The State of Crypto Leverage – Q3 2025 

Introduction  

Crypto-collateralized lending, reached a new all-time high at the end of the third quarter, largely driven by growth in onchain borrowing. While the long-held peak set in Q4 2021 has finally been eclipsed nearly four years later, the market today looks much different. For starters, onchain lending now occupies a much larger share of the total lending market; at the end of Q4 2021, onchain borrowing through lending applications such as Aave and collateralized debt position (CDP) stablecoins such as DAI made up 48.6% of the market. Today, these avenues own 66.9%. Within onchain borrowing, there has also been a notable shift in the balance between lending apps and CDP stablecoins. At the end of Q3 2025, lending applications accounted for more than 80% of the onchain market, with CDPs at just 16%, compared to 53% in Q4 2021. This is a notable shift away from synthetic, crypto-backed stablecoins towards the lending out of centralized stablecoins like USDT and USDC.  

Within centralized lending, we’ve seen a stark reversal of the conditions that prevailed during the last cycle. Un(der)collateralized lending has become far less common as CeFi lenders have chosen to self-regulate. After the 2022 credit implosions, the surviving firms recognized that opaque, relationship-based lending had become a reputational and commercial liability. Many are now pursuing public listings or more institutional sources of capital, which has forced tighter internal risk controls and full collateralization standards. Industry sentiment has also shifted, with uncollateralized lending frowned upon. The result is a smaller, more conservative CeFi credit sector where transparency and collateral quality (or its mere existence) are core to credibility. 

At face value, the new lending high should not be viewed with alarm; there is no magic outstanding borrows threshold that signals danger. What matters is the quality of the collateral base and the purpose of the borrowed funds. In both cases, this cycle looks much different from 2021. The previous high was defined by uncollateralized lending, speculative yield-chasing in short-lived protocols, and more volatile collateral assets. Today, collateralized lending is standard practice, the sector is more systemically visible (with onchain borrow share rising and public companies reporting their books), and collateral is more concentrated in BTC, ETH (which are both less volatile than in prior cycles), and stable, yield-bearing assets, such as Pendle Principal Tokens (PTs).  

Shortly after the third quarter ended, however, the futures market experienced its largest liquidation event in history. Over the course of just a single day on Oct. 10, more than $19 billion worth of perps positions were liquidated, and many more billion were auto deleveraged (ADL). While that event wiped out a massive amount of open interest, it is distinct from crypto-collateralized lending and not necessarily indicative of systemic overleverage. Prices fell so sharply that even modestly levered positions were forcibly closed, and many healthy accounts were swept up in cascading liquidations and ADLs triggered by exchange-level risk controls. The perps market got cleaned out, but largely by mechanical design and aggressive price decreases rather than widespread credit excess.

Key Takeaways  

• All told, crypto-collateralized lending expanded by $20.46 billion (+38.5%) in Q3 2025 to a new all-time high of $73.59 billion. This clears the previous all-time high of $69.37 billion at the end of Q4 2021 by $4.22 billion (6.09%). 

• The dollar-denominated value of outstanding loans on DeFi applications set another new quarter-end all-time high in Q3, growing by $14.52 billion (+54.84%) to $40.99 billion.  

• Galaxy Research is tracking more than $12 billion in debt outstanding used to directly buy or supplement the treasury strategies of DATs. The debt outstanding has been stagnant through most of the year, with just $422 million added in the third quarter. 

• Futures open interest (OI), including perpetual futures (perps), increased 41.46% QoQ from $132.75 billion to $187.79 billion on Sept. 30 before reaching an all-time high of $220.37 billion on Oct. 6. The perps unwind on Oct. 10 saw OI decline 30% overnight from $207.62 billion on Oct. 9 to $146.06 billion by end of Oct. 10. 

• The Oct. 10 liquidation cascade was the largest daily futures liquidation event in history, with long and short liquidations eclipsing more than $17 billion. Hyperliquid alone saw $10.08 billion in liquidations for the 24-hour period. Bybit and Binance followed, reporting $4.58 billion and $2.31 billion respectively. 

Crypto-Collateralized Lending  

The market map below highlights some of the major past and present players in the CeFi and DeFi crypto lending markets. Some of the largest CeFi lenders by loan book size crumbled in 2022 and 2023 as crypto asset prices tanked and liquidity dried up. These lenders are flagged with red caution dots in the map below. Since Galaxy’s last crypto leverage report, we have added one new DeFi app and six new instances of DeFi apps already covered in past reports.   The additional DeFi app is:    

• JupLend on Solana.  

