Sponsored by

THE SETUP

The Fed raised rates in September for the first time in this cycle, a quarter point to 3.75–4.00%, unanimously. It held in October, 10–2, with Governor Cook and Cleveland's Beth Hammack dissenting in favor of another increase.

Then two officials spoke, and hours later the data landed above both consensus and the threshold one of them had just named.

Governor Christopher Waller, in London on the morning of 10 November, said he was “leaning toward supporting an increase of 25 basis points” in December — but that if core CPI, due later that day, came in near the prior two readings, “a 50 basis point hike should be considered.” He added that going only a quarter “would put our credibility at risk, a risk I am not eager to take.” Core came in at 0.3%, above the 0.2% consensus and inside his trigger. He has not spoken since.

Dallas Fed President Lorie Logan, who had already told an audience on 30 October that the Committee “needs to implement at least two more 25-basis-point rate increases,” repeated it that same day, adding that “accommodative monetary policy is adding to inflationary pressures.” Two more, and then a pause.

So the hawks want one hike or two. The September dot plot says one. The market is priced above both: fed funds futures closed on Friday carrying about 62 basis points of expected tightening by the 17 March meeting — roughly two and a half quarter-point moves, with an implied rate near 4.50% against 3.88% today.

That is the problem on your desk this month. Not what happens on 9 December, but which of those numbers goes into a 2027 plan that a board will approve and then measure you against for fourteen months.

THE SIGNAL

The market is pricing more than anyone who votes is asking for

The Committee's own number is one. The September projection puts the median at 4.00–4.25% at the end of 2026 and in the same range at the end of 2027: one more quarter point, then a full year of nothing. New York Fed President and FOMC Vice Chair John Williams held that line on 13 October — one more hike may be appropriate, no reason for further increases in 2027.

Its hawks are at one or two. Waller's stated lean is a quarter in December, with a half conditional on data that has since arrived. Logan wants two more and then a pause, which tops out at 4.25–4.50%. Neither is the dot plot and neither is far from it.

The market is priced above both — but less confidently than the coverage suggests. December is as close to settled as the strip gets, at better than nine in ten. Beyond that it is probability rather than schedule: roughly two in three on a further increase at the 27 January meeting, and about the same again on 17 March. On Friday's close that came to about 62 basis points of expected tightening by mid-March, against the dot plot's 25 for the whole of 2027. Anyone telling you the market has priced three hikes is reading a distribution as a calendar.

That gap is the story. The strip is not pricing what the Committee is saying; it is pricing the risk that the Committee is wrong. Planning to the dots means betting against the market. Planning to the strip means betting against every official who has spoken.

And history says all of it may be light. T. Rowe Price counts six Fed hiking cycles since 1994: five ran at least 175 basis points, and the exception — 1997 — was one hike, quickly reversed. The record holds both outcomes, which is the point.

So: +25bp, +62bp, or more. The table runs one borrower down all three — the same company in every row, nothing changing but the rate it pays. It is KBRA's median direct-lending borrower at 30 September: 6.2x levered, interest covered 1.6 times, on $100m of EBITDA and $620m of debt, all floating, none hedged.

2027 path

Interest, run-rate

Increase

% of EBITDA

Interest cover

Today

$62.5m

—

—

1.60x

Dot plot, +25bp

$64.1m

$1.6m

1.6%

1.56x

Market strip, +62bp

$66.3m

$3.8m

3.8%

1.51x

Historical, +150bp

$71.8m

$9.3m

9.3%

1.39x

Scale it in one step: every 100 basis points costs 1% of EBITDA for each turn of leverage — 6.2% a year at 6.2x. Those are run-rates; a 2027 budget line is lower, because January and March are not in effect from 1 January.

Which covenant you have decides whether this is a cash problem or a compliance one

Leverage-based maintenance tests anchor the large majority of private credit structures. Sidley's March review puts the leverage ratio behind maintenance testing, incurrence baskets, pricing and cash sweeps, typically with a 25–35% cushion to the sponsor model EBITDA at closing — the model at closing, which for most borrowers sits above where EBITDA is now, so real headroom is tighter than the headline.

