Erica · PM Research MemoAll memos

PM RESEARCH MEMO

Title: The K-Shaped Credit Market: Non-Bank Transmission Blunts Hikes for AI and Large-Cap Borrowers, Consumers and CCC Take the Hit, Private-Credit Opacity, Software Recoveries Near Zero, and Risk-Off on Re-Steepening
Author / source: Henry Peabody (Senior Investment Strategist, GMO), interviewed by Jack Farley (Monetary Matters)
Source title: The Fed Is Losing Control of Credit | Henry Peabody
Source URL: https://www.youtube.com/watch?v=uE4KMgxdFFY
Video ID: uE4KMgxdFFY
Channel: The Monetary Matters Network
Published / upload: Thursday 24 Sep 2026 (upload_date 20260924). Recorded Thursday 17 Sep 2026 (per description), i.e. right after the September FOMC hike and before the Snider (22 Sep) and Bianco (24 Sep) uploads.
Duration: 56:24, continuous, no gaps
Memo date: Sunday, 27 September 2026 (America/Toronto)
Transcript: /workspace/youtube-transcripts/uE4KMgxdFFY.md · Brief: uE4KMgxdFFY_brief.md · Plain: uE4KMgxdFFY_plain.txt · Metadata: uE4KMgxdFFY.info.json
Referenced paper (not read for this memo): GMO "Triple Mandate", https://www.gmo.com/americas/research-library/triple-mandate_whitepaper/ (External)
Caption source: YouTube automatic English ASR only (en-orig json3; the en track returned HTTP 429). No manual captions. No speaker labels. Only >> turn markers. Speakers are inferred from context. Several turn boundaries are ambiguous (e.g. 09:00–09:30, where Farley's question runs into Peabody's answer in the same block).
Product: A credit-transmission, private-credit-cycle, curve and positioning research map with confirm/disconfirm signals. Not advice. Not a recommendation to buy or sell any security. GMO's positioning is reported as their view, not a desk recommendation.
Source discipline: The primary sources are this transcript, its brief, and its metadata only. Snider (23 Sep), Bianco (22 Sep and today) and other memos are cross-referenced only, with no number import. Any live level, price, probability or fact not spoken in the transcript is External check needed. Per the desk rule (decisions.md, 27 Sep), no market data was fetched.

How to read: Source = restatement with timestamps. Inference = analyst interpretation. Monica's angle = the desk's YouTube read, tested against the transcript.

ASR name / number locks

ASR heard Likely / note
"Wsh" (01:30), "Worsh" (05:00), "Walsh" (09:00), "wars talking about reducing the size of the balance sheet" (34:30) The Fed chair, who held a press conference after the hike with a unanimous vote. The brief assumes Warsh. This memo does NOT assert the chair's identity (External check needed). The Snider memo's ASR also rendered "Kevin Warsh" at Jackson Hole. That is corroborative of the ASR reading, not verification. The text refers to "the Chair" below.
"Worsh's commentary is completely irrelevant here. I said relevant, not irrelevant." (05:00) Speaker self-correction. The chair's commentary is relevant.
"Harry" (23:30) Henry (Farley addressing Peabody)
"Ebidon" EBITDA
"CLLO tanches" CLO tranches
"risk parody" risk parity
"restepeepens" re-steepens
"sacrosanked" sacrosanct
"put a PV on prayer as a terminal multiple" (02:30) Garbled. Possibly a quip about pricing hope into terminal multiples. Unresolved; not used.
"tapping emerging debt markets for housing AI data centers" (13:00) Probably "for hosting/housing AI data centers". The meaning is EM debt markets financing AI data centers. Wording unverified.
"Scott Bassant" Scott Bessent (the buyback claim itself is unverified)
"Micro Strategy" MicroStrategy (Strategy)
"$40 trillion and whatever 10% of GDP" (50:30) Ambiguous. Likely ~$40T debt and a deficit of ~10% of GDP, but the phrasing is unclear.
"AI related investments are… up pick a number 6% annualized" (17:00) Approximate, and hedged by the speaker. Nominal. "GDP contributions much lower."

Key caveats (read before using)

Stance (source-locked one-liner): Credit now flows largely outside banks and at fixed rates, so hikes barely touch large, termed-out and AI borrowers while consumers (mortgages in the 6s, auto loans in the 7s) and CCC take the damage. To cool demand the Fed "may need to move higher than they would otherwise", which adds risk. Private-credit opacity means "we could be in the midst of a correction already". GMO is de-risked: less duration and vanilla credit, more short seasoned structured credit and EM local debt, and a preference for the 3–5y belly as the risk-off hedge. Risk sells off when the curve re-steepens, not at inversion.

Conviction (memo overall): Medium. The K-shape and transmission argument is coherent and matches visible phenomena (AI issuance vs consumer rates). Much is qualitative and anecdotal. Key numbers are hedged or ASR-ambiguous, and GMO describes itself as a skeptical "value shop… sometimes to a fault."


