Price-sensitive
absorption
America cannot borrow cheaply, but can keep borrowing at a high price · private investors and funds absorb supply at sufficiently high yields · dealer burden stable · credit spreads stable
America’s
Costlier 30-Year Money
5% Is Today’s Price
If It Lingers, It Becomes Tomorrow’s Bill
RESEARCH NO.04THE PRICE OF LONG CAPITAL
RESEARCH NO.042026.08.23 · The Price of Long-Term Capital
The U.S. 30-Year
Why did 5.23% become necessary · Who is buying at that price · What changes if it lasts
First · What is the 30-year yield?
A benchmark long-term yield the market demands when the U.S. government borrows for a very long time.
Not merely a Treasury number, but the price of long-term capital · one of the benchmarks for reading corporate funding costs and asset discount rates.
This week · Long-term yields also rose outside the U.S.
| Country | 30-year | Highest in how long? |
|---|---|---|
| United Kingdom | 5.85% | Since May 2026 |
| United States | 5.33% | Since 2007 |
| Japan | 4.096% | Near an all-time high since issuance began |
| Germany | 3.78% | Highest in 15 years |
U.S. figure is intraday · others are official local closes · Japan’s 10-year also reached 2.934% a level not seen since September 1996, the highest in 30 years · The nature of each country’s record differs, so they are not collapsed into one headline.
Looking only at the U.S. 30-year makes this easy to read as a U.S. fiscal story · but in the same week, long-term yields in the U.K., Japan and Germany also moved higher · A scene that cannot be explained by U.S.-specific forces alone · this is where common repricing of long-term capital must be separated from each country’s own factors.
The strongest counterexample · Japan
Japan Ministry of Finance FY2026 JGB issuance plan · monthly issuance of super-long bonds (40Y·30Y·20Y) cut by ¥100bn. Supply was reduced, yet super-long yields rose toward record territory.
Supply fell, yet super-long yields still rose · a counterexample to the idea that Treasury supply alone can explain all long-term yields.
This was not just an American story.
So what moved inside the U.S.?
01 · What Makes Up 5%
The Same Bond, Six Months Later
30-year at 5.23% · Split inflation expectations from real yields · Only one side moved
If this is not solely a U.S. phenomenon, the next question is what actually moved within the U.S.
Start with three terms
Nominal yield · the 5.23% 30-year yield seen in the news · the total yield including inflation compensation.
Real yield · the return investors actually demand after stripping out inflation.
Expected inflation · the gap between the two · it includes not only pure inflation expectations but also risk and liquidity premia.
Nominal yield ≈ real yield + compensation for expected inflation
This is where we distinguish whether the rise came from inflation expectations or from the non-inflation side.
bp is the smallest common unit for measuring rates · 1bp = 0.01%p so 50bp equals 0.50 percentage points.
CUSIP912810US5same bond / six months
REAL+50bp
2.473% → 2.973%
BEI+3bp
2.23% → 2.26%
What moved
30-year TIPS CUSIP 912810US5 · one bond, two auctions
| Auction | Real yield | Expected inflation |
|---|---|---|
| 2026.02.19 new issue | 2.473% | 2.23% |
| 2026.08.20 reopening | 2.973% | 2.26% |
| Change | +50bp | +3bp |
CUSIP is the unique identifier for a specific bond · 912810US5 points to one and the same bond.
February 19 and August 20 · two auctions of the same 30-year TIPS · nearly all of the rise in long-term yields over this interval came from the real-yield side · the August 20 level of 2.973% was based on the 30-year TIPS auction the highest real yield since October 2001, a 25-year high · comparison quality is strong because maturity and coupon are identical, though market conditions differ between the original issue and reopening.
Expected inflation barely moved · only the real side stretched higher.
Comparing the same bond shows that the recent rise came far more from real yields than from expected inflation . The question now changes.
Why did the U.S. 30-year real yield rise to
nearly 3%?
August 20, 2026 · U.S. Treasury yields and policy rate
| Measure | Rate |
|---|---|
| Federal funds target midpoint | 3.625% |
| 2-year | 4.19% |
| 10-year | 4.69% |
| 20-year | 5.20% |
| 30-year | 5.23% |
The highest 30-year close was 5.31% on August 17 · the intraday 5.33% comes from market-data vendors, so it is not mixed with closing yields. The 20-year is shown because some circulating data confused the two maturities.
Four pressures that can feed into real yields
| Factor | Transmission channel |
|---|---|
| ① Fed path | The possibility that short-term real rates remain high for longer |
| ② Growth · capital demand | High U.S. returns on capital · AI, data centers, the power grid, reshoring and defense investment expand demand for long-term real funding |
| ③ Energy · inflation | Oil · war → expected inflation · or, through the Fed’s response, real yields as well |
| ④ Fiscal policy · Treasury supply | Structural deficits · long-duration issuance · the price of bearing long time |
The Fed says the same · the July minutes state directly that nominal Treasury yields rose 25–30bp and that the move was led by higher real yields (July 28–29, 2026 meeting · released August 19).
What the same-bond comparison confirms is only that this interval’s price change was concentrated in real yields rather than expected inflation · it does not decompose how much of that +50bp came from growth, the Fed, fiscal policy or AI · the four channels above can all enter the real side; they are not an allocation of causal shares.
What is established here
Over six months, in the same bond, the moving part was the real yield · expected inflation was almost flat.
