Crypto news and analysis
Advanced · Advanced analysis

Crypto credit markets explained

Decompose crypto credit risk across borrower, collateral, liquidation, custody, oracle, legal, and liquidity layers in onchain and bilateral markets.

18 min read3-question quizUp to 190 XP

A credit committee reviews a loan that is 150 percent collateralized and must decide whether that headline ratio survives a market gap, oracle delay, venue outage, and borrower default. The answer depends on who controls collateral, which rules are enforced by code or contract, and how quickly assets can be sold. Crypto credit spans onchain lending, exchange margin, secured bilateral loans, and unsecured facilities whose creditor, debtor, remedies, and recovery paths differ radically.

Risk decomposition prevents a reassuring feature from standing in for the whole structure. Overcollateralization can reduce expected loss and still fail under a price gap, oracle error, bridge failure, congestion, custody dispute, or correlated collateral decline. An institutional review maps exposure, probability of failure, severity, detection, and recovery for each dependency. It also separates current balances from estimates of liquidation proceeds and assumptions about legal enforceability or market depth.

What you will learn

  • Map creditor, borrower, collateral, control, and recovery rights
  • Decompose market, liquidity, oracle, operational, legal, and counterparty risks
  • Stress liquidation and loss severity without assuming continuous markets

Separate probability of default from loss severity

Credit risk has at least two questions: how likely is an obligation to become impaired, and how much is recoverable if it does? Overcollateralization lowers loss severity only when collateral remains accessible and saleable above debt plus costs. Borrower quality still matters when collateral can be withdrawn, disputed, rehypothecated, concentrated, or correlated with the borrower. A strong borrower with weak documentation can also create poor recovery outcomes.

Build a collateral haircut from price volatility, liquidity at the required size, liquidation latency, basis between venues, custody access, bridge or wrapper risk, legal priority, and wrong-way correlation. Do not infer executable proceeds from the last traded price. Model gaps and discontinuous liquidity, especially for weekend or cross-venue stress. Concentration limits should apply by ultimate risk source, not merely by token ticker, because several wrapped assets can depend on one issuer or bridge.

Trace liquidation as an operational process

A liquidation requires detection, a valid price, an authorized actor, transaction inclusion or contractual notice, financing, asset transfer, and a market exit. Onchain liquidators may compete for a discount, but network congestion or unprofitable gas can delay action. Offchain lenders may face margin-call cure periods, collateral-agent procedures, bank hours, or insolvency stays. Each step adds time during which collateral value can move.

Review incentives and capacity. A liquidation bonus must cover execution cost and price risk without imposing unnecessary loss on borrowers. Reserve funds may absorb shortfalls but are finite and can be invested in correlated assets. Backstop bidders can withdraw when capital is most valuable elsewhere. Stress tests should reduce available depth, widen basis, delay price updates, and impair several borrowers together rather than assuming independent defaults.

Create a risk register and monitoring system

Organize exposures into counterparty, collateral, market, liquidity, leverage, oracle, smart-contract, custody, legal, governance, concentration, and operational categories. For each, record the observable exposure, estimated probability range, loss mechanism, controls, owner, warning indicator, and recovery action. Avoid adding category scores into one precise number unless the aggregation method and dependence assumptions are defensible. A heat map supports discussion; it does not replace exposure data.

Monitoring should focus on causal deterioration: collateral coverage, utilization, maturity mismatch, withdrawal queues, oracle deviations, liquidation success, reserve capacity, borrower concentration, covenant compliance, and changes to administrator powers. Market prices can be early warnings but are not proof of insolvency. Define escalation and exit mechanics in advance, because stressed governance and thin liquidity make improvised responses slower and more expensive.

Reality check

Common misconceptions

Overcollateralization removes credit risk from crypto lending.

Recovery still depends on collateral price, market depth, access, liquidation speed, oracle integrity, legal priority, and correlated failure during stress.

Onchain liquidation is automatic, so operational risk is negligible.

Contracts require prices, transaction inclusion, economically motivated liquidators, functioning networks, and saleable collateral. Any link can fail or become uneconomic.

Before you act

Risks and limitations

  • Collateral can fall faster than margin calls or onchain liquidations can execute at sufficient depth.
  • Rehypothecation, bridge dependence, custody terms, or insolvency rules can prevent timely collateral access.
  • Oracle manipulation or stale prices can trigger improper liquidations or delay necessary ones.
  • Correlated borrowers and collateral can exhaust reserves and liquidation capital simultaneously.

Key takeaways

  1. Map the actual debtor, creditor, collateral controller, contracts, and remedies for each facility.
  2. Analyze probability of impairment separately from recovery and loss severity.
  3. Haircuts must include liquidity, latency, basis, wrapper, custody, and legal risks.
  4. Treat liquidation as a multi-step operating process rather than a guaranteed formula.
  5. Monitor causal exposures and dependencies instead of relying on a single composite score.

Primary and further reading

Knowledge check

Test your understanding

Score at least 2 out of 3 to complete this lesson. Explanations appear after you submit.

1. A loan begins at 180 percent collateralization, but the collateral trades on one shallow venue and is correlated with the borrower. Why can it remain risky?
2. A market maker borrows against its own exchange token, which falls when confidence in the borrower weakens. Which option diagnoses the risk?
3. A lender stresses collateral down 30 percent but assumes instant sales at the screen price. What must be added to make the liquidation test credible?