The Compass - July 2026

THE COMPASS July 2026

Every quarter tests a few assumptions nobody had gotten around to questioning. This one is testing several at once. Banks assumed deposit costs, not asset composition, would decide who wins on margin this year, and the data says otherwise. C&I growth just posted its best quarter in years, and the honest answer for why is still being determined. Two rating agencies quietly rewrote how they look at CLO collateral within days of each other. A bankruptcy court reminded a room full of sophisticated lenders that loan documents mean what they say. And a wave of new bank charters is about to test an assumption most banks never had to defend before: that deposit relationships are sticky. The Caird Team

From the Caird Team Deposits Are Repricing Faster Than Ever. That's Not the Good News It Sounds Like. Deposit betas, how much of a Fed rate move eventually shows up in deposit costs, have gotten both larger and faster with each cycle. Research from the Federal Reserve Bank of New York's Liberty Street Economics blog shows the 2015 to 2019 hiking cycle took three years to push the industry's cumulative deposit beta to roughly 40%. The 2022 to 2023 cycle reached that same 40% mark within about a year and finished near 49%, the highest cumulative beta of the tightening cycles researchers have tracked since. The competitive pressure behind that keeps escalating: money market funds first pulled deposits away in the 1970s and 1980s, online banks did it again in the 2000s and 2010s, and a new wave of de novo bank charters and stablecoin issuers is doing it now (this issue's Caird Perspective covers who). That acceleration is the real problem hiding inside good funding-side management. If deposits reprice to market within a quarter or two, the float that used to let banks tolerate a mismatched balance sheet is not there to buy time anymore. The pressure has moved to the asset side. Per the FDIC's Q1 2026 Quarterly Banking Profile and Treasury.gov's daily yield curve data, the 10-year Treasury rose from 4.30% at the end of Q1 to a range of 4.57% to 4.60% by mid-July. Unrealized securities losses, which had ground down from a $690 billion peak in 2022 to a $306 billion low late last year, ticked back up to $325 billion in Q1, the same mechanism now unfolding faster than in past cycles. There is a precedent for what happens when deposit-side competition intensifies faster than a bank's asset side adjusts to it. When Regulation Q's phase-out forced thrifts to compete harder for deposits in the early 1980s, per the Federal Reserve's own historical research, the institutions that failed were not undone by the deposit costs themselves. They were undone by what they did on the asset side to cover them, reaching for yield in commercial real estate and speculative credit without the capital behind it. The mechanism today is competition, not deregulation, but the lesson holds: the deposit side rarely breaks a bank on its own. What a bank does with its assets while the deposit side tightens decides the outcome. Caird View: We continue to favor floating-rate, short weighted average life, CLO AAA/AA tranches over incremental fixed-rate securities. The deposit-side cushion that let banks tolerate duration mismatch keeps shrinking with every cycle, and asset composition is increasingly the lever banks have left to protect margin. Investing in CLO tranches involves credit, structural and liquidity risks and is not suitable for all investors. Caird may hold positions in CLO securities or related instruments and may receive fees that create conflicts of interest. Sources: Federal Reserve Bank of New York, Liberty Street Economics; FDIC Quarterly Banking Profile, First Quarter 2026; Federal Reserve History (Regulation Q, the Savings and Loan Crisis); Treasury.gov daily par yield curve.

By the Numbers 2.36% Big Four Bank Average NIM, Q2 Individual banks ranged from 2.1% to the high 2% area. Bloomberg, Q2 2026 earnings releases. $261M Serta Simmons Excluded-Lender Award Bankruptcy court finds pro rata breach in landmark uptier case, July 7. Proskauer. 255bps SpaceX 2056 Bond Spread to Treasuries Widened from 175bps at issuance to 255bps as of 7/23. 54.4 U. Mich Consumer Sentiment, July (Prelim.) Highest since February, still ~12% below a year ago.

