Original author: @BlazingKevin_, Blockbooster researcher

1. There is no “base interest rate” for cryptocurrency

Leverage and financing in the cryptocurrency space—trillions of dollars in leveraged positions, mortgages, and income products—does not have a unified base rate curve.

According to the derivatives report for the first quarter of 2026 released by BitMEX, in the emerging "traditional asset perpetual contract" field alone, the average weekly trading volume soared from approximately US$525.8 million at the end of 2025 to US$30.7 billion in mid-March 2026, a quarter-on-quarter increase of approximately 5756%. Its monthly trading volume has also surged from US$7.9 billion in November 2025 to US$199.1 billion in March 2026, an increase of approximately 25 times in five months.

According to a 30-day snapshot provided by DefiLlama, Hyperliquid’s perpetual contract trading volume is approximately $172.63 billion, with open interest at approximately $9.13 billion. Commodity perpetual contracts accounted for approximately 30% of Hyperliquid’s open interest in the first quarter of 2026, driven primarily by demand for around-the-clock crude oil trading.

About the "TradFi Perpetual Contract" product line: Binance launched the TradFi Perpetual Contract on January 8, 2026. The first assets listed include gold (XAUUSDT) and silver (XAGUSDT). With its first-mover advantage, Binance accounts for approximately 62.7% of the TradFi perpetual contract market, followed by Hyperliquid with a 29.7% share.

Hyperliquid’s index data for these traditional asset perpetual contracts comes from a partnership with S&P Global, and this collaboration – which links crypto perpetual contracts directly to traditional indices – is drawing regulatory scrutiny from the U.S. Commodity Futures Trading Commission.

Meanwhile, Ethena’s U.S. dollar market capitalization as of early June 2026 was approximately $4.5 billion to $5.9 billion.

These products are all marked with "interest rates" or "yields" - perpetual contracts have financing rates, lending protocols have lending annual interest rates, sUSDe has staking yields, and tokenized treasury bonds have coupons - but so far, the cryptocurrency field still lacks its own SOFR (Standard Orderly Lending Reference Rate). There is currently no widely accepted benchmark curve that can serve as an anchor point for pricing. Each exchange and each protocol is becoming a micro-finance market, each quoting its own price, but there is a lack of an open and credible reference system between them.

2. What constitutes a cryptocurrency’s “base rate”?

Let’s first look at three different interest rate comparisons:

• The first group: benchmark financing interest rate, product yield and derivatives implied interest rate. **The annualized yield (APY) of sUSDe belongs to the product yield, that is, the return to the holder; the perpetual financing interest rate belongs to the derivatives implicit interest rate, that is, the fees paid by both long and short parties to link the perpetual contract price to the spot price; **The benchmark financing rate should be used as a public reference for countless other products to be quoted and used for pricing. **Product yields and derivatives implied interest rates are not the benchmark interest rate, but the "downstream" data of the benchmark interest rate, which are the result of various premiums and structures superimposed on the benchmark interest rate. **

• Group 2: Overnight interest rates and regular interest rates. **The perpetual financing rate is settled every 1 hour or 8 hours and is essentially an overnight rate - it only reflects the cost of capital "from now to the next settlement point" and has no term structure. It can't tell you the price difference between "borrow for 30 days" and "borrow for 90 days." Just as SOFR itself is an overnight rate, it relies on the futures market to construct periodic SOFR with a term structure. Interest rates without a term structure cannot support any medium- to long-term fixed income market.

• Group 3: Actual lending rates and algorithmic/implied rates. Actual bilateral lending transactions (e.g., Bitfinex’s margin financing book, where actual lenders and borrowers are matched) and algorithmic utilization-based pricing (e.g., Aave, where interest rates are automatically calculated through a formula based on pool utilization) are two distinct price-generating mechanisms. The former is voted on by market participants holding real funds, and the latter is a curve written by the protocol designers.

From these three differences, we can derive the standards that the "qualification benchmark" should meet:

Based on real transactions, the underlying market is sufficiently broad and deep (difficult to manipulate by a single participant), governance-agnostic (no conflicts of interest between managers and pricing markets), ideally having a term structure (can support mid- to long-term pricing).

(SOFR is based on actual trading volume in overnight repurchase agreements backed by U.S. Treasuries, which average daily trading volume "often exceeds $1 trillion." This is actual trading volume in overnight repos and is completely different from the notional trading volume in futures that underpins periodic SOFR.)

Using the logic of SOFR to examine cryptocurrencies, we can discover their structural isomorphism. In its study, the Bank for International Settlements likens the on-chain collateralized lending market to a “crypto-native currency market” that operates similarly to a traditional three-party repurchase agreement – ​​overcollateralization, mark-to-market, overnight rollover. Since on-chain lending is structured like a repo-like secured financing, it is an appropriate isomorphic reference to evaluate cryptocurrency benchmarks using the design of SOFR, a benchmark built on real repo transactions.

