A 2025 primer on crypto loans: how they work, DeFi vs CeFi mechanics, risks, costs, and how to decide if a crypto loan fits your situation.
Arkadii Kaminskyi
Head of Operations at Sats Terminal
Head of Operations at Sats Terminal with 5 years of experience in crypto. Specializes in DeFi, yield farming, and borrowing — has reviewed 50+ crypto products.

Crypto loans have moved from a niche corner of decentralized finance to a mainstream tool used by individuals, funds, and companies who want liquidity without selling their digital assets. In 2025, you can pledge bitcoin, ether, or a basket of tokens and walk away with stablecoins in minutes — no bank, no credit check, no paperwork. But the term "crypto loans" covers several different products, each with its own mechanics, risks, and audiences. This guide is the definitive 2025 primer. We'll define crypto loans precisely, separate the three main flavors, walk through how they work end to end, compare DeFi and CeFi venues, cover what changed in 2025, and finish with a clear decision tree for whether a crypto loan is right for you.
At their core, crypto loans are credit arrangements where cryptocurrency plays a role on at least one side of the transaction. The phrase is broad on purpose. It can mean borrowing cash or stablecoins by pledging crypto as collateral, or it can mean lending out crypto to earn yield, or it can describe niche institutional credit lines extended without collateral at all. Newcomers often conflate these, which leads to confusion about who bears risk, who pays whom, and what regulations apply.
The most common retail product is the collateralized crypto loan: you deposit a digital asset (typically bitcoin or ether) into a smart contract or with a custodian, and you receive a loan denominated in a stablecoin like USDC. The collateral usually exceeds the loan value — a property called over-collateralization — to protect the lender against price swings. You pay interest on what you borrow, and you reclaim your collateral when you repay. This is the model behind every major lending protocol on Ethereum and the model behind Borrow by Sats Terminal, which aggregates these offers across DeFi and CeFi venues.
For a more entry-level introduction, the what is cryptocurrency lending learn page covers the same territory at a slower pace. The collateral glossary entry explains why over-collateralization is the default for permissionless markets.
Before going deeper, it helps to draw clean lines between the three categories that fall under the umbrella term. Mixing them up is the single biggest source of confusion for newcomers.
This is the dominant retail use case in 2025. You own bitcoin and want dollars without triggering a sale. You deposit BTC (or wrapped equivalents like wBTC, BTCB, or cbBTC) into a smart contract, choose a stablecoin to borrow, and the protocol releases USDC up to a maximum loan-to-value ratio — commonly 50–75% depending on the venue and the asset's volatility. You keep the upside of your crypto holdings while accessing dollar liquidity. The 2025 bitcoin-backed loans explainer covers this category in depth.
The flip side of the same coin: instead of borrowing, you supply assets into a lending pool and earn interest from borrowers. The supply rate is usually below the borrow rate, with the spread covering reserves and protocol fees. People often call this "crypto lending" even though, technically, it's the same product viewed from the lender's seat. The complete beginner's guide to crypto lending walks through the supplier side carefully.
The third category is institutional. Funds, market makers, and trading desks sometimes borrow without posting full collateral, relying on legal agreements, reputation systems, or off-chain credit assessments. Products like Maple Finance pools or bilateral OTC lines fall here. Retail users almost never touch this segment, and after several high-profile failures in 2022, the institutional uncollateralized market has shrunk and become more conservative. The institutional crypto lending guide covers this niche.
| Type | Who uses it | Collateral | Typical rate (early 2025) | Main risk |
|---|---|---|---|---|
| Collateralized loan | Retail and institutional borrowers | Over-collateralized in BTC/ETH/stables | 4–10% APR on USDC | Liquidation if collateral falls |
| Lending out for yield | Retail and institutional suppliers | None (you're the lender) | 2–7% APY on USDC | Smart contract or counterparty risk |
| Flash loans | Sophisticated DeFi traders | None — repaid in same transaction | ~0.05–0.09% fee per loan | Tx fails if not repaid atomically |
| Uncollat institutional | Funds, market makers | Reputation / off-chain agreements | Negotiated, often 8–15% | Counterparty default |
For the purposes of this guide, when we say "crypto loans" without qualification, we mean the first category — collateralized retail loans — because that's what most readers come looking for.
