LTV and health factor: understand a DeFi loan’s position

Calculate the simple ratio first

Loan-to-value, or LTV, expresses debt relative to collateral. In the simplest teaching model, divide debt value by collateral value and multiply by 100 to express a percentage.

Choose collateral worth 1,000 currency units and debt worth 400. The simple LTV is 400/1,000, or 40%. You owe an amount worth 40% of the collateral under that valuation.

Real protocols can use collateral factors, debt borrow factors, eligible asset restrictions and position modes. Their effective ratio can differ from the unadjusted sum of wallet-style dollar values.

Kamino's documentation, for example, distinguishes current, maximum and liquidation LTV and describes risk-adjusted debt. Read the current market's inputs instead of transplanting one asset pair's example to another.

Borrowing capacity and liquidation are separate limits

The maximum borrow condition determines whether a position may open or add debt. A separate liquidation condition defines when the program may permit a liquidator to act.

Choose teaching limits of 60% maximum borrowing and liquidation above 80%. At 40%, the simple example is below both. At 65%, a further borrow may be disallowed even though the position is not yet above that chosen liquidation boundary.

These values illustrate two different decisions. They are not current settings for a named market. A product can have more complex adjusted values and mode-specific conditions.

The gap should be interpreted as room under a particular calculation, not a guaranteed reaction period. Prices and debt can change before the user observes or responds to them.

A price fall increases LTV without a new loan

Let the 1,000-unit collateral be 10 SOL at a chosen price of 100 per SOL. Keep the debt at 400 for the moment. If SOL falls to 80, collateral becomes 800 and simple LTV rises to 50%.

If collateral value reaches 500, the same 400 debt gives 80%. With 10 SOL, that corresponds to a chosen price of 50. Below that price the simple ratio exceeds 80%, under the example's assumptions.

Now add debt growth: if the obligation becomes 440 while collateral is 800, the ratio is 55%. Both interest and asset price affect the position even when no new borrowing occurs.

A different debt asset can move in price too. Use the protocol's actual valuation inputs for both sides rather than assuming the debt's currency value remains constant.

Debt growth adds another source of risk

Read the formula behind health factor

Some products show a health factor rather than or alongside LTV. A common teaching design divides threshold-adjusted collateral by debt, but the actual product's definition controls interpretation.

Under a chosen 80% collateral adjustment, 1,000 collateral and 400 debt produce a teaching health factor of 2. If collateral falls to 500 with debt unchanged, that model reaches 1.

This arithmetic explains one model. Another application can use different factors, debt weights or named metrics. Do not decide that a displayed value of 1 has the same meaning everywhere without reading the contract.

For several assets, examine how they are combined and whether a mode changes eligibility or factors. A simple single-asset equation is a starting explanation, not a replacement for that calculation.

Understand what each position action changes

Repaying debt reduces the numerator in a simple LTV. Adding eligible collateral increases the denominator. Withdrawing collateral or borrowing more moves the ratio in the opposite direction, subject to program permission.

Those operations can have their own costs and execution conditions. Repaying with collateral can involve a sale, so the resulting improvement depends on actual quote and amounts. A failed attempt does not change the position merely because the interface preview looked healthier.

Before an action, record the relevant inputs and proposed result. After execution, read the actual debt, collateral and health fields. This catches stale prices, amount rounding and differences between an estimate and accepted state.

Use liquidation mechanics to understand the consequence of crossing the condition, and oracle data to interpret the valuations that drive it.

Monitor inputs as well as the headline number

A useful position view shows quantities, prices, observation times, accrued debt and applicable factors. Its health number should link to those definitions.

Keep the market and position mode with any saved parameter record. Current settings can change through the protocol's governance or administrative process. A screenshot from opening does not establish today's configuration.

Set alerts around observed conditions and data freshness. A missing or old price should remain visible as a data issue rather than producing a reassuring new health value.

Common questions