What is not established
The causal shares behind the rise in that real yield · how much came from the Fed, growth, energy or fiscal policy.
Long-term yields rose around the world,
but in the U.S. the move was centered on real yields.
02A · How to Read a Treasury Auction
Demand Found at 5%
The fact that bonds sold is not the same as saying they sold at a low yield
First · price and yield are a seesaw
If a bond promises the same future cash flows · the cheaper you buy it, the higher the yield you receive.
When people say “yields rose,” it means buyers demanded more compensation · from the government’s perspective, the cost of new borrowing rose.
10Y4.683%Bid-to-cover 2.53×
20Y5.204%Bid-to-cover 2.53×
30Y5.216%Bid-to-cover 2.39× · indirect 66.85%
Orders came in at this yield
A Treasury auction ending without incident only means the full amount was sold. Demand strength differs completely depending on whether the same amount clears at 4% or 5.2%, and whether long-term holders take it down or intermediaries are left carrying it. That is why an auction should be read in three layers.
August 2026 · U.S. Treasury auction results
| Auction | Stop-out yield | Bid-to-cover | Indirect |
|---|---|---|---|
| 10-year · 8/12 | 4.683% | 2.53 | 76.73% |
| 20-year · 8/19 | 5.204% | 2.53 | 62.93% |
| 30-year · 8/13 | 5.216% | 2.39 | 66.85% |
The Treasury offered $25bn of 30-year bonds and received roughly $60bn of bids(bid-to-cover 2.39×) · the 30-year stop-out at 5.216% was the highest auction yield since 2001 · Bid-to-cover and the indirect bidder share are basic measures of demand strength.
It means bids arrived at 5.216% · it does not mean the same bids would have appeared at 4.5% or 4.0%.
30-year new issues · average award share by bidder typeNew issues only · reopenings excluded
| Period | Indirect | Dealers |
|---|---|---|
| 2023–2025 12 auctions | 65.31% | 15.42% |
| 2026 3 auctions | 67.80% | 9.68% |
When other investors bid weakly, primary dealers must take a larger share · the signal a fiscal-stress amplification story would require is lower indirect participation · higher dealer take-down but the current direction is the opposite · auctions are clearing without an expanding dealer take-down share— that is as far as the evidence goes · 2026 contains only three auctions, too small a sample to declare a trend.
The three darker bars are 2026 · among them February 5.88% was the lowest.
Borrowing has become more expensive, but in the structure that absorbs new 30-year issuance, there is still no sign of a buyers’ strike in dealer take-down shares · though 2026 has only three auctions, too thin a sample to claim a structural shift.
More important than whether it sold is
the yield at which it sold.
02B · The Buying Side
Who Buys America’s
Long Time?
The buyers are still there · but the hand setting the price is changing
Four types of Treasury buyers
| Buyer | Why they buy | Price sensitivity |
|---|---|---|
| Official institutions | Reserves · liquidity · safe assets · foreign central banks, governments and sovereign wealth funds | Potentially less sensitive |
| Private investors | Yield ·FX-hedging cost · relative value · asset managers, banks, insurers and pension funds | More sensitive |
| Investment funds | Mutual funds · money-market funds · ETFs · hedge funds | More sensitive |
| Dealers | Intermediaries in auctions and secondary markets · not the same thing as final long-term holders | Sensitive to intermediation margins |
Price-sensitive does not mean unwilling to buy Treasuries · it means buying more aggressively when yields are high enough · so the chain “less official demand → collapse in long-bond demand” is not assumed.
A smaller official-sector share does not mean Treasury demand vanished. It is closer to saying that price-sensitive private investors now play a larger role in determining yields.
In one line
The buyers remain · but the hand setting the price has become more sensitive.
View detailed foreign holdings
Foreign holdings · June 2026U.S. Treasury International Capital
| Measure | June holdings | 12-month change |
|---|---|---|
| Total foreign holdings | $9.299tn | +$20.54bn |
| Japan | $1.1167tn | −$3.81bn |
| United Kingdom | $939.9bn | +$8.43bn |
| China | $633.4bn | −$9.80bn |
The 73-to-27 split above describes the ownership structure of all U.S. Treasuries · the table below isolates only the foreign-owned portion using TIC data · the populations are different.
In June long-term Treasuries, official institutions were net sellers of $0.98bn · private investors were net buyers of $1.66bn · opposite directions.
TIC statistics are a window through which money passes, not a literal map of final beneficial owners · country holdings are recorded by custody location and may differ from the true owner country, so a one-month decline is not read as an exit signal.
The Fed is not in the seat that sets the 30-year price
Quantitative tightening ended on December 1, 2025 · reserve-management purchases and reinvestment of MBS principal payments are centered on short maturities.
But principal from maturing Treasury holdings is fully rolled over at auction and can be allocated to 30-year securities as well · this portion enters as noncompetitive bids and is therefore distinct from private demand that directly sets the clearing yield.
Treasuries are still selling ·
but they are selling to buyers who care about the price.
02C · Market Plumbing
This is where the plumbing
also matters
If price-sensitive hands matter more · what supports the market when a large holder sells all at once?
If price-sensitive buyers now matter more, the next question is what mechanisms can reduce market shock when large existing holders try to sell at once.
BUYBACK
The Treasury repurchases older long-dated securities that trade less actively to support market liquidity.
FIMA Repo
Foreign central banks can pledge Treasuries as collateral instead of selling them into the market to obtain dollars.