Walking through a real example clarifies the mechanics better than any abstract diagram. Suppose Maya holds 1 BTC, doesn't want to sell because she's bullish long-term, but needs USD 20,000 for a home renovation. Here is the path she takes when using a self-custodial venue.
Maya picks bitcoin as collateral. She then needs to decide whether to use a DeFi protocol directly, a CeFi lender, or an aggregator. Going direct to a protocol like Aave or Morpho Blue requires that her BTC first be wrapped into wBTC, BTCB, or cbBTC — bitcoin can't natively touch Ethereum-style smart contracts. An aggregator like Borrow by Sats Terminal handles bridging and wrapping automatically, so Maya can stay focused on terms instead of plumbing.
Maya signs a transaction depositing wBTC into the lending pool or isolated market. The smart contract records her deposit, increments her collateral balance, and immediately starts treating it as a borrowing power source. Crucially, this is non-custodial in DeFi: no human at any company is holding her bitcoin. The contract code is the custodian. Read more in the introduction to DeFi lending.
Maya enters her desired loan amount — say USD 20,000 in USDC. The protocol checks her health factor: does her collateral, at current BTC price, support the requested LTV? If yes, USDC lands in her wallet within a single block. She can now spend, swap, or bridge it as she likes. The decentralized finance glossary entry explains why this happens without any intermediary approval.
From the moment funds are received, Maya's loan exists in a live state. As BTC price moves, her LTV moves with it. If BTC rises, her LTV falls and she has more borrowing room. If BTC falls, her LTV climbs toward the liquidation threshold. Most protocols liquidate when LTV exceeds something like 80–85%, selling enough collateral to bring the position back into safety. The managing liquidation risk learn page covers practical defense techniques like topping up collateral or partial repayment.
When Maya is ready, she repays principal plus accrued interest in USDC. The smart contract burns the debt, unlocks her wBTC, and she withdraws — and if she used an aggregator, the wrapped asset is unwrapped and bridged back to native BTC. The full cycle, end to end, takes minutes for the transactions and however many days, weeks, or months Maya wanted to hold the loan. The repaying strategically learn page covers timing and partial-repayment tactics.
The two big institutional structures behind crypto loans are decentralized finance protocols and centralized finance lenders. Both end up putting stablecoins in your wallet against pledged crypto, but the operational details differ in ways that matter.
In DeFi, smart contracts hold collateral, set rates algorithmically based on pool utilization, and execute liquidations without human intervention. There is no KYC at the protocol level, balances are transparent on-chain, and the user keeps custody of keys throughout. Risks shift from counterparty default to smart-contract bugs, oracle failures, and stablecoin de-pegs. The centralized finance glossary and the parallel decentralized finance glossary entries summarize the two models.
In CeFi, a company like Ledn, Nexo, or Coinbase takes custody of your collateral, makes lending decisions, and settles loans on its books. KYC is mandatory in most jurisdictions. Rates are typically fixed and posted, often more borrower-friendly than DeFi at the cost of trusting a corporate custodian. The history of CeFi includes prominent failures (Celsius, BlockFi, Genesis), so questions about proof of reserves and rehypothecation matter. The CeFi vs DeFi pros and cons piece and the comparing DeFi vs CeFi learn page dive deeper.
| Dimension | DeFi crypto loans | CeFi crypto loans |
|---|---|---|
| Custody | Self-custody via smart contract | Custodian holds your collateral |
| Rates | Algorithmic, variable, utilization-driven | Posted, often fixed for a term |
| Transparency | Fully on-chain, public balances | Off-chain books; some publish proof of reserves |
| KYC | None at protocol level | Required in most jurisdictions |
| Minimum size | Effectively gas costs (sub-$100 viable) | Often $1,000+ minimums |
| Liquidation | Automatic, on-chain auction | Manual or algorithmic; varies by lender |
| Counterparty risk | Smart-contract and oracle risk | Lender insolvency risk |
| Recourse if dispute | None — code is final | Customer service, possibly courts |
Neither column is universally better. A retail user who wants no KYC, full custody, and to use a small loan of a few hundred dollars will probably prefer DeFi. A user who needs $250,000, wants a 12-month fixed rate, and is comfortable with a regulated counterparty will probably choose CeFi. Borrow by Sats Terminal aggregates both so users can compare side by side rather than picking ideologically.