EXECUTIVE SUMMARY


SOURCE-ACCURATE SUMMARY

Chronological cue timestamps (~30s blocks). Quotes are 20 words or fewer. P = Peabody; F = Farley (inferred, C1).

  1. (00:00) P, cold open: The credit cycle is "one of the single most dependable things in finance… human greed, incentives, and pushing product." "We could be in the midst of a correction already and we don't even really know it."
  2. (00:30–01:30) P, third mandate: Beyond stable prices and full employment, the Fed has a new, unofficial mandate, "the transmission of credit across the economy." A "less banked" system, private credit and "fixed rate bonds" have "loosened the connection between policy and the actual economy."
  3. (01:30–02:30) P, easy conditions: Financial conditions are "largely fairly easy"; "Wsh gave a nod to that… in his press conference" (C2). "Spreads are tight. Equities are… high." The dollar is at a "near-term local low," "rich on a real effective exchange rate basis." "Volatility is subdued." Easy for large, termed-out corporates, the AI buildout and private credit. Consumers face "mortgages… 6% something," and "home sales are slowing."
  4. (02:30–03:30) P, procyclical immunity: AI (electricity demand, labor demand) is "relatively immune" to hikes, which suggests "higher than we expected for probably longer." The Fed must "look at past cycle playbooks to manage through asset price inflation." "Triple C credit… marginal borrowers… maybe collateral damage."
  5. (03:30–05:00) P, why the bifurcation is acute: Rates moved off the zero bound. Short-term consumer borrowers are most exposed, and inflation hits the low end hardest. Breakfast-cost chart: top earners' affordability improved over 50+ years, while the weakest cohort's gains from the '60s and early '70s have "unwound… basically flat."
  6. (05:00–05:30) P, chair and vote: The chair's commentary is relevant (self-corrected): "concerned about the speed in which we come back down to target." "The unanimous vote behind him is telling."
  7. (05:30–07:00) P, bank vs non-bank: A bank risk manager facing hikes weighs capital and losses and widens spreads. That is "very rational in a regulated and banked system." Non-bank capital is "deployed at a zero cost or negative," earning yield plus fees. Banks have a "floating cost of capital." Losses now sit with institutions "that can withstand that loss."
  8. (07:00–08:30) F/P, no tightening so far; no price mechanism: F: banks tighten on hikes and non-banks don't. P: "They haven't thus far," with "anecdotal stories" of more conservative measures. Delay in realizing losses; poor "visibility of marks." Not "a systemic issue like we had in '08," and "lower levered," but "you don't have that observable mechanism… opaque." No "psychological reaction on the part of forced sellers."
  9. (08:30–09:30) P/F, how high? P: the risk is "we do keep rates higher longer. We do need to move higher," and the risk of "going too high… a mistake is absolutely real." The chair ("Walsh," C2) focuses on "speed to get back to target" and "financial conditions." A unanimous vote means "upward pressure on rates is very real." F: the Fed raised to "basically 3.9%," signaled "4.1%" by year-end and "4.1% until the end of 2027."
  10. (09:30–11:00) P, higher for longer vs fiscal: "The higher they go, the more they're going to have to ease after." Likely "higher for longer until we see cracks start to form." Transmission lags. "Triple C… clearing coupon… well north of 10% probably north of 12… We're in the midst of it." Fiscal sustainability needs "rates below nominal growth," so "financial repression… yield curve control" is possible later. Nominal growth is currently above rates.
  11. (11:00–12:30) F/P, elasticities: F: at 7% some AI borrowers still borrow, while "a seven or 8% mortgage rate would obviously crush the housing market." Is the hike sufficient? P: "Probably probably not." It hits "those most susceptible": "car loans are running north of 7% on used cars for high quality borrowers."
  12. (12:30–14:30) P, AI supply and generic AI risk: Hyperscalers borrow "regardless of what the rate… is." IG supply is forecast "about 200 billion higher over the next quarter than it was last year," and forecasts are rising. Interest in "tapping emerging debt markets for… AI data centers," in HY unsecured and secured. "No shortage of AI crowding out." IRR expectations may be "wildly off bounds." The market treats "generic AI risk… from the chips to Oracle to SpaceX to project finance" as one bucket, "trade relatively cheaply in the credit default swap market" (C3).
  13. (14:30–16:00) P, growth-asset cycles: Fiber, "mortgage brokers in 2007," today "Crocs… now they're an AI company" (C5). "Credit investors… only have the downside." "In every one of those cases, the Fed has had to move higher than they expected." Opacity and low elasticity mean "probably be tighter than we would otherwise."
  14. (16:00–17:30) P, the dog that didn't bark, and growth: Fifteen years ago, a "10-year flirting with 5%, which we didn't hit since 07" (C5) would have meant wider spreads, strained liquidity, "talking about an ease," and "the dollar… fading." That "hasn't come to pass." Growth is picking up: "ISM index is year-over-year improving… regional surveys… augers for a flatter curve." AI investment is up in nominal terms, "pick a number 6% annualized," with GDP contribution "much lower." To cut demand, the Fed must be "more aggressive than it has in recent cycles."
  15. (17:30–20:00) F/P, private credit and higher rates: Early in the cycle, "yields drew buyers in." Now buyers are "more picky." There is "increase in secondary trading," and "insurance companies buying secondary assets." Higher yields pulling in capital "feels… like that's the last marginal buyer."
  16. (20:00–22:30) F/P, floating-rate stress and opacity: "Coupons go up, fixed charge coverage… goes down." Borrowers are "levered generically six plus times." Restructurings "behind closed doors with one or two lenders" are less clear to the market than committee-led workouts. JPMorgan's default report has been "the bible." Banks used to report NPLs, nonaccruals and reserves. "We're all sort of guessing."
