Weekly Industry Update: Fed Raises Rates, Affirm Joins Unicorn Club, Introducing Sharpe Ratios

The Fed raised interest rates 25 bps this week to a target range of 1.25%-1.5%. On the inflation front, the BLS reported a higher than expected November CPI reading of 2.2% buttressing the Fed’s plan to raise rates three times in 2018. Whole loan investors face greater pressure on net returns as most rate increases have not been passed onto borrowers. PNC, taking a cue from both the FinTech playbook and traditional branch lending model, is now offering loans via their mobile wallet and their branch network. PNC expects returns from loans originated via mobile to outperform loans originated via the branch network. FinTech financings continue. The WSJ reports that Affirm raised $200 Mn led by Singaporean sovereign wealth fund GIC. The private market valuation of Affirm is estimated to be between $1.5 Bn and $2 Bn. Affirm provides point-of-sale financing to more than 1,200 merchants. The loan products have a 3 to 12-month payback periods and a median APR of 19%. The closest competitor to Affirm may be Sweden-based Klarna, which provides a POS financing solution for over 70,000 merchants. Klarna received a strategic investment from Visa and was recently valued at ~$2.25 Bn. Klarna recently secured a banking license, a path that other FinTechs including Square are pursuing. In securitization news, Asset-Backed Alert reports that Equifax completed its first securitization –a $200 Mn transaction backed by lender fee revenues. This deal is the first of its kind in the credit bureau space and was marketed and placed privately. Equifax generated $835 Mn in revenue last quarter, up 4% YOY. For this week’s special topic, we analyze the expected loss-adjusted returns and introduce Sharpe Ratio as one (of many) metrics to gauge risk-adjusted returns. Our goal here is to introduce how to properly apply the Sharpe ratio to incorporate volatility and the risk-free rate in portfolio construction. Sharpe Ratio – Motivation & Background Although investing in consumer credit remains attractive relative to other asset classes, the range of outcomes is increasing due to credit re-normalization, increased bank competition, and a rising rate environment. Over the last few months, we have received investor inquiry on how to incorporate both expectations and uncertainty in expectations to manage risk. The motivation is two-fold—expected returns are an insufficient statistic to drive investment decision making. For example, an investment with a 5% expected return but a higher range of outcomes may be less attractive than an investment with an expectation of 4% that has a tighter distribution of outcomes. Expected returns alone do not account for the volatility of possible outcomes. Additionally, the rising rate environment require investors to control for the risk-free rate when building and monitoring portfolio risk. The Sharpe ratio is designed to measure riskiness while controlling for risk-free rates and volatility. The ratio, introduced by Nobel Laureate Bill Sharpe, measure return per unit of total risk taken above the risk-free rate. [ Quant Note: The Sharpe ratio is typically used for assessing performance on liquid traded instruments (such as equities), and should not be used to evaluate ex-postperformance for illiquid collateral – including whole loans. However, the Sharpe ratiocan be a useful tool to analyze ex-ante forecasts are drawn from a credit model that generates independent and identically distributed draws (such as a monte carloprocess drawing from a roll-rate model). ] The Sharpe ratio is defined as:  represents the average return of the portfolio  is the average return of the risk-free rate for the time period under evaluation  is the average standard deviation of the portfolio The Sharpe ratio measures a portfolio’s added value relative to its total risk. A portfolio of risk-free assets or one with an excess return of zero would have a Sharpe ratio of zero. Methodology Using this framework, we analyzed a random sample of LendingClub loans across vintages, terms and grades and projected estimates of annualized loss-adjusted expected returns based on an arbitrary third-party credit model. We selected 7,000 random LendingClub loans broken out by their term – 36 or 60 months, their grade – A to G, and their year of origination – 2012 to 2017. Each bucket includes 100 randomly selected loans that fit the criteria, e.g. 100 randomly selected 36-month grade C loans originated in 2014. All these loans were run through a third-party credit model to generate annualized loss-adjusted expected returns over the life of the loans. Results and Conclusions Below we show an illustrative output that uses a Sharpe ratio framework to govern the investment decisioning process. The tables below show the annualized average expected loss-adjusted returns, standard deviations of those returns and Sharpe Ratios on Prime LendingClub loans in each bucket based on an arbitrary 3^rd party credit model. Based on this analytical framework and credit model, we note that A and B-grade loans deliver the highest risk-adjusted returns. Grade F and G loans have the lowest expected and risk-adjusted returns. Accordingly it makes sense that LendingClub suspended production of F & G loans. We remind the reader that we are using an arbitrary credit model and these results should not be used for investment decision making without consultation. Annualized average expected loss adjusted returns, standard deviations and Sharpe Ratios Source: PeerIQ In the chart below we illustrate the distribution of outcomes. Note that Grade A and B loans exhibit a narrower range of outcomes as compared to F & G loans. Also, we observe that the volatility of returns has increased in successive vintages. Overall, the chart shows that B-grade loans are outperforming other grades on a risk-adjusted basis. Distribution of annualized average expected loss adjusted returns Source: PeerIQ This analytical framework can be extended to include other higher moments in the probability distribution of returns – skewness and fat-tailness (kurtosis) as well. Also, the return framework can also incorporate prepayment expectations and financing costs. Finally, ABS investors can run cash flows using loan-level monte carlosimulations to identify the range of expected WAL and IRR outcomes on their investments. Reach out to PeerIQ learn more about how to implement this analytical framework and how 3^rd party credit models can assist in valuation and risk management of your loan portfolios. PeerIQ in the News: