Pricing Loans on Cryptocurrency Assets

The early stage of the cryptocurrency lending market has dramatically changed in the short time since the birth of the industry. These changes were driven by a combination of digital asset volatility, increased competition, an influx of financial products, and correlation to traditional market events. Investors and lenders are still struggling to find equilibrium pricing on interest rates against the backdrop of high price volatility and reduced demand during the protracted cryptocurrency bear market. Below, we review traditional loan pricing fundamentals and extrapolate what lessons can be applied to this dynamic, emerging asset class.

Loan Fundamentals Review1

Pricing

Loans are typically priced as a percentage of par value (per $1,000 in most cases). Discounts reflect percentages less than 100% and premiums reflect percentages greater than 100%. Discounts to par indicate that the buyer of the security requires some level of return when they are repaid their principal balance. For example, a buyer of a loan may pay 99% of par ($990) to get $1,000 at some point in the future when they are repaid by the borrower. This specific structure would reflect a return of 1% ($1,000 / $990 – 1) over that period. If the yield of 1% is generally attractive for this type of loan, it would be considered the yield to maturity; i.e., the expected rate of return. If the rate is too low, lenders will not underwrite the loan. If the rate is too high, competition will quickly bring prices down.

The yield to maturity (YTM) is considered the ‘discount rate’ for loans. Generally speaking, the lower the YTM, the lower the risk. The ‘risk free rate’ represents a theoretical return with zero credit risk. In the US, US Treasury rates are typically used as a proxy2. The difference in the YTM for other assets (assuming securities other than Treasuries have greater than 0 risk) above the risk-free rate is considered the credit spread (i.e., credit spread = YTM – risk free rate). 

Coupon Rate

Often confused with the YTM, the stated coupon rate is the interest income paid to the lender by the borrower. The payment can be made in multiple forms:

  • Cash – Reflected as a periodic payment paid as a percentage of the total principal value (i.e., the loan amount) to the lender. For example, an 8% coupon rate on a loan of $1,000 would reflect an $80 payment.
  • Payment-in-Kind – The payment is instead reflected as compounded interest and paid lump-sum on exit. In the example above, the $80 would equal $1,080 total balance due to the investor. These structures typically demand a premium to a cash payment structure given they are utilized by issuers with imminent cash flow needs.
  • Interest Only – Only interest is paid to the lender until the entire principal balance is paid at exit (referred to as a bullet payment).
  • Principal + Interest – Each payment to the lender includes a portion of interest and a portion of the lent capital. This reduces the burden of paying a lump-sum payment at exit. This structure is typical with real estate mortgages.
  • Variable versus Fixed – Interest rates can fluctuate (floating-rate securities are typically used to hedge inflation risk) or be paid in the form of a fixed amount. Historically, the interest rate on these loans are expressed as a rate on top of a floating instrument (e.g., LIBOR, TIPs, etc.).
  • Hybrid – Each of these instruments above may be combined for exotic instruments to fit various needs.

Origination Fees

Fees paid to the lender (or the originator of the deal, I will keep it simple for this post and assume they are the same) for work performed to consummate the deal. The funds go to pay the lender for time spent preparing documents, marketing the loan, covering expenses for legal and accounting, etc. In complex deals, the origination fees are greater.

Origination fees can be paid in multiple forms. They can be paid up front, as a portion of the notional balance (e.g., original issue discounts), or built into some exit fee (usually as a multiple of the amount).

A portion of the interest is paid to a servicer. A servicer performs administrative duties (such as recordkeeping, recording the ledger, etc.) related to the loan. More complex instruments generally require a greater fee. Larger banks may have deals where a lender teams up with a servicer and structures as a revenue-share or a JV model. 

Prepayment Penalty

A penalty imposed by the lender in case there an is voluntary early prepayment of a loan. The fees are assessed on exit; for example, a 1% prepayment fee would mean the lender would need to return 101% of the principal balance if paid before the stated maturity. Exit fees can also be assessed, where a premium or discount is reflected in the final payment. The purpose of these structures is to protect the lender against early repayments, which negatively impacts their ability to find a replacement borrower at the same (or greater) interest rate.

