KEY TAKEAWAYS
Two Houses Named the Capital Shortage in the Same Week. Merrill's August 31 letter calls the new regime an age of capital competition, with investment-grade issuers already selling nearly $1.5 trillion of bonds this year.
Every Dollar of EM Primary Supply Last Week Went to Investment Grade. Ten issuers priced roughly $8.6 billion equivalent, and the only sovereign borrower was Mexico.
Spreads Say Access Is Cheap While the Calendar Says It Was Shut. The EMBI Global Diversified sovereign spread sat at 234 basis points on August 28, inside the 253 it started the year at.
Seoul Joined the Tightening Turn and Split This Desk's Korea Call. The Bank of Korea hiked to 3.00% in a 6 to 1 vote and raised its 2026 growth forecast to 3.3%.
Two Houses Name the Same Shortage
The weight of sell-side conviction this week landed on a single diagnosis, that capital has stopped being abundant, and both houses that made the case did so without once pricing a frontier borrower. Merrill Lynch's August 31 Capital Market Outlook devotes its Market View to what Joseph Quinlan calls the end of capital abundance, enumerating five structural claimants on global savings, debt, defense, digitalization, demographics and deglobalization, and concluding that real interest rates may remain structurally higher than in the previous two decades. The evidence he marshals is a funding statistic, not a sentiment one: investment-grade companies have sold nearly $1.5 trillion of bonds this year, roughly 36% more than the comparable period last year. BlackRock's August 31 weekly reaches the identical conclusion by a different route, noting that the AI buildout and widening government deficits have intensified competition for capital, with U.S. 30-year yields at a 19-year high above 5% and more than 80% of the global bond universe now yielding above 4%.
Where the two houses part company is the portfolio conclusion, and neither conclusion touches the borrower that sits last in the queue. Merrill's Portfolio Considerations favor non-U.S. equities and name emerging markets explicitly, while its own BofA Global Research forecast table carries fed funds from 3.63% today to 4.38% by the fourth quarter. BlackRock stays pro-risk with an overweight to U.S. equities and a strategic preference for short to medium-term government bonds. Both treat scarce capital as an equity selection problem, who wins the capex cycle, and as a duration problem, how far out the curve to sit. Neither treats it as a credit access problem. Goldman Sachs Asset Management's August 29 Market Monitor supplies the number that makes the omission expensive: the EMBI Global Diversified sovereign spread closed August 28 at 234 basis points, tighter than the 253 that opened the year.
The Savings Pool Is Not the Queue
Merrill's own audit trail leads to the Fund, and the Fund's number proves the opposite of what the letter uses it for. Quinlan cites the International Monetary Fund to establish that gross global savings remain around 27% of world GDP, roughly in line with the average of this century, and treats that stability as evidence the supply of savings has not collapsed. He is right, and it is beside the point for a sovereign bondholder. An aggregate savings ratio measures the size of the pool. It says nothing about the order in which claims on that pool are settled. The five claimants Merrill names are, without exception, senior in that order: the U.S. Treasury financing a deficit that publicly held debt takes from roughly 101% of GDP in 2026 toward 120% by 2036, rearming developed-market defense ministries, hyperscalers whose investment-grade bond issuance has topped $100 billion this year by BlackRock's count, and aging developed-world pension systems. The frontier sovereign is not on the list.
The Fund's most recent framing of this question is a chapter title, and this week is the test it poses. The October 2025 World Economic Outlook devotes its Chapter 2 to "Emerging Market Resilience: Good Luck or Good Policies?", a question that only has an answer when the external environment stops being accommodating. This desk notes a genuine gap in its own evidence base this week: the retrieved Fund material returned reference apparatus rather than country baselines, so no specific Article IV financing gap is available to cite here, and none will be invented. Data Unavailable is the honest entry. What is available is the market's own arithmetic. With the 2-year Treasury at 4.34%, the 10-year at 4.72% and the 2-10 slope at just 37 basis points on August 28, the risk-free alternative to lending a frontier sovereign money has rarely been better compensated in this cycle.
