How Digital Credit Works: The Complete Guide to Bitcoin-Backed USD Loans

What is Digital Credit?

Digital Credit is the business of lending dollars against bitcoin. You pledge BTC as collateral, borrow USD (or stablecoins), and keep the upside: if bitcoin rises, your collateral is worth more; if it falls, you get a margin call and, eventually, liquidation. You are not selling your bitcoin — you are borrowing against it.

The mechanics are simple. The risk is not.

Every bitcoin-backed loan is really two products sold as one: a rate sheet (APR, LTV, term) and a custody arrangement (who holds your bitcoin, and what they are contractually allowed to do with it). Lenders advertise the rate sheet. The custody arrangement is what can actually hurt you. This guide walks through both — plus the eight-factor rubric Pledge uses to score lenders — so you can compare products on the axis that matters.

How a bitcoin-backed loan actually works

Three numbers define every loan, and you should read them in order.

Loan-to-value (LTV) is the ceiling on how much you can borrow against your bitcoin. Ledn caps at 50%, Unchained at up to 50%, SALT up to 70% (verified lender docs). A higher LTV lets you borrow more per bitcoin — and leaves you thinner margin against a price drop.

The margin call is the lender demanding a top-up when your LTV breaches a threshold — Ledn alerts at 70%/75% LTV, Unchained at 70%, SALT at 83.33% (verified lender docs). You either add more collateral or reduce the loan. The cure window is how long you get to do it: Strike gives 72 hours, Unchained 24 (verified lender docs). Some lenders publish none at all.

Liquidation is the forced sale of your collateral when LTV exceeds the liquidation threshold — Ledn 80%, Unchained 85%, Aave 78%, Coinbase/Morpho 86% (verified lender docs). It is the event that can actually hurt you, and it is a forced taxable sale in the US (more below).

The rate sheet is the headline. The LTV ladder — where the alerts fire, how long you get to cure, and whether liquidation is partial or full — is the fine print that decides how you lose money in a drawdown.

The three custody models

Every BTC loan is a custody arrangement wearing a rate sheet. There are three buckets, and they are not equal.

1. Self-custody / collaborative multisig — you keep a key

Unchained is the reference case: a 1-of-3 multisig where the borrower's company holds one key, and Unchained and Fortis Bank each hold one (verified — unchained.com/loans). Because no single party holds a majority of the keys, the lender cannot unilaterally move your bitcoin — and rehypothecation (the lender re-lending or selling your collateral) is structurally impossible. There is no contract clause to trust; the math forbids it. Unchained offers up to 50% max LTV on 3–60 month terms.

The tradeoff: you carry more operational responsibility (key management), and if you miss a margin call, the loan agreement provides for foreclosure on the collateral.

2. Ring-fenced custodian — a regulated third party holds it, contractually can't touch it

Ledn is the reference case. Your bitcoin sits with a qualified custodian — BitGo Trust, Anchorage, or Fidelity Digital Assets — ring-fenced from Ledn's own balance sheet, with no rehypothecation (verified — ledn.io). The custodian holds the keys; the lender has a contractual right to liquidate under defined terms, and nothing else. Ledn: 50% max LTV, APR 9.25–11.49% tiered by loan size, 12-month fixed terms, no prepayment penalty, no credit checks. Liquidation alerts at 70%/75% LTV, automatic partial liquidation at 80% (verified — ledn.io; borrowonbitcoin.com, Aug 10 2026).

This is the model Wall Street has started underwriting (more below).

3. Pooled / rehypothecation — the lender can use your collateral

Aave is the DeFi reference: collateral is pooled, and the protocol rehypothecates by design — your wrapped bitcoin (WBTC/cbBTC) backs other borrowers' positions. SALT, the CeFi reference, historically reserved the right to rehypothecate in its 2018–22 loan terms (DFPI California Consent Order, Dec 2024); today it says it won't — but the current agreement is the only thing that binds, and its APR (7.49–10.50%, Aug 2026) is cheaper than Ledn's. The rate says nothing about how your bitcoin is held.

Why custody is the red line

2022 is the reason this taxonomy exists. Celsius ($1.2B deficit, $4.7B frozen), BlockFi ($680M Alameda exposure), Voyager — all driven by rehypothecation, opaque leverage, and unsecured institutional lending. Mashinsky: 12 years. The sector rebuilt itself on overcollateralization — $73.59B in outstanding crypto-collateralized loans in Q3 2025, an all-time high (Galaxy Research). The risk didn't disappear; it moved from lenders' balance sheets into the custody structure.

That's why the first question to ask any lender is not "what's your APR" but "who holds my bitcoin, and what are you allowed to do with it?"

The 8-factor risk rubric

Pledge scores lenders on eight factors, each 0–10, weighted into a single composite (weights sum to 100). The rubric is public, reproducible, and evidence-bound — every scored claim traces to a primary source or is flagged as unverified. No paid rankings, no black box.