Market Signals WATCH Private Credit Redemption pressure keeps outrunning defaults Redemptions keep outpacing available liquidity across the non-traded BDC universe, with roughly $14 billion currently trapped behind redemption caps. PIK usage held flat in Q1, so near-term stress looks contained. Direct lending defaults near 5.6% already run more than double the historical 2.0% to 2.5% range, and Morgan Stanley puts 8.0% within reach. Banks with revolver or asset-backed exposure to BDCs are reviewing counterparty concentration name by name. The Fed's 2026 stress tests will formalize that review across the largest banks this fall. INFLECTING Syndicated Loans / TLB Defaults look modest. The distress ratio tells a different story June's default rate fell to 0.97% from 1.35% in May. The distress ratio moved the other way, rising to 6.87%, close to March's three-year high of 7.23%. The gap between the two measures points to dispersion, a small share of credits carrying most of the risk while the broad market looks calm. Heading into the 2028 maturity wall, which loans a portfolio holds will matter more than how much exposure it carries. CONSTRUCTIVE CLOs / Securities Two rating agencies are recalibrating CLOs within days of each other Fitch republished its CLO and CDO criteria effective June 1, sharpening recovery assumptions by seniority, security, and jurisdiction, though the underlying recovery table is largely unchanged since 2019 even as actual loan recoveries hit a 10-year low in 2025. Moody's followed four days later with an update letting Aa-and-below tranches reflect a CLO's actual portfolio quality rather than the worst case its covenants technically permit. Both agencies hold AAA to the old, stricter standard, so the relief concentrates at AA and below, and some market commentary reads the two moves landing four days apart as competitive relaxation rather than coincidence. Our take: a relative non-event for AAA, but likely to produce larger AAA or junior AAA tranches over time, with junior AAA taking on characteristics closer to the old AA tranche, which could lower equity's cost of capital and support incremental demand for TLBs. OPPORTUNISTIC Bank M&A Faster approvals, softer pace, stronger back half expected 83 bank M&A deals were announced in the first half of 2026, worth $15.1 billion, with median closing time down to 132 days from 187 a year ago after the OCC and FDIC rescinded prior merger policy statements. Announcement pace actually softened in H1. Faster approvals and pent-up demand point to a stronger back half and into 2027, consistent with a landscape that keeps separating scaling regionals from smaller banks. Sources: Syndicated Loans / TLB default rate and distress ratio, PitchBook LCD. CLOs / Securities, Fitch Ratings and Moody's Investors Service methodology publications, effective June 1 and June 5, 2026, respectively. Morgan Stanley research (Private Credit default-rate estimate).

Three Things We Are Talking About 01 C&I GROWTH: BORROWED FROM PRIVATE CREDIT, NOT YET EARNED C&I loans at commercial banks jumped 12.7% quarter over quarter in Q1 2026, on top of 4.3% growth for all of 2025. Individual large regional banks posted Q2 QoQ growth ranging from roughly 4% to 8% in their own earnings releases. Pantheon Macroeconomics credits the pickup almost entirely to private credit stress pushing borrowers back to banks, not independent investment demand. Morningstar DBRS and Truist Securities see genuine borrower appetite in the mix too, with the caveat that a single quarter doesn't make a year. Credit that migrated because a direct lender pulled back isn't the same risk as credit a bank actually won on price and structure, and the underwriting should treat it that way. The acceleration is also uneven by size. The names driving the 12.7% aggregate figure skew toward larger regionals with existing capital-markets and TLB capability. Smaller banks describe lending activity as resilient this year, not explosive. 02 CAMELS REWRITE: DURATION RISK GETS A BRIGHTER SPOTLIGHT The FFIEC's first comprehensive CAMELS revision in 30 years, proposed May 19 with comments due August 17, shifts supervision toward material financial risk over process and documentation. The Management component loses its outsized pull on composite scores. Sensitivity to Market Risk picks up explicit evaluation of NII performance against the rate environment and economic value sensitivity to rate volatility. For banks carrying unhedged duration, that's a sharper exam lens arriving right as the securities book is back under rate pressure. 03 TARIFFS: THE REGIME KEEPS CHANGING Temporary 10% reciprocal tariffs on roughly 60 trading partners expired at midnight July 24 and were immediately replaced with new 10% to 12.5% tariffs on largely the same countries, this time on forced-labor grounds instead of trade balance. USMCA talks with Mexico are still underway, and separate Canada tariffs don't take effect until August 19. This isn't a settled rate to underwrite around, it's a moving target that has changed its legal basis twice in a month. The risk for an import-exposed loan book isn't the tariff level, it's that the level and the justification behind it keep changing.

