Houlihan Lokey : Private Performing Credit Index

Houlihan Lokey is an expert in the valuation of private credit, including direct lending and BDC assets. We produce the index, studies on relevant topics, and other research and analyses:

Scroll down to read Study 1, jump to Study 2, or jump to the downloadable Current Index Report (PDF).

Study 1: How do IRRs vary by size of borrower?

We have been producing the Private Performing Credit Index (PPCI) for several quarters using a dataset of instruments we have valued since Q3 2017. Clients have asked us for a variety of data insights, and the question of whether different size loans persistently yield more has been common. To answer this, we turned to the same dataset we use to compute the quarterly index, but created quartile subindices for comparison purposes.

The answer to that question is yes: There are trends and conclusions, but it is not a simple response. We stratified the data into quartiles based on adjusted EBITDA-breakpoints of 10 million, 20 million, 50 million. The periods prior to and since COVID-19 display marked differences, but the smallest borrower quartile does persistently pay the highest IRRs. The largest borrowers pay the least, although that is most pronounced since Q2 2019.

The average of the entire dataset is presented as subindex average. For computational reasons, that average is not exactly the same as the PPCI, but the comparison between quartiles is accurate.

By examining the quartiles relative to the subindex average, that same observation is easier to see. However, there is a marked convergence of IRRs by quartile in Q4 2019, prior to COVID-19 lockdowns and market reaction. While the largest borrowers are routinely +/- 20bps from the average, the smallest borrowers have been 180bps above average in the past several quarters.

The overall divergence between quartiles is also dynamic and has moved wider and tighter over time. This graphic shows the dispersion as the highest quartile IRR minus the lowest quartile IRR. The trailing average of dispersion also shows an increase, as IRRs continuously reflect size premiums or discounts.

Conclusions and Observations

This is the simple answer to the question, "How do IRRs vary by size of borrower?": There is a wide variation in the dispersion of IRRs relative to size, but the smallest borrowers appear to consistently pay the higher IRRs. One conclusion for a direct lender may be to pursue smaller loans on the strategy that the increase in execution costs for a larger number of transactions is compensated by the increase in overall return. However, the opposite conclusion is also supported-that the pursuit of the largest transactions may reduce execution costs sufficient to gain an advantage, despite the lower IRRs.

It seems clear that the competing strategies can both be winners, but they vary over time and by market conditions.

Study 2: Can you break down the PPCI to explain changes?

Each quarter, we calculate the PPCI from the instruments valued in our automated systems. However, we also produce other aggregated statistics, such as weighted average spread or coupon. An examination of the relationship between these metrics and benchmark rate and average price can explain the market dynamics and the differences between public and private credit instruments.

Component Metrics

• Coupon is the coupon used to determine the cash flow for each instrument and does not include the benchmark rate. It is calculated as a principal weighted average. The entire instrument cash flow does include benchmark rate in the valuation calculation.
• Spread is the rate added to the benchmark rate used to discount the cash flows back to a price in the IRR function. For loan valuation, spread is determined by multiple factors by the valuation agent.
• Benchmark rate is typically SOFR or LIBOR and is computed in our models as a forward rate derived from the active swap curve. For our decomposition analysis, we use the swap rate best related to the weighted average life of the valued instruments-typically four years.
• Price is the instrument price after all factors have been incorporated. Typically expressed as a percentage of par, price is also shown as the amount of discount to par to facilitate display. We determine what portion of the change in IRR is attributable to a change in price.
• Composition is the measurement of the change in PPCI relative to the change in the portfolio of valued instruments.

In the chart above, we illustrate the relationships between these metrics in three highlighted periods. Spread and price have an inverse relationship; when the spread increases, the price decreases. During the valuation process, we take into consideration changes in market spreads and yields. We evaluate portfolio company performance, including relevant credit metrics, to determine an appropriate spread to apply to the cash flows. Despite the relationships, the calculation methodologies prevent a conclusive summation of the components to exactly explain the results of any period. The first bullet highlights this:

• Q4 '18: This period saw a large decrease in the coupon rate (-0.43%), which could have been the result of refinancings or resets as well as new issues. However, the spread decreased by less (-0.35%), which is consistent with a reduction in price. This reduction was obscured by the change in portfolio composition, which contributed to a decrease in IRR of -0.20 versus the entire change in price (-0.21, expressed as yield).
• Q1 '20: This period marked the beginning of COVID-19 and saw a huge decrease in benchmark rate (-1.22%), but it was almost fully offset by an increase in spread (1.31%). However, spread determines discount rate, so the price fell approximately 3.5 points (-0.79% in yield terms).
• Q1 '22: This period saw a large increase in the benchmark rate, but since there were only slight changes in spread and coupon, there was almost no change in price.

In the chart above, the price is expressed as a positive amount (the discount from par) and clearly shows that the strongest relationship is between spread and price. The increase in benchmark rate largely does not affect the price of these instruments. Coupon has remained remarkably stable despite economic volatility. The entire range of coupon variability has been 100 bps, where spread has ranged by 150 bps in the same period, and the price is now appropriately moderately lower (discount is a half point larger).

We note that the pricing of a loan embeds optionality, which is not typically isolated in its valuation. There are duration and volatility effects associated with any benchmark rate floors, the prepayment option (which is subject to market and credit conditions), and the concept of duration related to fixed spread (particularly when spread is a multiple of benchmark rate). The variance in prices can reflect these duration and volatility issues in a way that prevents the components above from explaining 100% of the aggregated IRR changes. Please contact us for a broader discussion of these valuation elements.

The answer then to the question about explaining changes in PPCI is, yes, the best explanation is to consider these metrics in light of the market conditions and the difference between outstanding and new issue instruments. The components viewed in relationship substantially explain changes but do not explain 100% due to secondary and related factors not included in these primary components.

Current Index Report

Houlihan Lokey's Portfolio Valuation and Fund Advisory Services practice is a leading advisor to many of the world's largest asset managers, who rely on our strong reputation with regulators, auditors, and investors; private company, structured product, and derivative valuation experience; and independent voice. As a function of this unique position within one of the world's largest capital markets, Houlihan Lokey routinely provides valuations of instruments in many of the least visible asset classes, particularly private credit. The Houlihan Lokey PPCI is an aggregate view drawn from our extensive dataset of private credit valuations.