lenders who securitized their loans and sold them to investors had reduced incentive to screen borrowers carefully, since the originators passed the default risk to the investors who purchased the mortgage-backed securities. The result was a dramatic deterioration in lending standards.

The Enron and WorldCom scandals demonstrate the principal-agent problem in extreme form: managers concealed billions in losses from shareholders and creditors, exploiting the information gap between insiders and outside investors. Even mandatory SEC disclosure requirements failed to prevent these information failures.

Importantly, when borrower net worth deteriorates — as happens during recessions — asymmetric information problems intensify. Borrowers with less equity at stake have less “skin in the game,” increasing both adverse selection (riskier borrowers dominate the applicant pool) and moral hazard (borrowers are more willing to gamble). This balance-sheet channel is a key mechanism through which economic downturns amplify financial stress, as explored in monetary policy transmission mechanisms.

How Asymmetric Information Shapes Financial Structure

Asymmetric information provides a unified explanation for the observable structure of financial systems worldwide. Mishkin identifies eight basic facts about financial structure — puzzles that seem surprising until information problems are considered:

For a broader overview of how financial intermediaries, markets, and regulation fit together, see our guide to the financial system.

Solutions to Adverse Selection

Financial markets have developed several mechanisms to reduce adverse selection. In Mishkin’s framework, four core solutions address the lemons problem in financial markets:

1. Private information production. Firms like Standard & Poor’s, Moody’s, and Value Line research and sell information about companies, helping investors distinguish good firms from bad. However, the free-rider problem limits this solution: investors who haven’t paid for the information observe the trades of informed investors and copy them. This bidding drives up prices before the original researcher can fully profit, reducing the incentive to produce information in the first place.

2. Government regulation. Governments require firms to disclose standardized financial information — the SEC mandates audited financial statements, standard accounting principles, and restrictions on insider trading. However, the Enron and WorldCom scandals demonstrated that mandatory disclosure does not eliminate asymmetric information. Firms have strong incentives to present information favorably, and managers will always retain more information than outside investors.

3. Financial intermediation. Banks make private, non-traded loans, which means other investors cannot free-ride on the bank’s information production. Because the bank captures the full benefit of its efforts, it has a strong ongoing incentive to produce accurate borrower information. This is the fundamental reason banks dominate over securities markets in corporate lending.

4. Collateral and net worth. Collateral (property the lender can seize upon default) and high borrower net worth both reduce the lender’s potential losses. Even if adverse selection leads to some bad loans, collateral limits the damage. This explains why collateral requirements are nearly universal in debt contracts.

Beyond Mishkin’s core framework, two broader economic concepts also help reduce adverse selection:

5. Screening. Lenders actively gather borrower-specific information before making lending decisions — reviewing credit histories, analyzing financial statements, conducting due diligence, and requiring business plans. Banks specialize in screening, which is a key part of why financial intermediation works. For how banks apply screening in practice, see bank credit risk management.

6. Signaling. Borrowers voluntarily reveal their quality through costly actions that low-quality borrowers cannot easily imitate. Obtaining independent credit ratings and providing audited financial statements serve as credible signals. The logic follows Michael Spence’s signaling framework: a signal is credible precisely because it is costly — a firm that is actually weak cannot afford the scrutiny of rigorous third-party auditing or sustain the standards required for a strong credit rating.

7. Credit rationing. When adverse selection is severe, lenders may refuse to lend — or restrict loan size below what the borrower requests — rather than raise rates. Why? Because higher interest rates drive away the best borrowers (who know they are low-risk and refuse to overpay) while attracting the worst borrowers (who know they are high-risk and accept any terms). The result is that raising rates worsens the quality of the borrower pool. To protect themselves, lenders ration credit — denying loans to some creditworthy borrowers even though those borrowers would willingly pay the going rate. For quantitative approaches to measuring credit risk, see credit risk and probability of default.

Solutions to Moral Hazard

Once funds have been transferred, the challenge shifts from identifying quality to controlling behavior. Several mechanisms address post-transaction moral hazard:

1. Monitoring and costly state verification. Lenders and investors audit borrowers to verify how funds are being used. Mishkin calls this costly state verification — the expense of verifying the true state of a borrower’s finances. Monitoring is effective but expensive, and in equity markets it suffers from a free-rider problem: each shareholder wants others to bear the monitoring costs.

2. Venture capital. Venture capital firms solve the equity moral hazard problem by taking board seats and actively overseeing management. Because VC equity is private (not publicly traded), other investors cannot free-ride on the VC firm’s monitoring. The VC firm captures the full benefit of its oversight, preserving its incentive to monitor closely.