A rate rise reaches a leverage test only through the cash balance. Most private credit leverage ratios net cash, so higher interest drains the balance and raises net debt — about a tenth of a turn on the historical path. Material for your sweep, rarely for your test.

An interest coverage test is reached directly. The table holds EBITDA where it is today and moves only the rate, so every figure is the rate path acting alone. If your test is a fixed-charge ratio, carrying amortization and sometimes capex in the denominator, read these as a ceiling on your headroom rather than the number.

If your test is set at

Cushion today

Dot plot

Market strip

Historical

1.10x

31.3%

29.5%

27.0%

21.0%

1.25x

21.9%

19.9%

17.1%

10.3%

1.40x

12.5%

10.3%

7.1%

breached

Cushion: how far EBITDA can fall before the test fails, as a share of EBITDA.

On the historical path a 1.25x test loses half its headroom and a 1.40x test runs out altogether — it binds at about 145 basis points, with EBITDA exactly where it is today.

The timing is not what it looks like. These tests run on trailing twelve months, so December's move barely registers at the 31 December test and arrives properly at the Q1 2027 one. December alone is fully in the denominator by Q4 2027; a path running through March is not until Q1 2028. That lag is why this is a planning problem rather than a December problem — and why a hedge has to be on before the move, not after.

The part that has nothing to do with your covenant

Section 163(j) caps the interest deduction at business interest income plus 30% of adjusted taxable income, and since 2025 that income is computed on an EBITDA basis. The cap starts to bind at an interest coverage ratio of about 3.3 times.

The median direct-lending borrower covers interest 1.6 times — half the level where it bites, with roughly $32.5m of this year's $62.5m interest bill already disallowed.

The disallowance carries forward indefinitely, so the shield is deferred rather than destroyed. For a sponsor-backed borrower that is a thinner distinction than it sounds: the carryforward is an attribute exposed to a section 382 limitation on an ownership change, and an ownership change is the exit everyone is working toward. Model it at or near zero, not at 21% — net the incremental interest down by a fifth and your true cost is about 27% above the number you wrote.

A CRM so smart, it updates itself.

HubSpot's CRM is so smart, it now updates itself. Calls get logged and summarized the moment they end. New leads get researched and a first outreach drafted, automatically. Deals move forward on their own, with next steps flagged before you have to ask. 

Even the one repetitive task you've been meaning to fix can now run itself, no code required. Customers using it are already seeing the difference. 

This isn't the CRM you have to work hard to figure out. It's the one already working before you log in.

THE RATE PATH

Current target range: 3.75–4.00%, after September's quarter-point increase from the 3.50–3.75% floor set in December 2025.

Into 2027: roughly two-in-three odds on a further increase at the 27 January meeting and about the same on 17 March — some 62 basis points cumulatively by mid-March on Friday's strip, which implies an effective rate near 4.50% against 3.88% today. The September dot plot carries one hike in total and nothing across 2027. Waller has put a half point on the table for December; Logan wants at least two more increases and then a pause.

Into 2027: roughly two-in-three odds on a further increase at the 27 January meeting and about the same on 17 March — some 62 basis points cumulatively by mid-March. The September dot plot carries one hike in total and nothing across 2027. Waller has put a half point on the table for December; Logan wants at least two more increases and then a pause.

SOFR: 3.90% (12 November).

Treasuries: Ten-year 5.15%, two-year 4.88% (13 November), a spread of +27bp. The two-year has added roughly 30 basis points since the end of October, most of it on the 10 November CPI — the policy end is doing the repricing.

Before the decision: October PCE lands 25 November and is the last major inflation read the Committee sees. November CPI is published 10 December, the day after the meeting — and 9 December also brings a fresh set of projections, which re-draws the baseline this issue is about. That is a reason to name your path now and a date to revisit it, not a reason to wait.