  Existing apps that have expanded chain coverage include:    

• Aave on Plasma.  

• Fluid on Plasma. 

• Dolomite on Botanix.  

• Euler on Linea, Plasma, and Arbitrum.   

Centralized Finance (CeFi)  

The table below compares the CeFi crypto lenders in our market analysis. Some of the companies offer multiple services to investors. Coinbase, for example, primarily operates as an exchange but also extends credit to investors through over-the-counter cryptocurrency loans and margin financing. The analysis shows only the size of their crypto-collateralized loan books, however. 

As of Sep. 30, Galaxy Research tracked $24.37 billion of open CeFi borrows. This represents quarter-over-quarter (QoQ) growth of 37.11%, or $6.6 billion, and $17.19 billion (+239.4%) growth since the bear market trough of $7.18 billion in Q4 2023. Still, CeFi borrows outstanding are 34.3% below their Q1 2022 all-time high of $37.08 billion. 

 The third quarter saw record growth and loan book sizes for a number of the CeFi lenders:  

• TwoPrime had its best quarter ever by absolute loan book growth (based on self-reported figures), expanding its book $394 million. 

• Arch had its best quarter ever by absolute loan book growth (based on self-reported figures), expanding its book $49.29 million. 

• Galaxy had its best quarter ever by absolute loan book growth (based on publicly disclosed financials), expanding its book $693 million. 

• Maple had its best quarter ever by absolute loan book growth (based on onchain data), expanding its book by $629.9 million. 

• Tether had its best quarter ever by absolute growth in its secured loans balance (based on its attestation report), expanding its book by $4.47 billion. 

 Tether, Nexo, and Galaxy are the top three lenders Galaxy Research tracks by outstanding loan values. As of Sept. 30, Tether maintained $14.6 billion of open loans, Nexo $2.04 billion, and Galaxy $1.8 billion. 

Tether is the dominant lender in our analysis, commanding a 59.91% share (up 289 basis points from last quarter) of the CeFi lending market. Add in Nexo (8.36% market share and down 265 basis points from last quarter) and Galaxy (7.38% market share and up 115 basis points from last quarter), and the top three tracked CeFi lenders control 75.66% of the market (up 140 basis points from last quarter). 

When comparing market shares, it’s important to note the distinctions between CeFi lenders. Some lenders only offer certain types of loans (e.g., BTC-collateralized only, altcoin-collateralized products, and cash loans that do not include stablecoins), only service certain types of clients (e.g., institutional vs. retail), and only operate in certain jurisdictions. The combination of these factors allows some lenders to scale more easily than others.  

The table below details the sources of Galaxy Research’s data about each CeFi lender and the logic we used to calculate the size of their books. While DeFi and onchain CeFi lending figures are retrievable from onchain data, which is transparent and easily accessible, retrieving CeFi data is tricky. This is due to inconsistencies in how CeFi lenders account for their outstanding loans and how often they make the information public, as well as the general difficulty of obtaining this information.  

Note: the values supplied by private third-party lenders have not been formally vetted by Galaxy Research. 

CeFi and DeFi Lending 

The dollar-denominated value of outstanding loans on DeFi applications set another new quarter-end all-time high in Q3, growing by $14.52 billion (+54.84%) to $40.99 billion. Combining DeFi apps with CeFi lending venues, there were $65.37 billion of outstanding crypto-collateralized borrows at quarter-end. This represents an expansion of $21.12 billion (+47.72%) QoQ, predominantly driven by the growth in open borrows across DeFi lending apps. On a combined basis, DeFi lending app and CeFi loans outstanding reached a new all-time high that eclipses the previous high of $53.44 billion in Q4 2021 by $11.93 billion (22.32%). 

Galaxy Research sees a few contributing factors to the continued growth in DeFi lending: 

• Points farming: airdrop farming and points programs incentivize users to keep borrows on even when they feel stress from market conditions or the economics of their loans. Such programs can be found across a number of applications and chains. 

• Improved collateral assets: collateral assets like Pendle PTs allow users to loop stablecoin strategies at favorable loan-to-values (LTVs), and thereby leverage efficient collateral assets that scale with stablecoin growth. 

• Price going up: borrow activity is reflexive to asset prices, as users can borrow more against the same collateral as price appreciates. BTC traded slightly higher through the third quarter, ETH appreciated ~50% and reached a new all-time high for the first time since November 2021, and SOL traded ~33% higher. 

Note: There is potential for double-counting between total CeFi loan book size and DeFi borrows. This is because some CeFi entities rely on DeFi applications to lend to offchain clients. For example, a hypothetical CeFi lender may pledge its idle BTC to borrow USDC onchain, then lend that USDC to a borrower offchain. In this scenario, the CeFi lender’s onchain borrow will be present in the DeFi open borrows and in the lender’s financial statements as an outstanding loan to its client. The lack of disclosures or onchain attribution makes filtering for this dynamic difficult.  

As a result of the quarter-over-quarter growth in outstanding borrows on DeFi lending applications, their relative lead over CeFi lending venues rebounded to a new all-time high. At the end of Q3 2025, DeFi lending app dominance over CeFi lending venues stood at 62.71%, up from 59.83% at the end of Q2 2025 and up 72 basis points from the Q4 2024 high when the share was 61.99%. 

The third leg of the stool, the crypto-collateralized portions of collateral debt position (CDP) stablecoin supply, decreased by $658 million (-7.4%) QoQ. Again, there is potential for double-counting between total CeFi loan book size and CDP stablecoin supply, because some CeFi entities might rely on minting CDP stablecoins with crypto collateral to fund loans to offchain clients. 

All told, crypto-collateralized lending expanded by $20.46 billion (+38.5%) in Q3 2025 to a new all-time high of $73.59 billion. This clears the previous all-time high of $69.37 billion at the end of Q4 2021 by $4.22 billion (6.09%). 

At the end of Q3 2025, DeFi lending applications represented 55.7% (+588 basis points from Q2 2025 share) of the crypto collateralized lending market, CeFi venues captured 33.12% (-36 basis points from Q2 2025 share) of the market, and the crypto-collateralized portion of CDP stablecoin supplies held 11.18% (-547 basis points from Q2 2025 share). Combining DeFi lending apps and CDP stablecoins, onchain lending venues held a 66.88% dominance (+33 basis points from Q2 2025 share) over the market, up two basis points from the all-time high of 66.86% at the end of Q4 2024. 

Additional Views of DeFi Lending   

Outstanding borrows on DeFi lending applications have remained resilient in the wake of turbulent market conditions. After setting a daily all-time high in loans outstanding at $43.82 billion on Oct. 7, loans outstanding have declined just $5.06 billion (11.55%) as of Oct. 31.  

Notably, the Plasma blockchain has attracted meaningful lending activity in just five weeks since launch, garnering more than $3 billion in outstanding borrows as of Oct. 31. The Aave protocol has a significant share of the lending market on Plasma, capturing 68.8% of all borrows on the chain as of Oct. 31. Plasma is now Aave’s second largest instance, dethroning Arbitrum from its long-held No. 2 spot. 

Liquidations 

The Oct. 10 auto-deleveraging (ADL) “armageddon,” as some X accounts hyperbolically called it, saw elevated levels of liquidations across lending applications. Aave V3 Core on Ethereum experienced its third highest day of liquidations ever at $192.86 million, lagging only behind the Yen carry trade unwind in August 2024 and the tariff tantrum that sent crypto prices tumbling in February 2025. Wrapped bitcoin tokens were the most liquidated asset type on Oct. 10, with $82.17 million worth liquidated. 

Kamino also experienced an elevated volume of liquidations from Oct. 8-10. Over this period $12.7 million worth of WSOL was liquidated and $2.6 million worth of FARTCOIN was liquidated. 

There was an additional $17.38 million of liquidations on Fluid (across five chains) and $27.7 million of liquidations on Morpho (across 12 chains) during the Oct. 10 leverage unwind. 

Onchain and Offchain Interest Rates  

The following compares borrow rates for stablecoins, BTC, and ETH in onchain lending markets and through offchain venues.  

Stablecoins   

The weighted average stablecoin borrow rate increased slightly from 4.23% on June 30 to 4.83% as of Oct. 31. This figure is calculated by blending the costs of borrowing from lending protocols and CDP stablecoin mint fees, weighted by outstanding borrows. 

The following breaks out the costs of borrowing stablecoins through lending applications and minting CDP stablecoins with crypto collateral. The two rates track each other closely, although CDP stablecoin mint rates are typically less volatile because they are manually set periodically and do not move in lockstep with the market. 

Benchmark over-the-counter (OTC) interest rates for USDC increased over the quarter, climbing from 5% on June 30 to 6.5% as of Sept. 30. Since then, however, the rate has declined slightly to 6%. The onchain rate for USDC tracked closer to the bottom (lowest risk) band of the OTC rate through the quarter and did not exceed the benchmark mid-rate through Q3 using the weekly average rate. 

