Good for ↑
Jobs strength raises near-term Fed hike odds, but the real signal is structural: AI capex and deficit spending are competing for long-end capital, steepening the yield curve. Banks with asset-sensitive balance sheets benefit from steepening more than from flat-curve hikes. The market is trading the short end; the trade is long US large-cap banks positioned for curve steepening.
Source: Bloomberg ↗
The obvious first-order read is: strong jobs → Fed hikes → financials benefit from higher short rates. That is fully priced. The 10Y yield is already at 4.78% (+1.4% on the week), and VIX at 14.5 signals the market is not pricing in stress — it's complacent about the duration risk building in the background. The non-obvious read is the one the strategist flagged: heavy AI infrastructure investment (hyperscalers are committing hundreds of billions in capex through 2027) and US government deficit spending are both issuing long-duration claims on capital simultaneously. This creates structural upward pressure on the 10Y and 30Y that is independent of Fed policy. A Fed hike compresses the short end while the long end drifts higher — that is a steepener, not a parallel shift. US large-cap banks — particularly those with large floating-rate loan books, significant deposit franchises, and limited long-duration bond holdings — are the structural beneficiaries of a steepening curve. Their net interest margin (NIM) expands when the spread between short-term funding costs and long-term lending rates widens. The market is pricing banks as 'rate-hike beneficiaries' (short-end story) when the real driver is curve steepening (long-end story). These are different trades: a flat-curve hike environment actually compresses NIM for banks that fund short and lend long. A steepening curve — which the AI capex + deficit dynamic structurally supports — is the genuine NIM expansion catalyst. Look for US large-cap money-center banks and regional banks with asset-sensitive balance sheets (loan books heavily weighted to variable-rate commercial and industrial loans).
The market believes a September Fed hike is the primary catalyst for bank outperformance — higher short rates lift deposit spreads and the sector re-rates on earnings per share upgrades. XLF is already +0.7% on the week, reflecting this consensus positioning.
Why that may be wrong
The market is over-extrapolating the Fed hike narrative into bank valuations without distinguishing between a parallel-shift rate environment (which can compress NIM if the curve stays flat) and a steepening environment (which structurally expands NIM). The AI capex + deficit supply dynamic that steepens the curve is a structural force that operates on a different timeline than Fed policy and is being systematically underweighted because it requires connecting the Treasury market, corporate capex cycles, and bank balance sheet mechanics — three separate analyst silos that rarely talk to each other.
If AI investment spending decelerates sharply (e.g., a major hyperscaler cuts capex guidance) or Congress passes meaningful deficit reduction legislation, the structural long-end supply pressure dissipates and the curve could flatten or invert further. Observable signal: 10Y yield falling back below 4.50% while the Fed holds or hikes would indicate the steepening thesis has failed and the flat-curve compression dynamic is reasserting.
The strongest counter is that AI capex is funded mainly through retained earnings and equity issuance by hyperscalers, not long-duration bond issuance. That would mean the duration supply impact on the 10Y is smaller than assumed. If so, the long-end yield rise is purely a Fed-expectations story and will reverse the moment the Fed signals a pause, ending the steepening case. The specific assumption most likely to be wrong is that corporate AI capex meaningfully adds to long-duration Treasury-equivalent supply pressure. If 10Y yields peak and start declining while the Fed holds rates flat (i.e., the curve flattens again), the case for wider bank lending margins fails and banks re-rate lower. Clear disproof: the 10Y yield declining 25+ basis points while the Fed holds rates steady.
The 2004-2006 Fed tightening cycle is the canonical cautionary tale: the Fed raised the fed funds rate from 1% to 5.25% in 17 consecutive hikes, yet the 10Y yield barely moved (Greenspan's 'conundrum'). Banks with flat-curve exposure saw their lending margins shrink despite rising short rates. The steepener trade only paid when the curve eventually normalized. Conversely, the 2021-2022 period showed that when fiscal deficit spending and private capex compete simultaneously for long-duration capital (pandemic stimulus + corporate bond issuance surge), the long end rises faster than models expect. The current setup, an AI capex cycle layered on top of a structural US deficit, rhymes more with 2021-2022 long-end repricing than with a clean Fed-hike cycle.
Strong jobs report raises probability of a September Fed hike, pushing short-end rates higher and lifting financials sector on the consensus 'higher rates = better NIM' narrative.
Simultaneously, AI infrastructure capex (hyperscalers issuing corporate debt) and US Treasury supply from deficit spending both compete for long-duration capital, pushing the 10Y yield structurally higher independent of Fed action. The 10Y at 4.78% is already reflecting this, but the market attributes it to Fed policy rather than structural supply.
A steepening yield curve (short end anchored by Fed, long end drifting up from supply) mechanically expands NIM for asset-sensitive US large-cap banks with floating-rate commercial loan books. These banks are being bought as 'rate-hike plays' but will actually outperform because of the steepening dynamic — a more durable and less consensus-recognized driver. The trade is long US large-cap money-center banks and asset-sensitive regional banks (S&P 500 and Dow 30 constituents) that have high variable-rate loan concentrations and limited long-duration HTM bond exposure.
Earlier connections in the same sector that share a theme with this one.
No comments yet.
Sign in to comment