DoubleLine's Jeffrey Gundlach said he doubts the roughly 60% probability of a September Fed hike priced into the curve, and that a hold on Wednesday would push long-end Treasury yields "fairly significantly" higher from the 30-year's 5.24%. He cited core PCE running above its 12-month rate, an import-export price gauge near 7%, AI credit spreads that have roughly doubled, and a Shiller CAPE of 42x as reasons the term premium stays elevated.
Jeffrey Gundlach is betting against the Treasury curve's pricing for a September rate increase, and he says a Fed decision to stand pat on Wednesday would drive long-end yields "fairly significantly" higher from the 30-year's current 5.24%.
"I wouldn't be surprised if the Fed doesn't hike next week," Gundlach, chief executive officer and chief investment officer at DoubleLine Capital, said on the firm's "Gundlach Unlocked" webcast Tuesday. "If that's the case, I expect long-term rates to rise fairly significantly after the Fed meeting. If they do hike, then perhaps the bond market stays about where it is right now."
The effective fed funds rate stood at 3.63% on Tuesday against a 2-year Treasury yield of 4.998%, a 137-basis-point gap that Gundlach said implies the policy rate should be about 50 basis points higher. CME FedWatch put the odds of a cumulative 25-basis-point increase at 59.4%, leaving a 40.6% probability of no change. The 30-year yield has climbed almost 500 basis points from its 2020 low of 27 basis points, producing mark-to-market losses of more than 50% on the long bond without a meaningful retracement — a pattern Gundlach reads as continuation, not exhaustion. The 10-year yield traded at 4.78% against a model fair value of 4.71% built from the German 10-year and a seven-year average of US nominal GDP, a fit with an R-squared of 0.93.
The stakes extend past the front end. DoubleLine funds hold outright short positions in 30-year Treasuries, a stance Gundlach said has "worked quite well," and he repeated a recommendation to avoid long-duration government debt "everywhere in the developed world." Japan's 30-year yield has risen to 3.97%, narrowing its gap to the US 30-year to under 150 basis points, evidence that the repricing is global rather than a US fiscal artifact. Friday's August consumer price index report lands before the Wednesday decision and will shape how the committee reads the inflation path.
Gundlach's case rests on inflation that is not decelerating. Core PCE on a six-month annualized basis is running above its 12-month rate, with core inflation at 3.3% and headline at 3.7% — both far from the Fed's 2% target. His preferred gauge, the import and export price index, shows export prices up 8.25% year over year and import prices up 5.95%, an average of roughly 7%. He overlaid the current CPI path on the 1960s through early 1980s experience, when headline inflation peaked at 12.5% and then approached 15%, and called the shapes "eerily similar."
Energy and commodity prices reinforce that reading. Brent crude is near $100 a barrel, the Bloomberg Commodity Index has gained 34% since the war began and is testing highs last seen more than a decade ago, and the US Strategic Petroleum Reserve has fallen to 287 million barrels from a peak of 750 million — a decline of more than 50% and the lowest level since the reserve was created. Global oil inventories sit at their lowest since 2018. Refilling the SPR would put a floor under crude, Gundlach said, keeping inflation stickier than the Fed wants.
The supply side of the bond market is doing the rest. The Treasury is borrowing against a deficit running at 6% to 7% of GDP, and corporate issuance — particularly from artificial intelligence borrowers — is adding to the pile. Gundlach flagged a divergence in credit that broader investment-grade spreads have masked: AI-linked investment-grade spreads have widened to about 125 basis points from 50, a 75-basis-point move, while the rest of the investment-grade market is unchanged. In high yield the gap is wider still, with AI spreads at roughly 325 basis points from about 180, even as non-AI high-yield spreads sit near their lows for the year.
"I'm really not sure who is buying these AI-related bonds," Gundlach said. "The market is clearly having a hard time digesting this amount of supply, and AI supply is going to continue to be an avalanche."
Equity valuations offer no offset. The S&P 500's Shiller cyclically adjusted price-to-earnings ratio is 42 times, above the level reached before the 1929 crash, and information technology accounts for a record 38% of the index — more concentrated than at the 1999 dot-com peak or on the eve of the 2008 crisis. Gundlach cited research covering 1965 to 2015 showing that at CAPE levels near 42, forward 10-year real returns have never been positive, typically landing between negative 5% and negative 9% a year. "This is an extremely concentrated market, which means it is an extremely dangerous market," he said. "I wouldn't recommend any cap-weighted equities."
The dollar is the other leg of the trade. The DXY index has fallen below 100 from 110 in late 2024, and Gundlach expects further weakness. The S&P 500 has trailed the MSCI Emerging Markets index by about 20% since late 2024, a gap he ties to the dollar's slide, and he favors emerging-market local-currency debt over US corporates on the same logic. He also rejected long-dated TIPS as a hedge, noting that 30-year TIPS yields have moved in lockstep with 30-year nominals since late 2021, leaving the spread between them essentially unchanged for five years.
Fiscal arithmetic frames the longer horizon. Total US public debt, including holdings by the Federal Reserve and Social Security, stands at $40 trillion and is on a path to $50 trillion by 2032. Congressional Budget Office projections that show the deficit expanding rest on assumptions of lower rates, smaller deficits and uninterrupted real GDP growth; stress those assumptions, Gundlach said, and the deficit reaches 7% to 8% of GDP within a decade. He was skeptical that the Treasury's newly announced buyback program would move long-end yields.
For bond investors, the near-term setup is binary. A hike Wednesday keeps the 30-year near 5.25% and validates the market's pricing; a hold removes the anchor and, on Gundlach's read, leaves the path of least resistance pointing higher in yield. Either way, the term premium that has driven the long bond's 500-basis-point repricing since 2020 shows little sign of compressing, and the August CPI print on Friday is the last major input before the committee votes.
This article is for informational purposes only and does not constitute investment advice.