The buyback program is the measure being expanded this time · FIMA is not new; it is an existing standing facility.
But where did the Treasury intervene?
Only the per-operation caps for the two long-duration buckets were lifted to match the others · the objective is less about total debt than keeping older, less-liquid long Treasuries circulating.
1 month–10 yearsFive bucketsUp to $4B
10–20 years$2B → at least $4B
20–30 years$2B → at least $4B
Not the debt total · managing long-end liquidity so trading does not seize up
What changes from September 9
Only the per-operation buyback limits for the two long buckets (10–20 years · 20–30 years) are raised to the same level as the other five buckets · effective September 9 through November 4.
What the two pieces of plumbing do — and do not do
| Measure | What they do | What they do not do |
|---|---|---|
| Buyback | Support long-end liquidity by taking out older, less-traded securities | Not QE · not debt reduction · not a guaranteed ceiling on yields |
| FIMA Repo | Give foreign central banks a way to obtain dollars without selling Treasuries into the market | Not a new buyer · not forced buying · not a ban on selling |
Neither piece of plumbing manufactures new Treasury demand. They are closer to mechanisms that reduce the risk of large existing holders dumping bonds all at once and help keep trading functioning during stress.
Not a policy that creates new net final private demand · plumbing that reduces selling shocks and liquidity friction.
View buyback · FIMA details
From September 9, 2026 · Buyback size per operation by maturity bucket
| Remaining maturity | Before Sep. 9 | After |
|---|---|---|
| 1 month–10 years Five buckets | Up to $4bn | Unchanged |
| 10–20 years | Up to $2bn | At least $4bn |
| 20–30 years | Up to $2bn | At least $4bn |
The change raises the two long buckets, previously capped at half the others at $2bn per operation, to the same $4bn level as the other five buckets · but the wording differs: the existing five buckets are “up to $4bn,” while the new long buckets are “at least $4bn” · effective September 9 through November 4.
Treasury’s official explanation
Support liquidity in long-dated nominal sectors · citing a steady supply of high-quality offers in long-end buybacks.
Hana Securities’ interpretation
A Treasury version of Operation Twist · fund secondary-market purchases of off-the-run securities with increased bill issuance · keep total debt unchanged while reducing the duration the market must absorb.
“Operation Twist” is Hana Securities’ interpretation (2026.08.20), not the Treasury’s official definition · this report does not force the conclusion in one direction.
Three pieces of evidence behind the market interpretation
- Announcement sequence · two weeks after the regular refunding announcement · only the two long buyback buckets were separately increased
- Effective window · September 9 through November 4 · spanning the November 3 midterm election
- Bill share · as of July 31 22.2% · already above the Treasury Borrowing Advisory Committee’s recommended 15–20% range
What buybacks do not do
Treasury source · TBAC presentation, FY2026 Q3, p.16.
“Additionally, buybacks are not expected to significantly affect privately-held net marketable borrowing as new issuance replaces securities that are bought back.”
Not QE · not debt reduction · not a guaranteed ceiling on yields. Scale matters too: $4bn per operation versus a $31tn U.S. Treasury market.
FIMA Repo terms
The FIMA Repo cap is $60bn per counterparty per day · overnight at 3.75%; seven-day at the average overnight policy rate plus 25bp (Yuanta Securities, 2026.08.14).
The auctions are still functioning ·
but the price and the hands doing the absorbing have already changed.
03 · The Lag
Average funding cost is
still 3.443%
The distance between the average rate on all marketable Treasuries and today’s 30-year market yield
3.443%Existing marketable Treasuries · average rate2026.07.31
→
5.23%New long-term capital · 30-year market yield2026.08.20
Left · 3.443%Average rate on the stock of debt already issuedAs of 2026.07.31 · bills 3.758% · notes 3.309% · bonds 3.442% · August data due in mid-September
Right · 5.23%The market yield for borrowing new 30-year money todayAs of 2026.08.20
These two figures are not a refinancing comparison at the same maturity. They are the distance between the accumulated average and today’s market rate.
A 5.23% 30-year yield today does not mean the entire existing federal debt stock starts paying 5.23% today. Debt is refinanced at prevailing market rates only as it matures, and the average funding cost rises later as that process accumulates.
Why the lag exists
Low-cost debt matures → Refinance at then-current market rates → Average funding cost rises → Net interest outlays increase
Federal debt held by the public $32.279tn does not all reset to 5% today · it feeds into the average funding cost in maturity order.
First 10 months of FY2026 · federal outlays2025.10~2026.07
| Category | Amount | YoY change |
|---|---|---|
| Net interest outlays | $963bn | +$117bn |
| Defense Department outlays | $763bn | +$39bn |
Interest exceeds Defense Department spending by $200bn · CBO attributes the increase to two factors: more debt and higher rates · the result of the lag is already visible here.
Counterargument · it would be wrong to read this as “5% means a fiscal crisis right now” · average funding cost is still 3.443% · the shock arrives with a lag.
Today’s bill is still priced cheaply.
The issue begins with the next bills.
03 · The Bill
The primary deficit improves,
so why does the total deficit worsen?
The primary deficit gets better while the total deficit deteriorates · the force reversing the direction is net interest
The government’s underlying deficit improves slightly, but interest costs rise faster than that improvement, so the total deficit grows instead.
How to read this table
Excluding interest, the deficit narrows · including interest, the total deficit widens.
