Fed funds target (post Sept. 16)
3.75–4.00%
12–0 unanimous vote; first hike since 2023
Dot plot median, end-2026 / end-2027
4.10% / 4.10%
SEP, Sept. 16, 2026 — implies at least one more hike
10-year Treasury yield
5.12%
Sept. 23, 2026, vs. 4.96% on Sept. 22
30-year Treasury yield
5.40%
Sept. 23, 2026; first crossed 5% on July 7
Hyperscaler bond issuance YTD
$220B
Through Aug. 10, 2026; vs. $108B for full-year 2025
The Fed didn't blink — it hiked, unanimously, and then signaled it again
On Sept. 16, 2026 the FOMC raised the federal funds target by 25 basis points to 3.75–4.00% — its first hike since 2023 — and the vote was12–0. The accompanying Summary of Economic Projections moved the median 2026 federal funds rate to 4.10%, with 4.10% still showing for end-2027 before drifting to 3.90% in 2028. That path implies at least one more hike this year and a higher-for-longer glide than markets had been pricing as recently as June, when economists still expected 70% odds of no change through year-end.
A week later, on Sept. 23, Governor Michael Barr removed any remaining ambiguity. In a Chicago Fed speech on housing he said the FOMC 'increased the policy rate last week' and that, in his base case, 'further policy adjustments are likely to be needed' to return it to 2% in a timely fashion. He framed the inflation problem as structural — explicitly naming 'tariffs' and 'increased investment demand to support the AI buildout' among the shocks keeping price pressures elevated — rather than transitory. CPI rent of primary residence is still 34% above its December 2019 level and the Atlanta Fed Home Affordability Monitor fell to 68 in July, the lowest reading in 21 years.
The five-hawk bench: who's actually voting, and how the chair is tilting
- Kevin Warsh (Chair) — Confirmed 54–45 on May 13, 2026 to succeed Powell; his Jackson Hole speech and Sept. 16 statement leaned explicitly into price stability, opening the door to a hike that markets had initially seen as 50/50.
- Beth Hammack (Cleveland) — Classified as a hawk in pre-meeting scorecards; voted for the September hike after three unnamed members had already favored one in July.
- Lorie Logan (Dallas) — Hawk; her regional-bank surveys fed the 'sticky services inflation' narrative that made a hike defensible even with payrolls softening.
- Christopher Waller (Governor) — Long-standing inflation hawk, voting member through 2026; the most-watched swing voter per CNBC's pre-meeting tally.
- Jeffrey Schmid (Kansas City) — Classified hawk; one of the three July dissenters that effectively pre-announced the September move.
- Michael Barr (Governor) — Voted yes on the hike and, per the Sept. 23 speech, is signaling more to come — the headline hawk the market heard most clearly this week.
Banks: the cleanest beneficiaries on the board
JPMorgan Chase printed Q2 2026 net interest income of $25.5 billion, up 10% year-over-year, on revenue of $82.5 billion. Bank of America followed with Q2 NII of $16.0 billion and quarterly earnings growth of 34%, while Wells Fargo reported NII of $12.3 billion, up 5% year-over-year, and investment-banking fees of $939 million — a 35% jump. With the dot plot signaling that policy rates stay above 4% through 2027, the curve has steepened: the 10-year added roughly 16 basis points on Sept. 23 alone, and asset-sensitive books at JPM, Morgan Stanley and Goldman Sachs keep compounding that yield pickup into net interest income.
BlackRock sits alongside the banks as a structural winner — Q2 2026 AUM hit a record $15.3 trillion on $868 billion of net inflows, with revenue of $7.1 billion up 31% year-over-year. Higher yields revive retail appetite for yield-bearing fixed-income products after a three-year drought, and BlackRock's iShares complex — including TLT, the long-duration Treasury ETF — captures the flows on the way up, even as the underlying bonds keep bleeding.
