Long-duration bonds Negative — supply, term premium and corporate competition all pushing the long end higher; but a first sized dissent says the demand side of the ledger is simply unmeasured

TLT bleeding since 2022; the 60/40 offset is broken — bonds no longer hedge equities. Gundlach (Jun 12): avoid long-term developed-market government bonds — the 30-year Treasury sits at 20-year highs and has broken above its 1990–2023 ±2σ band; he sees no significant downside in yields "even if the economy goes into recession," and an entitlement-funding cliff (a major retirement program admitting it runs dry by 2032, likely ~2029) plus possible yield-curve control argue the band-break is structural. Thomas (Jun 21, contrarian bullish): "Treasuries are NOT Trash" — investors sit at record-low treasury allocations after bonds' bad run. Bonds fail in inflation shocks (2022 — use commodities for defense there) but win in deflationary downturns (falling activity, shrinking inflation, Fed cuts + QE); he'd gradually build bond exposure funded by trimming risk (new money to bonds, rebalance profits out of stocks), with an AI-capex bust + AI-driven deflation a potential tailwind. Hay (Jul 11): his fiscal risk-line — a decisive break above 4.6% on the 10Y / north of 5% on the 30Y would signal America can no longer keep living and spending beyond its means ("defense" included) — the defined kill-switch he attaches to deficit-funded theses like the LMT alert. Finucane (Jul 16/17): since 2020 we're in a structural higher-long-yield regime (10s/30s biased up on geopolitical volatility, stickier inflation and fiscal arithmetic — poor for bond prices, possibly for decades outside short-term trades); Oxbow stays in ≤3-year Treasuries/IG corporates and locked ~15% of income portfolios into 2-year paper above 4% as the 2-year jumped back over 4% — betting the Fed cuts the short end once base effects pull inflation toward 2% by next March/April. Singh (SSR, Jul 19): soft June CPI/PPI left rates lower midweek (crypto and gold bounced), but the 2-yr hit a 15-month-high 4.23% on hike risk into the Jul-29 FOMC (Waller hinted; intraday odds spiked to ~45%) — the 10-yr at 4.57% pushing mortgage rates to 6.55%, the highest since Aug-2025. His base case: a Fed on hold is the best near-term outcome for risk; energy the wild card. Hay (Jul 21): the 30-year Treasury yield — less critical than the mortgage-driving 10-year, but a "breakout and spike to, say, 6% would be problematic" — "appears to be on the verge of an upside range expansion above multi-year resistance." Premature to call a breakout yet (a yield breakout = the underlying bond prices breaking down), but a second bond-market canary he's monitoring alongside the widening BB/CCC junk gap. Paulo Macro (Jul 23): bonds selling off with the 30Y knocking on 5.20% — "no Risk Off hiding spot there" as equities, gold and metals get hit together in the same tape. Gromen (MacroVoices #542, Jul 23): "defense stimmies" are the new structural seller — within three weeks the US, UK, Germany, Korea and Japan all pivoted to borrow-and-build defense, turning the last 50 years' structural creditors (Japan, Germany, Korea) into borrowers, into war, supply-chain breaks and inventory rundowns; war is "always inflationary — there's never been a deflationary war ever in history," so it all says sell bonds, sell bonds, sell bonds until something is triggered (Japan, UK, Germany or the US — doesn't matter which) and forces yield-curve control: "once one starts, they're all going to need to," with the currencies then debasing together against gold (and the yuan) and local-currency equity melt-ups. His $64,000 question: whether Warsh plays Mr.-Tough-Guy inflation-fighter and hikes first — the catalyst for the rate spike that forces the endgame. Eisman (Jul 24): the 10-year reached 4.7% on the Iran escalation and $100 oil. Hay (Jul 26): the 30-year at 5.18% "is a breakout," grinding higher despite a "creaky… fraying around the edges" economy and benign inflation prints — because foreign central banks "backing away from the treasury market" shrank the buyer pool, so "it doesn't take as much… selling pressure to pop these longer term yields." A run to 6% "would be really bad news for the stock market." Lacy Hunt — "about as close to a perma bull on bonds as you could find" — has flipped to a secular stagflation view, with the 6–7%-of-GDP deficit heading toward 10% in the next recession as the sword of Damocles. Singh (Jul 26): a pundit turning constructive — real 10-yr yields hit their highest since 2023 and initial claims their lowest since 1969 before the weekend (July hike odds touched 40%), but with the de-escalation "the 10-year is down to 4.63%; I think it might have peaked at 4.7 if this war is over" — and BlackRock's Q3 fixed-income outlook says it's time to buy bonds ("they would have been right today"). Counterweight: 11 major central banks flipped from cutting (2.7%→2.4%) to raising (2.6%), their 10-yr sovereign yields 3.2%→3.8%. His expression of the peak is the agency/mortgage REITs — DX, RWT under $5, the ~10% RWTS baby bond, AGNC, NLY after its dividend raise to $0.75 (13.4% yield): "if you believe rates have peaked, all of them are pretty much adds here." Jikh (Jul 28): names the buyer, not just the yield — the US 10-year near 4.7%, close to all-time highs, is "partially" a Japan story. Japan was "the most reliable customer at US bond auctions" and the #1 foreign holder of US debt; with the 30-yr JGB at ~4% offering a guaranteed domestic yield with no FX risk for the first time in 30 years, Japanese institutions are reversing — the July 10 GPIF repatriation directive, and insurers funding their biggest JGB buying in three years by selling US Treasuries. Fewer buyers means Washington must pay up to attract replacements: "even if you might not own any Japanese assets, your mortgage rate is partially set thanks to Japan." Same mechanism Hay flagged Jul 26 (foreign central banks backing away from the Treasury market), sourced to the specific flow. Paulo Macro (Jul 30): "speaking of submerged beachballs everywhere… let's not forget about the long bond" — after Warsh's non-hike, "bonds will take Warsh at their word: 'you're the market, you do the tightening'… good and hard." 2026-AUG-03 (David Hay / Haymaker portfolio update): the 30-year US T-bond yield made a "clear breakout to an 18-year high" at 5.27%, raising "the very real possibility of taking out the resistance at just under 5.5% that runs all the way back to 2003"; above that "a bit of friction at 5.75%" and then nothing structural until 6.75% (late 1999). Haymaker marks its multi-year avoid-long-Treasurys call — deficit-scale issuance while "the largest foreign buyers have been exiting stage left" — as scored: "Alas, that is no longer a prediction but a reality." 2026-AUG-15 — Rick Rieder: takes credit risk and refuses duration risk — interest-rate exposure kept "under 3 years" — because the long end is where the global fiscal + AI-related supply gets paid for. AUG-16 (Jay Singh, SSR): the long end is still the pressure point even as the front end rallies — last week's 30-year auction "not being that great" at 5.3%, the highest since 2007, and the 10-year holding ~4.69% while a 5% jump in oil steepened the curve into Friday. He puts a policy floor under it: if the 30-year approached 5.5% and the 10-year 5%, after a short-term selloff "[Bessent] would effectively force Trump to pull back" — the pref book's drawdown capped at 3-5% in the riskiest issues and less in utility names. 2026-AUG-17 — Jay Singh (David Lin Report), dissent: bought TLT last week "for the first time in years" on the temporary-peak-in-the-10-year call — flat-to-slightly-lower long rates into the midterms, with a major war escalation the only named break. 2026-AUG-19 — Halftime (Terranova / Liz Thomas / Simpson / Santoli): After the 30-year cleared 5.30%, the Treasury doubled the size of its long-end buybacks. Terranova calls it yield-curve control with an expiry date: "going into a midterm election, do you really want to see mortgage rates rising if you're the party in power?… It'll be 60 days, early September right up to the midterm election." Liz Thomas isolates the trigger — "apparently 5.3% on the 30-year was that level" — and notes the 30-year isn't where earnings are discounted (the 10-year is), with 2011's Operation Twist as the analogue that "effectively put a floor in for the S&P 500." Two limits: Simpson — "no matter what the Treasury does or what the Fed does, they can't control the long end"; Santoli — it worked because it caught "saturating bearishness on bonds," but "we have to see what the half-life of this measure is." Options confirmed the flow: TLT call buying double puts (Renick). 2026-AUG-15 (Mark Newton, Fundstrat, Jimmy Connor): negative on the long bond into year-end — "a treasury decline where yields creep back to the highs and actually get close to 5%." US still "the best house in a bad neighborhood," but long yields are rising globally and the US is "a newcomer to this party." 2026-AUG-25/26: Hay's anchor: the 30-year historically tracks nominal GDP growth — above it nearly all of 1986-2007, below it since the GFC, making the last five years “one of the worst return phases for U.S. Treasury investors on record”; bondholders are “the bag holders for America's fiscal and monetary recklessness.” Per Jim Grant, publicly-held Treasury supply is +8.4% Y/Y — the reported deficit is “overly flattering.” (David Hay, Haymaker Daily 2026-AUG-25/26) 2026-AUG-26: Rule: the Treasury has moved from managing the short end to intervening at the long end — the rising long rate was savers repricing “the underlying rate of US inflation, which is different than the CPI,” and the intervention came because the long bond “hurts other people who vote” (mortgages, consumer credit, prime, “the high yield or junk market is in disarray”). (Rick Rule, David Lin Report 2026-AUG-26) 2026-AUG-19: Faber's lone-wolf contrarian call: “there's not enough yield in fixed income… rates could and should go higher, which I think would surprise a lot of people.” (Meb Faber, Cambria webinar 2026-AUG-19) 2026-AUG-28: Treasury at least doubled liquidity-support buybacks in the 10–20yr and 20–30yr sectors ($2bn → ≥$4bn per operation, effective early September); 10s −6bp and 30s −9bp on the news, then inside two sessions 30s were back above 5.27% and the move fully handed back — "$4 billion an operation against $2.1 trillion