QUICK HITS

Anthropic prices on Wednesday, and the window is the story. The range is $70–78 a share for an offering that could reach $90.4bn including a $26bn private placement — on course to be the largest listing on record — valuing the company near a trillion dollars. Trading opens Thursday on Nasdaq under ANT, and Bloomberg put the book at roughly six times covered. The number that matters to the rest of us is not the valuation but where it prices in that range. As we wrote in October, 2026 beat its IPO forecast on dollars and missed badly on companies; a deal this size pricing at the top would be the clearest signal yet that the window is open to issuers who are not generational. It faces headwinds from both directions — the rate path above, and a sector absorbing AI regulation risk and Nvidia's earnings the day before pricing. Watch the print, not the debut.

Wholesale inflation hit 6.5% over twelve months, the fastest since late 2022, with truck freight up 3.9% in the month alone. Transportation and warehousing accounted for about three-quarters of the rise in final demand services, which rose only 0.2% overall — so the services move is concentrated rather than broad, but goods rose 0.5% and the annual rate is the number that matters for the path. For the 45% of you who told us in Issue #21 that landed cost was your least confident line, freight is the component moving.

Nvidia's CFO is now also its interim chief executive. Jensen Huang's resignation was announced 12 November, abruptly and without a runway; Colette Kress holds both roles while the board runs a search. Two questions it raises for everyone else. Whether your deputy finance bench is named anywhere — a plan that promotes the CFO only works if someone is already running the close. And who signs the certifications when one person holds both seats.

THE PLAYBOOK

Three things before the plan is signed off

1. Put a number in the plan and name which path it came from. Not a range that quietly averages the three — a number, with its source beside it. A plan built on the dot plot and one built on the futures strip are different plans, and the board should know which it approved. Put 9 December in the calendar: the new projections re-draw the baseline you chose against.

2. Run the top and the bottom, and run the top with a soft quarter. The spread between the dot-plot path and the historical path is 7.75% of EBITDA a year at 6.2 times levered. Run the no-more-hikes case too — 1997 is in the historical record and the Chair has said tightening is starting to work. A rate rise alone is survivable; it is the quarter where the path and weak EBITDA land in the same filing that moves a covenant.

3. Check the tax treatment before netting anything down. Below roughly 3.3 times cover you are at the section 163(j) cap and the incremental interest has no current shield, so your true cost is about 27% above the netted figure. It is the most common error in a leveraged plan, and it runs in the direction of understating the problem.

SPONSOR THIS NEWSLETTER

Your name, in front of the CFOs and senior finance leaders who read The CFO Desk every Tuesday. Write to [email protected] for rates and availability.

THE CLOSE

The December meeting will be covered as though it settles the question. It sets one rate for seven weeks.

What it does not settle is the path, and the path is what your plan commits to. The Committee's projection says one more hike and a quiet 2027. Its hawks want one or two. The market is priced above both, and less certainly than the headlines imply. One of those numbers is going into a document a board will hold you to.

Every plan written since 2023 assumed the next move was down. That assumption broke in September, and the plans being signed off over the next six weeks are the ones that have to say so in writing.

THE SMALL PRINT

For the skeptical reader, which is most of you.

The 10.1% all-in rate is a construct from two separate KBRA medians, not an observed rate, and is probably a shade high against where direct lending clears; the conclusions hold at 9%. Market-implied paths are probabilities, not schedules — the 62 basis points is cumulative expected tightening by mid-March, not three certain hikes. The 175bp historical figure is measured over three years, not the fourteen months a 2027 plan covers, and every prior cycle began from a floor or from zero where this one began from an already-restrictive 3.50–3.75% — which is why the table models 150bp rather than 175bp. The coverage table assumes an interest coverage test; a fixed-charge ratio carries amortization and sometimes capex in the denominator, so read those cushions as a ceiling. On tax: electing real property and farming businesses sit outside section 163(j) altogether, as do certain regulated utilities, and the small-business exemption runs to $32m of gross receipts for 2026. Most states conform, which makes the netting error larger than 21%; a few decouple, in which case the state shield survives. And a borrower already in a taxable loss gets to the same answer by a different route.

The Editors hold no position in any company mentioned.

— The CFO Desk
[email protected] · thecfodesk.com
CAPITAL · FINANCE · LEADERSHIP