Title: We’re in a Debt Bubble — Sticky Inflation, Valuation Discipline, Short Duration, Hard Assets, and AI CapEx Write-Off Risk
Author / source: Jonathan Wellum, CEO & CIO, RockLinc Investment Partners; interview with Maggie Lake, Wealthion
Source title: We’re in a Debt Bubble — And Higher Rates, Inflation and AI Spending Are Adding Pressure
Source URL: https://www.youtube.com/watch?v=IpLLx-ViaEM
Video ID: IpLLx-ViaEM
Published: 2026-09-21 (Mon Sep 21, 2026; YouTube upload_date 20260921)
Duration: 29:40 (1,780 seconds)
Tape context (show): Filmed Sep 11, 2026 (9/11 anniversary remark); Fed reaction “next week” = then mid-Sep FOMC week — not memo-date “today.”
Memo date: Tuesday, September 22, 2026 (America/Toronto)
Transcript path: /workspace/youtube-transcripts/IpLLx-ViaEM.md · Brief: /workspace/youtube-transcripts/IpLLx-ViaEM_brief.md · Meta: /workspace/youtube-transcripts/IpLLx-ViaEM.json
Caption / ASR caveats: YouTube automatic ASR only (en-orig timedtext json3 via yt-dlp). No manual English captions (only live_chat). Proper nouns and numbers are provisional spoken approximations — apply ASR locks below. Timestamps are approximate cue-group starts.
Product: Regime / systems map for allocation research. Not investment advice. Channel frames content as educational/informational. This memo contains no buy/sell recommendations — hypotheses and watchlists only. Wellum’s portfolio expressions are source expressions, not desk recommendations.
Source discipline: Primary sources are this transcript, desk brief, and slim JSON only. Brief used for ASR locks, orientation, and metadata. Do not silently merge companion desk memos (Bianco, Jensen Huang, Asianometry, Mike Green/Bessent, Excess Returns Janus, Alma/Jordi). Companion themes may be flagged as External / other desk memo without importing figures. Any fact not spoken here = External check needed. Wealthion Membership mid-roll (~10:26) and free portfolio-review pitch are sponsor / promo — ignored for desk analysis.
ASR name / number locks (from brief + transcript context):
| ASR heard | Intended |
|---|---|
| Jonathan Wellm / Wellm | Jonathan Wellum |
| Rocklink / Rock Link | RockLinc Investment Partners |
| Wealthy / Wealthy on | Wealthion |
| delobalization | deglobalization |
| distillance | distillates |
| Grunlack | Jeffrey Gundlach |
| Honey… Honey Badger | Honey Badger (silver / critical-minerals co.) |
| Eric Sprat | Eric Sprott |
| Komach | COMEX (Hunt silver-squeeze lore) |
| dd dollararize | de-dollarize |
| germanian | germanium |
| Filming Sep 11 → Fed “next week” | Context is mid-Sep 2026 FOMC week, not memo-date “today” |
| Curve / fiscal / debt figures (~$350T global; ~$5.5T US revenue; ~$40–42T debt; 3.75% × debt → ~$1.58–1.6T interest; ~26% of ~$6T revenue path; SS ~22%; 10y ~5% / 30y 5.3–5.4%; short end ~4%; Trump $5k dividend ~$1.2T+; energy demand ~3–3.5%; Prairie Creek >$300M equipment vs ~$300M road; Ag ~$50 early ’80s) | Spoken approximations — confirm vs primary fiscal / market sources before desk use |
How to read: Restatements of the talk = Source. Interpretive links and underwriting judgments = Analyst inference. Speaker-stated numbers are used as spoken and attributed; they are not independently verified.
Stance one-liner (source-locked): Debt bubble + sticky inflation → valuation discipline, short duration, hard assets / pricing-power; AI CapEx boom will write off; bonds only after rate peak.