The figure excluding interest is the primary deficit · the figure including interest is the total deficit.
CBO, “The Budget and Economic Outlook: 2026 to 2036” · share of GDP
| Category | 2026 | 2036 | Change |
|---|---|---|---|
| primary deficit | 2.6% | 2.1% | -0.5%p |
| Net interest outlays | 3.3% | 4.6% | +1.3%p |
| total deficit | 5.8% | 6.7% | +0.9%p |
| Federal debt held by the public | 100.6% | 120.2% | +19.6%p |
The primary deficit improves by 0.5%p while the total deficit worsens by 0.9%p · the force reversing the direction is net interest · the components are rounded, so the three figures do not add perfectly · the same path appears even under the current-law baseline, without assuming new fiscal expansion.
Upper line: total deficit · lower line: primary deficit · interest fills the widening gap.
Projected growth by category, 2026–2036 · CBO
| Category | Growth |
|---|---|
| Net interest outlays | +106% |
| Medicare | +85% |
| Social Security | +65% |
| Federal revenue | +48% |
Interest grows more than twice as fast as revenue · net interest rises from 19% to 26% of federal revenue. It is projected to exceed Medicare spending in 2028.
If rates are even slightly higher
CBO sensitivity · if all Treasury rates are 10bp above the baseline, the cumulative 2027–2036 deficit rises by about $379bn .
| Across-the-curve rate shock | Additional cumulative deficit, 2027–2036 |
|---|---|
| +50bp | About $1.9tn |
| +100bp | About $3.8tn |
Not an official CBO scenario · an arithmetic scaling of the 10bp sensitivity · CBO itself says linear approximation is reasonable for rate scenarios up to 1%p, and this calculation stays within that range · not used as a foundation for the main conclusion.
The fiscal risk to the U.S. is less an explosion than
a slow repricing.
04 · Timing
2027, the First Test
If more supply must be sold to increasingly price-sensitive buyers · how high a yield will the U.S. have to pay?
In 2027, two tests overlap. One is a market test of the yield at which investors will absorb more Treasury supply; the other is a political test of whether U.S. politics can actually deal with the burden.
MARKET TESTAt what yield will the market absorb the added supply?
FY2027Begins October 1, 2026 · where supply and politics meet
FY2027 begins October 1, 2026 · just over a month away · with refinancing pressure accumulating, it is the first period in which we can meaningfully test whether funding needs actually expand, which is why this report looks there first · this is where the lag described in Section 03 should begin to appear in borrowing volumes.
U.S. Treasury quarterly financing · privately held net marketable borrowing
| Quarter | Estimate |
|---|---|
| Jul–Sep 2026 | $739bn Raised $68bn from the May estimate |
| Oct–Dec 2026 | $628bn |
August 2026 refunding issue sizes
| Security | Size |
|---|---|
| 3-year | $58bn |
| 10-year | $42bn |
| 30-year | $25bn |
The Treasury plans to keep current coupon-security auction sizes unchanged for the next several quarters · the TBAC has noted the possibility of larger issuance from FY2027 as financing needs increase. FY2027 begins October 1, 2026.
POLITICAL TESTCan politics make the choices required to carry that cost?
On August 18, 2026, total U.S. federal debt crossed $40tn for the first time · as of August 20 it stood at $40.0333tn, of which $32.279tn was held by the public · a $10tn increase in four years and seven months since crossing $30tn in January 2022 · the debt ceiling is neither the cause of the 30-year yield nor fiscal sustainability itself · but it reveals whether the political system can deal with the issue.
Statutory debt limitMeasured on a different basis from total federal debt
| Category | Item |
|---|---|
| Statutory limit | $41.1tn |
| Basis | A law effective July 4, 2025 raised it by $5tn from $36.1tn |
| Amount used | Debt subject to limit since July 2025 +$2.9tn · more than half of the increase |
| CBO | Expects the limit to be reached again during 2027 · no specific date given |
| Bipartisan Policy Center | Late winter to midsummer 2027 possible re-hit · followed by accounting extraordinary measures |
Total federal debt and debt subject to the statutory limit use different scopes, so they should not be subtracted from each other · “debt limit minus total debt” is not a valid calculation. Nor is the X-date fixed yet.
The two lines use different accounting scopes, so no numeric gap is labeled · the band on the right marks 2027.
Variables that can swing both tests
Variables that will shape 2027
| Variable | Burden-expanding path | Burden-easing path |
|---|---|---|
| Middle East | Prolonged war · higher oil · more defense spending | Normalization · lower oil |
| Fed | Sticky inflation · persistently high real rates | Stable inflation · easing |
| AI | More demand for long-term capital and issuance | Productivity gains → growth · tax revenue |
| Politics | Tax cuts · spending expansion · fiscal gridlock | Tax increases · spending restraint · fiscal improvement |
FY2027 defense-related total budget resources requested: about $1.5tn · including about $1.1tn in base discretionary funding and $350bn in mandatory resources · the ultimate direction depends on outcomes, so it is not pre-assigned to either side.
Does AI compete in the same window?
Hyperscaler ultra-long corporate bonds · same company, six months apart
| Issue | 30-year Coupon | 40-year coupon |
|---|---|---|
| Meta 2025.10 | 5.625% | 5.750% |
| Meta 2026.04 | 6.300% | 6.450% |
The same company issued 30-year debt six months later at a coupon 68bp higher · the AI industry has also begun tapping long-term capital through ultra-long corporate bonds · another variable that could make the 2027 duration absorption test harder.