  17. (22:30–24:30) F/P, public credit and defaults: "Defaults have been relatively benign." HY quality has risen and CCC/B has been "on a downtrend for… over a decade." F: "the riskiest deals got financed away into private credit" (P: "absolutely right"). Some private-credit borrowers are re-engaging the bank-loan market. "Triple C spreads… substantially wider than the rest of the market." Default rate "higher in a year than it is today," but no spike call.
  18. (24:30–26:30) P, recoveries and software: Value = probability of default × recovery. Software has "no hard assets… relying on IP." You could "straight line an EBITDA number out 5 years and put more leverage on it." For "some of these companies… recovery won't be 50, 40, 30… you're going to look at donuts." "Someone… running a software company… is going to be redundant." "I'm not a private credit guy." He is a bear on asset coverage, competitive vulnerability, interest costs, margin compression and stock-based compensation.
  19. (26:30–29:00) P, GMO fixed-income stance: "At the low end of our spectrum… lower risk everywhere." A "value shop… sometimes to a fault." "Double B is relative[ly]" preferred in HY. The dollar is "a bleed valve in the out years" as policy eases, which supports "non-US assets, emerging local." "Avoiding beta for beta's sake"; more return from alpha. The "dependability of duration… to offset risk is certainly less stable." Not "a carbon copy of a risk parity strategy."
  20. (29:30–33:00) P, stock–bond regimes: Pre-1999 through 2020 was disinflationary, and hedges paid: "1 plus 1 equals 3." Since 2020, fiscal–monetary "cooperation" brought "higher inflation, higher capital cost." "Not just Powell's… transitory mistake." If inflation stays "3 1/2%… above 3%… next handful of years," he "would not feel great about that inverse relationship." He won't make the call: "You can't just bank on it anymore."
  21. (33:00–35:30) P, dollar: Japan was "a bit of a tell": the US was "nervous about… local dollar strength" and wanted "a bleed valve." End-of-dollar arguments are "compelling" but not "investable"; gold or options are "totally" sensible. Out-years lean to "dollar weakness," but near term there is risk of a "short covering bid" from "reshoring" and balance-sheet reduction ("wars talking about reducing the size of the balance sheet," C2), with "dollar shortage… funding markets… hiccup." Authorities "do not want a dollar substantially stronger," so strength "could be a tell."
  22. (35:30–36:30) P, EM local: "Low currency valuations combined with high real rates… pretty uncommon." "Stack the deck." "Start small and look for opportunities to add."
  23. (36:30–40:00) P, structured credit: "I don't love duration. I don't love straight up credit risk." Structured credit ranges from "car loan ABS that's a year or two long" to student loans, non-agency mortgages and "3% coupon agency mortgages." GMO focuses "up the capital structure": "AAA CLOs, student loans… seasoned… two-year and in duration." Multi-asset is "overweight there and underweight vanilla credit risk whether that's high yield investment grade or EM." "Predictability is a good thing." "Risk keeps trading at the moon."
  24. (40:00–43:30) F/P, BDCs: Public BDCs trade at discounts and private ones at NAV: "Well, what's NAV, right?" Lehman "2014 bonds" were unsellable in 2008. BDC 1.0 saw "return of capital… not return on capital" and winners sold, leaving "the dregs." "Skeptical of anything that is purely yield based," which is "inherently selling volatility." "I personally wouldn't be advocating for it." F: issuing at 1.3× NAV resembles MicroStrategy. P: similar to "Scott Bessent… buying discounted assets out the curve at 50–60 cents on the dollar… and reissuing" (C5).
  25. (44:00–46:00) P, curve and hedge: "Probably up for more flattening." "Real rates are staying high." "10-year, one-year forward is up. Break even inflation is not." "Not an inflation scare… even though inflation is rising." "It's not term premium" (else fiscal concerns). "Adjust to a higher rate regime." If rates keep rising, "it's going to hurt risk." GMO prefers "the belly." "Risk will perform well until the curve re-steepens… when it re-steepens that it happens." "We're flattening right now." Best when risk comes off: "3 to 5 year area."
  26. (46:00–47:30) P, what drives long yields: Breakevens have "not moved materially wider." "Energy is rising… inflation is not going the way the Fed wants." Real returns of "2 to 3%… over 10 or 30 years", which "we haven't seen… in a very long time." Then: "We've already had the term premium rise. We've already had the break evens expand" (C4). "Market adjusting… to a higher equilibrium." "Competition for capital… AI onshoring is… mammoth."
  27. (47:30–50:00) P, second act and fiscal: Once inflation falls, the Fed must "keep rates low… nominal growth over nominal interest rates" to buy fiscal time. Entitlements "have to be adjusted." Social Security "zero hour… sometime in the mid-30s." Fixes: higher nominal growth ("letting inflation run slightly higher"), tax reform and entitlement reform.
  28. (50:00–51:30) P, bond-market capacity: Funding Social Security from the general budget "would be a real challenge" at "$40 trillion and whatever 10% of GDP" (C3). "We make financial assets for other people to buy… our greatest export." A shift in marginal foreign demand matters. But "I would have said that at 75% debt to GDP," and "this thing is proving much more resilient."
  29. (51:30–53:30) P, long-run rates: "Long rates follow nominal GDP." "I don't think we're wildly out of bounds where we are." Long-end holders are often price-insensitive liability hedgers. The relevant rates are "10 years and the slope," because they tie to the affordability of leverage. "In a flat environment… that's when things get squeezed."
  30. (54:00–56:00) P, summary: A "K-shape in credit." Higher-quality, termed-out and myopic-return borrowers (large companies, "the AI complex") keep levering. Weaker borrowers are constrained ("weakness in triple C spreads"; refinancing from low coupons), and consumers pay "mortgage rates… in the sixes and auto loans… in the sevens." "Rates may need to move higher than they would otherwise," which "adds risk" and eventually hits risk assets as the wealth effect unwinds and "builds on itself."

