Municipal Quarterly Insights 2nd Quarter 2026

Q2 Review – A Plan Comes Together – Fed Cuts Turn to Fed Hikes

The conflict between the U.S. and Iran created an “on again, and off again” environment of confusion. At some point, declaring that a deal has been made, or that one side or the other is begging to make a deal, seems counterproductive. The “speak softly and carry a big stick” approach appears to be a necessary part of changing the direction of a nearly 50-year conflict. As bombing turned into blockades and eventually a Memorandum of Understanding, oil prices swung as high as $99/barrel in May and closed the quarter around $70/barrel. The spike in oil prices weighed on consumer sentiment. Although most hard data remained positive over the quarter, the soft survey information took a hit as high prices at the pump flowed into the consumer psyche. Consumer sentiment fell to the lowest reading on record and has not materially recovered from those low levels.

Persistent deficits, debt issuance, and continued inflationary pressures likely increased the term premium for the U.S. Treasury curve. Over the quarter, yields rose and the 30-year Treasury yield broke through the 5% level. The most interesting development was that, over the quarter, both the market’s expectations and the Fed’s indications showed that the expectation for Fed rate cuts in 2026 had turned into anticipation of rate hikes in 2026. Specifically, in March, the market expected two cuts this year. Then high oil prices and persistent signs of inflation eroded the forecast for cuts. By June, given the combination of tepid strength in the labor market and elevated inflation, the market did what it does and began anticipating future hikes. The Fed’s “Dot Plot” forward guidance followed a similar path.

AI has not wiped out all the jobs yet. In June, the unemployment rate turned lower and it currently sits at 4.2%. Recent JOLTS data show that the number of jobs available per job seeker has risen from below 1 job per seeker to 1.04 jobs per unemployed person. As the quarter played out, both the services side of the economy and the manufacturing sector showed growth. Services have shown strength for quite some time, but manufacturing has been in contraction for much of the past few years. We expect the AI buildout, reshoring, and OBBBA depreciation incentives have helped turn the manufacturing sector positive in 2026. The good news in the labor market is that jobs are available. The bad news for workers is that the recent puff of inflation has pushed real wage growth, measured as wage growth relative to inflation, into negative territory. Much of the inflationary pressure is related to energy and core services, and the hope is that the energy shock is “transitory” and tied to the war with Iran. If there is no near-term end to the tensions, workers will have lost ground from a purchasing-power perspective. If the situation persists, we expect workers will push for wage adjustments and the disparity will close. We have seen real wage growth for much of the past three years, so the development may be material and may feed into the weak consumer sentiment survey data. To this point, however, consumer spending continues to grow.

Over the course of the quarter, 10-year Treasury yields increased, moving from 4.31% to 4.46%, while 10-year high-quality municipal yields fell from 3.10% to 2.88%. Our forecast for the 10-year Treasury yield for the year was a range of 4.0% to 4.80%. The highest closing yield we saw in the second quarter was 4.66%, and that may turn out to be the peak for the year, depending on when the Fed starts tightening. Taxable bond indexes have generated year-to-date returns of roughly 1%, while municipal bonds have generally seen returns of 1% to 2%.

Q3 Interest Rate Outlook – There Is a New Fed Sheriff in Town

Much of what the market is concerned about we have covered in our past two Insights pieces. The economy is more resilient than many anticipated, likely due to OBBBA stimulus, tariff refunds, AI productivity gains, AI capital expenditures, a stable labor market, and the wealth effects of a broad-based stock market ascent. Clearly, energy prices and inflation persistence are top of mind for consumers and central bankers. Last quarter, we pointed out that the neutral rate (r*) is probably higher than where the current Fed Funds Rate sits. That suggests current Fed policy may be unexpectedly stimulative. As we have forecasted since the beginning of the year, we expect the next Fed policy move will be toward tightening.

The new Fed Chair, Kevin Warsh, has made it clear that he plans to do things differently than previous Fed Chairs. His first post-meeting statement was roughly half as long as the previous statement. Mr. Warsh also abstained from contributing to the Fed’s “Dot Plot” forecast of expectations for policy rates, unemployment, and inflation. He explained that he did not see value in the forecasts and did not want policy to be “hamstrung” by previous estimates. Of note, the new Chair has announced five new task forces to reinvent or alter some of the Fed’s traditions, including communication, the balance sheet, data sources and methods, productivity and labor markets, the impact of AI, and the Fed’s approach to inflation. One of the task forces may be a “tell” regarding the next restrictive Fed action. In a July 1, 2026, statement, Mr. Warsh indicated that he thought the Fed’s balance sheet was bloated, it took 18 years to reach its current size, and that a change in balance sheet policy would not be a surprise to the market. He went on to say that using the balance sheet borders on fiscal policy and that he wants interest rate policy to be the Fed’s main tool. We anticipate the next Fed move will be to engage in Quantitative Tightening (QT). It will shrink the balance sheet while providing a level of cover for more restrictive policy. A bonus would be that it would give the Fed room to address the next crisis from many fronts.