The chart below tracks the same rates as above, but for USDT lending. The onchain rate for USDT has been more in line with the OTC mid-rate than that of USDC. 

Bitcoin  

The chart below shows the weighted borrow rate for wrapped bitcoin (WBTC) on lending apps across several applications and chains. The cost of borrowing WBTC onchain is often low because wrapped bitcoin tokens are primarily used for collateral in onchain markets and are not in high demand for borrowing. In contrast to stablecoins, the cost of borrowing BTC onchain remains stable, because users borrow and repay it less frequently. The rate to borrow BTC onchain declined from 0.28% at the end of Q2 to 0.22% at the end of Q3. 

The historical divergence between onchain and offchain (over-the-counter) borrow rates for BTC persisted throughout the third quarter, though offchain BTC borrow rates came down substantially. In the OTC market, BTC borrowing demand is driven primarily by two factors: 1) the need to short BTC and 2) the use of BTC as collateral for stablecoin and cash loans. The former is a source of demand not commonly found in onchain lending markets, hence the spread between onchain and over-the-counter BTC borrow costs.  

OTC rates for BTC declined 175 basis points over the quarter from 3% to 1.25% as of Sept. 30. 

ETH and StETH  

The chart below shows the weighted borrow rate for ETH and stETH (staked ether on the Lido protocol) on lending apps across several applications and chains. The cost of borrowing ETH has historically been higher than that of stETH because users borrow ETH to fuel looping strategies to gain leveraged exposure to the Ethereum network staking APY – using stETH as collateral. As a consequence, the cost of borrowing ETH typically fluctuates within 30-50 basis points of the Ethereum network staking APY. This strategy becomes uneconomical when the cost to borrow exceeds the staking yield, so it is uncommon for the borrow APR to clear the staking APY for extended periods of time. However, there was a large spike in ETH borrow rates in July as large withdrawals were made from Aave V3 Core on Ethereum. More details on the impact of this will be covered below.    

As with WBTC, the cost of borrowing stETH is often low because the asset is primarily used as collateral and not much else.  

By using LSTs or liquid restaking tokens (LRTs), which earn yield, as collateral, users secure ETH loans at low, often negative, net borrow rates. This cost efficiency fuels a looping strategy where users repeatedly use LSTs as collateral to borrow unstaked ETH, stake it, and then recycle the resulting LSTs to borrow even more ETH, thereby amplifying their exposure to the ETH staking APY. This strategy only works so long as the borrow cost for ETH is below the staking APY achieved on stETH. Most of the time, users have been able to conduct this strategy without a hitch. However, from July 15-25, Aave V3 Core on Ethereum saw a flight of nearly 300,000 ETH. This sent borrow rates for ETH soaring, thereby making the looping strategy unprofitable (denoted by the net rate in the chart below persisting above 0%). Since then, the ETH rate environment has stabilized, and the looping strategy has become profitable to run again. However, traders appear to be looping at a smaller magnitude with the amount of ETH borrowed on Aave V3 core shrinking from ~2.3 million ETH in mid-July to ~2.1 million ETH as of Nov. 4. 

ETH Over-the-Counter Rates  

As with bitcoin, borrowing ETH through onchain lending apps is noticeably cheaper than borrowing it over the counter. This is driven by two factors: 1) as with BTC, there is demand from short sellers through offchain venues that is not as common onchain and 2) the Ethereum staking APY serves as a floor rate for offchain borrowing because there is little incentive for suppliers to deposit assets with offchain venues, or for offchain venues to lend assets out, at rates below the staking APY. So, with ETH, the floor rate for offchain lending is often the staking APY while onchain the staking APY is often the ceiling rate held together by loopers.  

Corporate Debt Strategies  

Digital asset treasury (DAT) companies were, again, a primary focus through the earlier part of the third quarter. The differentiating factor in the previous quarter was the proliferation of non-bitcoin related DATs. Ethereum DATs, such as SBET and Bitmine, gained greater mindshare as they began executing their treasury strategies and the Solana DAT canvas expanded with the launch of Forward Industries. 

Since the last report we have made some amendments to the logic defining what a DAT is. To get the most accurate representation we now define DATs as companies whose main business is not necessarily crypto-related but has a strategy to put crypto assets on their balance sheet. Given this amendment and greater access to public company data, the company mix used in the analysis is different from previous reports. Nonetheless, we are still tracking $12.055 billion in debt outstanding used to directly buy or supplement the treasury strategies of DATs. The debt outstanding has been stagnant through most of the year, with $422.5 million added in the third quarter. Strategy still carries the most debt at $8.214 billion (excluding perpetual debt). Strategy’s outstanding debt remains unchanged QoQ. 