Takeaway 1 — Regime is a debt bubble under sticky (not runaway) inflation. Source (00:00–00:19, 01:38–02:45, 07:09–07:43, 15:25–18:12): Wellum frames the binding constraint as global / sovereign debt (~$350T cited) plus US fiscal math that can push interest toward / past Social Security’s share of the budget. Inflation is expected to stay challenging / a bit higher, not a hyperinflation call — drivers: fiscal overspend, large deficits, money-supply residue, COVID supply scars, deglobalization, and Iran / oil / distillate / diesel pressure. Conviction on “debt + sticky inflation matter” = High (clearly repeated); conviction on precise fiscal percentages = Low–Med (ASR / spoken approximations).
Takeaway 2 — Sticky inflation → P/E compression + prefer pricing power / moats; blind buy-the-dip index is dangerous. Source (02:45–04:25, 04:57–06:35): If inflation stays elevated, price/earnings ratios should come down, rates stay higher, and margins get squeezed unless firms can pass through costs. Mag7 / AI / data-center names already carry high valuations; token/algorithm pricing, ROIC, and ROE are uncertain. Index concentration is “long in the tooth.” Non-obvious angle: this is a value-discipline memo more than a pure inflation-trade memo — capital preservation and purchasing-power sit above momentum.
Takeaway 3 — AI revolution is real; CapEx boom implies overinvestment, write-offs, and damage — worse if rates and inflation are still rising. Source (00:00–00:19, 06:03–07:09): Trillions of CapEx → overspending / overinvestment → write-offs. Overlay on rising funding costs and debt stock = system pressure. Why now: show filmed around Sep 11 with Fed “next week”; AI spend narrative and sticky US inflation readings were concurrent (Source framing). Horizon: multi-year structural (3–5y for commodities; duration decision keyed to rate peak, not Fed week).
Takeaway 4 — Bonds: stay short now; lengthen only after a peak in rates / economic slowdown; long bonds were among the worst trades ~10y and will reverse eventually. Source (03:18–04:25, 07:55–10:28, 11:01–12:07): Curve levels cited ~5% 10y, 5.3–5.4% 30y, short end ~4%. TLT “eviscerated”; analyst anecdote ~2004 levels / ~22y of no money if bought then. Canada tax angle: low-coupon discount bonds (1–2y) so discount→par = capital gains (half interest tax rate). Prefer making money in equities vs guessing the curve — not Bill Gross / Gundlach.
Takeaway 5 — Hard assets / scarce essential businesses as purchasing-power and repression hedge; selective tech only if not grossly overvalued. Source (13:16–15:30, 17:44–23:01): Structural deficits in Cu, Ag, U (nuclear), ag, energy; energy demand growth cited ~3–3.5% off a large base after decades flat; grid needs rebuild (60s–80s stock). Gold as monetary substitute under expected financial repression + ongoing inflation to exit debt (US, Canada, France, UK, Japan, China). CB gold repatriation (Netherlands, Germany cited) = trust / collateral. Prairie Creek / Honey Badger / Hunt brothers = anecdote / color, not the allocation thesis.
Non-obvious angle: The memo’s center of gravity is funding-cost × valuation × CapEx write-off under a debt bubble — not a gold-bug pamphlet and not a “bonds are dead forever” call. Bonds are conditionally attractive after a rate peak; AI is conditionally dangerous because CapEx + rates + inflation coincide.
Why now + horizon: Sticky US inflation prints + elevated rates + exploding AI spend + climbing global debt (show blurb / cold open). Research horizon: near-term Fed/path of rates for duration timing; medium 2026–28 for fiscal interest-vs-SS path and AI CapEx digestion; multi-year 3–5y for commodity structural deficits. Conviction overall on stance package = Med–High as a coherent source map; Low on any single spoken number without External fiscal/market confirmation.
What this is not: Not a hyperinflation forecast. Not a call that the dollar disappears (Source: dollar remains important; de-dollarization is a trend, not collapse). Not a Prairie Creek investment memo. No buy/sell.
Chronological bullets. Timestamps from ASR cue groups. Membership promo skipped.
(00:00–00:32) Cold open / framing. “We’re in a debt bubble.” Overspending and overinvestment → write-offs and damage. Investors need preservation of capital and preservation of purchasing power. Maggie Lake introduces Jonathan Wellum (ASR: Wellm), CEO & CIO of RockLinc (ASR: Rocklink).