SYSTEMS / VALUE-CHAIN MAP

FED HIKE (Sep; ~3.9% per Farley; dots 4.1% YE and through 2027, as spoken) │ ┌────────────────────────────────┴────────────────────────────────┐ ▼ ▼ BANKED CHANNEL (linear) NON-BANK / FIXED-RATE CHANNEL (weak) floating cost of capital private credit: yield + fees ≈ "zero or negative" CoC → spreads widen, lending tightens fixed-rate IG bonds: termed-out borrowers insensitive │ │ ▼ ▼ BOTTOM OF THE K ("collateral damage") TOP OF THE K (easy credit) consumers: mortgages ~6s, used auto >7% large caps, hyperscalers, "AI complex" CCC: clearing coupon >10–12%; spreads "substantially wider" IG supply +$200B q/q-yoy forecast; EM debt mkts for DCs floating-rate private-credit borrowers (6x+), coverage ↓ AI capex ~+6% nominal ("pick a number") software loans: no hard assets → recoveries "donuts" "generic AI risk" (chips→Oracle→SpaceX→project finance) │ trades cheaply in CDS (C3) │ │ └────────────── Demand not cooled enough at the top ──────────────┘ ▼ FED "may need to move higher than otherwise" (no number given) risk: overshoot = mistake; "higher they go, more they ease after" │ ┌─────────────────────────────────────┼──────────────────────────────────────┐ ▼ ▼ ▼ OPACITY LAYER CURVE DOLLAR / FISCAL no marks → no forced-seller signal flattening now (ISM/surveys) out-years: USD bleed valve ↓ bilateral restructurings hide 10y 1y-fwd ↑, breakevens flat near term: squeeze risk (reshoring, QT) defaults; secondaries ↑; insurers → "higher equilibrium" (C4) second act: rates < nominal GDP = "last marginal buyer" RISK-OFF ON RE-STEEPENING SS trust fund mid-2030s; repression/YCC option BDC discount vs private NAV 3–5y belly = best risk-off sector │ ▼ GMO POSITIONING (their view): low risk; BB in HY; OW short seasoned structured (AAA CLO, seasoned SL, non-agency, auto ABS, ~2y dur); UW vanilla HY/IG/EM beta; EM local (cheap FX + high real rates), start small; stock–bond hedge unreliable if infl >3%

Feedback loops (Inference from Source):

