CIBC Q1 2026 · First-Principles Analysis
Where did the
+28 bps
actually come from?
CIBC's Canadian P&C NIM (Slide 10, combined segment) went from 272 bps (Q1/25) to 300 bps (Q1/26). In a rate-cutting cycle. That's not supposed to happen.
Management gave a one-liner: "prolonged impact of higher rates, strategic shift in business mix, and pricing discipline." On the call the CFO quantified it as "about a third, a third, a third" across hedging/positioning, business mix, and product margin. That's a frame, not a mechanism.
The complete mechanism is not publicly disclosed. What follows is first-principles reasoning from published numbers, not a narrative sourced from the bank's press release.
Segment note: CIBC also reports a narrower Personal & Business Banking series on Slide 14 (277→311, +34 YoY). It is a subset of Slide 10's combined P&C. This analysis uses Slide 10 throughout because that's where the waterfall lives.
Step 1 — The Math
NIM is an identity.
There are only 4 levers.
Management narrative is always a label on top of one of these. Work backwards from the levers to the published numbers — not the other way around.
Published Directly quoted from CIBC source
Inferred Logical deduction from published data
Speculative Plausible but not traceable to evidence
Unknown Can't determine from public data
// NIM identity
NIM =
( Interest Income − Interest Expense )
/ Avg Earning Assets
A: raise yield on assets
B: lower cost of liabilities
C: mix shift in denominator
D: reclassification
A
Asset Yield Up
Cdn P&C loan portfolio: only +2 bps QoQ, +~18 bps YoY. Real — but small relative to the total. Published
B
Funding Cost Down
Deposit portfolio: +5 bps QoQ (Cdn P&C), +26 bps QoQ (U.S.). The dominant driver in the waterfall. Mechanism not specified by CIBC. PublishedMechanism Unknown
C
Mix Shift
Costco cards +51% YoY. Cards yield ~20% vs mortgages ~4.5%. Even a small portfolio shift produces outsized NIM movement. Inferred
D
Reclassification
Hedge book gains ("the hedges work") — allocated to segments via treasury transfer pricing. Episodic, rarely broken out. Speculative
Step 2 — The Numbers
What CIBC actually published.
Nothing more.
These waterfalls are directly from Slide 10 of the Q1/26 investor deck. The bucket labels are CIBC's. What's inside those buckets is not disclosed.
Canadian P&C NIM — QoQ bridge (bps) Published · Slide 10
Q4/25 → Q1/26 | +10 bps QoQ | +28 bps YoY
⚠ CIBC does not define what "Deposit Portfolio" and "Loan Portfolio" contain. These are CIBC's own aggregation buckets. The mechanism inside each bucket is unverifiable from public disclosure alone.
⚠ Scale note: delta bars are drawn at ~15× the scale of the base/result bars so small QoQ buckets stay readable. This is a deliberate visual compression, not a scaling error. 290 + 5 + 2 + 3 = 300 — math reconciles.
U.S. Commercial NIM — QoQ bridge (bps) Published · Slide 10
Q4/25 → Q1/26 | +17 bps QoQ | +23 bps YoY
The U.S. book is nearly entirely a deposit story: +26 bps from deposits alone. Fed rate cuts reduced the cost of U.S. client deposits. Loan yields fell concurrently (−8 bps). Net: CIBC's U.S. deposits repriced down faster than its loans. Inferred
Scale note: delta bars drawn at ~10× base-bar scale. 384 + 26 − 8 − 1 = 401 ✓
Step 3 — Candidate Mechanisms
What's actually inside
the "Deposit Portfolio" bucket.
Management will not say. Here are the mechanics that could produce the observed numbers, ranked by plausibility. Every claim is tagged.
Tier 1 — High Probability · Under-Disclosed
M1
CIBC's own term for this is SIRR (the UK term is "structural hedge" / "tractor"). AR25 p. 68 verbatim: "A variety of cash instruments and derivatives, primarily interest rate swaps, are used to manage these risks" — referring to non-maturity deposits and equity. The hedge locked in elevated rates during 2023–24 and continues to pay as policy rates fall.