of deficit financing is a rounding error dressed up as a rescue," possibly funded straight from the ~$950bn General Account and branded a "Treasury Twist." Swap spreads narrowed to the tightest since February (HF basis positioning ~$305bn vs <$50bn in 2022) and long-bond call skew ran up against puts while short maturities stayed neutral — "the fear has inverted." A trend model puts ultra-long duration at a −2.28 z-score (most stretched short since 2024); 64 analogs since 2003 show yields lower 71% of the time over 120 days (avg −25bp) — but "a crowded short describes who owns what today and says nothing about who funds $2.1 trillion next year… the structural problem is untouched either way." (Contrarian Codex 2026-AUG-28) 2026-AUG-23 — Jay Singh (SSR call, 2026-AUG-23): the Treasury's defence of the long end failed in public. Bessent pleaded with the bond market three times in three weeks and doubled long-end buybacks to $4B per operation (from Sept 9) against $97B of net long-bond issuance — "the buys are really de minimis" — and the initial rally "completely unwound by Friday," leaving the 30-year near 5.29%, its highest since 2007, and the 10-year at 4.7%, in the week US debt crossed $40 trillion. JPMorgan: a "band-aid on a bullet hole" over a $3.5T+ funding gap; Goldman's Mike Mitchell expects term premium to "drift higher" and sees 5.2-6% on the 10-year as the level that either attracts value buyers or damages growth. Singh withdrew his own peaking call: "I had expected yields to effectively start to peak, but the new expenditures of 600 billion on missile defense… has made it harder to find a peak in yields." Kedrosky (Meb Faber #648, Aug 28) adds a demand-side driver for the long end: AI/data-center financing is now large enough to distort sovereign flows — "one of the reasons behind the bond freakout we just saw, because it's pushing up longer-term rates." The textbook crowding-out has inverted: lending to a data center with a hyperscaler's prime credit behind a 12-year renewable lease is more appealing than "this flawed credit known as the United States or the UK," so sovereigns' fundraising is being challenged by data centers' fundraising rather than the reverse. 2026-AUG-30 (Jay Singh, SSR call): the constructive expression is municipals, not Treasuries — "short of an economic collapse, we don't expect bond yields to move dramatically lower in the near term, although yields could peak, which offers investors a window of opportunity to access some of the highest yields in a generation." On a taxable-equivalent basis "New York muni bonds, when you divide by your marginal tax rate, can provide yields that are much higher than corporate yields or US Treasury yields," with the 10-to-20-year range most attractive. The contrarian cushion: 10-year futures positioning is "about as net short as they've ever been in history… and that's including people like Druckenmiller," and relative valuations "point to the best entry point for longer-duration bonds versus stocks in nearly a generation." His own hedge on it: "another string of hot inflation data, more likely given the rise of gasoline and diesel prices in August, could send rates even higher." Steve Eisman (The David Lin Report, 2026-AUG-28) identifies a crowding-out channel rather than a fiscal one: "part of the reason why the interest rates have gone higher is that the amount of debt being raised for AI is sort of crowding out everything else. So rates could go higher, which would hurt the housing market, it would hurt the economy — but the calamity that some people like to pontificate about, I don't think is realistic." Note the asymmetry: he holds the rate risk while rejecting the debt-spiral narrative, and refuses a level call entirely ("I have absolutely no idea what the cap in the 10-year is"). (New Money clip, 2026-AUG-15) supplies the credit-market corollary: "most of the new issuance of bonds is AI related," so the bond sleeve of a 60/40 no longer diversifies against the equity sleeve. Luke Gromen (Goldfinger Capital, 2026-AUG-14) states it as near-certainty rather than forecast: "there is no way they can do this AI build-out, infrastructure build-out, reshore without it being massively inflationary. The bond market has to basically go down. Long-term bonds in the United States are down 90% against gold in the last 10 years. They're going to have to go down at least 99% more against gold over the next 10 to 20 years if we continue to try to reshore and build — and I think we're going to, and that's fine. You just got to load up the suckers holding the stuff and then do it." The wartime-footing version of the same point: with the Fed financing $8trn a year at 3/8 of a percent and inflation at 30–50%, "which part of the long end of the curve do you want to own? Where do mortgages trade? What happens to the entire US banking system that holds bonds as collateral?" 