Part of the rise in absolute coupon reflects higher Treasury yields over the same period · the corporate-specific burden must be checked separately through new-issue concession, bid-to-cover and spread.
Two layers of counterargument · the investment-grade corporate-bond spread index remains tight at 82bp and direct crowding-out has not been proven · however, new-issue concession has risen from a median 2.25bp to 12bp, while order-book coverage has fallen from 5× to below 2× · supply pressure is beginning to show beneath the index.
The Treasury is not the only borrower · combined 2026 capex guidance from major hyperscalers is about $730bn and they are also raising capital with ultra-long corporate bonds · it is still too early to call them the main cause of higher Treasury yields · the direction cuts both ways · they are competitors for capital and at the same time potential drivers of growth and tax revenue that could lower the debt ratio.
2027 is not the year of crisis,
but the year answers begin to emerge.
05 · Interpretation
Different Eyes on
the Same 5%
Why do top investors reach different conclusions from the same number?
The left side puts more weight on high yields as the price of strong U.S. growth and capital demand, while the right puts more weight on the extra risk compensation demanded for fiscal stress and confidence.
This page is not a vote on who is right · it places competing hypotheses for the same 5% side by side.
Where does each interpretation place more weight when explaining 5%?
Growth · real-economy capital demandFiscal policy · confidence · debt feedback
The Price of a Strong America
Rick RiederBlackRock · multi-asset allocationAI investment and real capex are creating growth and financing demand at the same time · he does not reduce high real yields to a single “loss of confidence” story · he also views the higher real-yield level itself as a potential bond entry opportunity
PIMCOBond manager · long-term outlookAcknowledges risk in the fiscal path · expects recurring market volatility around fiscal credibility · but a sudden inability of the U.S. to borrow in markets is not the base case
Expensive, but still functioning
Jeffrey ShermanDoubleLine · long-duration bond managerHyperscaler capex · fiscal excess across developed markets · inflation concerns are all weighing on the long end · but this remains distinct from a disorderly regime of 20–30bp daily jumps
Jamie DimonJPMorgan Chase · CEOStrongly warns that postponing the high-debt problem can end in crisis · but also identifies an escape condition: 3% real growth combined with a 100bp decline in funding costs could turn the debt ratio
The Price of a Heavier America
Jeffrey GundlachDoubleLine · long-duration bond managerReads the widening between 2-year and 30-year yields as a question of confidence in the Fed’s inflation credibility · prefers the short end and higher-quality assets · avoids long-duration bonds
Ray DalioMacro debt cycle · asset managementPlaces Japan’s reduction in holdings · new highs in long yields and a weaker dollar · Treasury Buyback announcements together and interprets them as signs of imbalance between debt supply and market demand · his general mechanism is when supply exceeds demand, yields rise sharply; or the central bank creates money to buy the debt, in which case the currency loses value.
The same 5% · interpretations split here
Howard Marks sits elsewhere on the map · he notes that America’s funding privilege depends on confidence, but says he does not know where the confidence threshold lies or when it will be reached.
The placement above is this report’s classification · the individuals did not use this exact axis themselves · it is not a popularity contest but a map of where each places the weight· Dalio’s timing estimates and causal readings of specific events contain elements of his own model and interpretation.
Looking at the same price,
the disagreement is about where the weight belongs.
05 · Regimes
Three Regimes
Where Are We Now?
Not three prophecies, but one question · which mechanism dominates the market?
The path that goes to none of them · gradual normalization
War and oil normalize → inflation slows → policy path shifts lower → real yields ease → 30-year returns to the 4% range
A normalization in which nothing breaks · in this case, the hypothesis that 5% becomes entrenched weakens · this report is not written on the assumption that 5% must persist.
Price-sensitive
Fiscal
amplification
High rates worsen the fiscal position · the weaker fiscal position pushes rates higher again · interest → deficit → more issuance → a higher price for long time · even if the Fed eases, the long end does not fall · a weaker dollar may accompany it
Expensive money
breaks the economy
High capital costs weaken housing · credit · consumption · employment → Fed easing → real yields fall sharply → long bonds rally
0 · A · B · C are states, not a sequence · A→B · A→C · B→C are all possible · B does not mean Treasuries fail to sell; it means the price begins to push itself higher
Not three mutually exclusive worlds, but states the market can move between.
Where are we now?
- Market functioning · A
- At what yield did it sell · who bought it · a more demanding version of A
- Early ingredients of B · present
- Confirmation that fiscal self-amplification dominates the market · still insufficient evidence
So far, the market remains closer to A: supply is absorbed, even at a high price. To say it has moved into B, high rates must repeatedly worsen the fiscal position, and that deterioration must in turn demand still higher rates.
Why this report does not yet call it B
Signals by regime · what separates them
| Signal | A · Absorption | B · Fiscal Amplification | C · Expensive Money Breaks the Economy |
|---|---|---|---|
| 30-year real yield | Stays high | Entrenched · rising | Falls sharply |
| 30-year expected inflation | Stable in the 2% range | Stable to potentially higher | Falls |
| Dealer take-down share | Stable | Rises | Falls |
| Indirect award share | Stable | Falls | Rises |
| Treasury supply-demand balance | Current supply absorbed as-is | More coupon issuance + weaker absorption capacity | Safe-haven demand rises |
| Dollar Index | Can remain firm if growth and yield advantage dominate | Warning if long yields rise while the dollar weakens | Initially stronger |
| Credit spreads | Low | Gradually widen | Widen sharply |
The funding and credit signals confirmed so far mostly resemble A · however, on August 20 one price combination associated with B appeared · the items do not carry equal weight, so they are not counted like votes; auction signals span multiple auctions, while the dollar signal is a one-day price combination.
