Private credit default rates are presenting a confusing picture, with estimates varying wildly from below 1% to as high as 19%, highlighting a significant disconnect in the $1.8 trillion market. Fitch Ratings recently reported a record 6.3% default rate for US private debt borrowers in August, surpassing the prior month's 6.1%. This figure, based on a trailing 12-month analysis of 1,300 borrowers, was also supported by research from KBRA.
In stark contrast, Houlihan Lokey presented a much lower default rate of 0.8% when weighted by loan principal, despite 2.5% of borrowers by count experiencing defaults. Their analysis, which includes both technical and payment defaults, indicates that default risk is primarily concentrated in smaller borrowers, with about 12% of those with less than $20 million in EBITDA seeing their loans marked below 90 cents on the dollar, a significant jump from 1% in 2023. This disparity suggests that the headline rate, heavily influenced by larger loans, may mask underlying stress in smaller segments of the market. Pimco, on the other hand, indicated a much higher "shadow" default rate of 19% for business development companies (BDCs).
The wide divergence in default figures is attributed to a lack of standardized disclosure and inconsistent definitions of distress within the private credit market. For instance, a mark below 90 cents on the dollar, while not a payment default, signifies a lender's opinion that a credit's original valuation is no longer accurate. This preliminary re-evaluation can precede actual payment issues. Credit Benchmark's research further emphasizes this disconnect, noting a 12% increase in default risk for the underlying holdings of BDCs, even as BDCs themselves might appear stable.
Industries like healthcare showed elevated defaults, both by count (4.2%) and size-weighted (2.7%), while software borrowers recorded some of the lowest default rates. This is partly due to conservative valuations in the software sector following earlier scrutiny. The increasing use of payment-in-kind (PIK) features, representing 6.3% of interest paid across the dataset, with 1.6% stemming from "bad PIK" introduced after origination, further complicates the assessment of true credit health.