The crypto-loan market in 2025 looks different from how it looked in 2022 or even 2023. Several structural changes have made the space more usable, more competitive, and a bit more regulated.
Morpho Blue, which launched in early 2024, has matured into a primary venue for crypto loans by 2025. Its design — minimal, immutable, with isolated markets per collateral/loan pair — addresses the contagion risk that plagued earlier shared-pool protocols. Each market has its own oracle, its own LTV cap, and its own interest-rate curve, so a problem with one asset can't drain another. The Morpho integration announcement details why this matters for borrowers, and the Morpho documentation explains the architecture.
Two years ago, serious lending was Ethereum mainnet plus a couple of L2s. By 2025, Aave v3 runs natively on Ethereum, Arbitrum, Polygon, Optimism, Base, and BNB Chain, with comparable depth on each. Aave's documentation tracks the full deployment list. Borrowers can pick the chain with the lowest gas costs for their loan size — small loans on Base, large loans where liquidity is deepest on Ethereum mainnet.
Operating a crypto loan used to require knowing how to bridge, how to wrap, how to compare rates manually, and how to monitor positions across multiple dashboards. Aggregators in 2025 collapse all of that into a single workflow. The crypto lending aggregator FAQ explains the pattern, and the how aggregators find best rates learn page walks through the algorithmic side.
The US, EU, and major Asian jurisdictions have moved cautiously on crypto-lending rules. MiCA in the EU is now in effect for stablecoin issuers and crypto-asset service providers. The US has seen some enforcement actions against centralized lenders that took retail deposits without proper licensing, while DeFi protocols largely remain in a grey zone. We're staying neutral here — see the regulatory landscape learn page for jurisdiction-by-jurisdiction context. The general trajectory is toward more clarity, especially for CeFi, while DeFi continues to argue that immutable code shouldn't be regulated as a financial intermediary.
Oracle networks have gotten more robust, with Chainlink, Pyth, and RedStone competing on price feed quality and update frequency. Liquidation infrastructure has more participants, leading to tighter execution and smaller liquidation discounts. Bug-bounty programs and audits have raised the bar for new protocols. None of this eliminates risk, but the marginal protocol is meaningfully safer than its 2022 equivalent.
The borrowable side of crypto loans has also matured. USDC has steadily extended its multi-chain footprint and improved its attestation cadence, while USDT remains dominant in raw liquidity especially on tron and BSC. Newer entrants like PYUSD and regulated euro-denominated stablecoins broaden the menu. For a borrower, this means more choice in what currency the loan is denominated in and more flexibility in where to spend it. The trade-off is that each stablecoin carries its own issuer and reserve risk, which the borrower inherits the moment they accept the loan proceeds and continue to hold them.
Wrapping bitcoin used to be the only way to use it as collateral on Ethereum-style protocols. In 2025, several products are pushing toward more bitcoin-native experiences — runes, taproot-based collateral schemes, and cross-chain bridges with cleaner trust assumptions. Aggregators continue to abstract this complexity for users, but the underlying choices about where bitcoin actually sits during a loan have meaningful security implications worth understanding. The bridging and wrapping bitcoin learn page goes into the trade-offs.
Three concepts sit at the operational heart of every crypto loan: the collateral you post, the loan-to-value ratio you operate at, and the health factor that summarizes how close you are to liquidation. Understanding these well is the difference between a borrower who sleeps fine and one who watches dashboards every hour.