(00:32–01:05) Volatility / promo bridge. Host notes crazy market volatility; free portfolio-review pitch (Wealthion / RockLinc network) — ignore for desk. Lean on Wellum for inflation discussion.
(01:05–01:38) Stubborn inflation readings; method. US inflation readings this week “stubbornly high.” Wellum: lean on long-term discipline / bedrock principles; none know the future — work with probabilities.
(01:38–02:45) Inflation drivers (not runaway call). Not calling for massive inflation spike; depends on Fed/policy response. Drivers: fiscal spending / overspending, large deficits, money-supply addition, COVID supply-chain remnants, deglobalization, Iran situation putting upward pressure on oil; distillate capacity destroyed → expensive diesel. Portfolio implication begins: inflation a bit higher → P/E should come down.
(02:45–03:53) Valuations, pricing power, long bonds as worst trade. Higher inflation → rates up; harder cost pass-through → prefer businesses with pricing power, moats, industry leaders. Careful on fixed income / longer duration. Long bonds = worst trade ~10 years / among worst in recorded history — will reverse. Stance: stay short; when rates seem to peak, lengthen duration for appropriate clients. Prefer hard assets; protect purchasing power.
(04:25–05:31) Buy-the-dip narrative pushback. Host: retail conditioned to buy dips / “giant index.” Wellum: index trade great but long in the tooth; inordinate concentration — small number of high-valuation names carrying indexes. Need discernment, not knee-jerk index buy.
(05:31–07:09) Mag7 / AI / CapEx write-off risk. Mag7 + AI / data-center stocks at high valuations. If rates go a bit further and inflation runs, pricing pressures regardless of growth delivery. Uncertain: pricing of algorithms/tokens/AI products; ROIC / ROE. Trillions spent → real AI revolution but boom dynamics → overspending / overinvestment → write-offs / damage. Worse amid rising rates + hotter inflation. Pressure on massive system debt / funding costs; Fed reaction “next week” flagged.
(07:09–07:55) Debt bubble named; powder dry. Covering interest on accumulated debt worldwide — “we’re in a debt bubble.” Package: know what you’re buying; careful valuations; keep powder dry; harder assets; reasonable diversification; stay nimble. Host: prepare proactively before drawdowns/liquidations.
(07:55–09:24) Bonds not permanently uninvestable; curve levels. People extrapolate recent success → chase to the top. Curve cited: 10y close to ~5%; 5.3–5.4% on 30y; short end ~4% almost down the front. High funding costs → indebted households, weak housing, private credit pressure → eventual economic anchor / slowdown → rates can come down → bonds could be a great trade then.
(09:24–10:28) TLT eviscerated; watch for peak rates. TLT “eviscerated”; years ago Wellum told viewers to stay short, collect coupon, avoid duration risk pending policy response. Analyst: TLT back to ~2004 levels → ~22 years of no money if invested then. Eventually great trade — not saying do it right now; watch for peak in interest rates.
(~10:26–11:01) Wealthion Membership mid-roll. Sponsor / membership pitch — skip.
(11:01–12:41) How RockLinc uses bonds; Canada tax angle; not Gross/Gundlach. Often buy 1–2y bonds; in Canada, low-coupon bonds at large discount → small coupon + gain to par = capital gains (half interest tax rate) for HNW. Use for near-term cash obligations, dampened volatility, liquid powder. Kept short end since starting RockLinc. Prefer making money in equities vs guessing yield-curve changes — not Bill Gross or “Grunlack” (Jeffrey Gundlach). Host: 60/40 “dead” narrative may be exactly when to pay attention to bonds / duration into a recession — ask smart questions of advisors.
(13:13–15:25) Digitization → energy → Cu/Ag/U structural deficits. Digitization = AI, data centers, robotics, EVs → large energy demand. Canada/US energy demand flat for decades via efficiency; now growth ~3–3.5% off large base; grid built 60s–70s–80s needs rebuild. Can own some tech (they do) but many grossly overvalued; tech won’t grow without metals/minerals (copper, silver, etc.). Prefer commodities with structural deficits; 3–5 year horizon, not Fed-week trading; uranium via nuclear demand.