2026-AUG-26 (Dowd/Phinance, WTFinance) — CONTRARIAN COUNTERPOINT to the theme's Negative view: "Most people think bonds are dead… I think one of the greatest asymmetric trades in the investment universe right now is the long end of the curve." Sovereign issuance and AI capex compete for the same capital so yields rise to attract it, then choke the economy; growth slows, the Fed cuts and long yields fall — "if I'm right on the recession and the AI bubble bursting, yields will come down in the long run." He concedes it "has gone against us recently," but "once this begins, it'll happen fast." Retail expression is cash; institutions can own the long end. 2026-SEP-01 (Stöferle/Incrementum): "I'm definitely not a buyer of European debt at these levels and also not US debt" — keep only ~15% in fixed income as a stabilizer and take it in EM local-currency and corporate debt instead. His mentor's post-it note, "scare your investors out of bonds," is finally playing out: sticky inflation (US above the 2% target ~65 months), unsustainable debt, and a French-vs-German spread showing "an enormous amount of distrust and volatility." Bessent's buyback "special operation" (raised from a $2bn to a $4bn+ maximum per operation to defend the 30-year) "achieved quite the opposite" of calming markets — Druckenmiller warns it damages Treasury-market credibility. The fix on the table is gold-backed sovereign debt: Judy Shelton's 50-year gold-convertible Treasury bond, floated at a $1–2bn notional as a market test (expected around the July 4th 250th anniversary; didn't happen). Precedent exists — at the end of the 1970s the US had to issue in Swiss francs and Deutsche Marks because nobody wanted dollar Treasuries. "The more turmoil we're seeing in fixed income markets globally, the more realistic it's going to become." Demand for gold's next leg comes precisely from this $150tn+ asset class once holders accept that bonds no longer hedge equities. 2026-SEP-01 (Sonders/Schwab, Master Investor): 30-year close to 5.3% and she does not expect Treasury's doubled long-end buybacks to cap it — the action treats the symptom while the disease (deficits, debt growth, a higher required compensation for financing it) is untouched, and Treasury is working against a Fed that wants the long end to do some of its tightening. New competition for the same buyer: massive AI-related investment-grade issuance pulling IG money out of Treasuries. She also frames the bigger risk as a Treasury market that becomes "unanchored from what either the Fed and/or the Treasury could do to try to contain that." (2026-SEP-03, John Polomny / AIA monthly) The bond sleeve cut on principle, not on price. Launching the AIA Permanent Portfolio off Harry Browne's 25/25/25/25 template, the one deliberate deviation is the deflation quadrant: "I want less exposure to bonds… especially true given that I expect persistent inflation and the government to institute yield-curve control." The argument is that the regime long bonds are supposed to hedge is precisely the one that gets administered — bonds as "certificates of confiscation" paying less than inflation. (2026-SEP-01, Paulo Macro) A precise non-confirmation at the long end: "the 5Y and 10Y future have both flushed the July low, but the 2Y and long bond have not (yet)" — a belly-only break, which reads as positioning rather than macro. The live test came the same afternoon: "despite oil rallying $5 on the latest Iranian 'love taps' this afternoon, the long bond is up a mere 2bps." With CTAs "pretty short bonds" — little left to sell, "a lot to buy if yields reversed lower," "a bug in search of a windshield" — he keeps his long-bond calls: "the long bond is not confirming the War trade and there appears to be a sticky rotational bid right here… I'll keep the calls. Hard trade." (2026-SEP-03, Dan Niles / Niles Investment Management, Excess Returns) "The 30-year is up at the highest level since 2007, that's a problem." Drivers: $40T of government debt against $33T of GDP, deficits at 6% of GDP ("the highest levels outside of a major war that we've ever had"), and a new one — "in the past, you didn't need to compete with AI debt spending by some of the biggest companies in the world, which in the past were massively cash flow generative like a Google." Dillian (2026-SEP-03): the loudest contrary voice yet, and he has put size behind it — “I am not worried about the bond market at all. I’m insanely bullish,” calling 5.2–5.3% on the 30-year and 4.7% on the 10-year “an incredible deal,” and disclosing that he “moved a huge portion of my money into bonds in the last month” for a three-to-five-year hold (no ETF or ticker named). His argument is a ledger correction rather than a forecast: supply is the observable half and everyone quotes it, but “it’s very easy to measure the supply of bonds & nobody ever talks about the demand for bonds.” The deficit is $2trn but only 6% of GDP against 12% in 2010, when auctions still cleared at three-times bid-to-cover; and demand is state-dependent — “if stocks were down 20%, trust me, interest rates would be much lower.” He treats the bear case as sentiment saturation, not analysis: Bloomberg running the sell-off on page one “every single day,” Ray Dalio in Time, and a market “really back to where we were in the late ’70s when people were calling bonds certificates of confiscation.” He concedes the new crowding-out force — a trillion of private-sector issuance including Google’s $40bn — but reads it as slowing the path, not reversing it. 2026-SEP-07 (Jay Singh): 30-year yields have risen across