Mr. Warsh appears to be a serious actor in his new role. His words have been used sparingly but seem to be thoughtfully chosen. It is unlikely to be an error that he announced he is more concerned about the number to the left of the decimal point than he is about the number to the right when it comes to inflation. That suggests a 2% goal is likely still the guidepost, but that a focus on an inflation print of exactly 2.0% is unproductively restrictive and may lead both the Fed and the market to become hyper-focused on the wrong target. The Fed tends to look at a mosaic of factors, and absolutes may anchor policy too heavily on meaningless differences in certain metrics. Called Donald Trump’s “Sock Puppet” by some politicians, one of his first comments out of the gate was that price stability was his current focus. Time will tell, but it seems President Trump may have chosen wisely, and that Mr. Warsh will choose position, and legacy, over party.

Chairman Warsh is only one vote out of 12 on the FOMC, so what can we expect going forward from this Fed? The Chair is important in the sense that he sets the policy agenda, controls communications, can influence markets away from rate setting, and is the architect of the balance sheet strategy. As it relates to rate setting, the Chair’s role is to build consensus; beyond that, he is only one vote. The market places the greatest chance of a hike in October, though there is some possibility of a September hike. We think the base case is QT, or shrinking the balance sheet, in the fall; continued rhetoric that price stability is the greatest focus; and a Fed that waits for the five task forces to release their findings toward the end of the year. QT shows a commitment to managing inflation while some possibly transitory sources of inflation subside. If inflation does not cool as the year progresses, a December hike may make sense. As we write this strategy piece, the 30-year Treasury yield is above 5%, something that has only been the case for a handful of short periods over the past 20 years. Long bond investors think that inflation is more hard-wired, the fiscal situation demands a yield premium, or both. With trillions in government debt to finance, and with economic activity and price stability flowing into debt interest costs, Mr. Warsh and the FOMC have work to do.

Staying on the central bank theme for a bit, the recent ECB Forum in Portugal seemed to develop unity among central bankers. That may be a good thing if they need to coordinate, but it has a remote chance of resulting in central banks acting uniformly, which could make for a less resilient global economy if policy is errant. A comment from the Bank for International Settlements (BIS) seemed to sum up the risk that AI could pose to the global economy nicely: “Disappointment in returns could trigger a sudden pullback in financing and turn the capex boom into a protracted investment bust” and “a major equity market correction could have larger economic consequences today than in the past.” CNBC recently provided numbers that place some weight on the BIS comments. As of June, the median monthly corporate AI spend per employee in the U.S. is $11. If we assume that all employees who work for corporations with 10 or more employees, or roughly 118 million people according to the Bureau of Labor Statistics, incur AI spending of $11 per month, that equates to approximately $15.5 billion in annual AI spending. In December 2025, Goldman Sachs estimated that AI hyperscalers in the U.S. would spend $527 billion in 2026 on capex, on top of the $1 trillion already invested through 2025. We sure hope that individuals, government entities, and not-for-profits have an insatiable willingness to spend on AI, because left to corporate spending, the AI breakeven would occur in roughly 97 years ($1.5 trillion capex divided by $15.5 billion in annual corporate AI spending). Do not forget that the useful life of the GPUs in the data centers is estimated to be approximately five to six years. According to Presenc AI, current AI revenue ranges from $60 billion to $90 billion. If we assume the midpoint revenue of $75 billion, that results in a payback period of approximately 20 years, ignoring the many operating costs and a few rounds of GPU updates over that horizon. Here is where things get interesting. According to CNBC, the costs per million output tokens for the major AI models are as follows: Anthropic Claude Fable 5 is $50, OpenAI GPT 5.5 is $30, Anthropic Claude Opus 4.8 is $25, OpenAI GPT 5.4 is $15, Google Gemini 3.5 Flash is $9, and the low-cost provider is DeepSeek’s V4 Pro at $0.87. It may be that the future of AI will look like the pricing scheme of the old Apple iPods, tablets, and phones, where people will, in a zombie-like fashion, pay dearly for the latest edition. But it sure seems like AI will face price competition, and the realized revenue could be insufficient to support the hyperscalers that get left in the dust.