(15:25–16:33) $350T debt; grow-out skepticism; $5k dividend mindset. Concerned about ~$350T global debt; US debt “over 40”; higher rates → funding without further currency debasement hard. Don’t believe grow-ourselves-out for 40 years (Reagan-era framing). Favors Trump economic policies generally, but ~$5,000 dividend to every American adult ≈ $1.2T+ — “fell out of seat”; represents endless-money mindset.
(16:33–17:44) US fiscal math vs Social Security. ~$5.5T revenue; ~$40T / $42T debt. At 3.75% (below current short rates) on ~$42T → ~$1.58–1.6T interest. If revenue path ~$6T next year → interest ≈ 26% of revenue vs historical interest 8–10% of budget. Social Security ~22% of US budget → interest may surpass SS in a year or two. Exit requires discipline/virtue currently “dormant.”
(17:44–18:44) Gold as monetary substitute; repression; developed-world debt. Want a monetary substitute. Expect financial repression + ongoing inflation to exit debt. Not just US: Canada (less severe), France “dreadful,” UK “dreadful,” Japan, China debt issues.
(18:25–19:50) Demographics overlay. Japan deaths:births ~2:1 — “collapsing socially.” China almost as bad. Russia births:deaths ratio ASR “63” (ambiguous — treat as unverified). US still slightly >1. India ~2:1. Vicious circle: can’t support next generation → more economic pressure. Changes forced by choice or force → build resilient portfolios now.
(19:50–22:05) Reshoring, mistrust, gold repatriation, purchasing power. Inevitables: debt + inability to deal with it + higher rates + deglobalization costs. Reshoring necessary but painful (US and Canada manufacturing gutted). Nations don’t trust each other → gold repatriation (Netherlands/“Holland,” Germany) = want collateral. Canada–US trade war = trust breakdown signal. Filming on Sep 11 anniversary contrast to past solidarity. De-dollarization trends (dollar still important; not collapse camp). Yen carry / less appetite for Treasuries piled on → preservation of capital and purchasing power. Prefer businesses that can reprice, scarce, essential: infrastructure, commodities (Au/Ag/Cu), agriculture, energy.
(22:05–23:01) Defensive / secular allocation posture. Allocate to powerful, defensive, growing secular areas; still own other industries selectively.
(23:01–27:31) Prairie Creek anecdote (not thesis). Site visit NWT via Fort St. John: Prairie Creek, developed by Hunt brothers (silver ~$50 early ’80s), abandoned pre-production with world-class equipment (one machine 1.6 hours on meter). Now Honey Badger / CEO Chad Williams studying restart; >$300M equipment value vs ~$300M road; Ag/Zn/tungsten/germanium; diamond economy pressure in NWT; indigenous/gov relationships. Wellum owns “a few shares”; some risk-tolerant clients; Eric Sprott / Rick Rule also investors (ASR). Illustration of commodity risk/reward and physical diligence — anecdote, not core allocation thesis.
(28:55–29:40) Close. Stay focused on fundamentals that don’t change with every rate/inflation print; slow adjustments; long-term focus; opportunities despite challenges.
A. Macro driver stack (Source core).
Fiscal deficits + money-supply residue + COVID supply scars + deglobalization + Iran/oil/distillates → sticky inflation (not runaway). Sticky inflation → higher-for-longer-ish rates → (i) P/E compression, (ii) margin squeeze unless pricing power, (iii) higher sovereign / private funding costs on a ~$350T global debt stock. Inference: Wellum’s “debt bubble” is a funding-cost solvency/repression frame more than a sudden default call.
B. Equity selection filter (Source).
Index / Mag7 / AI / data-center concentration at high valuations → blind buy-the-dip fails the filter. Survivors: moats, leaders, pass-through, scarce/essential businesses that can reprice. Tech held selectively; many names “grossly overvalued.” Inference: This is a quality + valuation overlay under inflation, not a sector short of tech as a category.
C. AI CapEx subsystem (Source).
Real productivity/revolution narrative → trillions CapEx → boom-typical overinvestment → write-offs / damage. Feedback: if rates still rising and inflation hotter, CapEx ROIC fails harder → equity valuation pressure + potential credit/funding stress. Inference: AI is simultaneously a demand driver for energy/metals (Cu/Ag/U/grid) and a financial-cycle risk (write-offs). Do not collapse those two channels into one trade idea.