Credit Ratings

A financial scoring system for issuers that provide standardization for risk analysis. S&P’s rating system, for example, goes from AAA down to C. Indices can be measures for loans with the same credit rating to get an average yield. Some of these indices can be observed on the St. Louis FRED system with daily reported yields and spreads3. The ratings typically reflect the probability of issuer default, although more recent models with the advent of P2P alternative lending may reflect a greater breadth of data points (although it is questionable as to which data sets contain statistically significant data with meaningful predictive power).

The intent of credit ratings is to standardize risk and allow for comparability. Typically, both the issuer and the actual loan receive a credit rating. No blockchain-based loan (nor issuer) currently has a credit rating associated with it.

Primary versus Secondary Markets

Primary markets involve the issuance of new securities to investors. Secondary markets involve the purchase and sale of securities after the borrower has sold its shares in primary markets. Secondary markets may reflect updated prices as the YTM changes based on a company’s credit rating over time, macroeconomic changes (such as a liquidity crunch), or demand for a particular issue. Primary market dealers may receive favorable pricing (up to a 50% discount in many cases) for providing liquidity.

Loan-to-Value Ratio

The Loan-to-Value ratio is expressed as the ration of the loan’s principal amount to the value of the firm or asset. For example, a company that takes out a loan of $500 and is worth $1,000 has an LTV of 50.0%. The inverse of the LTV (1 / 50% = 2x) represents the asset coverage ratio.

As the LTV increases, it becomes riskier for the lender and they may compensate by increasing the interest rate or adjusting the loan structure to acquire a greater rate of return.    

Loan Risk

There are four primary risk factors to consider as a lender:

  1. Interest Rate Risk – Loan prices are inversely related to market yields (the coupon rate certainly plays a part in determining price, but the coupon rate is distinct from a YTM). The impact of a shift in rate of return expectations are measured by factors such as duration and convexity (which estimate the magnitude of the change given a change in interest rate expectations). Also, when interest rates decline, lenders worry about the impact of prepayments on their portfolio.
  2. Inflation Risk – The risk that the interest you earn as a lender is less valuable considering future inflation. This risk factor helps us understand why a rate of return is expected even on ‘risk free assets’ such as T-Bills.
  3. Market Risk – Represents the systematic risk of the loan market (e.g., a liquidity crunch).
  4. Credit Risk – An asset specific measure, standardized risk buckets are estimated by credit rating agencies. This factor is more commonly referred to as default risk.   

In order to deal with the risk factors above, derivative contracts have become popular. These contracts are either embedded within loan structures (e.g., calls/puts) or traded in the form of OTC products (e.g., interest rate swaps or forward contracts). Borrowers utilize these contracts to either create more predictable cash flows (e.g., swapping floating for fixed payments) or speculating on future rates (e.g., swapping fixed for floating payments).  

Loan Valuation Methods

Loan valuation techniques are primarily based on cash flow analysis. Given that the structure of the cash flows is predetermined based on the credit agreement (as opposed to estimated in an equity valuation), modeling the cash flows is a relatively straightforward exercise. The key input used to value these loans is the discount rate (i.e., the yield to maturity). The YTM can be extrapolated based on the factors explained above; however, the general build-up to a discount rate will look like (not intended to be drawn to scale) the image below. Like the capital asset pricing model used to calculate the cost of equity, the cost of debt is built using the risk-free rate as the base. On top of that, layers of risk are added based on various, mutually-exclusive risk factors. In the chart below, the YTM is comprised of the risk-free rate, a premium for expected inflation, and an asset-specific risk factor. The combination of the tenor and the asset-specific risk is the ‘credit spread’.

The rates of return for bonds by credit rating are posted daily on the St. Louis Federal Reserve database3. In more private transactions, information on comparable yields are typically harder to acquire given investors benefit from the information asymmetry. However, some companies – such as SPP Capital – post average rates for recent transactions4. Either way, increasing yields (referred to as ‘widening’) point to lower credit quality and decreasing yields (referred to as ‘tightening’) point to improving credit quality.

The YTM deciphered for each asset represents an array of risk factors; however, the primary concern for any lender is default risk – i.e., 100% loss of capital.