Hedge notionals are disclosed (AR25 Note 13): cash-flow hedge IR swaps $53.1B (up from $41.2B FY24); fair-value hedge IR swaps $206.8B. AOCI cash-flow-hedge net change FY25 +$491M after $1,535M in FY24.
SIRR sensitivity moved materially: +100bps NII impact was $150M (Jan-25) → $163M (Oct-25) → $230M (Jan-26). EVE at −100bps = +$1,525M FY25. Asset sensitivity + positive rate-cut EVE is exactly the shape of a hedge book that gains from the rate path they just had.
What's not disclosed: how SIRR NII is allocated between the P&C segments and Corporate & Other via internal transfer pricing. Corporate & Other NII = $115MM in Q1/26 — elevated, with no attribution given.
Falsifier: An explicit segment-level SIRR NII attribution (à la Lloyds' £5.5bn "structural hedge income" line). CIBC does not publish one.
M2
A chequing account costs the bank ~0 bps. The bank invests that float in short-term government securities, interbank lending, or commercial loans at 4–5%.
CIBC's Q1/26 SFI (p.9) confirms the mix shift: Cdn P&BB average NIB demand deposits $12.4B Q1/26 (vs $12.0B Q1/25, +3.1% YoY) while interest-bearing deposits grew only +1.4% YoY. NIB share moved from 19.7% → 20.0% of Cdn P&BB deposits. U.S. C&WM NIB $8.9B, +6.4% YoY.
CIBC repeatedly emphasizes "mass affluent" acquisition and "transaction accounts" growth. This is what that actually means financially. The mass affluent client brings their operating account — that's the 0% funding the bank needs.
Falsifier: Product-level split (chequing vs savings vs GIC vs notice) would let you back out the exact funding-cost walk. SFI only shows NIB vs interest-bearing, not further.
M3
Published facts: Costco member count up +51% YoY (Slide 4). But the Cdn card book balance grew only +4.9% YoY ($20.9B → $21.9B Q1/25 → Q1/26, SFI p.9). Card balance growth ≠ member growth because new members revolve less. Canadian card APRs are ~20%, but realized portfolio yield is closer to ~10–13% because only ~40–50% of balances revolve. Mortgages yield approximately 4.5–5%.
Math on that $1B mortgage → card shift: NIM effect per-dollar is ~+600–800 bps (not the ~1,500 bps that headline APRs would imply). The +$1B is within range of the actual $1B Cdn card book YoY growth — so this mechanism is real, but the dollar size is modest.
This is reported as "volume" and "Costco partnership growth" — never as a NIM driver. But mathematically, card expansion is one of the highest-return NIM moves a retail bank can make. CIBC has been deliberately growing it.
Falsifier: CIBC's product-level loan yield breakdowns (mortgage vs card vs personal) in the supplementary would confirm. Not publicly granular enough.
Tier 2 — Real, But Likely Smaller
M4
When BoC raised rates 2022–2023, banks were slow to pass increases to savers (low deposit beta on the way up). When BoC cut 2024–2025, banks were fast to cut deposit rates (high deposit beta on the way down).
This is universal bank behavior. CIBC benefits from it like every other bank. The net effect: the bank captured more spread on the way up (customers got less than BoC gave) and retains more on the way down (cuts savers first).
No bank will state this plainly. It's called "pricing discipline" in every earnings deck in Canada.
M5
Canadian variable-rate mortgages at trigger rates stop paying down principal. The bank collects elevated interest for longer on a principal that barely shrinks — mechanically inflating interest income without any new origination.
This effect is wound down as BoC cuts, but residual balances are still in extended-amortization. It inflates the "Loan Portfolio" NIM contribution in the waterfall.