D. Duration / FI decision tree (Source).
Now: short duration (1–2y; Canada discount low-coupon for tax). Trigger to lengthen: perceived peak in rates and/or economic slowdown that re-opens rate cuts. Long end historically terrible (~10y / TLT ~2004 levels anecdote) but mean-reversion eventual. Inference: Bonds are a conditional call option on a growth/rate peak, not a permanent underweight. 60/40 “dead” media narrative treated as a contrarian timing risk by host (Source).
E. Hard-asset / repression channel (Source).
Debt math → cannot grow out easily → financial repression + ongoing inflation needed → gold as monetary substitute; CB repatriation = trust/collateral demand. Parallel structural: digitization energy demand ~3–3.5% + grid rebuild → Cu / Ag / U / ag / energy deficits on 3–5y horizon. Inference: Gold sleeve = monetary/regime hedge; industrial metals/uranium = physical bottleneck hedge to the same AI/digitization build that creates CapEx write-off risk in equities.
F. Demographics / geopolitics overlay (Source).
Japan/China/Russia demographic stress; US barely >1; India strong births. Reshoring + Canada–US trade war + wars/politics → mistrust → collateral preference. De-dollarization as trend, not dollar disappearance. Inference: These are slow constraints that raise the prior on repression/hard assets; they are not near-term trading triggers.
G. Anecdote node — Prairie Creek / Honey Badger (Source; non-thesis).
Hunt-era stranded Ag/Zn/W/Ge asset + restart study + named investors. Inference: Use only as process illustration (boots-on-ground diligence; commodity optionality with binary risk). Do not elevate to portfolio thesis or ticker recommendation.
H. Conflict / distribution layer.
Wealthion Membership + free portfolio review = promo (skipped). Wellum is CEO/CIO pitching a coherent RockLinc philosophy — treat as interested Source. Companion desk tapes (Bianco; Jensen; Asianometry silicon/compute; Green Bessent/Costanza; Janus 40y bull-over) = External / other desk memo — shared vocabulary (debt, AI CapEx, duration) ≠ shared model or numbers.
Chain 1 — Sticky inflation → P/E compression → index concentration pain (Source → Inference).
1st: Inflation stays “a little higher”; rates stay elevated (Source).
2nd: Required equity risk premia / discount rates up → multiples contract, especially where growth duration is longest (Mag7/AI) (Source).
3rd (Inference): Passive/index buyers who were rewarded for dip-buying face narrower leadership failure if a handful of high-valuation names re-rate together. Falsifier: inflation cools fast and Mag7 ROIC validates CapEx. Watch: sticky core services + AI name relative PE vs own history (External).
Chain 2 — AI CapEx boom + rising funding costs → write-offs → second-wave credit/funding stress (Source → Inference).
1st: Trillions spent into uncertain token/algorithm pricing and ROIC (Source).
2nd: Overinvestment → write-offs/damage, worse with rising rates (Source).
3rd (Inference): Write-offs can migrate from equity NAV to supplier/lessor/private-credit claims tied to data-center builds — amplifying the “debt bubble” funding stress Wellum flags. Falsifier: clear path to high ROIC on AI spend with stable funding. Watch: CapEx guidance vs incremental revenue; data-center financing stress markers (External).
Chain 3 — US interest share → SS-surpass path → repression / gold collateral demand (Source → Inference).
1st: Spoken math: ~3.75% on ~$42T ≈ $1.58–1.6T interest; ~26% of ~$6T revenue path; SS ~22% (Source — verify).
2nd: Political/moral capacity to tighten “dormant” → path of least resistance = inflate + repress (Source).
3rd (Inference): Policy mix that caps real yields / steers domestic savings into government paper raises the relative value of monetary substitutes (gold) and repricing real assets. Falsifier: credible multi-year primary surplus path without repression. Watch: net interest outlays vs SS in CBO/Treasury releases (External).
Chain 4 — Short-duration stance → missed rally if peak arrives; long duration too early → TLT-style pain (Source → Inference).
