# Early-Warning Framework for LMEs — Avoiding the Out-Group *Prepared 2026-05-25 · For Anchorage / Pat McGrath interview · Maxwell Nides* **The thesis.** Anchorage's defining recent struggle (per desk feedback) is being **in the out-group when an LME prints**. The out-group gets primed, the in-group gets new-money optionality + priming exemption. The difference is a 25-40-point recovery delta on the same instrument. Pat's desk doesn't need to know that — they need a **discipline for being early enough to be in the room**. This doc is the discipline. Two halves: 1. **The 10 signals**, ranked by lead time (how many months of warning each gives you) 2. **The Pat-doctrine response** — what you do when you see them --- ## THE 10 SIGNALS, RANKED BY LEAD TIME ### Tier 1 — 12+ months out (documentary / structural) **1. The trapdoor is open in the credit agreement (the J.Crew tell)** - **What to look for:** an *unrestricted subsidiary* designation capacity, **dropdown** permission language, IP / brand transfer carve-outs in the investment basket, ratio-based "general debt" capacity, builder basket size relative to EBITDA. - **Why it's Tier 1:** the trapdoor was drafted years before it gets used. If it exists, the optionality is sitting on the sponsor's shelf. - **Action:** the moment you take a position in a B2/B3 sponsor-owned credit, **score the credit agreement against the J.Crew / Chewy / Envision / Cabinetworks fact patterns**. If multiple trapdoors are open → assume LME is in the toolbox. **2. Uptier non-pro-rata amendment threshold is *only majority* (the Serta tell)** - **What to look for:** "all-affected-lender" consent carve-outs in the credit agreement. The shorter that list, the more vulnerable you are to a Serta-style 51% uptier. - **Why it's Tier 1:** post-Serta, the *good* AHC counsel (Gibson Dunn, Akin, DPW) negotiates "all-lender consent" expansions on every new credit. If the deal you bought into doesn't have them, you are pre-primed by document. - **Action:** before signing into any cooperation agreement on a new credit, **ask AHC counsel whether the document survived the Serta wave or not**. If not, the AHC needs to be larger than 51% to be defensive, not just 30%. ### Tier 2 — 6-12 months out (sponsor + capital structure) **3. Sponsor PE hold period approaches 5-7 years without an exit path** - **What to look for:** sponsor's fund vintage, portfolio company age, sponsor's stated hold horizon, recent dividend recap activity. - **Why it matters:** this is a fund-economics signal, not a credit-fundamentals signal. Four mechanisms stack: - *(a) Fund lifecycle clock.* Typical PE fund is 10 years (5yr invest + 5yr harvest). By year 5-7, GP needs **exits** — not marks — to crystallize carry and generate DPI for the next vintage's fundraise. - *(b) The markdown-vs-LME asymmetry is enormous.* A $500M markdown drops fund net multiple from say 1.4x → 1.1x, collapses GP carry from ~$80M to ~$10M, and degrades the track record needed for the next raise. **An LME that primes the lenders preserves the equity mark, preserves carry, preserves the fundraise.** GP loses nothing personally. So when the choice is markdown vs. LME, GP economics make LME strictly better — even when it's strictly worse for the lender. This asymmetry is the whole reason LMEs exist. - *(c) LP-advisory complicity.* The markdown hits LP internal portfolio reporting too. LP advisory boards rarely block LMEs that prime credit holders because the alternative — a written-down mark in their book — is worse for *them*. The lender is the only party hurt by the LME; everyone else upstream is aligned to do it. - *(d) The "stuck" trap.* Entry at 12x in 2019-2021 + multiple compression to 8x + flat EBITDA = equity value at current multiples is ≤ zero. Selling now = write-down. Holding to recovery = preserved option. **The LME buys 2-4 years of runway without crystallizing the loss.** - **The moral hazard:** lenders to sponsor-owned credit aren't underwriting a corporate credit — they're underwriting a structure where the GP and LP have an **asymmetric incentive to push losses to the debt** rather than absorb them. The credit-agreement protections are the only defense. - **Action — the vintage clock is forward-screenable.