What We Are Hearing From Our Clients “How many producers does it actually take to generate a dollar of loans? Is that ratio moving the wrong way?” This came up directly at the TBA CFO Conference, and it's the sharper version of a question every bank is asking. Gross origination activity is strong across most of the regional banks we talk to, consistent with the 12.7% aggregate C&I growth. The harder problem sits on the other side of the ledger. Early payoffs, line paydowns, and refinancings are running high enough that replacing the runoff, not generating new volume, is the real constraint on net growth. A bank adding producers to chase gross originations without solving the net-growth problem is buying activity, not efficiency. “We were finally seeing our unrealized losses unwind. This rate move undid it. What are our options for taking AOCI volatility off the table?” Unrealized losses on investment securities are back up, $325.1 billion industry-wide in Q1, and the July back-up in rates erased most of the improvement the spring brought. The real options are narrower than clients want them to be. Selling now crystallizes a loss but frees up the reinvestment at materially higher yields. Holding keeps the loss on paper but leaves the book exposed to the next move. Reclassifying from available-for-sale to held-to-maturity removes future swings from AOCI without removing the embedded loss, and the Basel III Endgame reproposal would extend AOCI capital treatment down to any bank over $100 billion in assets, a detail that matters more to some institutions than others. The lower-friction option sits at the margin: directing new purchases into floating-rate CLO AAAs instead of back into fixed-rate agency RMBS, a shift that avoids adding to the interest rate exposure already embedded in the book. That doesn't fix the existing book, but it stops making the problem bigger. Investing in CLO tranches involves credit, structural and liquidity risks and is not suitable for all investors. Caird may hold positions in CLO securities or related instruments and may receive fees that create conflicts of interest. “We're seeing softness in consumer-facing sectors. Does that bleed into our corporate book and slow our lending?” The consumer data is genuinely split rather than uniformly weak. Higher-income households are still spending, and University of Michigan sentiment just posted its second straight monthly gain. Lower-income households are under real and building strain. That split matters more than the headline number for a C&I book. Weakness transmits to corporate credit through companies selling to the strained cohort specifically, not through an entire sector. The names worth watching are consumer-facing borrowers with customer concentration in that lower-income segment, not every borrower tagged consumer discretionary.

Chart of the Month Unrealized losses on bank securities portfolios peaked at $690 billion in Q3 2022, then decreased to a $306 billion low in Q4 2025, and ticked back up to $325 billion in Q1 2026, the most recent quarter the FDIC has reported. The 10-year Treasury yield explains most of that path. It fell through most of 2024 and 2025, and the securities book recovered alongside it. Since Q1 2026 it has backed up again, from 4.30% to 4.55% as of July 17, the highest level since the end of 2024. The next Quarterly Banking Profile isn't out yet. This chart is what it will likely show. Sources: FDIC Quarterly Banking Profile releases, Q2 2022 through Q1 2026 (Q4 2023, Q3 2024, and Q4 2025 derived from reported dollar or percent change versus the adjacent verified quarter). 10-year Treasury yield: FRED series DGS10, quarter-end close, extended through July 17, 2026.