Tier 3 — Possible, Unverifiable
M6
NHA-MBS and CMB securitization of insured mortgages is real and CIBC uses it, but the primary benefit is funding cost and zero-risk-weight capital relief, not a mechanical NIM lift. When you sell an insured mortgage off-balance-sheet, you lose the asset yield along with the funding cost. Net NIM effect is ambiguous and depends on spread economics and reinvestment.
Consistent signal: Cdn P&BB loans grew only +2% YoY while NIM grew +34 bps (P&BB) / +28 bps (combined P&C). Divergence is real, but product mix (M3) is the better explanation than securitization.
Falsifier: CIBC's securitization footnotes in the annual report + the NII attribution inside "Treasury" vs P&C segments. Would require detailed balance-sheet archaeology.
Step 5 — The Synthesis
Best inference.
Clearly labeled.
Best-Inference Synthesis · Inferred — Not Confirmed
CIBC ran a quiet balance-sheet transformation during the rate spike. The NIM expansion is its delayed P&L recognition.
CIBC historically had the lowest NIM among the Big 6 — overexposed to commodity mortgages, underexposed to high-yield products and cheap transaction-account funding. In 2023–2024, while rates were high, they:
1. Built out a structural hedge book (tractor program) at elevated rates — locking in income that now flows through NII as policy rates fall.
2. Aggressively grew the Costco card portfolio — each card carries ~20% yield vs ~4.5% for a mortgage. The yield differential is enormous.
3. Pivoted client acquisition toward mass affluent, whose primary banking relationships come with chequing accounts — zero-cost deposits that fund the card book and commercial loans.
The result: funding costs went down structurally (more NIB deposits) while asset yields went up structurally (more cards, legacy hedges). Both moves were made during the rate spike. The P&L from them is arriving now, during rate cuts.
This is why analysts were wrong to model NIM compression from BoC cuts. The cuts were irrelevant to positions locked in 18–24 months earlier.
Caveats: This synthesis is built from publicly available numbers + knowledge of how bank balance sheets work. CIBC has not disclosed the size of its structural hedge book, the NIB deposit mix growth, or card yield data. These mechanisms are consistent with the published numbers — they are not confirmed by the bank. The actual reason may weight these factors differently, or involve mechanisms not listed here.
What we actually know
- +28 bps YoY combined Cdn P&C NIM (Slide 10) / +34 bps P&BB only (Slide 14) — different segments, both verified
- Slide 10 QoQ waterfall: 290 + 5 + 2 + 3 = 300 — reconciles
- U.S. deposit contribution: +26 bps — verified
- SIRR hedge book: $53.1B CFH + $206.8B FVH IR-swap notionals (AR25 Note 13)
- +100bps NII sensitivity rose to $230M (Q1/26) from $150M (Q1/25) — AR25 p.68 + Q1/26 Report p.33
- NIB deposit share 19.7% → 20.0% YoY in Cdn P&BB (SFI Q1/26 p.9)
- Cdn card book +4.9% YoY ($20.9B → $21.9B) — SFI p.9. Member count +51% is NOT the balance.
- Management quote on "prolonged rates, mix, pricing" — verified, Slide 14
- "The hedges work" — CFO Robert Sedran, earnings call
- "About a third / a third / a third" — CFO on call: hedging/positioning, business mix, product margin
- Roll-off: H1-2027 — CFO confirmed "those two lines cross" on Slide 33; structural tailwind ends
What we cannot determine
- What CIBC's "Deposit Portfolio" waterfall bucket actually contains (FTP allocation method)
- Segment-level SIRR NII attribution (like Lloyds' £5.5bn line)
- Product-level deposit split (chequing vs savings vs GIC vs notice)
- Product-level yield breakdown (mortgage vs card vs HELOC)
- Costco card $ balance separately (only member count disclosed)
- How much of "Other +3 bps" is treasury vs day count
- Whether securitization affected earning-asset yield mechanically