  1. Transmission-dilution loop: Weak transmission at the top of the K requires more hikes, which cause more damage at the bottom. The Fed overshoots relative to what a banked system would need, and the policy mistake risk rises ("absolutely real").
  2. Opacity-delay loop: No marks means no forced sellers, which means losses are recognized late. Private-credit inflows continue ("last marginal buyer"), which extends the cycle and makes the eventual adjustment sharper when marks or secondaries reveal prices.
  3. AI crowding loop: Price-insensitive AI borrowers absorb capital ("mammoth"). Real rates and capital costs rise for everyone else, which amplifies the K. It also raises the whole AI complex's correlation ("generic AI risk"), which cuts idiosyncratic diversification in credit.
  4. Growth-flattening-then-steepening loop: Improving ISM and hikes flatten the curve while risk "trades at the moon". Cracks form, the Fed eases, the curve re-steepens and risk sells off. This is the same sequence as Snider's bear case.
  5. Fiscal second-act loop: Once disinflation arrives, the Fed has to hold rates below nominal growth. That creates financial-repression and dollar-weakness risk in the out-years (GMO's EM local and non-US tilt).

SECOND AND THIRD-ORDER EFFECTS

Chain A: Weak transmission → higher terminal → bottom-of-K credit events

  1. First order (Source): Non-banks and fixed-rate borrowers are insensitive, the Fed may need to go "higher than otherwise", and consumers and CCC take the hit.
  2. Second order (Inference): Hikes above the ~4.1% dots (External: live dots and pricing) would hit floating-rate private-credit borrowers (6x+) and CCC refinancings first. Default rates rise from "benign" and are "higher in a year" (Source).
  3. Third order (Inference): Consumer-credit stress (auto >7%, mortgages in the 6s) and a small-cap, floating-rate equity drag (the IWM read in the brief is Inference) precede any IG or AI stress. The K widens before it breaks. Watch: CCC vs BB spread gap; used-auto and card delinquencies; LME and default counts (External). Falsifier: CCC gap narrows and consumer delinquencies stabilize despite hikes, which would mean transmission is not as uneven as claimed.

Chain B: Private-credit opacity → delayed recognition → secondary and BDC price discovery

  1. First order (Source): No marks; bilateral restructurings; rising secondaries; insurers buying; public BDCs at discounts while private BDCs sit at NAV ("what's NAV?").
  2. Second order (Inference): The public-BDC discount to NAV and secondary pricing are the only quasi-market marks. They become the leading indicator of whether a "correction already" is under way.
  3. Third order (Inference): If discounts widen and secondary bids fall, expect redemption and gating pressure on non-traded vehicles and a 2008-style "sell the winners, keep the dregs" dynamic (Source's BDC 1.0 lesson). Watch: public BDC price/NAV; non-traded BDC redemption requests; secondary fund pricing; PIK income share (External). Falsifier: discounts close and PIK and non-accruals stay flat, which argues against a hidden correction.

Chain C: Software loan recoveries near zero → sector repricing (IGV read-through, Inference)

  1. First order (Source): No hard assets, straight-lined EBITDA, over-levered; recoveries "donuts" for some; redundancy risk from AI (Claude named); stock-based compensation not counted.
  2. Second order (Inference): Software is often cited as the largest sector in private-credit and leveraged-loan books (External check needed; not spoken). Loss-given-default assumptions near zero recovery would reprice the cohort and lender vehicles more than headline default rates suggest.
  3. Third order (Inference): Equity read-through to IGV is via (i) the AI-redundancy narrative applied to mid-cap and PE-owned software, (ii) take-private and exit comps and multiples, and (iii) a stock-based-compensation scrutiny spillover. It is not via IGV constituents' own balance sheets, which are largely large-cap and lower-levered (External check needed). Watch: software BSL prices; software restructurings and recoveries; take-private multiples; IGV vs SPX relative performance (External, IBKR/TV only). Falsifier: software loan recoveries print in normal ranges (30–50%+) in upcoming workouts.

Chain D: Flattening → re-steepening = risk-off (joint signal with Snider and Bianco)

  1. First order (Source): "Probably up for more flattening." Risk "will perform well until the curve re-steepens." The 3–5y belly is the best risk-off sector.
  2. Second order (Inference): An Oct 28 hike (Bianco: welcomed by bonds) and Snider's long-end-led flatten both extend the flattening phase. In Peabody's frame that phase is risk-on or benign, not the danger zone.
  3. Third order (Inference): The desk's alert should key on re-steepening authorship. A bull steepener (front end rallying as cuts get priced) is Snider's bear case and Peabody's risk-off. A bear steepener (long end selling off) is Bianco's credibility failure. Both are negative for risk, through different channels. Watch: 2s10s and 5s10s direction and Δ2y vs Δ10y attribution; 3–5y sector relative performance (External). Falsifier: the curve re-steepens and risk rallies, as in a soft-landing reflation.

Chain E: Stock–bond hedge unreliability → portfolio construction shifts

  1. First order (Source): If inflation is >3–3.5% for years, "you can't just bank on" duration hedging. GMO shifts to short structured credit, long-short and EM local.
  2. Second order (Inference): Demand rises for short, high-quality spread product (AAA CLOs, seasoned ABS) as the "hide-out" asset. Spreads there could compress, eroding the valuation case GMO cites. Crowding risk.
  3. Third order (Inference): Allocator rotation away from long duration reinforces the "higher equilibrium" in long real yields (Source: 2–3% real), which feeds back into higher capital costs for the bottom of the K. Watch: AAA CLO spreads vs history; stock–bond rolling correlation; real-yield level (External). Falsifier: inflation falls toward 2% and the stock–bond correlation turns negative again (Peabody: "some of that should come back").

SCENARIO FRAMEWORK

Probabilities are analyst inference for research prioritization. They are not Peabody's and not desk allocations.

Bull, soft transmission catch-up (Inference ~20%)

Base, Peabody's map (Inference ~45%): higher for longer, K widens, defaults grind higher, eventual re-steepening

Bear, overshoot and opacity unwind (Inference ~20%)

Alternate, disinflation arrives early and the "second act" starts (Inference ~15%)

Invalidation markers for Peabody's map: (i) non-bank lenders visibly tighten in step with banks, so transmission is not broken; (ii) AI and large-cap issuance slows sharply on price, so the top of the K is rate-sensitive after all; (iii) CCC gap and consumer delinquencies improve during hikes; (iv) software loan recoveries normal; (v) risk sells off during flattening or inversion rather than re-steepening; (vi) breakevens surge, making the long-end move an inflation scare after all.


COMPANY / ASSET WATCHLIST

No buy/sell. No target weights. Monitoring hypotheses only. Named by the speaker: Oracle, SpaceX, Crocs, MicroStrategy, JPMorgan (default report), Lehman, pets.com, Claude (AI model). ETF mappings are Inference.

Cluster Names / instruments Thesis (Source anchor) Metrics Catalysts Risks / falsifiers
Bottom-of-K credit CCC/HY indices; floating-rate leveraged loans CCC coupon >10–12%; spreads "substantially wider"; defaults higher in a year (10:00; 24:00) CCC–BB gap; default and LME counts; JPM default report (External) Oct 28 FOMC; quarterly defaults Gap narrows during hikes
Private credit / BDCs Public BDCs; non-traded BDCs; secondary funds Opacity; 6x+ leverage; "last marginal buyer"; "what's NAV?" (19:00–22:30; 40:00–42:30) Price/NAV discount; PIK share; non-accruals; redemption requests (External) Q3 BDC earnings (External dates) Discounts close; non-accruals flat
Software credit → IGV (Inference) Software BSL/private loans; IGV Recoveries "donuts" for some; AI redundancy; SBC (24:30–26:30) Software loan prices; workout recoveries; take-private multiples; IGV vs SPX (External) Restructuring announcements Normal recoveries; indirect link to IGV constituents
AI complex credit Oracle, SpaceX (named); hyperscaler IG; data-center project finance; chips Price-insensitive borrowers; IG supply +$200B; "generic AI risk" cheap in CDS (12:30–14:30) AI issuer CDS co-movement; IG AI-issuance share; new-issue concessions (External) Big AI bond deals Idiosyncratic pricing returns; issuance slows on price
Consumer credit Mortgage rates; used-auto loans; home sales Mortgages ~6s; used auto >7% for HQ borrowers; home sales slowing (02:00; 12:00; 54:30) Rates; delinquencies; existing-home sales (External) Data releases Delinquencies stabilize
Small caps (Inference, from brief) IWM Floating-rate, lower-quality side of the K Interest coverage; IWM vs SPX (External) FOMC Transmission proves even
Curve / hedge 2s10s; 5s10s; 3–5y belly; 10y 1y-fwd; breakevens More flattening; risk-off on re-steepening; belly best; 10y fwd up, breakevens flat (44:00–46:30) Curve direction and attribution; breakeven vs real-yield split (External, IBKR/TV) Oct 28 FOMC; ISM Risk-off during flattening
Structured credit (GMO OW) AAA CLOs; seasoned student loans; non-agency RMBS; auto ABS Short (~2y), predictable, less correlated (37:30–39:30) AAA CLO spreads; seasoned ABS spreads (External) Issuance cycles Crowding compresses the value case
EM local (GMO) EM local-currency sovereign debt Cheap FX plus high real rates, "uncommon"; start small (35:30–36:30) EM real yields; FX valuation (External) Fed path; USD Near-term USD squeeze (Source)
Dollar DXY; USDJPY Out-year weakness; near-term squeeze from reshoring/QT; strength → policy tell (33:00–35:00) DXY; funding spreads (External) QT decisions Sustained USD strength without policy response
Fiscal / Treasury ops Treasury buybacks; long-end supply Bessent buyback claim (C5); rates < nominal GDP "second act" (43:30; 47:30) Buyback operations; nominal GDP vs 10y (External) Refunding announcements Claim not verified