3. Debt contracts. Debt’s fixed contractual payments make it inherently less information-intensive than equity. With a debt contract, the lender needs to verify the borrower’s financial state only when the borrower cannot make payments — at default. With equity, investors must continuously verify all profits to ensure they receive their fair share. This dramatic difference in monitoring costs is a primary reason debt is the dominant form of external corporate finance worldwide.

4. Restrictive covenants. Debt contracts typically include detailed restrictions on borrower behavior: requirements to maintain adequate insurance, restrictions on selling or pledging assets, limits on additional borrowing, and obligations to provide regular financial reports. These covenants directly constrain the risky activities that moral hazard encourages.

5. Net worth and collateral. When borrowers have substantial equity at stake — “skin in the game” — their incentives align more closely with lenders. A borrower who would lose significant personal wealth in a default is less likely to take excessive risks. This is why lenders require down payments on mortgages and why banks assess borrower net worth before lending.

For how credit default swaps serve as tools for transferring and managing credit risk in the context of information asymmetry, see our dedicated guide.

Adverse Selection vs Moral Hazard

These two problems are often confused, but they differ in timing, mechanism, and required solutions:

The key insight is that adverse selection and moral hazard require different solutions. Private information production, government regulation, financial intermediation, and collateral address the first; monitoring, restrictive covenants, debt contracts, and incentive alignment address the second. A comprehensive lending framework must tackle both.

Common Mistakes

1. Confusing adverse selection with moral hazard. Adverse selection is a pre-transaction problem caused by hidden information — the lender does not know the borrower’s true quality. Moral hazard is a post-transaction problem caused by hidden actions — the borrower changes behavior after receiving funds. A bank screening loan applicants is addressing adverse selection. A bank monitoring existing borrowers through covenant enforcement is addressing moral hazard. The distinction matters because the solutions differ fundamentally.

2. Assuming government disclosure fully solves adverse selection. Mandatory disclosure (SEC filings, audited financial statements) reduces asymmetric information but does not eliminate it. The Enron and WorldCom scandals proved that firms can manipulate disclosed information for years before detection. Managers always retain more information than outside investors, and bad firms have strong incentives to present themselves favorably. This is why financial intermediaries remain essential even in well-regulated markets.

3. Thinking asymmetric information only affects equity markets. Bond markets suffer equally. In credit markets, adverse selection causes high-risk borrowers to crowd out low-risk borrowers: when interest rates rise, the best borrowers (who know they are low-risk) refuse to pay inflated rates, while the riskiest borrowers (who have little to lose) accept any terms. Debt contracts reduce moral hazard relative to equity, but they do not eliminate it.

4. Believing higher interest rates always clear credit markets. In standard supply-and-demand models, prices adjust to equilibrate quantity. But in credit markets, raising interest rates can make the lender worse off by worsening the borrower pool (adverse selection) and encouraging riskier behavior (moral hazard). This is why lenders sometimes ration credit — refusing to lend to some borrowers at any rate — rather than simply charging more. Credit rationing is a direct consequence of asymmetric information.

5. Ignoring the free-rider problem in information production. Private firms that produce and sell borrower information (credit rating agencies, research firms) face a fundamental limitation: non-paying investors observe the trades of informed investors and copy them without bearing the research cost. This free-riding drives up asset prices before the original researcher can profit, destroying the incentive to produce information. The free-rider problem is the core reason financial intermediaries (holding private, non-traded loans) dominate over securities markets.

Limitations of Asymmetric Information Models

1. Assumes fully rational actors. The lemons model and related frameworks assume that all participants process information rationally and act in their self-interest. In reality, behavioral biases — overconfidence, herding, loss aversion — can produce market outcomes that asymmetric information models alone cannot explain.

2. Signaling can be prohibitively costly. Not all high-quality firms can afford to credibly signal their type. A startup with genuinely strong prospects may lack the collateral or track record to differentiate itself from weaker competitors, even though its underlying quality is high. Signaling theory assumes costless communication of the signal’s existence — in practice, both producing and interpreting signals involve real costs.

3. Not all market failures are information-driven. Liquidity crises, bank runs driven by pure panic (rather than fundamentals), and contagion across interconnected institutions can all cause market failures independent of asymmetric information. Attributing every financial disruption to information problems oversimplifies the causes of instability.

4. Repeated interactions and reputation partially mitigate information problems. Static models (one-shot games) tend to overstate the severity of adverse selection and moral hazard. In practice, long-term banking relationships, reputation concerns, and repeated lending interactions reduce information gaps over time — a dynamic that simple adverse selection models do not capture.