The mechanism the Street is missing is not spread widening, it is rationing, and rationing is invisible in a spread series. A sovereign that cannot issue does not print a wider spread. It prints nothing at all, and its existing bonds continue to trade wherever the last marginal buyer left them, which in a week of tightening index spreads looks like strength. Gramercy's August 29 EM Weekly records hard-currency sovereigns returning 0.48% at the index level, with the CCC bucket up 1.11% against 0.33% for single-B credits. Distressed and event-driven names led, Venezuela up 2.89%, Ukraine 1.79%, Lebanon 1.28%. That is a secondary market pricing recovery stories, not a primary market extending credit. The two are routinely confused, and the confusion is most dangerous precisely when index spreads are near their tights.
Where Rationing Shows Up Before Spread Does
The clearest evidence that the queue has lengthened comes from the central banks that no longer get to choose their own policy, and Hungary's governor named it explicitly. The MNB cut 25 basis points to 5.50% in a third straight reduction, with July inflation at just 1.2%, and Governor Varga paired the cut with an explicit warning that the high global yield environment poses capital-outflow risks for emerging markets. That is a monetary authority with domestic room to ease naming the external constraint as the binding one. The Bank of Korea moved the other way entirely, hiking 25 basis points to 3.00% in a 6 to 1 vote, its second consecutive increase, while raising its 2026 growth forecast to 3.3% and upgrading its core inflation projections. Gramercy reads the dispersion as an argument for country selection over index beta. This desk reads it as the first stage of a sort between sovereigns that can fund themselves and sovereigns that must be funded.
The dollar is doing the sorting, and it accelerated after Jackson Hole. Chair Warsh used his first keynote to break with two decades of practice by declining to signal a policy path, saying the economy is at full employment while inflation figures are more concerning, and concluding that prices should be the Committee's predominant focus. Markets repriced September hike odds to around 60%, the short end of the Treasury curve widened by as much as 10 basis points, and the dollar index rose 0.91% on the week. July core PCE printed 3.3% for a second straight month, leaving the Fed's preferred gauge stuck in a 3.3% to 3.4% band for four consecutive months, while real consumer spending went flat after a 0.4% June gain. A Fed biased toward tightening and withholding guidance keeps an uncertainty premium embedded in the global discount rate, and the frontier borrower pays that premium last and most.
Beneath the index, the sort is already visible in the names this desk has been Cautious on. Angola fell 0.86%, Senegal 0.49% and Gabon 0.41%, the weakest hard-currency performers of the week, with Africa returning 0.28% against Latin America's 0.65%. Local currency was the only sub-asset class in negative territory at minus 0.26%, dragged by Colombia's 3.10% decline, of which 2.68% was currency. Meanwhile the sanctions perimeter widened in a way that maps directly onto funding access: Washington's "Operation Economic Outcast" designated roughly 60 entities and named five lifeline categories, shipping, digital assets, gold, aviation and technology, allowing exposure to be mapped by jurisdiction. Türkiye screens in three of the five, alongside its reliance on Iranian gas imports, in the same week its local market added 1.25% on an almost entirely price-driven move. A market pricing a credit higher while its designation surface widens has not read the annex.
Ten Deals, One Thing in Common
The single most consequential line published about emerging markets last week was not a view, it was a sentence in a data table. Gramercy's August 29 market section records that primary market activity was steady but entirely investment grade, with ten issuers pricing approximately $8.6 billion equivalent in hard-currency supply, roughly half of it euro denominated. Mexico was the only sovereign borrower, and it printed a four-tranche yen deal of roughly $1.8 billion equivalent at yields between 3.16% and 5.49%, with three of the four tranches pricing wider than initial talk. The rest was bank financials, from ICICI Bank to Bank Leumi. Read that against the tape. In a week when the CCC bucket gained 1.11% and the index spread sat at 234 basis points, not one sub-investment-grade borrower raised a dollar. Secondary performance is a poll of existing holders. The primary calendar is the only vote that settles whether a sovereign can roll a maturity, and last week the frontier did not get to cast one.