every major developed sovereign. Japan's 30Y JGB above 4.18% (an all-time modern record — regular issuance only began in 1999) with the 10Y near 3% for the first time since 1996; France's 30Y at 4.9274%, highest since September 2008, on a 5% deficit without growth plus a €15bn heat-wave shock; UK gilts +17bp intraday on Sep 2. Norway's $2.3trn sovereign wealth fund is cutting all government-bond holdings — potentially $80bn of Treasuries. OECD sovereign borrowing is $18trn this year, $14trn of it refinancing. Contrarian Codex (2026-SEP-07) states the destination outright: "the long end rises whether the committee hikes or cuts, so the only choice is how, and the road ends in de facto yield curve control, then the outright version." The 10-year closed near 4.72% after Jackson Hole and the 30-year near 5.21%, against 3.94% when the US attacked Iran. The buyer side is the argument: swap-spread funds are now the marginal buyer of the long bond at ~$305bn vs under $50bn in 2022, and Cayman relative-value funds levered 20–100x on overnight repo took a net $1.2tn of Treasuries over 2022–24 — ~37% of net note-and-bond issuance, more than every other foreign buyer combined — with ~$200bn already unwound. "A buyer built like that behaves as a market-maker rather than an investor," so an equity selloff shrinks the whole book and "the flight-to-safety bid turns up for a day and then flips into yields grinding higher while stocks are still falling." He also rejects the standard arbiter: term premium "is a residual, and a residual cannot distinguish between a market pricing a higher neutral rate and a market pricing fiscal risk while the Treasury is standing there buying the instrument with a General Account — like taking your temperature with your hand in warm water." And the capital-rule hope is dead: leverage-ratio relief was finalised so the constraint would stop binding for any GSIB with a primary dealer, "and the 30-year then went to a 19-year high on a buyers' strike that started in June," so "the capital rule was never the binding constraint." 2026-SEP-07 (Alden & Gromen, BTC Sessions): Alden's buyer-by-buyer census — foreigners buying in dollars but “not nearly enough” as a share of issuance, a self-described balance-sheet hawk at the Fed, banks with capacity only behind further SLR relief, and insurers and pensions as “fairly honest balance sheets” that must sell something to buy something while stuck in private credit — “so I do think that they're getting squeezed,” though with no MOVE-index stress she grades it “a pretty orderly degradation of the global bond market.” Gromen takes the same census to a different shape: foreign central banks net-absent at the long end for 12–13 years, foreign hedge funds gated on low vol, insurers unable to mark down private credit — “he's got a nonlinearity facing him at the long end.” 2026-SEP-08 (Jared Dillian, Excess Returns): the sentiment case restated in one line — "after 2022, everybody hated bonds and they've continued to hate bonds and they hate bonds with a burning passion today, which makes me like them a lot." He also names rapidly rising rates as the single vulnerability of his own five-sleeve allocation, since it kills bonds, stocks and gold together — which is exactly what 2022 was (−11.8%). 2026-SEP-08 (Larry McDonald, Julia La Roche) — a second, sized dissent. He turns constructive on duration for the first time: bears-versus-bulls at record territory, CFTC positioning the mirror image of the 2017–21 max-long extreme, and the bear case "80 to 90% priced in" — "everyone knows the bear case. Nobody's thinking about the bull case." The trigger is his supernova sequence: a commodity-led inflation spike wounds an already-broken bottom-70% consumer and pulls recession forward fast, re-rating the long end. Expressed through TLT and ZROZ, plus IVOL on a curve steepening. The convexity is the argument — a 6⅛% Google issue at 88 goes to "120 or 130" if the 30-year returns to 3–3.5%, against maybe 85 on one more punch. 2026-SEP-04 — Fraser Jenkins, asked whether 4%-yielding bonds are the diversifier again: "No, I have a strong view on this… it's likely they will not perform the diversifying role that they've performed historically." He splits the position by role rather than zeroing it — long-duration governments keep liquidity, drawdown mitigation and cash-flow matching, and lose only diversification. Underneath sits his standing objection: "there is absolutely no such thing as a risk-free asset. People just use that term partly because it makes the maths easier… It's contingent on political and economic states of the world, and I don't think we're in those states anymore." 2026-SEP-09 — Grandich (David Lin Report): out of Treasuries since end-2021 for a ~$2bn planning group, which he calls "one of the best calls of my career" — the premise was simply that rates could only go up, a lot. His total-return test: "anybody that's purchased Treasury bonds since the end of '21 to now, including the dividend yield, has actually lost money. It's unheard of," breaking the rule the industry taught 42 years ago ("you buy stocks to make money and you buy bonds to save your money"). Line in the sand: 5% on the 10-year, held more than a couple of days (4.83% at recording) — above it he expects "a huge, huge bond crisis," with the weight transferring to equities because the AI build-out is funding itself in debt instruments. He says the losses continue. 