3. What are the characteristics of SOFR? Why was LIBOR discontinued?

The London Interbank Offered Rate (LIBOR) was once the cornerstone of global finance. At its peak, approximately $300 trillion in financial contracts (including interest rate swaps, mortgages, student loans, corporate bonds, etc.) in the five major currency areas relied on LIBOR. But LIBOR suffers from a fatal design flaw: It is not based on actual transactions but on daily “self-reported” borrowing cost estimates from a handful of banks.

After the 2008 financial crisis, this flaw was completely exposed. Regulatory investigations have revealed that traders at several large global banks systematically manipulated LIBOR quotes to profit from their derivatives positions.

The rigging scandal directly led to the abolition of LIBOR.

Replacing LIBOR is SOFR (Secured Overnight Financing Rate). The design of SOFR is almost a "reverse engineering" of all the shortcomings of LIBOR: it does not use self-reported valuations, but is based on real transactions in the U.S. Treasury repo market; it is taken from the volume-weighted median of three repo markets (tri-party repo, GCF repo and bilateral repo cleared through FICC's DVP service), thereby providing broad and deep coverage and effectively preventing manipulation by a single participant; it is managed by the Federal Reserve Bank of New York and follows IOSCO principles. There is no conflict of interest between managers and pricing markets.

However, SOFR has an "inherent flaw": it is an overnight rate with no term structure. The market not only needs "the overnight cost of the day" but also "expected financing costs in the next three months" to price medium and long-term loans. As a result, CME Group launched CME Term SOFR – a set of forward-looking rates covering 1-month, 3-month, 6-month and 12-month maturities.

It uses SOFR futures trading data to reversely calculate the market's expectations for future SOFR trends, thereby "constructing" a forward-looking term curve. (Representative notional trading volume in SOFR futures used to construct term SOFR was approximately $2.3 trillion per day in the fourth quarter of 2023.)

4. Some candidate interest rates for discussion

There are many candidates for interest rates known as "rates" or "yields" on the market. Let’s look at it one by one; the focus of the discussion can be on why some rates are clearly not suitable as benchmarks and which rates have room for improvement.

One thread that runs through all analyzes is - "Who has the right to make decisions?": Is it market weight, algorithm use, or governance mechanisms?

4.1 Perpetual Funding Rate (Hyperliquid / Binance)

The perpetual financing rate is the implicit price of leverage, determined by the basis between the spot price and the perpetual price: it is essentially an overnight rate with no term structure.

When the spot market of the underlying asset of Transactional Finance (TradFi) is closed (for example, stocks and precious metals are closed on weekends), the exchange cannot obtain the real spot price to calculate the funding rate. Binance’s approach is to freeze the index price at the last spot price and switch to an exponentially weighted moving average (EWMA) mark price with a volatility cap of ±3%; Hyperliquid also switched to EWMA over the weekend and set a volatility cap for each product. During the market break, the "anchor" of the perpetual contract price is essentially a predicted value, not the actual trading price. A price limit occurs when the market reopens and actual prices gap above that limit. Therefore, the price during the market break is a forecast value and not a true anchor value for the arbitrage trade.

On May 29, 2026, the U.S. Commodity Futures Trading Commission (CFTC) approved KalshiEX’s Bitcoin Perpetual Contract (BTCPERP), which is the first truly regulated Bitcoin perpetual contract in the United States. At the same time, the CFTC issued a policy statement on perpetual contracts, staff guidance on round-the-clock trading and clearing, and a position of taking no action against Coinbase’s offering of perpetual contracts through Deribit. The significance is that a regulated and centrally cleared perpetual contract means that its funding rate and basis are generated in a compliant and clearing-capable environment - which may become a candidate for a future "cryptocurrency SOFR". In addition, the CFTC’s previous review of Hyperliquid’s cooperation with S&P Global Indexes also showed that “regulation is moving closer to the cryptocurrency benchmark.”

4.2 Bitfinex Margin Financing + FRR

This is the native USD term funding market for cryptocurrencies.

The mechanics are as follows: Bitfinex operates a peer-to-peer margin financing market, where lenders provide funds to margin traders in order to earn interest. The key design is that financing terms range from 2 to 120 days (typically 2, 7, and 30 days), and matching requires that both interest rate and term match. This means that Bitfinex’s funding book naturally constitutes a real lending curve from short to long term: 30-day funds and 120-day funds are priced at different prices, determined by the actual supply and demand match. This is one of the very few real lending markets in the cryptocurrency space that has a term structure natively.

And the FRR (Flash Rate of Return) is the reference rate for this market: FRR is the average rate across all active fixed-rate financings, weighted by their size, and updated hourly. Essentially, it is “Bitfinex’s version of the Benchmark Reference Rate” – an index that reflects the current average borrowing costs in the market. Lenders can directly choose to lend at FRR, allowing their interest rates to automatically follow market changes.

Bitfinex charges approximately 15% on lending revenue (18% on hidden orders); minimum order amount is $150. FRR is quoted at a daily rate and annualized based on that daily rate: Bitfinex USD FRR is around 0.0136%/day, which is around 5.1% annualized – on the same level as candidates such as Tokenized Treasury, Aave, and SSR.