1st: Stay short; lengthen only at peak (Source).
2nd: Economy eventually anchors under high debt service (housing, private credit) (Source).
3rd (Inference): Timing error is two-sided — early duration extension repeats 2022–style damage; late extension misses the “great trade” Wellum anticipates. Desk research value is peak-detection checklist, not a static underweight. Watch: real activity softening + private credit stress + front-end pricing of cuts (External).
Chain 5 — Digitization energy demand + grid rebuild → structural metal deficits even if AI equity narrative disappoints (Source → Inference).
1st: Energy demand growth ~3–3.5%; grid aged (Source).
2nd: Cu/Ag/U needed whether or not AI equity ROIC works (Source logic: tech “won’t survive/grow” without metals).
3rd (Inference): Partial decoupling: physical bottleneck assets can remain supported by buildout while AI equity suffers write-offs — coherent with Wellum owning commodities more comfortably than rich tech. Risk: CapEx air-pocket hits both. Horizon explicitly 3–5y, not Fed week (Source).
Do not merge: Mike Green passive-marginal-buyer / Bessent buyback / Costanza long-end; Excess Returns Janus “40y bond bull over” instrument rebuild; Alma/Jordi Bessent clocks; Bianco rate/path calls; Jensen Huang AI supply narrative; Asianometry compute-scarcity = External / other desk memo. Shared words (debt, AI, duration, gold) ≠ shared arithmetic.
Horizons: near = weeks–one quarter from memo date (Tue 22 Sep 2026); medium = through 2027 and into the 3–5y commodity horizon Wellum emphasizes. Probabilities are Analyst inference for research prioritization — guest did not assign numeric odds.
Assumptions: Inflation stays sticky; rates stay high or grind higher before any peak; AI CapEx continues without clear ROIC proof; fiscal interest burden rises toward spoken SS-comparison path; geopolitics/mistrust support gold collateral demand; reshoring costs keep goods inflation alive.
Winners (hypotheses, not tickets): Pricing-power compounders; Au as monetary substitute; Cu/Ag/U and related energy/grid enablers on multi-year deficits; short-duration liquidity as dry powder.
Losers (hypotheses): Long-duration bonds entered early; richest Mag7/AI/data-center multiples; blind index dip-buying.
Leading indicators: Sticky US inflation prints; rising net interest outlays; AI CapEx guidance up while incremental monetization stays vague; CB gold buying/repatriation continues; private credit / housing stress without full risk-off in mega-cap tech (External confirmation needed).
Assumptions: Inflation “a little higher” and choppy, not runaway; Fed path uncertain (show’s “next week” already stale by memo date — External live path); AI remains real but write-offs begin in pockets while buildout still lifts energy/metals demand; RockLinc-style book stays short duration, selective equities, hard-asset overweight; bonds become interesting only when growth clearly slows and rates peak.
Winners: Balanced preservation book: short FI + scarce real assets + moat equities; patience on duration.
Losers: All-or-nothing “bonds dead forever” or “AI can’t go down” narratives.
Leading indicators: Soft patch in housing/private credit; first notable AI-related impairment / ROIC disappointment headlines; 10y unable to sustain breakout without growth break; commodity deficits persist on 3–5y project timelines (External).
Assumptions: Inflation cools quickly; Fed cuts; AI CapEx ROIC surprises positively and multiples hold/expand; fiscal panic narrative fades as revenues surprise; gold/commodities lag as real yields rise for “good” growth reasons or as dollar strengthens; duration rally starts before Wellum-style managers extend.
Winners: Long-duration bonds; high-valuation growth/AI leadership; index buy-the-dip.
Losers: Early hard-asset overweight if real yields spike for growth-positive reasons; short-duration opportunity cost into a sharp bull steepener/rally in TLT.
Leading indicators: Sustained disinflation; AI unit economics clarifying higher; CDS/fiscal stress markers quiet; commodity inventory builds (External).
Scenario use: Research prioritization only. Re-score after live inflation, fiscal, and CapEx prints — do not freeze spoken Sep 11 tape levels as current marks.