August 2026 · days when the 30-year yield and dollar diverged
| Date | 30-year | Dollar Index |
|---|---|---|
| 08.19 | 5.19% Day of buyback announcement | Falls |
| 08.20 | 5.23% Rebound | Fell · below 200-day moving average |
| 08.21 | Not confirmed | 98.80 Stayed below 200-day moving average |
The warning combination defined for fiscal amplification — “higher 30-year yield + weaker dollar” — appeared clearly once on August 20 · Buyback The 30-year yield, which had fallen immediately after the announcement, rebounded to 5.23% the next day · meanwhile the dollar kept falling · gold, silver and Bitcoin rebounded together while the Nasdaq was flat · The exact 200-day moving average differs by data provider, so only direction is shown rather than a precise level · the confirmed point is that the Dollar Index was below it.
A one-day or few-day price combination, not confirmation of a regime · February–July 2026 showed the exact opposite (yields up + dollar stronger + gold down + Bitcoin down) · this report did not verify whether the same August combination occurred earlier in the year.
Structural changes that appear first · stress signals to confirm later
| Structural changes that appear first | Stress signals to confirm later |
|---|---|
| Background condition · a larger share of price-sensitive private buyers | Indirect award share keeps falling |
| Long yields rise while the dollar weakens | Dealer take-down share keeps rising |
| Plumbing adjustments such as buybacks · FIMA | Bid-to-cover Trend breaks |
| The price of bearing long time rises | Credit spreads widen sharply · price pressure around auctions intensifies |
So far, some of the earlier structural changes are visible · later-stage stress signals remain limited.
Why these two signal groups should not be weighted equally
| Signal | Observation window |
|---|---|
| Auction results · dealer take-down share | Multiple auctions |
| Indirect award share | Multiple auctions |
| 30-year yield and dollar combination | One day |
The funding signal reflects repeated outcomes while the price signal reflects a one-day combination · counting items like votes would give one day the same weight as multiple auctions · what can be said now is only that the regime is closer to A, while the warning combination for B is retained as an observation.
It is not a comfortable A ·
but there is still not enough evidence to call it B.
06 · Assets
What Survives
High Rates?
Do not invest in the rate number; invest in the reason rates are high
Gold in 2026 · LBMA AM priceOne benchmark only
| Date | Gold | Real yields · dollar over the same interval |
|---|---|---|
| 01.29 YTD high | $5,501.70 | Dollar Index · 1/27 YTD low 96.22 |
| 07.01 YTD low | $3,978.55 −27.7% from the high | 30-year real yield and dollar rose together |
| 08.21 | $4,581.95 +15.2% from the low | Dollar Index 98.80 · below 200-day moving average |
Act I · January to early July · despite fiscal concern and high 30-year yields, gold corrected while real yields and the dollar rose together. A deteriorating U.S. fiscal position did not automatically make gold rise.
Act II · August · long yields stayed high while the dollar weakened, overlapping with rebounds in gold and Bitcoin.
All three observations use only the LBMA AM price · even on the same day, prices and daily return signs can differ across providers, so spot closes and futures prices are not mixed in the same sentence.
The combination most favorable to gold
High long-term yields + real yields stop rising or fall + weaker dollar. One of the main differences between the two 2026 phases was the combination of real yields and the dollar.
“Weaker policy credibility” is one possible interpretation of that combination · what the data here directly show are real yields and the dollar, not credibility itself.
DalioDebt · currency value
GundlachFed · fiscal credibility
PIMCOStore of value · diversification
This reportReal-yield · dollar combination
The destination asset may be the same · the starting logic is not
Relative asset implications by regime
| Asset | A · Absorption | B · Fiscal Amplification | C · Expensive Money Breaks the Economy |
|---|---|---|---|
| Gold | Neutral to favorable | Potential beneficiary If accompanied by weaker dollar · stable real yields | Mixed initially · favorable after policy turn |
| U.S. long-duration Treasuries | High carry, but duration risk remains | Most disadvantaged | Largest reversal upside |
| High-quality equities | Differentiation around cash flow | Pressure from higher discount rates | Sharp initial decline · later recovery |
| Cash · short-term Treasuries | High carry | Strong defensiveness | Defensive initially · later reallocation |
| Bitcoin | Favorable if liquidity remains ample | Potentially favorable with a weaker dollar Non-sovereign scarcity narrative · but still a high-beta asset sensitive to the dollar and liquidity | Can fall sharply at first Not treated as the same safe-haven asset as gold |
| Commodities | Some benefit if growth holds | Energy benefits if war · inflation is the driver | Generally unfavorable |
| Dollar | Can stay firm if growth · rate advantage dominates | Initial safe haven · later path depends on fiscal confidence | Initially stronger |
Not a buy or sell recommendation · a conditional map of which macro forces each asset is exposed to.