** Map vintages backward from today: - **2019 vintages:** year 7 right now — peak LME pressure (Springs Windows = Clearlake 2021 buy, LME'd Dec 2024) - **2020 vintages:** year 6 — heavy pressure through 2026-2027 - **2021 vintages:** year 5 — **2026 is when 2021 vintages start LMEing.** KKR's Internet Brands 2022 recap = 2021-cohort, now in trouble - **2022 vintages:** year 4 — quiet now, watch in 2027-2028 - Pull Preqin/PitchBook vintage data, cross against the B2/B3 book by sponsor, generate a **forward LME pipeline** sorted by vintage clock × stuck-multiple × leverage. This is alpha most desks don't have. **4. Maturity wall within 18 months at distressed trading levels** - **What to look for:** TLB or bond maturity in the next 18 months + trading level <85 + leverage above 6x. - **Why it matters:** this is the **trifecta of an LME catalyst**. The sponsor can't refi at par, can't extend without a deal, and can't file because there's no operational distress. - **Action:** **screen every credit in your book for the 18-month / sub-85 / >6x trifecta monthly**. Anchorage's CLO sleeve alone gives them ~$8B of exposure; the screen runs in 20 minutes. **5. EBITDA add-backs growing as % of reported** - **What to look for:** lender reporting package showing add-backs >30% of reported EBITDA, growing quarter-over-quarter; "run-rate" / "synergies" / "transformation" buckets expanding. - **Why it matters:** the moment the sponsor starts manufacturing covenant headroom via add-backs, the next step is a covenant relief amendment, which is **often the vehicle for opening the trapdoor**. - **Action:** **track EBITDA bridge slide quarter-over-quarter** in every lender presentation. Bridge growth = leading indicator. ### Tier 3 — 3-6 months out (governance + advisor) **6. CFO turnover or CRO appointment** - **What to look for:** abrupt CFO resignation (especially mid-quarter), or board appointment of a Chief Restructuring Officer (PJT, AlixPartners, Alvarez & Marsal, FTI named in the press release). - **Why it matters:** boards don't hire CROs as a hobby. Once a CRO is in seat, the LME workstream is six weeks behind them. - **Action:** **monitor 8-K and S-1/A filings + restructuring advisor press** for these names. Octus, Reorg, Debtwire all catch them. Time to **call AHC counsel within 48 hours of the press release**. **7. Sponsor retains LME-specific counsel** - **What to look for:** Kirkland's *liability management* team (not the M&A team), Davis Polk's RX team, Paul Weiss RX team, Latham RX. These are different teams than M&A counsel for the same firm. - **Why it matters:** when the sponsor moves from M&A to RX counsel, they've decided. The LME is in design. - **Action:** **build a desk-level relationship at each of those four firms.** They can't tell you what they're doing, but the fact that they're representing X is itself news that *they* let leak. **8. CUSIP transfer denial or anti-organization pressure (the Internet Brands / KKR tell)** - **What to look for:** sponsor blocks secondary loan transfers to specific buyers, or warns existing lenders that organizing with restructuring advisors will affect future primary deal allocations. - **Why it matters:** this is the sponsor **specifically trying to suppress AHC formation**. It's the loudest "LME is coming" tell short of an actual filing. - **Action:** if you're already a holder, **organize anyway** — the sponsor can't legally prevent it, only make it costly. Get on Akin / Gibson Dunn's call list **the day** you see this signal. If you're not yet a holder, **buy** through synthetic/TRS routes if cash-loan transfers are blocked. ### Tier 4 — 1-3 months out (market mechanics) **9. Specific tranche trading volumes spike + holder concentration shifts** - **What to look for:** unusual block volume on the bond/TLB; a known distressed buyer (Diameter, Silver Point, Sculptor, Centerbridge) accumulating; CDS-cash