Not all collateral is created equal. Bitcoin and ether are the gold standards because of their deep liquidity and 24/7 markets — liquidators can always find a buyer at a price close to mid-market. Wrapped versions of bitcoin (wBTC, BTCB, cbBTC) inherit those characteristics with the added wrinkle of bridge or custodian risk on the wrapping itself. Stablecoins as collateral are also common; they provide stable borrowing power but earn lower supply yields. Volatile altcoins are accepted in some isolated markets but typically with much lower LTV caps to compensate for the risk that price might gap through liquidation thresholds during fast moves. The understanding collateral and LTV learn page walks through the choices.
Each market specifies two LTV numbers that matter: the maximum LTV at which you can open or expand a position, and the liquidation LTV at which forced sale begins. The gap between them is your safety margin. For BTC collateral on Aave v3, max LTV is commonly around 73% with liquidation around 78%, leaving roughly five percentage points of buffer. On Morpho Blue isolated markets, the parameters are set per market and can vary more widely. The smaller the gap, the more efficient the borrowing but the less margin for sudden price moves.
Most protocols expose a health factor: a unitless number where 1.0 represents the liquidation threshold and higher values mean safer. A health factor of 2.0 means your collateral could halve in value before liquidation; a health factor of 1.2 means a 20% drop puts you in trouble. Treating the health factor as a speed limit you respect — say, never letting it fall below 1.5 except briefly — converts liquidation risk from an abstract worry into a concrete signal. The monitoring loan health learn page covers tools and alerts.
No discussion of crypto loans is honest without a clear-eyed look at what can go wrong. Risks and costs come in five buckets.
The single biggest risk to a borrower is forced liquidation when collateral price falls. If you borrow USD 50,000 against 1 BTC at 70% LTV and BTC drops 25%, you may breach the threshold and have part of your collateral sold at a discount, often plus a liquidation penalty of 5–10%. The fix is to borrow conservatively (40–55% LTV) and to monitor health regularly. The risks every borrower should know piece goes deep.
DeFi loans depend on code. A bug, an oracle attack, or an economic exploit can drain a market. Audits and immutable design (Morpho Blue) reduce but don't eliminate this. Diversification across protocols, sticking to high-TVL venues, and avoiding new untested markets are practical mitigations.
CeFi lenders can fail. Celsius, BlockFi, and Genesis all entered bankruptcy in 2022–2023, and customer funds were either frozen or partially lost. Proof-of-reserves attestations help but are not a complete substitute for transparent on-chain accounting. The is crypto lending safe piece covers due diligence.
If you borrow USDC or USDT, you carry the issuer's risk. USDC briefly de-pegged in March 2023 during the Silicon Valley Bank crisis. USDT has faced repeated questions about reserves. Holding the borrowed stablecoin in your wallet means you bear de-peg risk; spending it immediately transfers that risk to the recipient.
Interest is the obvious cost, but you should also consider gas fees on chain, bridging fees if you cross networks, the spread on swapping into stablecoins, and any aggregator or protocol fees. The all-in cost on a small loan can be meaningfully higher than the headline APR. The crypto lending rates piece breaks down what good looks like in 2025.
Theory is one thing; what people actually do with crypto loans is another. A few patterns dominate the user base in 2025.
The cleanest use case: a long-term holder of bitcoin needs cash for a real-world expense — a property purchase, tuition, a major medical bill — and would rather not sell. Selling crypto crystallizes capital gains tax in most jurisdictions. Borrowing against the same crypto provides liquidity without triggering that tax bill, which can be material for holders who bought years ago at much lower prices. The strategy isn't free — interest is owed and liquidation is a risk — but the after-tax math often favors borrowing for medium-duration needs.
Companies whose treasury is denominated in crypto often face mismatched cash flows: they receive in BTC or ETH but pay vendors and employees in fiat. Standing crypto-loan facilities convert that volatility into predictable working capital. The business doesn't need to time the market on conversions; it borrows stablecoins as needed and repays when revenue allows.
Experienced traders use crypto loans to gain leverage. Pledge BTC, borrow USDC, swap to BTC, pledge again, borrow again, and so on. Each round increases exposure to bitcoin price. This is profitable when prices rise and brutal when they fall — leveraged liquidation cascades have been a regular feature of every crypto downturn. We mention it because it exists, not because we recommend it for newcomers.
