Multi-Interval OLS Regression Analysis on the Uncovered Interest Rate Parities of Bitcoin and Ethereum
We are confident that the short-term lending rates for Bitcoin and Ethereum both follow an Uncovered Interest Rate Parity (UIP); therefore, as a consequence, they vary over time and converge with their ‘‘dominant’’ currency.
Short-term interest/lending rates of any arbitrary commodity/asset is highly correlated with it’s price volatility, liquidity, market cap, as well as market sentiment. However, these interest rates are generally capped at around 10% and 2% Annual Percentage Yield (APY) for retail and institutional borrowers respectively.
Lending rates for securities are typically aligned with the repo rates at around 2% APY, provided that other risks are omitted. For instance, stocks that are under the General Collateral borrow bracket are considered to be highly liquid and tolerant to volatility, and subsequently, have reasonable borrowing conditions (repo rates + fees). In addition, the lending rates of Bitcoin and Ethereum to institutions typically float at 2% APY, close to what we see in the repo markets. Conversely, Bitcoin and Ethereum have a much higher associated APY for retail-facing CFD brokers. However, unlike ordinary CFD swap contracts offered by traditional forex brokers, major cryptocurrency exchanges do not arbitrage on the interest rate differentials of their perpetual contracts. Instead, leading cryptocurrency exchanges such as Binance, OKex, BitMex, FTX, Bybit, and Bitflyer collectively charge almost 11% APY, in order to adjust the price differentials of their perpetual contracts to that of spot market prices. Similar lending rate practices are also commonplace among gold and silver loans at approximately 10% APY (excluding various associated transaction fees). Finally, the lending rates of highly liquid tokenized Bitcoin and Ethereum also yield sub-2% APY amongst many popular Decentralized Finance (DeFi) platforms.
The futures market highlights why interest rates on crypto assets among institutional borrowers are inherently lower than seen for traditional commodities; notably, the physical delivery option for EFPs (Exchange of futures for physicals) is much more efficient for cryptocurrencies, as opposed to commodity futures. This is predominantly due to the absence of delivery fees and logistical inefficiencies seen for crypto assets, as well as the uncovered risks associated with traditional commodities.
Dangers of commodity futures were highlighted in March 2020 when HSBC lost $200M, as a consequence of mistrading their gold delivery contracts. In addition, April 2020 saw the near term futures of West Texas Intermediate (WTI) crash by 300%, falling to a negative price (-55.90%, eventually settling at -$37.63 per barrel) for the first time since records began. This was due to logistical inefficiencies and an oversupply resulting from COVID-19. Therefore, the repercussions of commodity futures are enormous for institutions.
We can confidently conclude that whilst keeping factors such as liquidity, volatility and market cap constant, crypto assets have lower interest rates than traditional commodities.
By obtaining interest rate data for various commodities, fiat and crypto currencies over time, we will assess whether a mutual dependent relationship exists, to ultimately determine whether or not Uncovered Interest Rate Parity is present.
Co-author: Jack Warbrick