basis blowing out. - **Why it matters:** the smart money is **buying** to be in the AHC. If you're not buying, you're being out-bid for the seat at the table. - **Action:** **CDS-cash basis + dealer-flow color is the highest-frequency signal you have**. Anchorage's PMs should be getting end-of-day dealer trade color on every name in the book. If they're not, that's an infrastructure gap to fix. **10. Indenture trustee solicitation or "Dear Holder" letter** - **What to look for:** the bond trustee (Wilmington Trust, U.S. Bank Trust) sends a direction-request letter; or sponsor counsel sends an "invitation to discuss" to a sub-set of holders ("**the deal away**"). - **Why it matters:** this is **2-6 weeks of warning at most**. If you're not in the bloc this letter went to, you're in the out-group by design. - **Action:** if you receive the letter — **sign the cooperation agreement same day** and call AHC counsel. If you didn't receive one but you're a holder — **call counsel to confirm you're on the distribution list**. If you're not, you're already on the wrong side. --- ## THE PAT-DOCTRINE RESPONSE — WHY THIS IS AN INFRASTRUCTURE PROBLEM, NOT A SIGNAL PROBLEM The 10 signals are easy. **The reason Anchorage gets out-grouped is not because they miss signals — it's because they don't have the infrastructure to act on them fast enough.** Out-grouping happens at the speed of phone calls. ### What "infrastructure" actually means **1. Counsel relationships, not Octus subscriptions.** - Octus tells you the AHC is forming. **A standing relationship with Gibson Dunn / Akin / DPW / Wachtell / Milbank tells you what the AHC is going to *do*.** - The right desk infrastructure: standing introductions to the 5-7 RX partners across each firm who run AHCs in the $200M-$1.5B lane. **Call once a quarter** even when nothing is happening. - Pat said it on the Bloomberg podcast: *"there isn't one of these LMEs to deal with, there's 40 of them."* If there are 40 AHCs forming per year, you need to be in 5-10 of them. That's 5-10 partner relationships, not 50. **2. FA relationships at PJT / Houlihan / Lazard / Perella / Centerview.** - Sponsor FA tells you what the sponsor *thinks* is on the table. Creditor FA tells you what the creditor group *will accept*. - Anchorage needs to be on the call list for new mandates the day they're scoped. This is **relationship throughput, not deal flow**. **3. A position-level monitoring discipline.** - The 18-month / sub-85 / >6x screen run weekly across the CLO portfolio + the drawdown book + the SMA book = automated. - Output: a 10-name watch list each week. - Each name has a designated PM running point. If the name moves to Tier 2 (sub-80 + maturity inside 12 months), it gets escalated to weekly 1:1. **4. Pre-positioned cooperation agreement templates.** - The first NDA + co-op agreement gets drafted in a hurry under sponsor time pressure. The desks that win are the ones with **standing templates** they can adapt in 24 hours, not draft from scratch in 96. - Pat's desk should have **five pre-cleared co-op templates** keyed to: priming defense, new-money provision, fulcrum-equity exchange, double-dip defense, drop-down defense. **Pulled from the shelf, not built from scratch.** **5. Capital ready to deploy at the AHC-formation moment.** - AHCs require new-money commitments. If you join the co-op without ring-fenced dry powder, you're a passive holder, not an in-group lender. - ACO IX's $1.5B + the evergreen vehicle + the CLO sleeve gives Anchorage the war chest. **The discipline is keeping it allocated and ready, not just committed.** ### The honest framing if Pat asks "how do we stop being in the out-group" > *"It's not a signal problem — Octus picks up the AHC formation a week after it forms, but everyone has Octus. The thing that gets you in the room is being the lender who **called Gibson Dunn or Akin 90 days before** the formation, because you'd already run the trapdoor scan on the credit agreement, you'd seen the EBITDA add-back creep, and you'd noticed the CFO change. You're not buying a signal — you're investing in the **standing relationship throughput** that turns a signal into a seat. The fix is infrastructure: five partner relationships at each of the AHC counsel firms, ten standing introductions at the sponsor FAs, a weekly screen of the book that produces a watchlist with an owner per name, and a co-op template shelf you can pull from. The signals are downstream of all of that."