| Company | Q2 2026 NII | YoY growth | Source |
|---|---|---|---|
| JPMorgan Chase | $25.5B | +10% | Q2 2026 press release, filed Aug. 6, 2026 |
| Bank of America | $16.0B | +5% (NII + noninterest) | Q2 2026 press release, filed Jul. 31, 2026 |
| Wells Fargo | $12.3B | +5% | Q2 2026 press release, filed Jul. 14, 2026 |
Hyperscalers: the AI buildout is now the inflation story, and the bond market knows it
The same five-hawk Fed that is good for bank NIMs is bad — mechanically — for the AI hyperscalers funding their data-center build with long-dated debt. Barr's Sept. 23 speech specifically named 'increased investment demand to support the AI buildout' as a contributor to inflation that is 'above the Fed's 2% target and not clearly trending toward target in a timely way.' That framing is what kept the dot plot at 4.1% for 2027 and what drove the 30-year to 5.40% on Sept. 23 — its highest print since 2007.
Hyperscaler bond issuance tells you how the funding math has changed. Alphabet, Amazon, Meta, Microsoft and Oracle placed roughly $108 billion in 2025; through Aug. 10, 2026 they had already issued $220 billion, per BNP Paribas — a pace that puts2026 issuance near $280 billion once the year closes. Critically, 41% of that paper was placed beyond 15 years, against 16% for the broader U.S. corporate index — the AI build is locking in long duration at the worst possible moment in the rate cycle.
Oracle is the cleanest case study. Its 2040 unsecured paper now yields 7.48% on a 5.375% coupon — a $25 billion offering the company launched in February after announcing a $45–50 billion gross funding plan for calendar 2026. Net debt to EBITDA at Oracle sits at 3.17x on trailing data, capex-to-revenue is 105% (TTM), and free cash flow is negative — the company is paying the marginal cost of debt to fund growth that is itself part of the inflation problem the Fed is trying to contain. Meta Platforms had its own version of this in Q2: free cash flow collapsed91% year-over-year to $784 million as2026 capex guidance widened to $130–145 billion. Alphabet printed $44.9 billion of capex in a single quarter — up 100% year-over-year — and negative $5.9 billion of free cash flow, while pushing full-year capex guidance to $195–205 billion.
The 60/40 re-test: bonds no longer hedge, AI-funded refi math no longer closes
The 60/40 portfolio is up 8.62% YTD as of Sept. 22, 2026 — fine in absolute terms, but the mechanism has flipped. When the 30-year crossed 5% on July 7 and reached 5.31% on Aug. 17, the long end did its job as a counterweight to equity drawdowns: stock-bond correlation shifted back into negative territory in 2025 after the 2022 co-moving disaster. But at 5.40% on the 30-year, the marginal buyer is no longer a60/40 allocator reaching for duration — it is the U.S. Treasury itself funding the deficit, plus the hyperscalers above. Barclays framed the 5% threshold as a 'historically important' inflection; beyond it, rates become a 'persistent headwind for equities' that limits valuation growth.
The AI refi math is the part investors are missing. Hyperscaler combined free cash flow is on track to fall roughly 70% in 2026 — to about $100 billion from a higher base — while capex heads toward $635 billion. That gap, per Goldman Sachs, will be filled with debt, and is on track to exceed $1 trillion of cumulative AI-related bond issuance by 2027. Every basis point of long-end yield above 5% makes the math worse: at 5.40% on a 30-year, Microsoft, Meta, Alphabet, Amazon and Oracle are paying tens of billions in incremental annual interest versus the same debt priced at 4%. That delta does not show up in this quarter's earnings — it shows up in 2028 free cash flow per share, which is what determines the multiple.
The bond-proxy parts of the market — utilities and REITs — are losing their relative-yield moat for the same reason. With the 10-year at 5.12%, the marginal income investor can buy a Treasury and match the dividend yield of a regulated utility without taking duration or equity risk; that is why the utilities complex has under-performed sharply since the 30-year first crossed 5% in July. Realty Income and similar net-lease REITs face the same headwind: their cost of capital is now benchmarked to a 5.10–5.60% investment-grade REIT curve, which compresses acquisition yields.
The yield curve in one week (basis points, Sept. 22 → Sept. 23, 2026)
Long-end yields led the move higher after Barr's Sept. 23 speech.