DILIGENCE QUESTIONS / RESEARCH AGENDA

  1. Chair identity (External): Confirm who chaired the September FOMC press conference and whether the vote was unanimous. Do not use the ASR "Wsh/Worsh/Walsh".
  2. Dots and the policy rate (External): Verify Farley's "3.9% now; 4.1% YE; 4.1% through end-2027" against the September SEP (target range vs effective rate).
  3. Live pricing (External, own sources): Oct 28 and December hike odds. Cross-check with the Bianco memo's ">50%".
  4. Read the GMO "Triple Mandate" paper (External) for its quantitative backing of the transmission argument, bank vs non-bank share of credit, and elasticity estimates.
  5. CCC metrics (External): CCC yield vs spread, reconciling Peabody's ">10–12% coupon" with Snider's "~1,100 bp spread" as separate metrics; CCC–BB gap history.
  6. Private-credit stress (External): Public BDC price/NAV discounts, PIK share, non-accruals, non-traded redemption requests, secondary pricing, insurer buying.
  7. Software loan recoveries (External): Recent software workouts and recovery rates; software share of private-credit and BSL books; map to the IGV narrative. Keep this separate from IGV constituents' own credit.
  8. AI issuance (External): Source of the "+$200B IG supply next quarter vs last year" forecast; the AI share; EM-market data-center financing; AI-issuer CDS co-movement (tests "generic AI risk").
  9. Curve (External, IBKR/TV): Confirm flattening "right now" as of 17 Sep and after; 10y 1y-forward vs breakevens; set a re-steepening alert shared with the Snider scorecard.
  10. Term-premium contradiction (C4): Decide which statement to underwrite. Check term-premium models (ACM/KW) for the recent move.
  11. Consumer rates (External): Mortgage rates "in the 6s", used-auto >7% for high-quality borrowers, home-sales trend, delinquency data.
  12. Anecdotes (C5): Crocs "AI company"; Bessent buybacks at 50–60c "adding to the debt ceiling"; "10y… 5%… not since 07". Verify or drop.
  13. Fiscal: Social Security trust fund depletion date (mid-2030s, as spoken); debt and deficit figures behind "$40T / 10% of GDP" (C3).
  14. Cross-memo table: Maintain one table of Bianco, Snider and Peabody predictions for breakevens, real yields, CCC, curve authorship and AI financing after Oct 28.

RISK ANALYSIS

Risk Type Notes
No speaker labels (C1) Source integrity Host summaries (3.9/4.1, 7% AI, riskiest deals to private credit) could be mis-attributed to Peabody.
Chair name garbled (C2) Source integrity Not asserted. Corroboration from Snider's ASR is not verification.
Ambiguous figures (C3) Measurement $40T / 10% of GDP; "pick a number 6%"; CDS "cheaply"; unnamed IG forecast source.
Term-premium contradiction (C4) Thesis Undermines precision of the "higher equilibrium, not term premium" claim.
Opacity cuts both ways Thesis / timing "Correction already" is unfalsifiable near term, so it can be neither confirmed nor refuted quickly.
Value-shop bias Thesis Self-described skeptic ("sometimes to a fault"), and early before: "risk keeps trading at the moon."
Software → IGV linkage Thesis Loan-market commentary; the equity transmission is indirect.
Crowding in defensive structured credit Execution / external The hide-out trade may compress its own value.
Dollar two-way risk External The near-term squeeze contradicts the out-year weakness thesis for EM local timing.
Recording lag (C6) Timing Recorded 17 Sep. Pre-dates the Snider (22 Sep) and Bianco (24 Sep) tapes and any market moves since.
Companion contamination Process Do not merge Snider's CCC bp, Bianco's odds or 22 Sep inflation averages into Peabody's scoreboard.
No trade mandate Mandate GMO positioning ≠ desk recommendation. No buy/sell.

Thesis risk: Transmission may be more even than claimed (non-banks tighten, AI slows on price). Timing risk: The K can widen for quarters before a re-steepening. Execution risk: None; no tickets. External risk: dots, chair identity, credit indices, BDC data, AI issuance.


