← InsightsBorrower-side commercial finance analysis · US & Canada

Inflation Cooled. A Hike Is Still a 42 Percent Bet.

Capital Markets & TrendsBorrower Advisory

Cooling inflation data did not settle the September question. It moved the odds and left the direction ambiguous, which is the harder problem for anyone carrying floating debt.

The Bureau of Labor Statistics released July CPI on August 12. Headline prices rose 0.1 percent for the month and 3.4 percent over the year. Core, excluding food and energy, rose 0.2 percent and 2.5 percent annually. Both annual rates came in a tenth lower than June, and every reading matched the consensus forecast. Treasury yields fell across the board.

Then the futures market priced a 42 percent chance that the Federal Reserve raises rates in September.

An in-line print alongside 42 percent hike odds is the thing worth planning around. Not because 42 percent is high, but because most mid-market capital plans are built on the assumption that the next move, whenever it comes, is downward.

Forty-two percent is a price, not a prediction

The distinction between a price and a forecast matters more than it first appears. A futures-implied probability is not a forecast anyone is offering; it is what the market is charging for one side of a trade. As of the release, that price said a September hike was closer to a coin flip than to a tail risk.

A coin-flip probability has a direct consequence. If you are borrowing at a floating rate, you are on the other side of that price. The fixed-rate quote available to the same borrower already contains the market’s view of where rates go, which means choosing to float is choosing to take the risk the fixed quote was pricing. That can be the right call. It is rarely a deliberate one.

The probability itself moves daily, and by the time you read this it will have moved. The number is not the point. The point is that the distribution has two sides and most plans are drawn with one.

What the swing is worth

Annual interest cost of a rate move across three floating balances A shaded grid with two rows and three columns. A 25 basis point move costs $25,000 a year on a $10 million floating balance, $50,000 on $20 million and $125,000 on $50 million. A 50 basis point move costs $50,000, $100,000 and $250,000 respectively, with the largest figure shown in the darkest navy as the value argued. CFO & EXECUTIVE STRATEGY Inflation Cooled. A Hike Is Still 42%. Annual interest cost of a rate move, by floating balance. July CPI eased on both measures. September hike odds: 42%. $10M BALANCE $20M BALANCE $50M BALANCE 25 basis point move 50 basis point move $25,000 $50,000 $125,000 $50,000 $100,000 $250,000 The plan says cut. The market prices 42% the other way. Source: Bureau of Labor Statistics July 2026 CPI, futures pricing Aug 12 | Thalos Capital Research Thalos Capital ©

Take an illustrative case. A business carries $20 million of floating-rate debt and its 2027 plan assumes a 25 basis point cut. The risk case is a 25 basis point hike. The distance between those two outcomes is 50 basis points.

On $20 million, 25 basis points is $50,000 a year. Fifty basis points is $100,000 a year. So the gap between the plan and the risk case is $100,000 annually, and across a three-year facility it is $300,000.

The $300,000 is not the product of a dramatic rate move. It is the product of a plan pointing one direction while the market prices meaningful odds of the other. A company that had modelled both ends would have the same debt at the same price and simply know what it was carrying.

Scale it and the arithmetic holds: a 25 basis point move is $25,000 a year on $10 million and $125,000 on $50 million. Nothing exotic, just multiplication most plans never run because they contain a single rate path.

What the print did and did not resolve

What the July print settled
Every reading in line, and September still unresolved
The July CPI released on August 12 matched the consensus forecast on each line. Both annual measures eased by a tenth against June, and Treasury yields fell across the board. The September question did not close.
Headline CPI, annual rate for July
3.4%
Down 0.1 point from June
Core CPI, excluding food and energy
2.5%
Down 0.1 point from June
Odds of a September rate hike, on futures pricing
42%
A price, not a forecast
The first two numbers are backward-looking and settled. The third is a live price that moves daily, and it is the only one of the three that says anything about what a facility written next month will cost. A plan carrying a single rate assumption has taken a position on it without recording that it did.
Source: Bureau of Labor Statistics July 2026 CPI, futures pricing Aug 12 | Thalos Capital Research Thalos Capital ©

It is worth being precise about what happened on August 12, because in-line data is easy to over-read in both directions.

Inflation did ease. Both annual measures fell a tenth, and core at 2.5 percent sits far closer to the Federal Reserve’s target than headline at 3.4 percent does. Yields falling across the board says the bond market read the print as mildly friendly.

None of that resolved September. Every number arrived exactly where forecasters expected, and the hike probability still sat at 42 percent afterwards. When a perfectly in-line print leaves the next decision this open, the uncertainty is no longer in the data. It is in the reaction function, and better forecasting does not help a borrower there.

The plan most companies are carrying

The common structure is a single rate assumption, usually mildly optimistic, applied across a multi-year plan. It is not careless. It is what happens when the rate line in a model has one cell.

The correction is mechanical rather than clever. Model the facility at the current rate, at the plan case, and at the risk case, and read what breaks in the third column. Then check whether the covenant headroom that looks comfortable in the plan case survives a move the market currently prices near even odds. If it does not, the structure is the thing to change, not the forecast.

How Thalos Capital Approaches This

Thalos Capital prices fixed, floating and blended structures against the borrower’s own cash flows rather than against a house view on rates, because a structure that only works in one rate environment has outsourced the outcome. The test applied is whether the plan holds at both ends of the range that is currently priced, not whether it holds at the midpoint.

In practice that means running the same need through several structures and several sources, then comparing what each costs in the plan case and in the risk case. The comparison is what lets a borrower choose to float deliberately, with the cost of being wrong already quantified, rather than by default.

The work runs borrower-side across a range from $50 thousand to $100 million and above, and it is worth doing before a facility is needed rather than while it is being negotiated.

None of this requires a rate call. It requires accepting that the market is currently pricing two directions and that only one of them is in most plans. On $20 million that asymmetry is $100,000 a year of difference, already knowable, and cheaper to model this week than to discover next September.

Most financing situations have more options than the borrower initially sees. A conversation is enough to map them. Submit your financing request at https://thaloscapital.com/contact.

Ready to map your options?
Most financing situations have more options than the borrower initially sees.
A conversation is enough to map them.
Submit your financing request →
Explore our financing solutions
Equipment FinancingWorking CapitalAsset-Based LendingStrategic Debt