Unit: bps change
2-year
Approx., per Treasury daily yield curve
8
10-year
5.029% → 5.116% on Sept. 23
16
20-year
Last seen at 5.434% on Sept. 15
14
30-year
5.29% → 5.40% on Sept. 23
10
What moves next: near-term catalysts and the2028 horizon
- Days–quarters (catalysts): The Oct. 28–29 FOMC meeting is the next binary event. A second hike would take policy to 4.00–4.25% and likely push the 30-year toward 5.60–5.75%; a hold would re-anchor the curve but the dot plot already says4.1% is the year-end median, so a 'skip' is the dovish surprise.
- Days–quarters (stocks): Bank earnings into Q3 2026 will show whether the NII tailwind compounds — at JPM a25 bps higher rate path adds roughly $1.5–2.0 billion of annualized NII on a static balance sheet; BAC and WFC are more sensitive because their deposit bases are more rate-sensitive.
- Days–quarters (bonds): Hyperscaler issuance into year-end is the swing factor for the long end. Another $50–80 billion of long-duration supply between now and December, on top of Treasury refunding, would test whether 5.40% on the 30-year is the ceiling or the floor.
- 1–3 years (structural): AI capex is now a documented input to Fed inflation models. If hyperscaler issuance tracks Goldman's $1 trillion-plus forecast through 2027, the marginal funding cost for AI capacity will be 5–7%, and equity multiples have to price that in.
- 1–3 years (banks): JPMorgan Chase and Bank of America compound the advantage as long as the curve stays upward-sloping; the inflection risk is a fast inversion, not a fast cut. Watch the 2-year/10-year spread: a re-inversion would erase the asset-sensitive tailwind within two quarters.
- 1–3 years (long-duration pain): TLT is up 3.26% YTD as of Sept. 22, 2026 — the coupon is doing the work, not the price. With the dot plot signaling 4.10% through 2027, the price path of long-duration paper is capped.
Investable takeaways from the five-hawk Fed
- Q2 2026 NII of $25.5B (+10% YoY) is the cleanest read on the rate path the five hawks just locked in
- Dot plot signaling 4.10% through 2027 extends the NIM tailwind into 2028 earnings
- Key risk: a curve inversion — not a cut — is the bigger threat to the asset-sensitive book
- Q2 2026 NII of $12.3B (+5% YoY); IB fees +35% as the rate path resets M&A activity
- CFO has flagged a 'step up' in interest income into H2 2026, ahead of the second dot-plot hike
- Deposit beta has lagged peers, so further rate moves will translate more cleanly into NII
- 2040 paper yields 7.48% vs. 5.375% coupon — 230 bps wider than the 30Y benchmark it funds against
- Capex-to-revenue at 105% (TTM); free cash flow is negative and the company is targeting $45–50B of gross issuance in 2026
- If the 30Y stays above 5%, every refinancing cycle adds ~$1B of annual interest expense versus a 4% world
- Anchor hyperscaler in the $220B-of-2026-issuance club; 41% of paper placed beyond 15 years
- FCF per share faces a 2027–2028 headwind if the 30Y stays above 5%, but AI revenue offsets in the near term
- The 'AI buildout is inflation' framing from Barr is the policy risk that matters most for the long end of the model
- Q2 2026 FCF collapsed 91% YoY to $784M while capex guidance widened to $130–145B for 2026
- The ad business is funding the AI capex for now, but a sustained 5%+ long end forces the company to dilute or lever further
- Watch 2027 capex guidance — a cut would be the first sign the AI math is starting to bend
- The 5.10–5.60% investment-grade REIT curve is now the cost of capital for net-lease acquisitions
- Treasury yields at 5.12% (10Y) match or exceed O's dividend yield, eliminating the relative-income bid
- Acquisition cap rate discipline is what protects the dividend; multiple expansion is no longer on the table
- Q2 2026 AUM hit $15.3T on $868B of net inflows; yield-starved retail is back in fixed income
- Higher yields revive fee-yield on the bond side while iShares captures the flows on the way up
- The mix shift toward fixed income is the durable tailwind even if equity AUM stalls