APPENDIX A: SOURCE VS INFERENCE QUICK KEY

Claim Tag
Third mandate = credit transmission; less banked; fixed-rate bonds loosen policy link Source (00:30–01:30)
Easy conditions: spreads tight, equities high, vol subdued, USD local low / REER rich Source (01:30–02:00). Levels as of 17 Sep, External
AI/large caps immune; consumers (6s mortgages, >7% used auto) and CCC collateral damage Source (02:00–03:30; 11:30–12:30; 54:00–55:00)
Non-banks' "zero or negative" cost of capital; haven't tightened "thus far" Source (06:30–07:30)
Fed at ~3.9%; 4.1% YE; 4.1% through 2027 Source, host statement (09:00–09:30). External
Fed may need to go higher; risk of overshoot "absolutely real" Source (08:30–10:30; 15:30; 55:00)
">4.1%" as a number Inference (Monica's angle), not spoken
"Could be in the midst of a correction already" Source, hedged (00:00; 19:00)
Private credit 6x+; bilateral restructurings; secondaries; insurers; last marginal buyer Source (19:00–22:30)
CCC coupon >10–12%; CCC spreads substantially wider; defaults higher in a year Source (10:00; 24:00–24:30)
Software recoveries "donuts" for some; redundancy; SBC Source (25:00–26:30)
IGV negative read-through Inference
IG supply +$200B; generic AI risk (chips/Oracle/SpaceX/project finance) cheap in CDS Source (13:00–14:30). Forecast source unnamed
GMO: low risk; BB; OW short structured (AAA CLO etc.); UW vanilla beta; EM local Source (27:00–39:30)
Stock–bond hedge unreliable if inflation >3–3.5% Source (32:00–33:00)
More flattening; risk-off on re-steepening; 3–5y best risk-off Source (44:00–46:00)
Not inflation scare / not term premium vs "term premium rise… breakevens expand" Source, internally inconsistent (44:30 vs 47:00)
USD out-year weakness; near-term squeeze; policy tell Source (33:00–35:00)
Second act; SS mid-2030s; long rates ≈ nominal GDP Source (47:30–53:30)
Scenario probabilities; chains; cross-memo links Inference

APPENDIX B: AS-SPOKEN NUMERIC LOCK LIST

Item As spoken
Policy rate "basically 3.9%" (host)
Dots "4.1%" by year-end; "4.1% until the end of 2027" (host)
Mortgages "6% something" / "in the sixes"
Used-auto loans "north of 7%" for high-quality borrowers / "in the sevens"
Hypothetical AI borrowers at 7%; mortgages at 7–8% "crush" housing (host)
CCC clearing coupon "well north of 10% probably north of 12"
IG supply "about 200 billion higher over the next quarter than… last year" (forecast)
AI investment "pick a number 6% annualized," nominal
10y "flirting with 5%, which we didn't hit since 07"
Private-credit leverage "six plus times"
Recoveries "won't be 50, 40, 30… donuts" (some companies)
Inflation threshold "3 1/2%… above 3%… next handful of years"
Real returns "2 to 3%… over 10 or 30 years"
Structured credit duration "two-year and in"; auto ABS "a year or two"
BDC premium "1.3 to NAV" (host)
Buybacks "50 60 cents on the dollar" (Bessent, claim)
Fiscal "$40 trillion and whatever 10% of GDP"; "75% debt to GDP" (past reference)
Social Security trust fund "zero hour… mid-30s"
Curve sector "3 to 5 year" best when risk comes off

APPENDIX C: CROSS-MEMO LINKS

Topic This memo (Peabody, rec. 17 Sep) Bianco 27 Sep (2026-09-27-bianco-economy-changed-forever.md) Snider 23 Sep (2026-09-23-eurodollar-snider-bond-market-flatten.md)
Next hike Dots ≈ one more to 4.1% (host); risk of more Oct 28 >50% priced; hike welcomed by bonds "Very likely… maybe next month"; dots ~one more
Long-end driver Higher equilibrium / real rates / capital demand; not an inflation scare (C4) Policy too easy Post-hike long-end dip = demand destruction
Curve More flattening; risk-off on re-steepening; 3–5y belly Hike implies flattening 2s10s ~20 bp via long end; invert → front-end steepener = bear
Credit stress CCC coupon >10–12%; K-shape; private-credit opacity n/a CCC ~1,100 bp spread; AI-infra financing harder to distribute
AI financing Borrowers price-insensitive; IG supply +$200B; generic AI risk cheap in CDS n/a Financing pricier and harder to distribute → reflation flips to tightening
Inflation / hedge Stock–bond hedge unreliable if >3–3.5% Goods won't deflate (22 Sep: stuck 3–4%) TIPS "comparatively benign"
Chair ASR "Wsh/Worsh/Walsh" (unverified) Powell "was" chair (implies a successor) ASR "Kevin Warsh" Jackson Hole

Shared tie-breakers for all three: (1) the breakeven vs real-yield decomposition of long-end moves; (2) re-steepening authorship; (3) whether CCC and AI financing stress widens together or diverges. No number import across memos.


End of memo. Markdown only. Saved to /workspace/pm-memos/2026-09-27-monetary-matters-peabody-fed-losing-credit.md. Published to the desk library on 27 Sep 2026: https://andrepow.here.now/memos/peabody-gmo-fed-losing-credit/. Not emailed. Not messaged. No market data fetched. Not advice.

Desk copy · not a trade recommendation · Erica · 27 Sep 2026