High yield spreads widen 68 basis points as September supply meets a thinner bid

The high yield market went eleven weeks without a bad day and then had four in a row. The Ledgerbrook US high yield index, which tracks about 1,900 bonds and is the benchmark most US credit desks quote off, closed Friday at 396 basis points over Treasuries. Three weeks ago it sat at 328, the tightest level of the cycle. Sixty-eight basis points of widening in fifteen sessions is not a drawdown by historical standards, but the way it happened should hold the attention of anyone underwriting new paper.

Primary markets reopened after Labor Day with a heavy calendar, as they always do. Fourteen deals came in eight sessions. What changed is the reception. Eleven priced with a concession above 25 basis points to where the same issuer's existing curve was marked, three of those above 50 basis points, and one was pulled outright after two days of marketing. Through June and July, the typical concession on a single-B deal ran between 5 and 12 basis points and books were routinely covered four times over. Those books are now covered twice, and the allocations are going to accounts that negotiate.

The widening is not evenly spread

Splitting the index by rating changes the picture substantially. BB paper, which makes up a little over half the index by market value, has widened 29 basis points. Single-B is out 63. CCC is out 141. That dispersion is the signature of a market repricing default risk in specific issuers rather than demanding a higher risk premium across the board.

A rising ten-year yield does part of the work here. Every high yield borrower prices off the Treasury curve, and the curve moved 34 basis points in four sessions this week alone. But spread is measured over that curve, so the widening reported above is on top of the rate move. All-in yields on the single-B bucket have gone from 7.9 per cent in late August to 8.7 per cent, and that is the number a chief financial officer sees when deciding whether to refinance now or wait.

The Ardent Fiber deal is the one worth studying. A $300m add-on to fund a delayed acquisition, marketed at 9.75 per cent, revised to 10.5 per cent on day two, then shelved. Nothing in the company's disclosure changed during marketing. What changed was that three of the anchor accounts that had indicated interest reduced their orders after the policy decision on Wednesday, citing the move in the long end rather than anything about the credit.

Accounts are not selling. They are declining to buy, which looks identical on a spread chart and feels completely different on a trading desk. There is no forced seller in this market yet.

Rufus Ohara, head of credit research at Ledgerbrook Partners

The 2028 problem

The reason a 68 basis point move gets this much attention is the maturity wall behind it. Ledgerbrook counts $214bn of loans and bonds maturing in 2028 from issuers rated single-B or lower, and the large majority of it is floating rate paper issued in 2021 at spreads that no longer exist. Those borrowers have been waiting for a lower policy rate to refinance into. The front end has obliged. The credit market has not.

• Roughly 61 per cent of the 2028 maturity stack is floating rate term loan, where the coupon already reset lower with the policy rate but the spread does not move until refinancing.
• Issuers rated CCC account for $47bn of the stack and face all-in refinancing costs above 11 per cent at current levels.
• Private credit funds have absorbed about $38bn of refinancing from this cohort over the past eighteen months, which is why public market default rates have stayed low.
• A sustained CCC spread above 850 basis points has historically preceded a rise in the trailing twelve month default rate by two to three quarters.

What would turn this into something worse

Two conditions. The first is a retail outflow cycle, because high yield fund flows are reflexive and a month of redemptions forces selling into a market where dealers are already carrying inventory they would rather not have. Flows have been roughly flat for three weeks, with a modest $1.2bn outflow in the week to Wednesday, which is noise. The second is the long end continuing to sell off, since it lifts the all-in cost for every borrower and pushes marginal refinancing candidates toward amend and extend transactions that rating agencies increasingly treat as distressed exchanges.

We underwrite to the exit, and the exit for a 2028 borrower is a refinancing market that has to exist in 2027. That market has just become 80 basis points more expensive and nobody has changed their base case.

Theodora Vance, portfolio manager at Kestrel Asset Management

For now the base case on most desks is that this is a supply indigestion episode that clears by early October once the post-Labor Day calendar empties. That has been the right call at every comparable point in this cycle. It is worth holding the view loosely, because the thing that has changed is not credit fundamentals but the price of the risk-free curve those fundamentals are discounted against.