
Key Points
- Higher-for-longer rates should keep bank earnings resilient, but deposit costs are repricing faster, leaving limited room for further expansion in net interest margins.
- Loan growth is now doing more of the work for net interest income, but higher financing costs and a slowdown in private-credit lending could moderate momentum. Credit remains manageable at large banks, with risks more concentrated among smaller banks and commercial property loans.
- Investment banks and life insurers have the clearest upside, supported by stronger capital-markets activity and reinvestment yields that remain above existing portfolio yields. Japanese banks benefited from widening lending spreads and trading income.
- Based on our fair P/B of 1.5x, the sector still looks expensive and implies approximately -5% downside. We maintain our 2.0 Stars rating.
Strong earnings have yet to win over investors
Financials have lagged the broader market this year, even though earnings have been stronger than we expected in January. The MSCI World Financials Index has gained 2.5% YTD, compared with 10.8% for the MSCI World Index, after giving back most of its gains since early September. Fundamentally, loan growth has recovered, capital-markets activity has accelerated, and life insurers continue to reinvest maturing assets at yields above their existing portfolio yields.
Figure 1: Global financial stocks lag broader market

Can these drivers carry earnings into 2027? Our answer is that they can keep earnings supported, but the way higher rates feed into financial-sector earnings is becoming less favourable. Bank funding costs are repricing more quickly, loan demand should gradually respond to higher borrowing costs, and P&C pricing has turned softer. Credit remains manageable among large banks, but that also leaves limited scope for positive credit surprises.
Higher-for-longer rates should support NIM, but further expansion looks limited
Another rate hike should help banks defend current margins, but it is unlikely to restart a meaningful margin-expansion cycle. Following September’s hike to 3.75-4.00%, we expect the Federal Reserve to raise rates again by year-end, with policy remaining restrictive into 2027. That should keep asset yields elevated and reduce the risk of a sharp contraction in NIM.
The benefit from another hike is likely to be much smaller than in 2022. Back then, deposit rates lagged policy rates for several quarters, so asset yields rose far faster than funding costs. Deposit costs now respond far more quickly to policy-rate changes, as depositors have become more rate-aware and deposit growth has tilted towards more rate-sensitive uninsured balances. NIM could still widen gradually if asset yields continue to rise faster than funding costs, but we do not expect it to become a major source of earnings support. Much of the earlier asset repricing has already occurred, while deposit costs adjust more quickly and absorb more of the benefit from any further rate increase. In 2Q26, the industry’s earning-asset yield rose 2.3 basis points QoQ, but funding costs rose 1.6 basis points, leaving NIM only one basis point higher at 3.32%. NIM was six basis points above a year earlier but still below the 3.39% reached in 4Q25.
Figure 2: US bank NIM has flattened since late 2025

Rate-sensitivity estimates from the five largest US banks show that a 100bp rate increase would lift net interest income by 1.8% over 12 months, down from 3.4% two years ago. A -100bp cut would reduce it by -3.5%. Higher-for-longer rates should therefore help protect earnings, but another hike is unlikely to drive significant upgrades. Because large diversified banks make up a substantial part of the index, their lower rate sensitivity limits the benefit to the broader sector.
Figure 3: Large US banks now gain far less from a rate hike than they would lose from a cut

Loan growth is now doing more of the work for NII, but higher financing costs could slow it
With margins broadly stable, loan growth has become the main source of NII growth. Total loans at US commercial banks grew 6.9% YoY in 2Q26, above the long-term average of around 5.4%, while commercial and industrial loans grew 6.2%, the strongest pace in around three and a half years. As a result, industry NII rose 2.8% QoQ even though NIM barely moved.
A large share has come from lending to non-bank financial institutions, including private-credit funds, which rose 22.4% YoY in 2Q26. Since 2013, US regulators have discouraged banks from lending directly to heavily indebted companies, and capital rules make such loans costly to hold. Private-credit funds, which face fewer restrictions, have taken over much of this lending. Banks now lend to the funds instead, a safer position because the funds' own investors absorb losses first. As a result, private credit grows outside the banking system while banks' exposure grows indirectly. These loans carry thinner spreads, so they add volume more than margin, while tying banks more closely to the health of private credit.
That channel is now cooling. Private-credit loans at the six largest US banks rose only from USD256 billion to USD260 billion in 2Q26, after increasing by about USD20 billion a quarter through 2025. US direct lending fell 40% in the three months to May, as redemption caps at Blackstone and Cliffwater and software-loan concerns reduced fund deployment, so funds drew less on their bank credit lines.
Figure 4: Private-credit lending has been a key source of large-bank loan growth