View detailed data for gold · Bitcoin · bonds
Central-bank gold purchases · separate actual buying from intentions
| Actual purchases | Tonnes |
|---|---|
| Q1 Initial estimate | 244 tonnes |
| Q1 Revised | 57 tonnes |
| Q2 | 289 tonnes Record high for a second quarter |
| H1 total | 345 tonnes Lowest first-half total since 2022 |
The Q1 revision from 244 tonnes to 57 tonnes must be read together with the headline data · quarterly figures are rounded, so simple addition differs by 1 tonne · the official H1 total is 345 tonnes.
| Intentions · 2026 survey | Share |
|---|---|
| Expect global holdings to rise over the next 12 months | 89% |
| Plan to increase own holdings | 45% Record-high share |
Actual buying is not unambiguously strong right now, while intent to diversify reserves remains high · the two describe different things and are not blended into one sentence.
Bitcoin in 2026 · two phases
| Phase | Rates · dollar | Bitcoin |
|---|---|---|
| Feb–Jul | Yields up + dollar stronger | About $93,000 at the start of the year → 6/30 $57,735 |
| August | Yields near highs + dollar weaker | 8/21 $77,308 Highest since May |
Even after the August rebound, down 36.1% YoY. That is as far as the 2026 observations go · because Bitcoin trades 24 hours, the August 21 price is as of 09:08 ET.
Not all bonds are the same
| Period | In this regime |
|---|---|
| Cash · short-term Treasuries | High carry · low price volatility |
| 5–10 year intermediate Treasuries | Middle ground between carry and price risk |
| 20–30 year long Treasuries | The segment where fiscal and supply risk loads most heavily |
| TIPS | Not inflation, but real yields are the exposure here |
| Corporate bonds | Treasury yield + credit risk · spreads are currently near historical lows |
Being cautious on fiscal risk in long-duration Treasuries is not the same as avoiding all bonds · PIMCO argues that higher starting yields have restored meaningful carry to high-quality bonds · this report does not use vague “bond allocation” language without specifying which maturity segment it means.
Where the chain broke
Transmission path · simple assumption versus reality
| Link | Simple assumption | August 2026 reality |
|---|---|---|
| U.S. 30-year | Rises | Highest in 19 years |
| Dollar | Stronger | Weaker · below 200-day moving average |
| USD/KRW | Rises | KRW 1,386.5 · won stronger for a fourth day |
The chain broke at the Dollar · three Korea-side factors were: semiconductor exports of $26.03bn from Aug. 1–20, equal to 47.2% of total exports · market consensus for a 25bp Bank of Korea hike · and broad dollar weakness.
U.S. long yields do not transmit to Korea in a straight line · the dollar and domestic fundamentals can reverse the direction in between.
Not the Number, but the Composition
Even at the same 5% · the result differs depending on whether it is
the price of growth or the price of fiscal stress.
The price of a strong America · or the price of a heavier America
Six steps · what was established and what remains open
01
The recent rise in the 30-year came more from the real-yield side than from expected inflation
Still open · the causal contribution to the +50bp real-yield move02
Auction functioning remains intact · but the yield required to place the supply has risen
Still open · how many basis points hyperscaler bond issuance has added to Treasury yields03
Price-sensitive buyers now carry more weight, while dealer take-down share has actually fallen
Still open · whether this mix persists04
Through refinancing, high yields feed through in maturity order into fiscal costs
Still open · how quickly the pass-through completes05
Rising financing needs and another encounter with the debt limit overlap in FY2027
Still open · whether B has already begun · and when06
Even with the same yield, if the composition differs, the winning assets differ
Still open · whether the long-run responses of the dollar, gold and Bitcoin stay fixed in one directionAppendix
Glossary and Sources
Glossary
- noncompetitive bids
- A bid that accepts the auction’s stop-out yield without submitting a yield. Distinguished from competitive bids, which help determine the auction price.
- bp
- The smallest common unit for rates. 1bp = 0.01 percentage points · 50bp = 0.50 percentage points.
- Real yield
- The yield left after stripping out inflation. The yield on TIPS corresponds to this real return.
- Expected inflation
- The difference between the nominal yield and real yield at the same maturity. It includes not only pure inflation expectations but also inflation-risk premia and liquidity effects.
- TIPS
- Treasuries whose principal adjusts with the Consumer Price Index. The coupon is set in real-yield terms.
- Bid-to-cover
- Total bids divided by the amount offered. A basic measure of the depth of auction demand.
- indirect bidder share
- The share awarded through indirect bids. It includes foreign central banks, foreign private investors and domestic institutions, so it is not synonymous with foreign demand.
- primary dealers
- Financial institutions designated to transact with the New York Fed, the Federal Reserve’s market desk. They are core intermediaries expected to participate competitively in every Treasury auction. If other investors bid weakly, dealer take-down shares can rise.
- FIMA Repo
- A standing facility through which foreign monetary authorities pledge U.S. Treasuries to the New York Fed and borrow dollars overnight or for seven days. The cap is $60bn per counterparty per day. It provides dollar liquidity without forcing Treasury sales into the market.
- Buyback
- Treasury repurchases of already-issued government securities in the secondary market. Because the purchases are funded with new issuance, total debt does not decline.
- off-the-run securities
- Outstanding Treasuries that are no longer the most recently issued benchmark securities. They usually trade less actively and can carry wider bid-ask spreads.
- duration
- A measure of how sensitive a bond’s price is to changes in yields. Duration generally increases with maturity.