Extrapolation to Blockchain-Based Loans

The initial model for blockchain-based loans has been focused on providing liquidity for popular cryptocurrency assets such as Bitcoin, Ether, and Litecoin. This conversation will focus on loan products targeted at these assets; however, I note the expansive universe of securitized assets will soon hit the industry. In determining an appropriate yield for blockchain-based loans, it is important consider a few factors:

First, this is a completely new industry. Granted that Bitcoin has been around for more than 10 years, the initial retail ‘splash’ happened late 2017. A true understand of market risk (from the standpoint of traditional financial theories) has not been developed.

Today’s popular blockchain-based loans are effectively margin loans, the same as those you would have access to in a typical brokerage account5. Interest rates are developed based on consideration of trading liquidity (which is exponentially deeper than cryptocurrency markets), algorithmic / high-frequency trading with sophisticated electronic networks, operating expense margins, and an appropriate margin to the cost of servicing the loan. Given that price volatilities today on crypto assets are 3-4x broad equity indices, the need for a lender to have high-quality OTC and exchange relationships with sophisticated trading functions are a priori.

Another consideration involves estimating supply & demand for various security structures, a process typically known as price discovery (i.e., the notorious invisible hand of the market). The market has evolved so quickly that APRs have declined from greater than 20% at the top of the market to below 10% within a year. This decline is the result of greater competition in the space, low interest rate pricing strategies, and further development of portfolio surveillance and monitoring systems. At some point in the future, capital pricing may begin to reflect equilibrium. Until then, pricing competition between firms may reflect hyper competitive behavior as the market continues to discover itself.

There is also a focus on ‘paying down’ points at origination. In real estate mortgage loan structures, the borrower usually has the option to pay an additional amount to reduce the interest rate on their loan. On average, paying the lender approximately 1.0% of the notional balance of the loan will reduce the interest rate by 0.25%. For crypto loans, there are different methods for a borrower to reduce their interest rate such as the traditional ‘paying down’ points method, membership / token purchase staking, or through increasing AUM on a company’s platform. 

The lack of a securitization and secondary trading market contributes to a slower growth profile of blockchain-based loans. Traditional unsecured credit markets were previously disrupted by a slew of new fintech P2P lending models that could still quickly sell their loans given they had the same look and feel of traditional bank loans. Currently, no market exists for lenders to quickly sell / securitize blockchain-based loans, forcing them to be conservative with underwriting and in developing their risk models. Despite being 100% secured and ~2x collateralized, loans against these cryptos are sold at a premium given the nascent stage of the industry.   

The intent of many cryptocurrency advocates is to reduce reliance on traditional central banking influenced economic models. Given the lack of a central authority influencing the cost of the largest cryptocurrencies (although one could argue that the exchanges, operating costs of miners, and primary markets are currently influencing prices for their benefit6), debates about the concept of a ‘risk free rate’ are now emerging. As the market evolves, it will be critical to observe correlations in traditional interest rates with those in emerging cryptocurrency assets.

Finally, the lack of custody and asset insurance solutions has been discussed in depth since inception of blockchain-based loans. Given the lack of these products, potentially billions of dollars are currently on the sidelines7. Moreover, effective custody solutions are just now beginning to emerge, but it will be years before institutional investors begin to see blockchain-based financial products as a common piece of their investment portfolio. Security audits can and should be mission critical items to review during due diligence.

Final Words

Margin loans are quite risky. If underlying assets fall, it could lead to an uncomfortable situation for the borrower as they either need to come up with the money to add LTV cushion, sell assets, or get liquidated – all at a magnified loss given the interest paid on the loan. The future of blockchain-based loans will depend on a decline in price volatility and maturation of institutional investment administration products.     

While no exact science exists today, the bright future for the cryptocurrency industry will compel the creation of new financial products to accelerate liquidity in this burgeoning market. We see an array of products built this year on the back of crypto technology, such as interest accounts, derivatives contracts, custody solutions, and the addition of many more types of digital assets on these lending platforms. Reaching equilibrium will require an optimal balance of supply & demand, and I believe we are on a long path to its discovery.

Finally, I leave you with a short comparison of blockchain-based loans and two similar loan structures: auto loans and stock portfolio margin loans.