The Sort Between Funded and Waiting
The Standing Book, now due, carries four theses this desk has held for more than a month, and the evidence this week did not treat them equally. The preference for the front end of the EM hard-currency curve over duration has been carried since The Seven-Day Regime in mid-August and rolled forward every dispatch since, and this week the evidence ran against it: Gramercy records the 10-plus year segment returning 0.73% against 0.27% for the 3 to 5 year bucket, with the same pattern in corporates. This desk is not repositioning on one week's curve behaviour while September hike odds sit near 60%, but the counter-evidence is recorded here rather than omitted. The Cautious stance on Sub-Saharan African commodity exporters, held since mid-July, was supported by Angola, Senegal and Gabon leading the week's decliners. The Korea call taken on August 30 requires an honest split, set out below. The Asymmetry on Pakistan persists unchanged, with nothing this week to test it. The Constructive stance on the Indian rupee and external accounts, taken on August 30, is rolled forward deliberately rather than dropped, and this week's access test is the reason why.
We are Cautious on frontier hard-currency sovereigns carrying 2027 refinancing needs. A calendar that granted zero sub-investment-grade access in a week of 234 basis point index spreads is a funding signal, and it will not stay invisible in secondary pricing indefinitely.
We Prefer EM investment-grade issuers with demonstrated primary access over frontier high yield. The ten borrowers who cleared last week, roughly half in euros, are the credits that have proven they can refinance into a competitive capital market rather than merely trade well in one.
We are Constructive on the Korean won and are walking back the duration leg of that call. This desk went Constructive on Korean duration and the won on August 30, and the Bank of Korea's second consecutive hike to 3.00% alongside upgraded core inflation projections contradicts the duration half directly; the currency half, with the won near its strongest in a year, stands.
We Prefer the front end of the EM hard-currency curve over duration, with the counter-evidence noted. A 37 basis point 2-10 slope offers no compensation for extending, though a third consecutive week of long-end leadership would force this desk to change its mind rather than restate it.
We are Cautious on Sub-Saharan African commodity exporters, consistent with the stance held since mid-July. Africa's 0.28% return against Latin America's 0.65% is the ninth week this positioning has been carried, and the China import-demand floor has still not broadened into volumes.
We are Constructive on the Indian rupee and external accounts, rolled forward from August 30. Two Indian lenders, ICICI Bank and Union Bank of India, were among the ten borrowers that cleared primary last week, which is exactly the access this desk's framework says to price.
We are Cautious on Türkiye across external and local exposure. Appearing in three of the Treasury's five lifeline categories is a designation surface, not a designation, but it is not priced into a local market that just added 1.25% on price alone.
Solvency Is a Calendar Problem
A sovereign does not default on a spread, it defaults on a date. The market spent this week debating whether capital is scarce, and two of the largest houses in the world concluded that it is, then expressed that conclusion entirely within the asset classes where scarcity is an opportunity rather than a constraint. Merrill put the regime in one line in its August 31 letter: "Prologue will be defined by borrowers chasing capital." The letter does not ask which borrowers are unable to give chase. Its own house forecast answers it by taking fed funds to 4.38% by year end, and its Chief Investment Officer notes that a genuine hiking cycle hits the most levered areas hardest. The frontier sovereign is the most levered borrower in the queue and the least able to wait. Spreads at 234 basis points describe the confidence of people who already own the bonds. The primary calendar describes who can still borrow. Last week those two numbers told opposite stories. Only one of them has to be refinanced.
What Would Change Our Mind
The Fed blinks on September 16. A hold accompanied by explicit guidance that the tightening bias has ended would reopen the risk appetite that rations frontier primary access, and would undercut the entire funding-queue thesis.
A sub-investment-grade sovereign clears a benchmark deal at or inside talk. One frontier borrower pricing without concession would prove last week's all-investment-grade calendar was a late-August seasonal artefact rather than a rationing signal.
The August payrolls print on Friday September 4. A firm number after a run of sub-100k prints hardens the September hike into consensus and lengthens the queue; a sharp miss turns the Fed's problem into a growth problem and changes which sovereigns are vulnerable.
Regards,
Sovereign Dispatcher