- primary deficit
- The fiscal deficit excluding interest costs. By stripping out the burden of past debt, it shows the current year’s underlying fiscal operations.
- Federal debt held by the public
- The portion of federal debt held outside government accounts — by markets, foreign holders, central banks and others. The Treasury’s official term is debt held by the public.
- net marketable borrowing
- Net Treasury issuance that private investors must actually absorb. Excludes Federal Reserve holdings and intragovernmental holdings.
- Statutory debt limit
- The legal ceiling Congress places on federal debt. It is measured on a different basis from total federal debt, so the two should not be directly subtracted.
- X-date
- The date when Treasury cash and accounting extraordinary measures are exhausted. It depends on the flow of tax receipts and cannot be fixed in advance.
- Dollar Index
- An index of the dollar’s relative value against a basket of major currencies.
- Credit spreads
- The difference between a corporate bond yield and the Treasury yield at the same maturity. The price of credit risk.
- new-issue concession
- The extra yield a newly issued bond offers over comparable outstanding debt. It generally rises when demand is weaker.
- Coupon
- The stated interest rate fixed when a bond is issued and paid through maturity. Existing coupons do not change when market yields rise.
- Coupon-bearing securities
- A bond that pays a fixed coupon periodically until maturity. Distinguished from inflation-linked Treasuries.
- FX-hedging
- Removing FX risk on foreign-currency assets in advance using forwards or swaps. Higher hedging costs reduce the effective return.
- discount rates
- The rate used to convert future cash flows into present value. A higher discount rate lowers today’s value.
- Remaining maturity
- The time remaining from today to maturity. It shortens as time passes.
Sources
- U.S. Treasury Fiscal Data · Treasury Securities Auctions Data Auction results · award shares by bidder type · Average Interest Rates on U.S. Treasury Securities 2026.07.31 · Debt to the Penny 2026.08.20
- U.S. Treasury · Treasury Presentation to TBAC, FY2026 Q3 Buyback language on p.16 · Press release sb0607 2026.08.19 buyback expansion · Quarterly borrowing estimate 2026.08.03 · Refunding statement 2026.08.05
- U.S. Treasury TIC · Major Foreign Holders of Treasury Securities Table 5 June 2026 data · released August 17
- Federal Reserve · H.15 Selected Interest Rates 2026.08.21 release · July FOMC minutes
- CBO · The Budget and Economic Outlook: 2026 to 2036 2026.02 · How Changes in Economic Conditions Might Affect the Federal Budget Interest-rate sensitivity · Monthly Budget Review: July 2026
- Bipartisan Policy Center · Estimate of when the debt limit will be reached again 2026.06.04
- Japan Ministry of Finance · Treasury interest-rate data 2026.08.18 · FY2026 JGB issuance plan
- World Gold Council · Gold Demand Trends Q1·Q2 2026 · Central Bank Gold Reserves Survey 2026
- Meta · Form 424B2 prospectus supplement 2025.10 · 2026.04 issuance terms
- Hanwha Investment & Securities Kim Seong-su · Han Si-hwa FI Weekly · 2026.08.18 · Global sovereign and corporate bond issuance · hyperscaler share of issuance
- Hana Securities Heo Seong-woo Global Bonds · 2026.08.20 · Interpretation of buyback expansion · size of the U.S. Treasury market
- SK Securities Won Yu-seung Bond Comment · 2026.08.20 · Buyback caps by maturity bucket · structural upward pressure
- IBK Investment & Securities Kim Ye-seul Global Strategy · 2026.08.21 · Federal debt crossing $40tn · net-interest path
- iM Securities Park Sang-hyun Economy Brief · 2026.08.19 · 08.21 · U.S.-Japan co-movement in long yields · share of interest outlays
- Eugene Investment & Securities Bang In-seong FinBERT-FOMC · 2026.08.21 · Analysis of July minutes · probability of a September hold
- Ray Dalio · How Countries Go Broke: The Dynamic Behind What is Happening Now Author post · 2026.08.21
- DoubleLine · Jeffrey Gundlach on Channel 11 2026.08.03 · Summary of Jeffrey Sherman interview 2026.08.18
- PIMCO · Secular Outlook 「Rupture and Resilience」 2026.06
- BlackRock · Rick Rieder, Summer Markets and AI Investing 2026.07.15
- JPMorgan Chase · Jamie Dimon shareholder letter 2025 Annual Report
- Oaktree · Howard Marks memo, Nobody Knows (Yet Again) 2025.04.09
- Apollo · Torsten Sløk, The Daily Spark 2026.07.31
- Barclays · Private share of U.S. Treasury holdings 2026.08.11
- Yuanta Securities Kim Ho-jeong · Lim Ji-yoon · 15 Questions on Japan’s Financial Markets 2026.08.14 · FIMA Repo cap
- Korea Customs Service Preliminary exports and imports, Aug. 1–20 · Korean FX market Regular trading session 15:30 close
This report does not recommend buying or selling any specific security · asset-level comments are conditional descriptions of the macro factors to which each asset is exposed · investment decisions remain the responsibility of the investor · market moves after the reference date are not reflected.
Reference dates · U.S. indicators: August 20, 2026 close · Korea, gold and dollar: August 21 close · Bitcoin: August 21, 09:08 ET. Intraday prices and closing prices are not mixed in the same table · gold uses only the LBMA AM price.
RESEARCH NO.04 · 2026.08.23 · www.snowshagal.com