2026-SEP-10 (Ciampaglia, Sprott): investors are watching 'the showdown going on in the US bond market between Bessent and the bond vigilantes' as bond yields rise with inflation pressure and globally high debt; he cites it as a driver of the returning western gold flows. Lundin (2026-SEP-10): yields are rising because vigilantes are demanding compensation for dollar depreciation, which is the same reason to buy gold. Bessent's doubled, then tripled, long-bond buybacks are a drop in the bucket, and the market is taking up the challenge. With the 10-year approaching 5%, the widely watched tripwire for stocks, he calls the situation very fragile. E.B. Tucker (2026-SEP-11) dissent on the demand side: the tripled $6B long-end buyback plus a structural, GENIUS-Act-mandated stablecoin bid for bills (heading to trillions) gives Treasury the room to manage long rates down, not lose control of them. McGlone (2026-SEP-10) dissents bullish: T-bonds at ~5.34% are "the next big trade" and "that 5% bond, treasuries is a place to be"; TLT is "essentially a put on the S&P 500 with positive carry, no time decay," waiting for the stock-market break that turns energy inflation into deflation (10-year at 4.92%, highest since 2023). Lebowitz (RIA, Sep 10) dissents: the fundamentals say yields are already too high. Core CPI 2.5%, trimmed-mean PCE 2.3% and breakevens 2.4-2.5% are all back at pre-Iran-war levels, and growth runs at about 1/3 of trend. The move is narrative-driven (oil, deficits, AI debt crowding out, memory-chip prices, BoJ intervention, downgrades). The 10-yr at 5% is 'a line in the sand' with an institutional put (insurers, pensions, endowments) likely before the central planners. Trailing 10-yr bond returns at historic lows are the inverse of the CAPE chart. Accumulate patiently: 5-7-yr bonds held to maturity are a 'free option'. Wigglesworth (The Meb Faber Show, 2026-SEP-11) — the duration trap in "safe" paper. TIPS "got taken to the woodshed" in 2022 because "the duration on those suckers is immense"; Austria's 2021 century bond (a never-defaulted sovereign) fell ~80% at one point, while Argentina's century bond defaulted after about three years yet did better on its coupon; a 40-year UK inflation-linked gilt "did even worse than Bitcoin." Credit quality and inflation linkage don't protect against duration. 2026-SEP-02 Pomboy: still bearish on rates — yields keep going higher — with one caveat: the largest speculative short in the long end she has seen, which fiscal-discipline talk or Treasury jawboning could turn into a sizable but unsustainable short-covering rally. Her one chart: crises strike at successively lower yields as leverage builds, so a 30-year above 5% is already past where she expected something to break. Paulo Macro (Sep-12): now "uncomfortable in the long bond position" — a flattener stop-out if the Fed holds could blow out long yields, which equities would have a very hard time ignoring; his bond calls "will join so many other options trades in heaven." Jay Singh (SEP-13): the 10-year at 4.96% (a three-year high) and a $22bn 30-year auction tailing at 5.308%, the 30-year to 5.34% even after a good auction; Bessent's long-end buyback capped at $6bn with only $5.19bn executed for lack of sellers — "a rounding error in a $30 trillion-plus market"; "once you show markets that you're willing to adjust policy when they move against you, they will always demand more." G7 average 10-year above 4% for the first time since 2008 (BMO). Anna Wong (2026-SEP-11): long yields rising together with oil is global risk aversion pulling money out of Treasuries instead of into them as a safe haven, which she calls 'a bit concerning'. The long end is illiquid: Bessent's $4bn buyback moved yields far more than 1/100th of Operation Twist's $400bn effect. 2026-SEP-10 Gundlach (DoubleLine, Gundlach Unlocked ep. 3): the 30-year went from 27bp (2020) to 5.24% and "didn't retrace hardly at all" - a failed retrace of a ~500bp move means continuation higher. His 10Y model (German 10Y + 7-yr avg US nominal GDP, R² 0.93) reads 4.71% vs 4.78% spot, but "the path of least resistance could very well be higher." Long-term TIPS are no hedge: the 30Y TIPS/nominal spread has been flat for 5 years through the same rate rise. Astrid Wilde (2026-SEP-11): AI labs model at least ~20% near-contracted returns on data-centre builds, so 'the price of lending for other use cases is going to continue to go up because they're competing with building a data center... Why would you buy bonds?' Luke Gromen & Darius Dale (2026-SEP-13): Dale's five-model mean puts 10-year fair value at 5.87% (yield curve 5.2, inflation expectations 5.74, term premium 5.99, real yield 6.13, nominal-GDP spread 6.27) against 'Bessent panicking at 4.7 something.' Gromen: the move may be convex ('48, 52, 58, 62 happen fairly quickly') because life insurers can't buy at any yield, and 'they're going to lose the long end no matter what they do.' Whether ~6% is survivable depends on DXY (workable in the low-to-mid 80s, a debt