Footnotes

  1. The overview is not meant to be robust, but see the following article on Investopedia for a great overview: https://www.investopedia.com/university/Loans/
  2. Federal Reserve Selected Interest Rate Data – H.15. https://www.federalreserve.gov/releases/h15/
  3. FRED (Federal Reserve Bank of St. Louis). https://fred.stlouisfed.org/
  4. SPP Capital, see the “Market Update” section on the top right: http://sppcapital.com/
  5. Data aggregation of various margin Loan rates today: https://investorjunkie.com/12389/best-margin-rates/
  6. See the Bitwise report on the difference between reported and actual trading volumes. https://www.bitwiseinvestments.com/
  7. Coinbase Reportedly Secures $20 Billion Hedge Fund Through Its Prime Brokerage Services, July 18, 2018. Coin Telegraph. https://cointelegraph.com/news/coinbase-reportedly-secures-20-billion-hedge-fund-through-its-prime-brokerage-services

Crypto Millionaires and the Estate Planning Quandary: Using Traditional Life Insurance Policies to Hedge Downside Risk

The rapid inflation in blockchain-based asset values has contributed to a new generation of “crypto-millionaires.” As of the date of this article, there were 7,046 active addresses with greater than $1.0 million in BTC assets. Back in August 2018, there were approximately 24,000 addresses with greater than $1.0 million in BTC assets. Estimates at the height of the market range from 10,000 to 200,000 addresses with greater than $1.0 million of BTC1. Assuming BTC values grow over the next 2 decades, I expect for this figure to exponentially grow.

A key question for these millionaires is how to create liquidity using BTC (in a world where common goods cannot be purchased without BTC) in a tax efficient manner. Today, utilizing BTC for the purchase of goods and services would require two levels of taxation: (i) income / capital gains tax on the appreciation of BTC and (ii) sales tax. Another unforeseen tax is the estate tax, which is a unique consideration given that many crypto-millionaires had a low cost basis when entering their positions.   

While many structures will likely exist in the future to facilitate tax expediency with blockchain-based assets, I propose a strategy using a traditional universal life insurance policy and loans acquired through a crypto service provider as a means for an investor to both protect against steep declines in blockchain-based assets and boost returns.

Life Insurance Structure

Before going through the math, I wanted to start with an overview of the structure. The graph below depicts the mechanism in 3 stages. The blue boxes represent a crypto service provider activities, the green boxes represent activities of the investor / grantor, and the pink boxes represent activities of a life insurance company.

The first step in the process is for the investor to post their crypto collateral to a crypto service provider. A crypto service provider then secures the assets in their cold storage wallet with multi-signature functionality and lends fiat currency against the crypto collateral. The fiat currency is then used to facilitate a single-premium universal life insurance policy and an 8-year annuity (this is the max allowable LTV based on the math below). The selected universal life insurance policy is structured to provide a return linked to the S&P 500 with a floor / cap of 0.0% / 8.0%.

The investment thesis is to provide a greater rate of return on $5.0 million of BTC while protecting against downside risk. Below, we detail the steps in the process and the sensitize the resulting economics to the investor.

Exhibit 1: The Structure

The Math

The first step in the process is to obtain a loan from a service provider. Currently, you can take out a fiat-based loan against 50.0% of the value of your BTC assets. For example, if you have $5.0 million in BTC, a crypto service provider would offer a loan up to $2.5 million. While the lowest rates offered are around 6-7%, we used 8% in our analysis. The loan proceeds are granted by an investor to an ILIT to avoid the proceeds being counted as part of the owner’s estate.

The amount of the loan required to facilitate the strategy would be $2,347,852. This amount is calculated as 14.5% LTV for the fiat loan (in order to finance the premium payment) plus the amount required to finance an annuity that covers interest payments on the total loan balance ($2.4 million) for the longest period possible (in this case, 8 years). We estimated that the insurance premium for a $5.0 million universal life insurance policy to be roughly $725,0002. Assuming a time to liquidity of 20 years (i.e., death), the annual loan payment would be roughly $240,000.

Exhibit 2: Loan Payment Calculation

Exhibit 3: Financing the Annuity3

Note that the total debt ($2.4 million) was established at a 47.0% LTV.

