The Song Remains the Same

August gave us a hawkish Fed chairman, a Treasury secretary buying his own long bonds and an earnings season better than almost anyone forecasted. All three matter, but markets are paying the most attention to the third one.

Second quarter earnings grew about 31.8% excluding a large one-time gain¹, and revenue growth was the strongest in five years. Analysts look for another 28.2% in the third quarter¹, and companies are guiding higher at roughly twice the normal rate. Unemployment is low and consumer spending accelerated through the spring. The data mostly supports a constructive view.

What we are watching is how much of this year’s strength came from items that will not repeat. An unusually large tax refund season, more than $100 billion in tariff refunds and an investment markup that overstated index earnings all helped. None of them recur.

That is not a reason to avoid risk markets, although it does mean paying closer attention to what happens when the one-time items stop.

Western Winds

Chairman Warsh spent his first Jackson Hole telling markets what he will not do: treat two good inflation prints as evidence of progress. He reaffirmed 2%, measured by the PCE deflator, as a fixed target, after leaving it in doubt in July. Recent inflation numbers are concerning, and the better summer readings did not persuade him the underlying trend has improved. Moderate wage growth no longer reassures him, since wages have not forecast inflation well in a long time. Financial conditions, he added, would be hard to describe as restrictive.

Put those together and the default changed. Before Friday, the FOMC likely stood pat on the policy rate unless the data argued for a move. Now it is forecast to move higher unless the data argues against it. Several former Fed officials came away reading it the same way, and the market repriced, with odds of a rate increase higher across the curve.

Warsh seems to be after credibility rather than a particular funds rate, and a target investors will price without being told what the committee intends. He has work to do. A bit more than half the PCE basket is rising faster than 3% a year. That is well off the post-pandemic peak, but not yet normal.

On the most important part, he is doing better than the commentary suggests. Long-term inflation expectations are priced about where they should be. Breakevens and inflation swaps sit close to the FOMC’s target across the curve, and they sit there because the market believes this committee will raise rates if the data dictates it.

And long-term interest rates are doing much of the tightening already. The 30-year bond touched a 19-year high this month¹, and contrary to some commentary, much of that move looks like a higher expected path for short rates rather than investors demanding compensation for owning the country’s debt. Real yields have moved higher, which generally means the market is expecting higher growth and policy rates together, and the economy has given it reason to reflect that. Capital spending is running hot, with more than half of this year’s growth in capex traced to the AI build. Revenue growth at the index level is the strongest in over five years.

So financial conditions have tightened without the FOMC doing too much. Warsh described that outcome approvingly in July, and he has again now. He gets the tightening but does not have to own a hike.

The catch is that borrowed credibility has an endpoint. August CPI arrives on September 11, four days before a FOMC meeting where some dissenters are still waiting for a hike. A soft print makes holding easy to explain. A firm number puts Warsh back where July left him, defending a hold without a good reason. That increases the risk of a hike.

A Bigger Stick

On August 19, with the 30-year bond near a 19-year high¹, the Treasury announced that it would double its long-dated buybacks to at least $4 billion an operation, running from early September through November ⁴.

Yields fell immediately, but by the next afternoon they were back above where they started. The amount of buyback was never the point. Marketable Treasury debt is about $30 trillion², and $4 billion an operation against that is almost a rounding error. But the threat is the point. Within a day the Treasury Secretary indicated the operations could get larger, and officials let it be known that the department could reach into the Treasury General Account if needed. Anyone considering selling long-dated bonds would be going against a buyer with no stated size limit and no obligation to explain itself, which is a real deterrent, and it may work for a while.

The problem is what comes after a while.

Governments that defend a price in their own bond market have a poor record. Even in the US, the Fed capped long Treasury yields from 1942 to 1951 to fund the war. The cap outlived the war by six years, helped produce double-digit inflation and took the 1951 accord to undo. The wall between debt management and price management was built on purpose.

The mechanism is the issue. Once the market decides the Treasury is defending a level, every move toward that level becomes a test of resolve, and the operations have to grow to keep working. What starts as cash management ends as a commitment that must be constantly supported.

Stanley Druckenmiller made this case in the Wall Street Journal last week, and made it better than we will. His argument is that the intervention responded to a price rather than to a problem. There were no failed auctions, no dealer balance sheets seizing up, nothing resembling the Treasury market in March 2020 or the gilt market in the fall of 2022. Trading was orderly⁶.

