BofA's Hartnett warns: If Baisent fails to contain long-term interest rates, a US dollar crash and asset sell-off will follow

Last week, U.S. Treasury Secretary Besant announced that the size of long-term Treasury repurchase operations would be doubled. Bank of America’s Chief Investment Strategist, Michael Hartnett, characterized this move as “quasi-quantitative easing” (quasi-QE) in his latest Flow Show report. Meanwhile, the total U.S. national debt exceeded $40 trillion for the first time and continues to climb by $1 trillion each quarter.

Hartnett stated bluntly that all three indicators of Besant’s “3-3-3” economic framework have failed to meet their targets, and policy credibility is being reflected in the simultaneous weakening of both the bond and FX markets—“Rising yields and a weakening currency indicate that credibility is eroding.”

He concluded: This round of “quasi-QE” may temporarily suppress interest rates, but it will be unable to genuinely drive U.S. Treasury yields lower. If the 30-year Treasury yield cannot fall below 5%, the U.S. dollar faces a sharp decline, and the market will enter a systemic risk-off phase with a sell-off in risk assets. This will prompt shorting of risk assets, ultra-large-scale AI compute, private credit, and shorting of financial stocks. At the same time, Hartnett maintains a long position in gold and out-of-favor long-duration assets.

The “3-3-3” Plan: Three Prongs, Three Misses

At the beginning of his tenure at the end of 2024, Besant proposed the “3-3-3” economic framework with a target date of 2028. The three core goals are: an average annual real GDP growth rate of 3%, reducing the federal fiscal deficit to 3% of GDP, and increasing domestic daily oil production by 3 million barrels.

According to data disclosed by Hartnett in his report, all three indicators are currently lagging:

• GDP growth rate: The average rate over the past six quarters has been less than 2%.
• Fiscal deficit: Currently about 6% of GDP, double the target value.
• Oil production: Since 2024, has risen by only about 300,000 barrels/day, just a tenth of the goal.

Hartnett noted that policy credibility has always been judged by the bond and FX markets. The collective failure to meet the three targets is the deep-rooted reason behind currently elevated Treasury yields and pressure on the dollar.

Three “Maginot Lines”: $4 Gasoline, 160 USD/JPY, 5% Treasury Yield

Hartnett described current policy as centered around defending three “Maginot lines”: Gasoline prices not exceeding $4/gallon, the dollar-yen rate not exceeding 160, and 10- and 30-year Treasury yields not exceeding 5%. He believes breaching any of these three lines will directly threaten economic growth, the AI investment boom, and asset bubbles.

The reality, however, is far from optimistic. Hartnett’s report notes that American gasoline prices have once again risen above $4/gallon (a sharp increase from the prewar $3/gallon). With the ongoing U.S.-Iran “economic war” and both U.S. crude inventories and strategic petroleum reserves at 40- to 50-year lows, the potential for oil prices to drop is extremely limited.

On the FX front, Hartnett believes FX intervention to support the yen will require the Bank of Japan to deliver a significant rate hike on September 18 in order to stabilize Japanese long-term yields and reduce the risk of Japan dumping U.S. Treasuries.

The most critical line, however, is keeping long-term Treasury yields below 5%—which is directly linked to avoiding U.S. government credit events (first, downgrades by rating agencies, and second, potential Treasury auction failures), as well as keeping AI financing costs manageable.

Why “Quasi-QE”: The Dual Pressures of $40 Trillion Debt & AI Financing

Hartnett’s report outlines three main reasons behind the Treasury’s forced intervention:

First, the debt level has hit a historical threshold. The total U.S. national debt just surpassed $40 trillion and continues to expand by about $1 trillion each quarter.

Second, net Treasury issuance squeezes out corporate financing. In both 2026 and 2027, net Treasury issuance is projected to reach $2 trillion per year, directly encroaching on the corporate bond market. Meanwhile, year-to-date issuance of bonds related to AI already totals between $200 and $300 billion, and the AI arms race has become the U.S. government’s top national security and macro priority.

Third, a disconnect between data and markets is triggering a crisis of confidence. Hartnett points out that when non-farm payroll growth is zero and inflation data is also zero, yet the 30-year Treasury yield has climbed to a 20-year high, it indicates that the market’s trust in fiscal sustainability has cracked—“Immediate credibility must be restored.”

Against this backdrop, Besant has successively rolled out a series of “put option”-style policy interventions: signing dollar swap agreements with Asian and Gulf states, intervening in the FX market to support the yen, and now—under tacit approval from the Fed—doubling long-end Treasury repo operations. Hartnett sees this series of actions as “a policy panic to repair the fixed income market.”

He also highlights that this panic-style fiscal intervention still needs Federal Reserve Chairman Walsh to deliver just the right amount of hawkishness at the Jackson Hole symposium on August 28—not too dovish to avoid further pressure on the dollar, but not too hawkish lest it spook the market.

“Success is Expected, Failure Is Unimaginable”

Hartnett summed up the extreme current situation in one sentence: “Success is expected, failure is unimaginable.”

His logic is: QE was the starting point of the great bull market in the post-Lehman era and the foundation for Wall Street’s “too big to fail” narrative. Extraordinary monetary stimulus over the past 20 years has led to extraordinary increases in asset prices. Therefore, a new round of “quasi-QE” is expected to succeed.

But he then laid out a clear roadmap for a failure scenario:

“If Besant’s ‘panic-driven twist operations’ can’t push the 30-year yield below 5%, the dollar will sharply depreciate, and asset allocation will shift—to shorting risk assets, shorting leverage (AI ultra-large-scale compute companies, private credit), and shorting cyclical assets (financial stocks), all the way to the midterm elections.”

Against this backdrop, Hartnett’s long positions are concentrated in gold and overlooked long-duration assets, including REITs, biotech ETFs, regional banks, small-cap shares, and Hong Kong real estate.