Hypotheses for diligence — not buy/sell tickets. Prefer baskets / factors over single names unless Source named them as color.
| Asset / factor | Thesis link (Source) | Metrics to watch | Catalysts | Risks |
|---|---|---|---|---|
| US short-duration Treasuries / T-bills / 1–2y | Stay short; liquid powder; Canada analog uses discount low-coupon | Front-end yields ~spoken 4%; roll yield; real short rates | Growth slowdown → later lengthening decision | Reinvestment risk if cuts arrive faster than peak call |
| Long Treasuries / TLT (monitor, not own-now per Source) | Eventually great after peak; currently “eviscerated” | 10y ~5% / 30y ~5.3–5.4% spoken; total return since 2004 anecdote | Clear rate peak + recession odds up | Extending too early; policy path uncertainty |
| Canada low-coupon discount bonds (1–2y) | Tax: discount→par as capital gains (half interest rate) | After-tax HNW spread vs interest-bearing | Client cash-outlay matching | Tax rule change (External); credit residual if not govvies |
| Pricing-power / moat equity basket | Pass-through under sticky inflation; avoid margin squeeze | Gross margin resilience; pricing surveys; leverage | Sticky inflation prints | Multiple compression still hits if paid too rich |
| Mag7 / AI / data-center complex (underwrite, don’t chase) | High valuations; ROIC/token pricing uncertain; write-off risk | CapEx/revenue delta; ROIC; impairments; funding costs | CapEx guidance vs monetization | Momentum can extend; timing of write-offs uncertain |
| Gold (monetary substitute) | Repression + inflation exit path; CB collateral/trust | Real yields; CB purchases; repatriation anecdotes | Fiscal stress / geopolitics | Opportunity cost if disinflation + strong dollar |
| Silver / copper / uranium | Structural deficits from digitization/AI/energy/nuclear; 3–5y | Inventories; mine supply; grid/nuclear project FIDs | Energy demand stay ~3%+ | Cyclical growth air-pocket; policy/permit delays |
| Ag / energy / infrastructure “repricers” | Scarce, essential, can reprice in messy world | Pricing power; regulation; volume | Reshoring / grid spend | Political windfall taxes; demand shock |
| Prairie Creek / Honey Badger (anecdote only) | Color: Hunt stranded asset; restart study; Wellum “few shares” | Restart study outcomes; road funding (~$300M spoken); Ag/Zn/W/Ge | Road / indigenous / gov progress | Binary mining risk; illiquidity; ASR name/ticker verify — not thesis |
Fiscal arithmetic audit (External): Reconcile spoken ~$5.5T revenue, ~$40–42T debt, 3.75% blended rate → ~$1.58–1.6T interest, ~26% of ~$6T revenue, SS ~22% vs latest CBO/Treasury / OMB tables. Which year, which interest concept (net interest vs gross)?
Global $350T debt (External): What definition (IFF, IIF, BIS, all sectors)? Trajectory and currency mix?
Curve marks at tape vs now (External): Confirm 10y ~5%, 30y 5.3–5.4%, front ~4% as of filming Sep 11 vs memo date Sep 22 — do not use ASR as live marks.
AI CapEx → write-off leading indicators: Which hyperscaler / semiconductor / data-center lessor metrics first show ROIC disappointment or impairments? Map private-credit exposure (External).
Peak-rate checklist: What combination of housing, private credit, unemployment, and breakevens would falsify “stay short” and trigger duration research (without making a trade call)?
Gold vs industrial metals split: Under Wellum’s two channels (repression monetary vs digitization bottleneck), which regimes favor Au over Cu/Ag/U and vice versa? (Analyst framing)
Demographics claims: Verify Japan ~2:1 deaths:births; China; Russia “63” ASR ambiguity; US/India ratios vs UN/national stats (External).
Prairie Creek facts (optional color only): Confirm Honey Badger entity/ticker, Chad Williams role, Sprott/Rule holdings, equipment/road figures — keep out of core thesis (External).
Iran / distillate / diesel channel: How much of sticky goods inflation in the tape week was energy/distillate vs services (External)?
Cross-desk non-merge: If comparing to Bianco, Jensen, Asianometry, Green/Bessent, or Janus Excess Returns memos, list overlaps as External / other desk only — no number import.