We have discussed the concerns with private debt at length this year, but the exposure that some private debt funds have to the AI theme is one aspect that has caused some funds to experience significant investor liquidation requests. In 2026, hyperscalers issued more than $100 billion in debt, and forecasts suggest that total issuance in 2026 will approach $250 billion. Their percentage of total investment-grade corporate debt issuance has grown from approximately 6% of the market in 2022 to more than 13% on a year-to-date basis in 2026. Although we have had a savings glut for many years following the Great Financial Crisis, that may become tested. The economic “crowding out” theory says that when governments run deficits and have to borrow money, it impacts corporations through higher borrowing costs, resulting in less corporate investment and slower economic growth. This moment feels like corporations may be doing some of the crowding out. Q1 2026 corporate profit growth was 29.4% for the S&P 500 companies, so seemingly the U.S. government’s level of borrowing is not impacting corporate profits too badly. According to SIFMA and the St. Louis Fed, total corporate debt and obligations as of Q1 2026 are approximately $16 trillion to $18 trillion. That compares to total outstanding U.S. Treasury and agency debt of $32 trillion to $33 trillion. The government may be the dominant debtor, but at some point, they may crowd out each other. We are starting to see risk spreads widen on the debt of some of the big tech borrowers, even where the additional debt is still fairly meaningless to the balance sheet of some of the strongest tech borrowers. The insatiable appetite for capital feels like it will increase borrowing costs, since investors will need to be enticed with stronger returns or yields to fund what seems to be a high-risk venture. Will investors start to materially assign different spreads to the dominant, or most cost-effective, AI model providers, hyperscalers, and infrastructure companies? What if the victor in this competition becomes the “one AI to rule them all?” We expect the “hunch” that has some investors showing a level of concern and caution in this space will be the subject of many market commentaries over the next several years.

AI will be changing the face of the labor market, so stable employment may give way to a challenging career environment. We like to think that job losses will be focused on call center positions, data entry, basic programming, or jobs made entirely of one predictable digital task. Career advancement in professional fields will likely become more difficult. We are already seeing this in the decline in availability of some entry-level positions. Seasoned professionals, those whose jobs involve judgment, relationships, handling exceptions, and responsibilities, will augment their teams with AI research and capabilities. Getting that first job will be a major accomplishment. We may be about to see what AI will do to professional tasks. Anthropic is releasing AI agents designed to handle a broader mix of financial services tasks. The new agents will draft pitch decks for client meetings, review financial statements, and bring cases to the attention of compliance for review. The new capability is aimed at banking, insurance, asset management, and financial technology efforts. Nicolas Lin, Anthropic’s head of product for financial services, said, “Finance is a great blueprint for the rest of knowledge work.” In February, they released plug-ins for their Claude software to assist with financial analysis, equity research, private equity, and wealth management. The face of work will change for many. Is it any wonder that many young people are sour on the influence that AI will impose upon their lives?

We anticipate that the BIS warning will become relevant in the years to come, but our base-case scenario for the rest of 2026 is essentially the status quo. AI spending, corporate profitability, and OBBBA stimulus should continue to support enthusiasm for both equities and the economy more broadly. Bonds remain range-bound, with our forecast for the 10-year Treasury remaining between 4.0% and 4.80%. Massive debt issuance should put corporate risk spreads in the crosshairs to widen. The skirmish with Iran will likely persist. Iran survived our blockade and sanctions in the past, and continuing the on-again, off-again war through the mid-term elections makes too much sense. If you were Iran, wouldn’t you prefer a gridlocked U.S. government? As a result, we expect oil prices to be elevated but stable, at a level that causes U.S. consumers some stress. Negative real wage growth and high prices at the pump should weigh on consumer sentiment. The Fed will incrementally fight inflation, first through QT, as the mixed economic messages do not seem to call for rate hikes immediately. Longer term, persistent deficits and the cost to finance our debt burden should keep interest rates higher for longer. But if history is a guide, high real rates, and even a low equity risk premium, suggest strong future bond index returns are likely.      

Municipal Market Developments – The Song Remains the Same

So far this year, we have covered at length how we are investing in our process to use AI to make municipal credit analysis more forward-looking and insightful. That concept received a lot of financial news headlines this past quarter. It seems we are not alone in our efforts to make the models better at data extraction, and we have also discovered that humans are needed to monitor data extraction for accuracy. AI offers time savings, but an experienced understanding of the differences between each type of credit is currently necessary. The mosaic of information that an analyst currently accesses to determine the nuanced difference between an issuer’s ability to pay and its willingness to pay remains uniquely a “human” thing. Where AI gets interesting is ripping through massive amounts of data to help us identify the changes in credit and financial metrics that have the greatest relationship to a subsequent rating downgrade or upgrade. The surprising thing is that the changing metrics that often precipitate an upgrade or downgrade are not the same.

