Loan-to-Value: Choosing a Number You Can Live With

Loan to value is the size of a debt measured against the collateral that secures it. Borrow 6,000 against 10,000 of ETH and your loan to value is 60%. That single ratio decides how far the collateral can fall before the position is closed, which makes it more important than the interest rate you negotiate.

Borrowers usually pick a number by looking at the maximum the market allows and subtracting a little. That is not a method. A workable LTV comes from the volatility of the collateral, the length of time you plan to hold the debt, and the cost of being closed out at the worst possible moment.

• The maximum is a limit, not a recommendation. It is set to protect the lender, and it assumes the collateral can be sold quickly.
• Your real buffer is the gap to liquidation, not the gap to the maximum. Those are different numbers on almost every market.
• Interest moves the ratio. An LTV that looks stable can drift upward for a year without any price change.
• Volatility sets the number. Two assets with the same maximum loan to value can deserve very different debts.

What loan to value measures

Loan to value compares debt to collateral at current prices. It is a snapshot, recalculated as either side moves, and it says nothing about your ability to repay.

The arithmetic

Divide the outstanding debt by the collateral’s market value. A 9,000 debt against 20,000 of collateral is a 45% LTV. Add 500 of accrued interest and the same collateral now supports a 47.5% ratio. Nothing about the asset changed; the debt grew.

Note what the ratio leaves out. It ignores the interest rate, the expected holding period, the liquidity of the collateral and any tax or exit cost. It is one number describing one relationship at one instant.

Why the ratio is not the whole risk

Two positions can share an LTV and carry completely different risk. One holds a large-cap asset with deep spot markets; the other holds a thin token trading a few hundred thousand dollars a day. The first closes at a predictable price. The second may gap through the liquidation level and leave bad debt for the protocol to socialise.

That asymmetry is why markets assign different parameters per asset rather than one global number. The ratio is comparable only within a single collateral type.

Why the maximum loan to value is a ceiling

The maximum exists so the lender has room to seize and sell collateral before the debt exceeds it. It is calibrated for the protocol’s protection, and it is deliberately set close to the point where the position becomes dangerous for you.

The gap between maximum and threshold

Most markets publish two numbers. The maximum loan to value caps what you can borrow at opening. The liquidation threshold is the higher ratio at which the position becomes eligible for closure. The space between them is the protocol’s margin, not yours.

Borrowing at the maximum means you start with that entire margin already consumed. A small adverse move puts you at the threshold immediately. Borrowing well below the maximum is the only way to own any of that space yourself.

What the ceiling assumes

The maximum assumes continuous pricing and available liquidity. In a fast market both weaken at once: prices gap, books thin, and the liquidation meant for the threshold lands somewhere below it.

A maximum loan to value is the point at which the protocol stops trusting the position. It is not the point at which you should stop trusting your own.

Choosing a loan to value you can live with

Start from a drawdown you consider plausible, not from a percentage someone else uses. The question is not how much you can borrow, but how much you can borrow and still be solvent after a bad month.

The drawdown test

Pick a fall in collateral price that you would not call a catastrophe — for a large-cap asset, a third is a reasonable stress case; for a smaller one, half. Then solve backwards for the debt that survives it with margin to spare.

If the collateral can lose 40% and you want to sit comfortably below the threshold afterwards, your starting loan to value should leave that much room plus a buffer for interest. On most markets that lands a volatile asset in the 30–50% range, and stablecoin collateral higher.

Correlated collateral

Diversifying collateral does not help if everything in the basket falls together. Correlated assets behave as one larger position, and the ratio computed across them looks calmer than the risk actually is. Count correlated collateral as a single exposure when you size the debt.

Running the numbers

Take 20,000 of ETH at 3,000 per unit, a maximum loan to value of 80%, a liquidation threshold of 82.5% and a 5% liquidation penalty.

Borrow 12,000 — a 60% LTV. The position reaches the threshold when collateral value falls to 12,000 divided by 0.825, or 14,545. That is a price of about 2,182, roughly 27% below the entry price. A single bad week can get there.

Borrow 9,000 — a 45% LTV. The threshold is reached at 10,909 of collateral value, a price near 1,636, or 45% below entry. The position survives a drawdown that would have closed the first one.

Borrow 6,000 — a 30% loan to value. The threshold sits near 1,091, a 64% fall. This is the version that survives a bear market rather than a correction.

Now add time. At 7% annual interest on the 9,000 debt, a year of no repayments adds 630. The threshold then arrives at 11,673 of collateral value, a price near 1,751 instead of 1,636. Interest quietly consumed about 7% of the buffer, and the effect compounds the longer the position runs.

Loan to value by collateral type

Parameters differ by venue and by asset, and the ranges below are typical rather than fixed.

A stablecoin at 85% loan to value is not safer than a volatile asset at 45%. The stablecoin has no volatility to absorb, so the only thing that matters is the peg holding — and when it does not, the buffer is already gone.