* This framing is what Pat actually wants to hear — it's **not what most candidates would say**. Most candidates would list the 10 signals (which Pat already knows) and stop. The answer that distinguishes you is **acknowledging the signals are the easy part and the infrastructure is the hard part**. That's the senior-associate seat. --- ## INTERVIEW DEPLOYMENT — TWO QUESTION-SHAPES ### If Pat asks the open-ended question > *"What's the hardest thing about being in the right side of an LME?"* Your answer (90 seconds): > *"In my view it's not the signals — by the time the AHC has formed, you can read about it on Octus. The hardest thing is being **early enough by relationship to be in the room when the co-op gets drafted**. The signal taxonomy is well-understood — trapdoor capacity in the credit agreement, EBITDA add-back creep, CFO turnover, the sponsor moving from M&A counsel to RX counsel, CUSIP transfer denials. But everyone sees those at the same time. The differentiation is the standing call list at Gibson Dunn, Akin, Davis Polk, and Wachtell — and being the desk they call within 48 hours of getting the mandate. From there it's about having a co-op template you can adapt in a day, not draft in a week, and having dry powder ring-fenced to commit to new money the moment the AHC needs it. That's why your platform's been able to be in the steerco on Altice and the AHC on At Home and Gabe's and Newfold simultaneously — it's relationship throughput, not signal-detection."* This (a) demonstrates the framework, (b) names Anchorage's recent wins to validate your understanding of *their* infrastructure, (c) signals that you understand the senior-associate role is to *operate* the relationships, not just read the signals. ### If Pat asks the closed question > *"Walk me through what indicators you'd watch on a B2 credit to flag LME risk."* You list the **Tier 1 + Tier 2 + Tier 3 signals** in order of lead time (not in random order). Then close with: *"the question is when you act, not when you see them — for me the action moment is signal #6 or #7, not #9 or #10."* That shows you understand the asymmetric value of being early. --- ## RISKS IF YOU MISSPEAK - **Don't claim Anchorage is out-grouped on specific names.** You don't know which ones. The user-flagged "Anchorage struggles with the out-group" is internal feedback, not a public fact pattern. - **Don't recite the J.Crew / Serta / Wesco fact patterns** as if they're news. Pat lived through them. Reference them as shorthand only. - **Don't pretend the infrastructure exists at most shops.** Most don't have it. Pat knows this. Saying "Anchorage should build it" is wrong — *Anchorage already has more of it than most*. Frame the infrastructure as something **you'd want to help operate**, not something you'd be installing. - **Don't say "I'd build the system."** A senior associate doesn't build the system — they execute against it and improve it at the margin. Match the role. --- ## SOURCES + RELATED PREP DOCS - `anchorage_master_prep_FINAL.md` — Pat's Bloomberg podcast worldview, the "40 LMEs" line - `anchorage-distressed-credit-deep-technical.md` — J.Crew trapdoor / Serta uptier / Wesco double-dip mechanics - `anchorage_sourcing_playbook.md` — the upstream version of this doc: how to source positions in the first place - `anchorage_pat_desk_name_pitches.md` — Internet Brands has the **live CUSIP-transfer-block + primary-allocation-coercion** example you can cite - [Octus / Reorg AHC tracker](https://octus.com) — for the actual signal-monitoring infrastructure - [Wharton WRDIC 2025 LME panel — McGrath](https://2025.wrdic.org/speakers-sessions/panel-lmes/patrick-mcgrath/)