The perfect-10 standard: "A 10/10 Bitcoin loan would be nearly boring." The borrower keeps meaningful control of collateral; the collateral cannot be reused; reserves are independently verifiable; legal status is clear; liquidation is conservative and explicit; the lender has survived multiple cycles with clean public evidence. Most real products should not reach 10 — every borrowing path adds a tradeoff. A 7–8 is strong, not weak.

Live range (2026-06-21 audit snapshot): Unchained 9.02 (top — 1-of-3 multisig on loans, no rehypothecation), Xapo 8.83, Aave 8.33, Maker/Sky 8.15, Ledn 7.41, Figure 7.10, Arch 7.03, Strike 6.35, Coinbase/Morpho 6.32, Debifi 6.08, SALT 5.53, Nexo 5.25, Moon Mortgage 4.91, YouHodler 4.02, Lava 3.39 (custody model unconfirmable).

The rubric's honest caveats:

• Loss protection and transparency are the weakest-structured factors — prose-backed, thin on structured evidence — yet they carry real safety weight. The methodology says so; the guide should too.
• Scores are lender-level, not product-level. Two lenders with equal scores can carry very different products (APR, LTV, liquidation terms). The number tracks the lender; the agreement tracks the loan.
• Scores are a point-in-time snapshot (2026-06-21). Rates and LTV tables move — timestamp anything you quote.

How to compare products

Step 1 — Custody first. Identify the bucket: self-custody multisig, ring-fenced custodian, or pooled/rehypothecation. This is the 20-weight factor and the one that can actually lose you bitcoin. If a lender's terms reserve the right to rehypothecate, nothing else on the rate sheet matters as much.

Step 2 — Then the rate. APR is the annualized cost including fees. Compare like-for-like: Ledn 9.25–11.49% vs SALT 7.49–10.50% (Aug 2026) — the cheaper lender is the one with the rehypothecation history. A low APR is a lure, not assurance.

Step 3 — The liquidation math. Margin-call threshold, cure window, partial vs full liquidation, penalty. Examples (verified lender docs): Ledn alerts at 70/75% LTV, partial liquidation at 80%; Unchained 24-hour cure window, collateral liquidated on an uncured margin call; Strike 72-hour cure; Aave liquidation at 78% LTV with a ~5% penalty; Coinbase/Morpho 86% LLTV with 4.38% penalty. A lender with a longer cure window and a partial-liquidation step is friendlier than one that goes straight to full foreclosure.

Step 4 — Fees and tax. Origination fees (Figure 1%, Arch 0.49–1.49%) add to the real cost. Borrowing against BTC is not a taxable event at origination; liquidation is a forced taxable sale (US: up to 37% short-term / 20% long-term capital gains). Jurisdiction-specific — consult a tax professional.

Step 5 — Proof, not promises. Check for independently verifiable reserves (Ledn's Open Book Report via The Network Firm LLP, latest 2026-03-31), a real regulatory charter, and a track record that includes a down market. "No customer losses" claims are self-reported — treat them as marketing until independently verified.

Step 6 — Read the actual agreement. Not the marketing page. The current terms are the only thing that binds. SALT's current terms disclaim rehypothecation; its historical terms reserved it. What's on the rate sheet today can change in the fine print tomorrow.

A worked comparison. Two lenders, same goal — borrow dollars against bitcoin without selling. Ledn quotes 9.25–11.49% APR, 50% max LTV, no rehypothecation, collateral ring-fenced at a qualified custodian, partial liquidation at 80%. SALT quotes a cheaper 7.49–10.50% APR and a higher 70% LTV — but it historically reserved the right to rehypothecate your collateral, and the current agreement is the only thing that binds. The cheaper rate is the marketing; the custody structure is the risk. The rubric's 20-weight rehypothecation factor exists precisely because that gap — cheap rate, different custody — is where borrowers get hurt.

Wall Street is underwriting this now

The custody question has become a ratings question. S&P assigned BBB- — the first investment-grade ABS in digital assets — to Ledn Issuer Trust 2026-1 ($188M, closed Feb/Mar 2026, 2x oversubscribed; $160M senior Class A at BBB-(sf), $28M Class B at B-(sf), backed by ~4,078 BTC ≈ $357M at review). Read the stress test twice: S&P's A-stress scenario assumed 79% default / 68% recovery on Class A. Cantor Fitzgerald launched a $2B institutional BTC lending program (May 2025, Copper/Anchorage custody). Tether took a strategic stake in Ledn (Nov 18 2025; Tether cites $2.8B+ BTC-backed originations since inception, $1B+ in 2025).

The sector has institutionalized. That is progress — and it is also why the rubric matters more, not less.

The honest caveat

No rehypothecation ≠ no risk. Custodian failure, liquidation mechanics, and tax still bite. Native-BTC rails are moving fast — Zest's BitVM vaults were announced May 2026 but are not yet live (verified 2026-08-20); Aave+Babylon is in an Aave DAO temp-check. The landscape will look different in six months.

The take

The APR is what they charge you. The custody structure is what they can do to you. Compare on the second one first. Risk is the product.