spiral at 95–98). WSJ (Telis Demos, 2026-SEP-14): agency mortgage bonds are negatively convex again - current-coupon MBS ~5.8% vs just under 5% on the 10-year, but with 5%+ loans now over 40% of balances the refinancing option is live. Rising rates extend duration and hurt price; falling rates (e.g. after an AI-sector downturn) return principal to reinvest lower. Harley Bassman: more yield than Treasurys, but you lose more when rates rise and make less when they fall. FHN Financial calls the market very complacent about faster prepayment speeds (bigger balances, better credit, nonbank servicers, AI-accelerated refinancing per Morgan Stanley). Near term, Treasurys under pressure and the Fed poised to tighten. David Woo (2026-SEP-14, David Lin): the US 10Y hit 5% (highest since 2023); surging yields are "largely the result of this out of control AI capex" — FCF-negative hyperscalers with capex outgrowing earnings and $3T+ of off-balance-sheet commitments mean every capex dollar is new debt, while China sells Treasuries monthly, Japan urged pension funds to sell Treasuries and buy JGBs, and Norges Bank is cutting US bonds by $80B. Bessent's ~$8B buyback cap is "a joke"; the AI trade blows up around 5.25–5.30% on the 10Y. 2026-SEP-14 — David Rosenberg: bullish dissent — 10-year at 5% is Oct-2023 again, 'block your nose and buy it'. Catalysts: Nov-3 midterm gridlock ends the fiscal impulse; the Nov-4 refunding is the Treasury's supply lever (Oct-2023 bill tilt → 10Y −100bp in a quarter); slower AI capex cuts corporate issuance; near-record CBOT net spec short = squeeze fuel (4.80% → 4.50%). Yield cushion 500bp vs 60bp in 2021; expects bonds to beat stocks in 1–3 months. 2026-SEP-14 (David Hay): global long-bond breakdown — TLT's 10-yr chart 'clearly illustrates the price breakdown' to new lows; the 30-yr UST yield breaks out to 5.33%; 'a global phenomenon with a few exceptions, such as the Chinese government bond market.' France 'on the cusp of an actual sovereign debt crisis' (10-yr OAT 4.47%; spending 57% of GDP, debt service +25% y/y vs ~3% nominal GDP — 'the dreaded debt trap'). 2026-SEP-15: bearish on long Treasuries since the 10-yr hit 0.5% in 2020; now 5%, a tenfold rise. 2026-SEP-14/15 (CNBC): Liz Thomas buys the 10-year at 4.94–5% as her final trade ("I'll take that bet" the Treasury/Fed cap it near 5%); Rieder is "dabbling" in the long end but still significantly underweight on Treasury and credit supply, preferring the very front end (7.2% yield, <3yr duration, A-). Mike Taylor (Hedgeye, 2026-SEP-15): sovereigns 'can no longer really borrow on the long end… yields are up everywhere' — JGB 10-year got a 3 handle; he's shorting Swiss stocks because Switzerland's 0.58% 10-year 'can double twice.' US offset: if Europe blows up, 'the problem with the long end of our yield curve goes away.' Sep-10 (Paulo Macro, Fly on the Wall w/ Le Shrub): with 30s “rounding 5.3” and 10s “knocking on 5,” Le Shrub holds a small contrarian TLT option trade (“risking 30 bips” for ~10×) as the Street turns bond-bearish and Bessent is “two out of two” on crude and the yen. Paulo adds a structural bid: over-funded US corporate pensions can defease into ~5.5% long bonds/TIPS. 2026-SEP-05 — Don Durrett: the long end is fragile enough that the Treasury is intervening to stop foreign selling; the next tell is a sovereign holder dumping its bonds outright. Peter Lukacs (2026-SEP-16): government borrowing and AI borrowing crowd out and push yields higher, and energy prices add to it; 'everything is pushing yields higher. That's generally not a good sign.' Gundlach (2026-SEP-16): stay in the belly (≤7 years) until a 2% real yield — ~6% on the long end with inflation near 4%; his 7-yr NGDP + Bund model now reads the 10-year ~20bp too high, not enough to extend; 30-year TIPS are no hedge (moved in lockstep with nominals), 5-years-and-in TIPS are. Gundlach (2026-SEP-16, post-FOMC): not a buyer of the 30-year at 5⅓% or the 10-year at 5% — wants a 300bp real yield and would buy the 30-year "aggressively" at 6½%, but expects yield-curve control before it gets there; prefers 2–7-year securitized credit (+2% YTD vs the Agg −1.5%). Peterffy, 2026-SEP-16: a sanguine dissent: the 10-year through 5% and a higher 30-year are "okay with me" as long as the rise stays gradual; only a very sharp, sudden jump would "cause all kinds of problems." Hunt (Risk Takers, 2026-SEP-16): technician's blowoff call — US 10-year in an upside Hunt Volatility Funnel with the next rest after 5.334%, the 30-year triggered sooner with a 6.339% target and open space above; the 40-year bond bull ended in a 2020 capitulation and with debt far above Volcker-era levels "we don't need to get that high for everything to break." Expects a Lehman-style fiscal event "within 3 years and quite possibly within the next week." Howell (2026-SEP-09): yields are rising because nominal GDP growth is the fastest since the mid-1980s (US, Japan, eurozone; China the exception with falling yields), not mainly because of debt, and sit ~100bp below where NGDP implies, so upward pressure persists despite suppression. Debt-to-liquidity, not debt-to-GDP, triggers crises. Harrington (CNBC Halftime, 2026-SEP-16): seven cuts totalling 1.75pp over two years took the 