His larger point matters more. The 30-year Treasury yield is the last fiscal disciplinarian the country has. Neither party is likely to reform Social Security or Medicare because an agency published a projection. Democracies usually repair their finances when the cost of not repairing them shows up somewhere voters can feel it, in mortgage rates or a weak bond auction. Suppressing this only subsidizes delay, and Druckenmiller would rather go ahead and pay the market’s inevitable price.

Funding long-bond buybacks with T-bills is quantitative easing in small doses, run out of Treasury rather than the Fed, while inflation sits above target and Warsh is trying to keep conditions tight. The operations also end the day after the midterms. Motives aside, that spends credibility built over a long time.

There is another side to this. A reasonable case holds that the rise in long-term yields reflects better expected growth rather than fiscal stress, and that the term premium has barely moved. We agree with that much. Where it goes wrong is the next step, that the long end in August is the least liquid corner of the least liquid month and supplying liquidity there is simply the job. If the move really is about growth, it does not need defending.

Governments that defend a price against the fundamentals usually lose. The only question is how much they spend before they learn that.

Consumer Check-In

TSome consumer data is upbeat. Johnson Redbook year-over-year sales growth is running almost 10%³. Same store sales accelerated at retailers serving lower-income households and at those serving higher-income ones, two trends that do not usually happen together. Unemployment remains low.

But a couple of one-time tailwinds have helped. An unusually large tax refund season lifted household cash flow in the spring, right as those sales were being rung up, and more than $100 billion of tariff refunds went back to companies over three months⁴, some of it reaching the shelf via lower prices and heavier promotion. Both of these benefits are beginning to fade.

The labor market also remains solid, but there are some issues. Trend job growth is only running around 20,000 a month¹. The participation rate has fallen another point this year¹, which is part of why the unemployment rate still looks contained. July payrolls missed badly and the market shrugged, because the FOMC is focused on inflation rather than jobs¹.

Consumer confidence remains a question. The Conference Board index sits near its Covid lows, and the Michigan survey is close to historically weak readings¹, while some management teams act like the consumer is doing better than they have since 2022. Sentiment has generally been a poor forecaster, so we would not lead with it. But the offsetting income that let households forget some of their headaches is going away.

Some important variables are not cooperating. Brent near $90 ¹,with the Strait of Hormuz in limbo, acts as a hidden tax, and mortgage rates near 6.7%⁵ are doing the same for housing. Neither has really moved in the direction the household needs. And tariffs remain a question mark. Tariffs of 50% on Canadian vehicles, parts and steel arrive January 1, 2027, and those will show up in auto prices absent a deal before then.

None of this says the consumer is breaking, especially given earnings expectations, but a good quarter probably pulled forward some demand. Consumption is about two-thirds of the economy, so this is an area we are watching into year end.

Earnings Path

Second quarter earnings were excellent, but what comes next?

S&P 500 2Q earnings tracked close to 50% growth against expectations near 22% when the quarter began¹. Roughly 20 points of that came from a few large technology companies marking up private company stakes they hold. Strip that out and the index still grew about 31%¹. The median company grew 14%.

And estimates keep rising, which is one reason the market has absorbed higher long rates without much complaint. Analysts have 3Q earnings growing 28% on 12% higher revenue¹, and the guidance behind that is unusually friendly. Managements are not behaving as if they are bracing for a slowdown.

The shape of next year is a little harder. Consensus estimates have index growth slowing to about 14% in 2027, with the deceleration concentrated in communication services and energy¹.

Valuation is less of an obstacle than it gets made out to be. The forward multiple is 19.6, a shade under its five year average and modestly above its ten-year¹, which is not a market priced for perfection. The cap-weighted index does keep getting more expensive relative to the average stock, which is something to keep an eye on.

Where This Leaves Us

We come out of August with some constructive data points.

Earnings are growing, estimates are still rising and managements are guiding higher rather than lower. The forward multiple near 19.6 sits close to its long-run averages, which in our view is not a stretched starting point. The Fed gained credibility without raising rates. The long end tightened conditions on its own, and long-term inflation expectations stayed anchored through it.

The work from here is more in the details than the direction. August CPI is released September 11, and the committee meets four days later. The Treasury’s buybacks either stay a liquidity tool or become a defended price, and the size of the operations will tell us which. The consumer needs to stay healthy once the refunds are gone, which is important since consumer spending is such a large part of the economy. And next year’s earnings growth needs to decelerate to a still solid rate to keep its momentum as a market driver. That is where our attention will be.

Sources:

¹FactSet/Factset.com

²Fred.stlouisfed.org

³tradingeconomics.com

⁴storage.courtlistener.com

⁵Freddiemac.com

⁶WSJ.com