The Caird Perspective New Charters Are Testing How Sticky Deposits Really Are The OCC received 14 de novo charter applications in 2025, nearly matching the prior four years combined, and is on pace for roughly 25 in 2026. Most are national trust charters, which permit custody and tokenization but not deposit-taking or lending. A few go further. Circle won its charter July 8, Sony Bank followed two days later, and Mercury Technologies received a broader charter in April that lets it build Zelle integration, lending products, and its own payments infrastructure, a more direct run at the deposit relationship than a trust charter alone. A second, older charter pathway is moving just as fast, and it can take deposits directly, no trust wrapper required. The FDIC has conditionally approved four industrial loan company charters since January: Ford Credit Bank, GM Financial Bank, Edward Jones Bank, and Stellantis Bank, with Nissan's application pending since June 2025. An ILC charter lets an automaker's captive finance arm or a brokerage's cash-management business hold deposits and lend against them directly, a more direct run at the relationship than anything crypto-adjacent has managed so far. The charters are not the whole story. What sits behind them matters more. USDC circulation is up 19% this year to $73.2 billion despite falling rates, and the stablecoin market overall has grown to roughly $310 billion. Estimates of what that could mean for bank deposits vary widely. Standard Chartered has put a figure as high as $500 billion by the end of 2028, while the Fed's own research spans $65 billion to $1.26 trillion in potential lending impact, depending on adoption and where issuers hold reserves. That range is itself worth noting: nobody has landed on a precise number yet, but the direction of the risk is not really in question. The competition is not hypothetical either. X Money launched in beta in late June offering 6% APY on deposits held through a partner bank, an explicit teaser rate that is not sustainable at current rates but is doing exactly what a teaser rate is designed to do. It is one data point, not a trend, but it is the clearest sign yet of how aggressively new entrants are willing to price for share. Deposit relationships have historically relied, to some extent, on friction, the small cost of moving money that keeps a balance sitting where it is. Real-time settlement rails, a federally chartered stablecoin issuer, and a captive finance arm that can now hold deposits directly are all, in their own way, an attempt at removing that friction. It may be worth adding this scenario to the usual rate-driven funding stress tests, as a distinct variable rather than a replacement for them. Sources: OCC charter and licensing decisions (2025-2026); CNBC and IBTimes, Circle and Sony Bank OCC approvals, July 2026; Mercury Technologies OCC conditional approval, April 27, 2026; FDIC ILC charter approvals via Banking Dive (Ford Credit Bank and GM Financial Bank, January 22, 2026; Edward Jones Bank; Stellantis Bank); Standard Chartered research; Federal Reserve FEDS Notes, December 2025; X Money product launch coverage, June 2026.

Caird Team Update Where We Have Been and Where We Are Headed The Caird team attended the Texas Bankers Association CFO Conference in Lakeway last month. Three themes stood out. CFOs are pressing their teams on efficiency in C&I growth specifically, how many producers it takes to generate a dollar of loans, and whether that ratio is moving the wrong way. The regulatory panel, OCC, FDIC, and the Texas Department of Banking, struck a constructive tone, focused on material risk identification and prudent AI governance, with the rescission of leveraged lending guidance coming up repeatedly as a tool banks can use to compete more effectively. And on deposits, bankers were candid that liability-side management, not asset yields, has driven most of the recent NIM improvement, a theme this issue's lead essay picks up in more detail. On the team front, our summer interns Lauren Yu and Sebastian DiBenedetto have been doing real work this quarter: underwriting new TLBs, digging into securities analysis, and learning how to evaluate a bank balance sheet from the inside rather than the outside. Their last day with us is at the end of the month. It has been a genuine pleasure having them on the team, and whatever they take on next, they leave with a real base of insight and hands-on experience to build from. We will be rooting for both of them as their careers develop and look forward to staying in touch! From the Team: Lauren lost the office World Cup bracket. Ben won, and per house rules, that means he gets to set her phone background for the next year. As a UNC fencer, her new background touting Duke fencing will go over well, we are sure.

The Caird Team info@cairdinvestments.com · 469-729-0530 cairdinvestments.com · LinkedIn Caird Investment Partners, LLC is a SEC registered investment adviser. This material is provided for informational purposes, and it is not, and may not be relied on in any manner as, legal, tax or investment advice or as an offer to sell or a solicitation of an offer to buy an interest in any fund or investment vehicle managed by Caird Investment Partners, LLC (“Caird”). The information and data are as of the publication date unless otherwise noted. Past performance is not indicative of future results. Content obtained from third-party sources, although believed to be reliable, has not been independently verified as to its accuracy or completeness and cannot be guaranteed. Caird has no obligation to update, modify or amend the content of this post nor notify its readers in the event that any information, opinion, projection, forecast or estimate included, changes or subsequently becomes inaccurate.