Freelancers, founders, and contractors with irregular income use crypto loans as a flexible credit line. Instead of forced sales during dry months, they tap into borrowing power and repay when income returns. This is a less risky use of leverage because the borrowed amount is calibrated to known future cash flow, not to a speculative thesis.
An investor who is over-weight bitcoin but bullish long-term might borrow against BTC to buy other assets — say, equities or real estate — without selling the underlying position. The strategy concentrates risk (you now have leveraged exposure to two asset classes) but allows diversification without realizing gains. It's a legitimate use case in moderation.
Crypto loans aren't universally good or bad. They're a tool that fits some situations and not others. Use the following decision tree honestly.
If yes, continue. If no, a crypto loan probably isn't relevant — you'd be buying crypto only to borrow against it, which adds layers of risk and cost. There are usually simpler ways to access dollars.
Borrowing to consume something productive (renovating a property, paying down a higher-cost debt, funding a business) is more defensible than borrowing for speculation. If you're borrowing to lever up further into crypto, understand that you're now exposed twice — once on collateral price, once on whatever you bought with the proceeds.
If a 30% drop in BTC would force you to sell your house to top up collateral, your LTV is too high or your loan is too big. Healthy borrowers can weather a 40–50% drawdown without panic.
Crypto-loan rates and terms vary widely across venues. A 1.5% difference in APR over a year on a USD 100,000 loan is USD 1,500 — enough to matter. Aggregators like Borrow by Sats Terminal exist precisely to compress this comparison into one screen.
Whether DeFi or CeFi, you should be able to explain in your own words who has custody of your collateral, what triggers liquidation, what happens to your collateral if the venue fails, and what your recourse is. If you can't, slow down.
If you answered all five honestly and a crypto loan still fits, you're in the population this product was designed for. The 2025 complete guide to bitcoin borrowing and the full breakdown of how crypto lending works are the next steps for going deeper. A useful exercise before borrowing for the first time is to pick a small test amount — a few hundred dollars — and run the entire loop end to end. Deposit, borrow, hold for a week, repay, withdraw. The point isn't the money; it's the learning. By the time you do a serious loan, the mechanics should feel routine rather than novel, which is exactly when financial decisions should be made.
Borrow by Sats Terminal is an aggregator built for one specific use case: borrowing stablecoins against bitcoin without selling. It compares loan offers across DeFi protocols (Aave v3 and Morpho Blue) and selected CeFi lenders, presenting the most competitive terms in a single interface. Behind the scenes, Borrow handles bridging and wrapping automatically — your native BTC becomes wBTC, BTCB, or cbBTC depending on which destination chain offers the best rate, and reverses the process when you repay.
The product is self-custodial throughout. There's no central account holding your bitcoin; instead, a Privy wallet is created from your email and acts as the signer for transactions you authorize. There's no KYC at the protocol level, and Borrow itself doesn't take custody of any user funds. Supported chains include Ethereum, Arbitrum, Base, Polygon, Optimism, and BNB Chain. Borrowable assets are USDC primarily and USDT on selected chains.
What this means for a typical user: you connect with email, deposit BTC, see a list of comparable offers ranked by effective rate, choose one, and walk away with USDC. Repayment runs in reverse with the same simplicity. The what is Borrow by Sats Terminal FAQ, the what are bitcoin-backed loans FAQ, and the how bitcoin-backed loans work learn page together provide the full operating manual.
Common Questions
The two phrases are often used interchangeably but technically describe opposite sides of the same market. A crypto loan is taken from the borrower's perspective: you pledge collateral and receive borrowed funds. Crypto lending is taken from the supplier's perspective: you deposit assets into a pool and earn interest from borrowers. The same protocol facilitates both — one user's borrow is another user's supply. In casual writing, "crypto lending" sometimes broadens to include both sides, which is why this distinction occasionally creates confusion. When in doubt, ask whether the speaker is the one paying interest or the one earning it.