We expect that insightful credit analysis and issuer/security selection will come to drive performance in the coming years. Years of federal Covid stimulus, along with strong tax receipts, may come to an end. The status quo for municipal bonds is that municipalities may come to expect less funding from Washington. A study we should conduct would examine issuers that received Covid dollars and managed expenses before the spending ran out. It would help sort out the fiscally responsible issuers from the less serious ones. Earlier in the year, we talked about school districts being the poster child for fiscal mismanagement when one-time federal dollars turn into ongoing expenses. We have already covered the topic at length, but school districts face mounting financial pressure. Enrollment is shrinking, labor costs are rising, and federal pandemic aid is subsiding. Meanwhile, the number of teachers is at an all-time high, and the mismatch between student headcount and teacher headcount has contributed to the development that nearly a third of school districts operate at a deficit as of FY 2024. When the superintendent of Milwaukee Public Schools says about 91% of their students are “not yet scoring proficient on grade level in literacy,” it seems we have a major problem on our hands. Many school districts have tough choices ahead. Since negative ratings or outlook revisions increased 40% in the year through June 2025, this is a sector we are watching closely and developing new surveillance tools to help navigate.

The municipal bond market just absorbed a large amount of issuance while outperforming other sectors, as yields ended up falling for munis while Treasury yields rose. Investors have been increasingly using mutual funds over ETFs, and active ETFs generally outperformed their passive peers. Proposed and passed tax law changes should increase demand for municipal bonds. The state of Washington just passed a 9.9% income tax for high-income earners. Maine just passed a 2% income tax on those making $1 million or more in a year. Some Minnesota lawmakers are proposing a wealth tax on decamillionaires, or those with a net worth of more than $10 million. The tax changes and proposals should marginally generate interest in both tax-advantaged income and moving vans. With strong demand and decent returns, against the backdrop of our “status quo” forecast, we see coupon income as the most likely source of return for the balance of the year, and we feel pretty good about our “coupon-like” total return estimate offered at the start of the year.

Strategy and Summary – Steady as She Goes

Our strategy for this quarter is uncomfortably similar to last quarter. The U.S. economy remains fundamentally resilient, unemployment is low, corporate earnings remain robust, AI is delivering productivity gains, and fiscal stimulus is aiding economic activity. Those strengths are counterbalanced by persistent inflation, elevated deficits, AI job-market impacts, and geopolitical uncertainty. Recent Fed meeting minutes make it seem like the FOMC believes it is somewhere between a neutral and an accommodative position, so a tightening move is the most probable outcome. The new Fed Chair, Warsh, is not a “Manchurian Candidate,” and price stability is his first stated concern. All signs point to tightening by the Fed. Initially, that can put bond prices under pressure, but the longer-term influence that a tightening cycle can have on bond prices has to do with where the market expects yields to be at the end of the cycle, or the terminal rate. We expect that the 10-year Treasury yield will explore a range from 4% to 4.80% in 2026. We think traditional hikes are more of a late-2026 issue and more probably a 2027 concern. Given all the moving parts in the economy and geopolitical uncertainty, it feels odd to say, but our forecast for 2026 is “Steady as She Goes.”

This Newsletter is impersonal and does not provide individual advice or recommendations for any specific subscriber, reader or portfolio.  This Newsletter is not and should not be construed by any user and/or prospective user as, 1) a solicitation or 2) provision of investment related advice or services tailored to any particular individual’s or entity’s financial situation or investment objective(s).   Investment involves substantial risk.  Neither the Author, nor Advanced Capital Group, Inc. makes any guarantee or other promise as to any results that may be obtained from using the Newsletter.  No reader should make any investment decision without first consulting his or her own personal financial advisor and conducting his or her own research and due diligence.  To the maximum extent permitted by law, the Author and Advanced Capital Group, Inc., disclaim any and all liability in the event any information, commentary, analysis, opinions, advice and/or recommendations in the Newsletter prove to be inaccurate, incomplete or unreliable, or result in any investment or other losses.  The Newsletter’s commentary, analysis, options, advice and recommendations present the personal and subjective views of the Author and are subject to change at any time without notice.  The information provided in this Newsletter is obtained from sources which the Author and Advanced Capital Group, Inc. believe to be reliable.  However, neither the Author nor Advanced Capital Group, Inc.  has independently verified or otherwise investigated all such information.  Neither the Author nor Advanced Capital Group, Inc. guarantee the accuracy or completeness of any such information.