Phased Approach

In terms of the change in value of BTC, we assumed no growth. Later, I sensitized the results on both the upside and downside. In order to understand the strategy, I broke it down in the following phases:

  • Phase I: Annuity Payments cover interest payments through the initial 8 years. Increasing the number of years would take above the LTV target. These payments would be made at the beginning of the year / period. The annuity payments have another function in that they defer the risk for the underlying collateral base – if BTC has 8 more years to gain liquidity (and one is a believer in the growth story), the annuity offers a valuable strategy to defer the holding cost of holding BTC.
  • Phase II: Annuity payments have ended; therefore, future interest payments would be funded either through liquidation of BTC (given LTV is more favorable as a result of the PI structure of interest payments) or accessing the cash value (discussed below).
  • Phase III: Exit and distribute proceeds to trustees of the ILIT. For the purpose of the below analysis, we assumed that the ILIT would immediately distribute the proceeds on the date of death.

Cash Value

A key component of life insurance policies is the cash value. The cash value represents an internal account within a whole / universal life insurance policy that could be accessed by the owner on a tax-free basis. For a whole life policy, the cash value usually turns positive beginning year 4 or 5. For a universal life insurance policy with a single premium, approximately 25.0% would be kept by the insurance company for fees, and the remaining balance would be deposited as an immediate cash value balance. That amount could be accessed by the grantor for whatever reason, like paying interest due. Day 1, the cash value would cover about 2.3 years of interest ($725k * 75.0% = $534.8k / annual interest). Assuming a long-run growth rate of 5.0% for the S&P 500 over the next 8 years on an annualized basis, the cash value account would be worth $803.4k, covering 3.4 years of payments.

Exhibit 4: Cash Value Calculations

Remember, any loans taken from the balance would be drawn 1 for 1 against the ultimate life insurance payout. In addition, there is no need to take out any funds from this account for the strategy to work, it just provides additional value.

Exit

We assumed an exit at the end of 20 years to estimate the net distribution. At exit, the total distribution would be $5.0 million. The remaining payments after the annuity include the loan’s interest ($2.9 million) and tax liability (for the purpose of this analysis, we assume that the cost basis was $25,000 – resulting in a large tax liability of $995,000).

At exit, the trustees would receive the net distributions plus the BTC. Assuming the mechanics above and no growth in BTC, an additional $1.1 million would be added to the value of the estate.

Exhibit 5: Exit Waterfall4

Sensitivities

I sensitized the performance of the investment strategy based on growth rates of BTC. Assuming a decline in the value of BTC, the portfolio outperforms given the uncorrelated hedge built into the life insurance policy. Furthermore, the tax liability would be reduced. The net benefit of the strategy works better as we sensitized lower performance. In a worst-case scenario where BTC goes to $0, the insurance policy is still fully paid for.

Assuming an increase in the value of BTC, I note that the portfolio return is greater due to the tax-free distribution from the life insurance policy covering a portion of the increased tax liability. The gains are diminishing if greater rates of return are achieved, without performance drag as a result of the policy.

Exhibit 6: Sensitivity Tables

Conclusion

The wave of crypto-investing we all experienced the last couple of years will need to be monetized in some fashion as investors consider their exit strategies or look to diversify their portfolio. The structure described above can be combined with other mechanisms to either increase the hedge or increase exposure to the underlying crypto. We note that improvements in crypto liquidity in futures markets may one day enable a low risk hedging option for investors that want less asset volatility but still wish to participate on the upside. Additional liquidity would also enable dynamic tactical hedging strategies informed by burgeoning asset management services in the crypto space.

Any strategy focused on providing liquidity to crypto holders should consider optionality for repayments. The approach above considers 4 sources of liquidity: (i) growth in the BTC asset, (ii) cash value principal + growth, (iii) annuity stream, and (iv) deferring a large tax liability5.

Footnotes:

  1. Source: https://bitinfocharts.com/top-100-richest-bitcoin-addresses.html
  2. Source: https://termlifeadvice.com/single-premium-life-insurance-the-pros-and-cons-2018/
  3. The annuity was “back-solved” so that the resulting annuity distribution equals the interest on total debt.
  4. Interest after annuity represent the remaining interest payments on debt. We assume they would be paid lump-sum on exit but note they would be paid periodically over the life of the loan. No PIK options currently exist on crypto-based lending platforms.
  5. No transaction fees were considered in the analysis (except the large haircut required for the annuity stream).

Disclaimer: I am not a financial advisor, none of the information contained in this blog post should be considered investment advice.