Securing “Serta” protections—which fortify affected lenders’ voting rights—has become a focus for lenders over the past three years as managers assess the damage that liquidity management transactions (LMTs) pose for non–participating lenders. For the record, the covenant feature that opened the door for LMTs is simple majority voting that allows some lenders to gain control of a loan and vote themselves the right to (1) provide fresh financing via LMTs that primes prior first–lien claims while (2) elevating the pre–existing claims of lenders that participated in the LMT over non–participating lenders. Known as up–tiering LMT transactions—or investor–on–investor violence in the parlance of our time—this technique has been used on a growing number of transactions in recent years including NYDJ, Travelport, Envision and of course Serta. Covenant Review provides thorough primers and a running list of such deals here and here.
The LSTA asked Covenant Review to put together a perspective on (1) the potential incremental credit loss for lenders that are primed by up–tiering LMTs during a default cycle and (2) how the market is adapting to this issue for its annual conference. What follows is a summary of our comments.
Setting the stage
To puzzle out the implications of LMTs, we start with two baseline facts.
- Serta loophole: As of Sept. 30, 62% of the Credit Suisse Leveraged Loan Index allowed majority consent voting (i.e. have a Serta loophole), based on Covenant Review data.
- Recovery gap: In recent years, among loans that have had up–tiering transactions associated with LMTs, our colleagues at Fitch Ratings found that priming lenders recovered roughly 50 cents on the dollar versus roughly 10 cents for primed lenders. Taking the analysis a step further, the weighted–average recovery across primed and non–primed lenders—that is the theorical recovery in the absence of an LMT—was 33 cents of the dollar and thus primed lenders suffered 23 percentage points of incremental loss
Scenarios
- Historical average: Between 2007 and the 12 months ended Sept. 30, the average annual default rate for the Fitch Leveraged Loan Index was 2.47%. Apply 14 bps of additional loss-given-default associated with up-tiering LMTs for primed lenders, the average incremental loss is roughly 35 bps a year based on the historical average.
- Great Financial Crisis-level defaults: Between 2007 and 2009, the cumulative default rate was 15.3% for the Fitch Leveraged Loan Index. In this peak default scenario, the incremental loss to primed lenders would be 214 bps over three years, or 71 bps a year.
- Fitch’s 2024 default outlook: Fitch’s outlook of 3.5%-4.0% implies an incremental loss to primed lenders of 49-56 bps.
Of course, none of these estimates will be on the mark. For one thing, it’s highly unlikely that a lender would be on the primed end of an LMT transaction 100% of the time. Moreover, the markets will surprise us as they always do. Still, the calculations provide some context based on historical averages of what the future might look like.
How is the market adapting?
In light of the LMT theme that Serta kicked off in 2020, managers have become more intently focused on Serta protections. As this chart shows, the percentage of Credit Suisse Index loans with Serta protection—that is those that do not allow majority voting—expanded to 38% as of Sept. 30 from just 9% as recently as year–end 2019.
This trend reflects:
- New–issue: The percentage of new loans clearing with Serta protections stood at record 52% in the year to Sept. 30 (YTD), up from 46 % in 2022.
- Amend–to–extend: Among YTD A–to–E exercise, the share with Serta protection increased to 49% post–amendment from 19% pre.
Finally, lenders note anecdotally a downshift in the use of up-tiering LMTs over the past 12 months. Players say that in general lenders are trying to work more collaboratively toward solutions that are inclusive of all lenders that are looking to play rather than isolating and disadvantaging some lenders—or worse still allowing new lenders to buy at a discount and then use LMTs to up tier into a higher recovery claim.
Broader context: are lenders being paid for looser protections?
Of course, the decades–long trend toward looser covenant protections that allowed up–tiering LMTs didn’t happen in a vacuum. In the 1990s when documents were tight and lenders protected by a tangle of five or six maintenance financial tests that tracked financial models with a quarter–to–a–half turn of headroom, the market was small and relatively illiquid. Since then, amount of institutional loans held outside of banks has soared to $1.5 trillion, according to the Credit Suisse Leveraged Loan Index, from $106 billion in 1999. Moreover, trading loan activity has exploded to $824 billion in 2022, according to Ted Basta at the LSTA. As result, managers are free to trade out of names that contain document vulnerabilities today in a way they weren’t able to in the distant past.
- Most important, spreads have expanded considerably over the years. Among single B rated, the average nominal spread of Index loans stood at 400 bps, up from 388 bps at the end of 2019, 310 bps at the year–end 2009, 315 bps at yearend 1999 and 245 bps at yearend 1991 at the dawn of the Index. Managers point out these spreads understate the actual increase because OIDs today are wider than they’ve been historically.
Does this mean that investors are being compensated for the potential value leakage caused by LMTs and other covenant loopholes? Managers say the answer is no and yes. On an individual deal, the incremental spread is clearly not going to cover the potential incremental loss from being on the wrong side of an up–tiering LMT. On a portfolio basis, however, managers say higher spreads do likely cover the higher potential loss–given–default associated with LMTs and other document erosion.
Steve Miller
CEO of Fulcrum Financial Data
Thanks to Ian Walker, Jessica Reiss, Jonathan Blau and, from Fitch Ratings, Joshua Clark and Elizabeth Han for their contributions to this analysis and to Meredith Coffey for raising the topic in the first place.
For additional covenant stats or other cuts of the data, please contact research@covenantreview.com.
