As we do each quarter, we would like to share our perspective on the digital asset lending market to provide transparency, insight, and trends to our counterparties.
At the end of 2018, we had crossed $1 billion in total originations with an active loan book of $153 million. Since then, we originated an additional $425 million in Q1, bringing our total originations to $1.53 billion since launching the lending business in March 2018. Our current active loan book rose to $181 million at the end of Q1, a 17% increase from the end of 2018.
Based on the first few weeks of April, we expect to see sustained growth in the lending market in Q2. As crypto prices have rallied, we have seen an acceleration of crypto and cash loans, as well as new lending participants entering the market. It is promising to see increased access to financing in the marketplace which we believe fosters greater liquidity and market efficiency.
Quarterly Asset Composition
At the end of 2018 our loan book was still BTC dominant and that trend continues today with 68% of the loan portfolio comprised of BTC. There is consistent BTC borrowing demand among market-makers and high frequency trading firms that need working capital for arbitrage opportunities. We saw new firms with this profile enter the market in Q1 and foresee this trend continuing. As the data reveals, most BTC loans are originated for purposes such as hedging or market-neutral trading strategies, not for price speculation, a common misconception.
In Q4 2018, we began lending USD as a pilot program. Since then, we’ve seen significant growth in USD loans and today, USD lending represents 10% of our active loan portfolio. While we are still in the early stages of expanding our cash loan business, we believe there will be significant growth in the market this year.
Because we have been operating in this industry for several years and understand the liquidity and volatility characteristics of digital assets, we are happy to accept digital assets as collateral for cash financing. We believe this experience adds value to institutions looking for liquidity without having to sell and realize capital gains, miners looking to pay electricity bills without selling inventory, and hedge funds seeking to leverage long positions.
2019 Monthly Asset Composition
The story of intra-quarter asset composition is basically the same. We saw BTC loans continue to increase as a percent of the overall loan book, and altcoin loans mostly decrease. USD composition stayed roughly the same.
Shorting Dynamics: Avoiding the Squeeze
Originating over $425 million in loans this quarter, we saw a handful of interesting trends and patterns relating to short-selling and covering across different assets. Below, we take a closer look at the prominence of loans/returns tracked against BTC, LTC and ETH prices over Q1. The differences highlighted in the charts spur questions around market psychology, timing of bets, and access to critical information.
Due to the large rally across assets earlier this month, we also included data through April 15, 2019 in our analysis to paint a more complete picture and provide greater transparency.
“Prominence of ETH Loans” is defined as the number of ETH on loan in a given day divided by the maximum ETH on loan in Q1. The above figure plots Prominence of ETH Loans vs. ETH price. The light blue line is at 100% when we are lending out the maximum units of ETH for the quarter, and less at all other times. The metric serves to visualize when loans are issued and returned relative to all loan activity in the quarter.
Taking a closer look at the chart above, in early February we saw a large uptick in ETH short-interest when the price was $100, and loans outstanding ballooned to the highest level over the quarter. This interest fell 30% after the price ticked up on February 7 to $120. A week later, the price rallied again to $140, squeezing out another batch of shorts and bringing short interest to less than 60% of the maximum. Our borrowers likely covered shorts after price had already moved against their positions.
On the contrary, between March 20 and April 2, we saw a 33% decline in short interest well before the run up to the $180 level on April 15, implying borrowers got ahead of the rising market and decreased short positions just in time. As prices held into April 15, short-interest continued to fall and closed at the lowest levels YTD.
Like Figure 4, Figure 5 above shows Prominence of LTC Loans vs. LTC price over Q1. LTC rallied consistently throughout this quarter, providing a unique look at the psychology of short selling. In early February, as LTC rallied from $30 to $45, short interest increased from 50% prominence to 75% prominence. However, most of those shorts were covered before the price rallied again on Feb 22 above $50.
In mid-March, shorts piled on to the highest levels all quarter as LTC hovered around $60, another level speculators believed to be a top. But that belief was not correct and yet again, the shorts covered before the price rallied swiftly to $90. Like clockwork, they reengaged after the price established itself in the $80-90 range. The data implies that as price rallies, shorts are added, yet these participants are disciplined enough to close before substantial subsequent rallies. Unlike ETH, LTC borrow returns preceded major rallies rather than trailed – indicating better information or better understanding of momentum in the asset.
Closing out the period, LTC short interest remains near the maximum for the quarter.
For reference, we included the same chart used for ETH and LTC in BTC to show the difference in use case. BTC borrow is used primarily for trading working capital, market neutral arbitrage strategies, remittance, and other non-speculative related purposes. The chart shows little correlation between prominence and price. BTC on loan has steadily increased throughout the quarter and we’re currently lending over 30,000 units of BTC.
Year-In Trends: What We’re Noticing at Our One-Year Anniversary
One of the common criticisms we receive is that we are fueling the naysayers, so-called “no coiners,” and bearish market participants by enabling large block OTC short-selling. However, we haven’t seen this type of counterparty since we launched the lending business in March 2018. Those participants who are shorting are often long something else, only short for a small window, and/or managing hedge funds that are long-term bullish on the space. In other words, Dimon, Buffett, and Roubini are not shorting. It is easy to say something is worthless but it is difficult to put capital behind that position and justify the risk of being net short in this market.
Rates, Competition, and Risk Adjusted Returns
Rates have compressed over time as lending matures and more supply enters the market. Since Genesis faces both borrowers and lenders principally, credit risk is a critical component of pricing rates. One rate certainly does not fit all, and we consider a multitude of factors when pricing, including credit, overcollateralization level, loan structure, duration, and more.
When we launched the business last year, few venues offered large spot digital asset borrowing. Since then, several platforms have emerged including decentralized protocol-based platforms that match lenders to borrowers on an agency basis (i.e. the protocol does not take any risk and simply crosses liquidity pools), as well as centralized exchanges and lending desks offering large scale borrowing.
This increased competition has naturally led to lower borrowing rates and tighter spreads, but we believe the speed at which rates are declining will slow as the return-to-risk moves closer to zero or negative. Rates can still be positive, but after adjusting for default risk, the loan might not be worth it to the lender. Each lender must consider its own required return relative to risk.
Collateralization level and counterparty credit are the two main drivers of the probability of default and when the expected default loss becomes greater than the expected return on the loan, it is no longer worthwhile. Because the credit profiles of borrowers vary greatly and the volatility of assets can change dramatically over time, it will likely take a while before digital asset lending rates become efficient.