XPO: Owning the First Lien Instead of the Notes

The Call

Buy the first-lien Term Loan B-3 due 2031 at ~100.4 for S+175 and a 5.39% yield to worst. Avoid the unsecured notes. Wait for the 2026 to 2027 refinancing cycle to own unsecured paper, when there should be proper call protection and a new-issue concession to be paid for taking it.

In November 2025 XPO was a straightforward credit, and the part that wasn't obvious was which tranche to own. The business was deleveraging on operating self-help through the weakest freight environment in a decade, nothing matured before 2028 and the balance sheet was clean, but the yields on their own didn't tell you the right place to own it.

Why the Loan and Not the Notes

Net secured leverage was 1.22x and net total leverage 2.28x, cash was $335mm against $1,263mm of LTM EBITDA, and there was $935mm of liquidity with the $600mm revolver fully undrawn.

You get the argument by reading the yield column against the lien position. The 2031 term loan yielded 5.39% from first position in the structure, while the unsecured 2031 notes yielded 5.21% and the 2032s 5.31%, both below the loan and from behind $1.9B of secured claims, so buying either one meant giving up seniority and yield at the same time. Only the 2034 debentures paid more than the loan, at 5.81%, and that's compensation for holding to 2034 in a credit where the thesis is falling leverage.

The notes screened tight because they were trading at 104 to 105 into their call schedules, so any further tightening ran into the call price and total return was capped by the document rather than by the credit. The loan had the opposite profile, with a floating rate, minimal duration, a 101 soft-call rolling off in February 2026, and a near-par basis with a first lien on terminals, trailers, tractors and receivables.

What the Credit Rests On

In LTL the operating ratio is the main profitability metric, and each 100bp of OR improvement goes straight into EBITDA margin and tends to hold across cycles.

In 3Q25, XPO ran a mid-82% OR, roughly 350bp better than two years earlier, while LTL pounds per day fell 6.1% year over year and shipments per day fell 3.5%. When margin expands during a volume decline, that indicates structural improvement rather than a cyclical rebound that's waiting to be given back. The supporting metrics point the same way. Purchased transportation was down 48% as linehaul moved in-house, the claims ratio was 0.2% against 1.2% at the start of the program, yield ex-fuel had grown sequentially for eleven consecutive quarters and on-time performance had improved for fourteen consecutive quarters, and average tractor age was down to 3.6 years, which indicates the $2.5B investment cycle is largely complete.

ODFL, in the low 70s, is still the benchmark. XPO had moved ahead of every non-union carrier except ODFL, and for a credit investor the direction and where it comes from matter more than the absolute level.

Downside

My base case takes leverage to 1.7x in 2026, 1.4x by 2027 and 0.6x by 2030, on a modest volume recovery and stable LTL margins in the mid-20s. Free cash flow above a $600mm minimum cash balance pays for $150mm of buybacks a year, in line with the 2024 and 2025 run rate, and the rest goes to prepaying the term loans, TLB-2 first, since it matures earliest and the borrower gets to direct voluntary prepayments. That leaves the $830mm of 2028 secured notes as the whole maturity wall, and I assumed they get refinanced at size and at similar spreads rather than repaid. The loans are retired in 2029, and after that the company holds more cash than secured debt, $1,118mm against $830mm.

The downside case assumes a 7% revenue decline, LTL margins compressing to 18.5 to 19.5%, and working capital usage of 5% of revenue in 2026. In that case leverage peaks at 3.5x in 2027 and improves after that. It starts from $450mm of cash at year-end 2025 and spends it, with free cash flow negative for three years (roughly -$320mm, -$138mm and -$26mm), so cash reaches zero in 2027, and the revolver funds the gap, peaking at $34mm against a $600mm facility. Liquidity troughs at $566mm in 2028. Either way, a first-lien claim attaching at 1.2x isn't impaired.

The memo lists five risks. The first is covenant flexibility, with capacity under the credit agreement for roughly $2.6B of incremental first-lien debt, which is enough to take first-lien net leverage from 1.2x to 3.3x. The others are a deeper or longer freight trough, execution risk on the margin program, the European segment as a lower-margin, FX-exposed drag pending a sale, and the $1.5B secured maturity wall in mid-2028.

What This Memo Does Not Do

The recovery work stops at a single point estimate, and there's no waterfall. I halved EBITDA, applied a 4.5x distressed multiple and concluded that the secured debt recovers in full, which is almost certainly the right answer at 1.2x attachment, but it's a conclusion I stated rather than one I built. There's no going-concern case, no liquidation marks on the terminals, trailers, tractors and receivables the lien actually attaches to, and no sensitivity around the 4.5x.

The covenant section names the baskets without testing the definition underneath them. It has the $400mm free-and-clear basket and roughly $2.2B of ratio capacity against a 3.0x first-lien net leverage test at 1.2x, roughly $2.6B in all, and it confirms that incremental debt would come in pari passu rather than priming. Whether that $2.6B is real depends on the EBITDA definition and its add-backs, and I didn't examine those, or the transfer restrictions that decide whether collateral can leave the credit group, which the memo asserts without citing the agreement.

The base case also assumes no acquisitions for five years, which is a strong assumption to make about a company Brad Jacobs runs. It's the likeliest reason the deleveraging path doesn't happen, and I didn't put it in the risk list.

A second pass at the memo would cover all three of those.

This is independent work, written solo. The levels, metrics and refinancing view are as of 30-Nov-2025 and haven't been refreshed.