The timeline for the earliest maturity, call, or put date of DAT issued debt has changed with the updated logic. Most of the DAT debt is due, at the earliest, between June 2027 and October 2028. 

The effective quarterly interest due from Strategy and Semler Scientific remains unchanged from the previous report, while the new additions, VFH Parent and DeFi Development Corp., owe $68.98 million and $1.95 million respectively. 

Total debt outstanding, including the debt taken on by DATs, reached an all-time high of $86.26 billion in the third quarter. This represents relative growth of 31.33% QoQ and $20.58 billion of absolute growth. The new all-time is $13.59 billion higher than the previous high set in Q4 2021 of $72.67 billion. 

Futures Market  

Futures open interest (OI), including perpetual futures (perps), increased 41.46% QoQ from $132.75 billion to $187.79 billion on Sept. 30 before reaching an all-time high of $220.37 billion on Oct. 6. The perps unwind on Oct. 10 saw OI decline 30% overnight from $207.62 billion on Oct. 9 to $146.06 billion by Oct. 10 market close. Over the course of Q3:  

• Bitcoin futures open interest increased $9.41 billion (+14.45%) to $74.52 billion and stood at $77.82 billion as of Oct. 29. 

• Ethereum futures open interest increased $22.37 billion (+76.61%) to $51.52 billion and stood at $44.78 billion as of Oct. 29. 

• Solana futures open interest increased $6.05 billion (+91.94%) to $12.63 billion and stood at $11.52 billion as of Oct. 29. 

• The futures open interest of all other crypto assets increased $17.21 billion to $49.12 billion (+53.93%) and stood at $30.68 billion as of Oct. 29. 

It’s important to understand that total futures open interest headline figure doesn’t tell you whether traders are using 2× leverage or 100× leverage, nor does it reveal how much of the open interest comes from directional bets versus hedged, delta-neutral structures. Some portion of OI is created by traders who hold the underlying asset and short perps or futures to hedge it. These positions inflate open interest without adding meaningful directional risk, so the raw OI figure alone says little about how aggressively levered the market actually is.

The Oct. 10 liquidation cascade was the largest daily futures liquidation event in history, with long and short liquidations eclipsing more than $19 billion. Hyperliquid alone saw $10.27 billion in liquidations for the 24-hour period. Bybit and Binance followed, reporting $4.6 billion and $2.3 billion respectively. 

CME’s share of open interest, including perps and non-perps, stood at 14.4% on Sept. 30; down 109 basis points QoQ. CME’s share of total Ethereum OI (calculated as CME Ethereum OI divided by total market OI) stood at 17.28% as of Sept. 30 (up 651 basis points QoQ); and CME’s share of total bitcoin OI stood at 20.78% as of Sept. 30 (down 554 basis points QoQ). The Oct. 10 leverage unwind sent the CME’s share of futures OI 398 basis points higher overnight from 14.78% to 18.76%. 

Perpetual Futures  

Perps OI stood at $147.55 billion as of Sept. 30, growing by $38.62 billion (+36.05%) from the end of Q2. As of Oct. 29, perps OI was sitting 27.14% below its all-time high of $173.41 billion reached on Oct. 6. Bitcoin perps’ market share stood at 45.15%, Ethereum 24.24%, Solana 7.47%, and all others 23.15%. 

Perps OI dominance was 76.67% as of Oct. 29 and has declined 538 basis points from the end of Q2. The liquidation cascade on Oct. 10 sent perps dominance tumbling 442 basis points overnight as users were liquidated and ADL’d while non-perpetual futures remained steady. 

Conclusion 

Loans outstanding, futures open interest, and DAT debt all reached new highs in Q3 2025, but the composition of leverage looks different from past cycles. Growth was driven primarily by collateralized onchain lending and CeFi lending rather than by unbacked credit and highly speculative strategies. 

While the Oct. 10 liquidation cascade exposed the scale and reflexivity of derivatives leverage, it did not reveal systemic credit risk. Most forced unwinds were mechanical, not the result of insolvency. Still, the event highlights that leverage remains deeply embedded in market structure, even if its transmission channels have changed.  

Overall, crypto leverage is expanding again, but under tighter collateral standards, greater transparency, and clearer separation between credit and speculation. The system is not risk-free, but it is built on a stronger base than it was in 2021.