The key is its volatility: Historical annual interest rates on U.S. dollar loans have fluctuated between 3% and 20%, closely tied to the need for leverage.

This daily interest rate curve spans different maturities from 2 days to 120 days, forming a native USD funding curve with the actual term structure of the cryptocurrency.

Bitfinex and Tether belong to the same parent company, iFinex, and have overlapping management. This allows Bitfinex to have the most abundant USDT liquidity in the entire cryptocurrency field - which is one of the reasons why its financing market is so developed; but it also concentrates counterparty risk and stablecoin issuer risk in the same system. Borrow from Bitfinex, use Bitfinex's Tether placement service, and in extreme cases be backed by the same parent company - it's a highly self-consistent structure.

Although Bitfinex’s funding market is the oldest and largest native U.S. dollar term funding market in cryptocurrency, its absolute size (size of funding order book and daily matching volume) is still much smaller than the trillions of dollars of trading volume in the perpetual market mentioned above.

Compare FRR with LIBOR and SOFR: In terms of the dimension of "based on real transactions", FRR is actually clearer than LIBOR. FRR is calculated based on actual execution of fixed-rate financing transactions and weighted by size, reflecting real market behavior. However, FRR arises from the order book of a single exchange (concentration), operated by the same parent company iFinex, which also controls the largest stablecoin, Tether (conflict of interest), and which operator is also the lender of last resort in its own market (further concentration and conflict of interest). Therefore, in the two dimensions of concentration and conflict of interest, FRR touches on exactly the problems that SOFR aims to eliminate.

4.3 DeFi lending interest rate (Aave/Morpho)

This is a typical example of algorithm-based utilization pricing: the interest rate is not determined through bilateral matching, but is automatically calculated by a preset formula based on the utilization of the fund pool - the higher the utilization, the higher the interest rate. Interest rates fluctuate in real time based on borrowing demand.

The Aave mainnet USDC deposit interest rate fluctuates with usage, roughly between 3.5%–6%; the USDC vault hosted on Morpho has a yield of approximately 5%–7% after deducting custody fees.

4.4 MakerDAO / Sky Savings Rate (DSR of DAI / SSR of USDS)

This is a "quasi-policy rate" set directly by protocol governance. DAI’s DSR (Dai Savings Rate) and USDS’s SSR (Sky Savings Rate) are widely cited and function similarly to policy rates set by central banks – they are not triggered by market matching or algorithms, but rather by Sky’s governance votes.

DSR/SSR’s governance settings, FRR’s market weights, and Aave’s algorithmic utilization represent a comparison of three fundamentally different rate generation mechanisms.

Governance mechanisms, market weighting and algorithmic use - each of these three mechanisms has credibility issues and manipulation risks, and the benchmark for mature markets should ideally come from the mechanism least susceptible to manipulation (market-weighted real transactions with sufficient breadth and depth). As far as the current value is concerned, due to the adjustment of the governance mechanism, the SSR (USD Depository Receipt) has been reduced from 4.75% at the end of April 2026 to about 3.6%-3.75% at the beginning of June (the adjustment of the "governance mechanism" is in sync with the Federal Reserve's policy path); the circulating supply of USDS is approximately US$11 billion.

4.5 Tokenized Treasury Yield (BUIDL / BENJI, etc.)

The yield rate of this part is about 4% to 5%, which belongs to the "risk-free part" and meets the standards of the "cryptocurrency risk-free benchmark". Tokens such as BlackRock’s BUIDL and Franklin Templeton’s BENJI put the coupon yield of U.S. Treasury bonds on the chain. Calculated at current prices, the annualized yield of major Treasury tokens (BUIDL, USDY, USDM, USYC, etc.) in April 2026 is approximately 4.1% to 4.7%, which is closely related to the 3-month U.S. Treasury bond yield. Its yield can almost directly serve as a benchmark for traditional risk-free rates.

The secondary market pricing of the "risk-free portion" of tokenized treasury is very concentrated - taking Ondo's tokenized treasury as an example, between February and April 2026, the median deviation of its transaction price was only about 2 basis points, and 95% of the transaction prices fell within 5 basis points. This shows that when the underlying asset is sufficiently standardized and risk-free, on-chain price discovery can be very accurate; in contrast, the "price" of high-risk instruments such as perpetual contracts during non-trading hours is highly speculative - the lower the risk, the closer the price is to the true price; the higher the risk, the more like guessing the pricing.

4.6 Athena

This is a securitization product that combines perpetual financing rates and collateral yields. Its annualized rate of return is highly dependent on the level of perpetual market financing interest rates, so it is essentially a repackaging of the implicit interest rate, rather than the base interest rate itself.

Compare these seven candidate indicators together: they each measure a different indicator (leverage sentiment, actual lending, algorithmic application, governance policy, risk-free coupon, institutional arbitrage).