10-year up from 3.73% to 5%, so policy moves barely set the rates that matter. Terranova watches whether yields fall on the hike as the market's verdict that one is enough (Yellen's December 2015 template). Joe Brown (Heresy Financial, 2026-SEP-17): "the bond market is right now calling [Bessent's] bluff" — yields rose across the curve despite long-end buybacks raised $2B → $4B → $6B (Sep 10, 15–20-yr maturities) against a $32T marketable stock; the 30-year sits at levels last seen briefly in May 2004 and the 10-year nears 5% (highest since 2006–07). Long yields simply price growth vs inflation — oil back at Iran-war highs, tariffs, spending — so he expects them "to continue to go up at least for the foreseeable future." Deluard (2026-SEP-17): after five years of "zero duration" he would "for the first time… start reallocating a little bit to Treasuries" — the 3-year and the long end "certainly more attractive than equities"; 5% nominal is "not generous" vs ~2% real growth + ≥3% inflation; bought as a hedge for a winter correction, "not out of bullishness." 2026-SEP-17 — Tom McClellan: bearish bonds — every QE round (2009, 2011, 2012–13, post-COVID, now QE5) saw bond prices tank; gold leads the 30-year yield by ~20.5 months, mapping to a steep yield advance from late 2026 into ~August 2028 (direction and timing, not magnitude). A 30-year above 6% and 10-year near 5.5% are "reasonable"; the pain lands on mortgages. Only a shift to QT would mitigate it. Polomny (AIA free weekly 9.16.26): "ten-year rates have broken their 40-year downtrend and are in a defined uptrend" (monthly $TNX chart: the secular bond bull from 1984 ended in 2020, confirmed break above the channel = new secular bond bear market) while "the root cause of the problem is only getting worse" — $1T of federal debt now added every ~70 days vs 716 days in the 2000s. Jay Singh (2026-SEP-20): a Western bond rout since Jan-2025. JGB 10yr +193bp to 3.03% (highest since 1996), France +135bp to 4.55%, Germany +120bp, Italy +95bp, UK +86bp to 5.43%, US +46bp and above 5% (first time since 2007). Drivers: energy, deficits (US 6%+, France ~6%), about $3tn of AI corporate issuance crowding out sovereigns, QT, and a buyer retreat: Norway sold about $80bn and Japan about as much, '160 billion dollars of demand that has evaporated.' The UK reversed QT and suspended 20- and 30-year gilt issuance ('imagine if the US tomorrow said, we're just not even going to sell 20 and 30-year bonds'). Yet he judges middle and long-end rates 'likely close to their peaks' and is adding to PIMCO CEFs. 2026-SEP-18 (Eisman): the 10-year spent time above 5% on oil near $110; "for now, 5% on the 10-year does seem to be the demarcation line" for equities (he again concedes 4.5% "was wrong"). A supply-side leg beyond war and inflation: ~$500B of AI-related debt raised this year is "creating a crowding out effect" - "some investors would rather buy AI long-term debt than long-term US treasuries." 2026-SEP-19 (Jeff Weniger, Corgi Invest): a dissent. His "bizarre thesis" is that inflation surprises to the upside while the bond market "stays cool as a cucumber" and the long-end selloff "stops puking". Even a 5% 10-year against ~3.5% CPI is a ~1.5% real yield, "not cheap and not expensive", and bond volatility has been tamed since COVID (+/-3-4bp days). The risk he flags is corporate debt-plus-equity issuance crowding out Treasuries (long bond already ~5.25%). Gentile (2026-SEP-19): "the bond markets are revolting" - UK 10/30-year near 6%, US at 5%, France at records, Japan blowing out - because investors expect repayment "with massively devalued paper currency"; a 5% coupon is a negative real return if money loses half its value in 10 years. George Noble on The Real Eisman Playbook Ep 76, 2026-SEP-21: Eisman says upward pressure on rates is "mostly because of AI, not so much the deficit"; AI debt issuance is crowding out Treasuries, and without it rates would be "probably 40 basis points lower" (his rough number). Noble: the market prices the 10-year US rate 10 years forward at ~6%, a comment on "the sustainability of the path," and "the Japanese are going to get there first." With Japan and China net sellers and foreigners put off by asset seizures, tax and trade policy: "who's going to buy the bonds?" Harley Bassman on MacroVoices #550, 2026-SEP-17: nominal yields are up ~150bp while the 10-year breakeven closed at 2.34% (four-year average 2.35%). The long end is pricing trust in the US as a going concern (the yield required to hold dollars), not inflation. Hyperscalers' price-insensitive AI borrowing (slide 9, ~30-35% of Treasury's need; units unclear as captioned) competes with the Treasury for money. Contrarian Codex (2026-SEP-22): the 10-year hit 5.04% and the 30-year topped 5.42% (~19-year highs) and the pension/life-insurer bid never showed up. Private placements are ~23% of life insurers' admitted bonds, and selling unmarked private credit to buy 5% Treasuries would force capital-eating marks: "a Mexican standoff between private credit, insurers and the long end." Foreign central-bank holdings have been flat for a decade, so the Treasury itself is becoming the buyer. A $5.19bn buyback against a $6bn cap sent the 10-year up 11bp in a day.