Thesis risk: Sticky-inflation + debt-bubble frame could be wrong if productivity (including AI) raises real growth enough to stabilize debt/GDP without repression, or if political discipline returns faster than “dormant” rhetoric implies (Source acknowledges grow-out attempts). Inflation could also spike beyond “a little higher,” breaking the moderate P/E-compression narrative into a sharper risk-off that hurts even hard-asset equities short-term.
Timing risk: Duration extension keyed to “peak rates” is inherently late-cycle and easy to miss or anticipate early — Source’s own TLT history is the cautionary tale. AI write-offs may arrive years after CapEx peaks. Commodity 3–5y horizon can underperform for long stretches of Fed-week volatility (Source accepts this).
Execution / expression risk: “Pricing power” and “hard assets” are easy to overpay for; Source repeatedly stresses valuation discipline. Canada capital-gains bond structure is jurisdiction-specific. Single-asset mining anecdotes (Prairie Creek) are high-risk expressions that Source himself treats as small/optional — desk must not promote them to core.
External / identification risk: ASR-only transcript — names (Wellum, RockLinc, Gundlach, Sprott, COMEX, germanium, Honey Badger) and numbers (fiscal %, curve, demographics, $350T) require verification before any client-facing or risk-system use. Tape filmed Sep 11 with Fed “next week”; memo date Sep 22 — policy path may have moved (External reopen). Sponsor content ignored but channel incentives remain.
Process risk: Merging this Source with Bianco/Jensen/Asianometry/Green/Janus models would create false precision — prohibited without explicit External labeling.
| Item | Spoken claim | Caveat |
|---|---|---|
| Global debt | ~$350T | Definition unclear — External |
| US revenue | ~$5.5T | Year/basis unclear |
| US debt | ~$40T / $42T | Gross vs other measures |
| Illustrative rate | 3.75% (below then short rates) | Not market quote |
| Interest $ | ~$1.58–1.6T | Arithmetic on above |
| Revenue path | ~$6T “next year” | Projection |
| Interest / revenue | ~26% | vs hist 8–10% of budget |
| Social Security | ~22% of US budget | Compare apples-to-apples |
| Trump $5k dividend | ~$1.2T+ if every adult | Mindset illustration |
| 10y / 30y / front | ~5% / 5.3–5.4% / ~4% | As of filming; ASR |
| TLT history | ~2004 levels; ~22y flat | Analyst anecdote |
| Energy demand growth | ~3–3.5% | Off large previously flat base |
| Prairie Creek | >$300M equipment vs ~$300M road; Ag ~$50 early ’80s; 1.6h on meter | Anecdote |
External checks needed (not verified in this memo):
- Live US Treasury curve and TLT total-return history vs 2004 anecdote
- CBO/Treasury net interest, revenue, debt stock, SS outlay shares
- IIF/BIS/other global debt $350T definition
- Iran/distillate/diesel price contribution to then-current inflation
- Demographic ratios (Japan, China, Russia ASR “63”, US, India)
- Honey Badger / Prairie Creek corporate facts (only if desk pursues color)
- Any cross-read to Bianco, Jensen Huang, Asianometry, Mike Green/Bessent, Janus Excess Returns — label External / other desk memo; do not import figures
- Post–Sep 11 FOMC outcome and path vs “Fed next week” tape line
ASR / identification flags (apply locks before quoting):
- Wellm → Wellum; Rocklink → RockLinc; Wealthy → Wealthion
- delobalization → deglobalization; distillance → distillates
- Grunlack → Jeffrey Gundlach
- Honey Badger / Chad Williams; Eric Sprat → Eric Sprott; Komach → COMEX
- dd dollararize → de-dollarize; germanian → germanium
- Russia “63” births:deaths — ambiguous; do not use without External confirm
- Membership mid-roll ~10:26 — non-analytic
Document control: Saved only to /workspace/pm-memos/2026-09-22-wealthion-wellum-debt-bubble.md. Do not publish / email unless AP explicitly approves a separate publish job.
End of memo.
Desk copy · not a trade recommendation · Erica · 22 Sep 2026