Higher financing costs are a growing constraint on future loan demand. The 10-year Treasury yield has risen above 5%, its highest level since 2002, so corporates and households refinancing debt face higher payments, which may delay investment and weigh on housing affordability. Hyperscalers' bond issuance to fund AI investment has continued despite higher yields and we do not expect a near-term cut as the spending is treat as strategic. The bigger risk is AI revenue fail to justify the spending, which would slow both capex and the bond issuance that supports banks' debt-underwriting revenue.
Credit remains healthy for large banks, while risks are concentrated in smaller lenders and CRE
Consumer credit at the largest banks is improving. Bank of America’s credit-card net charge-off rate fell to 3.55% in 2Q26 from 3.82% a year earlier, while 30-day card delinquencies eased to 2.23% from 2.36%. Spending data point in the same direction. Card spending per household rose 4.5% YoY in August, more than four times the 2025 average, while card utilization returned to roughly pre-pandemic levels and lower- and middle-income households continued to hold deposits well above 2019 levels. The one soft spot is big-ticket, rate-sensitive spending such as cars, where weaker households are still holding back. An early sign that higher borrowing costs are weighing on loan demand.
Industry-wide’s past-due and nonaccrual ratio fell to 1.44% in 2Q26 from 1.53% in 1Q26, provision expense declined to USD19.3 billion from USD21.4 billion and reserve coverage rose to around 173% from 167%.Stress remains concentrated outside the largest banks. Credit-card delinquency was 2.7% at the 100 largest US banks in 2Q26, compared with 6.5% at smaller banks. Commercial real estate (CRE) delinquency remained around 1.5%, above its pre-pandemic level of roughly 0.9%. CRE exposure is also concentrated among mid-sized lenders: banks with USD1-50 billion in assets hold CRE loans worth around 180-190% of their equity, compared with around 50% at the largest banks.
Figure 5: Credit-card stress remains far higher outside the largest US
banks
Figure 6: CRE exposure is concentrated among mid-sized banks

For the index, this cuts both ways. Weaker borrowers are most likely to feel restrictive rates first, but they are concentrated at smaller lenders outside the index, so credit is unlikely to become a meaningful drag. Management guidance points to stable credit costs, JPMorgan recently lowered its 2026 card charge-off guidance to around 3.2% from 3.4%, and industry provisions are running roughly in line with charge-offs. However, with provisions already low, there is little room for reserve releases to lift earnings further, and growth will need to come from revenue instead.
Investment banks remain the clearest earnings-growth pocket within financials
Capital markets offer the strongest incremental earnings catalyst in the sector. Combined capital-markets revenue across Bank of America, Citi, Goldman Sachs, JPMorgan and Morgan Stanley reached USD59.8 billion in 2Q26, up 40% YoY, with trading revenue up 38% and investment-banking fees up 46%. The strength was broad-based, with equities trading revenue up 69-86% at each of the four largest US dealers, helped by mega-IPOs and the semiconductor rally.
Figure
7: 2Q26 revenue growth remained strong across selected capital-markets
businesses
Debt businesses grew far more slowly, for three reasons. Much of the corporate refinancing need was met in 2024–25, so debt issuance started 2026 from a high base, while equity issuance was rebounding from depressed levels. Rising yields also discouraged opportunistic refinancing: with the 10-year Treasury yield above 5%, companies refinanced only what they had to. And the standout source of new issuance, hyperscalers' bond sales to fund AI is concentrated in a handful of highly rated borrowers, whose investment-grade deals pay far lower underwriting fees than high-yield bonds or leveraged loans.
The third quarter was more challenging. Global M&A fell 41% QoQ to USD993 billion, and equity issuance dropped 26% QoQ, although it remained 39% higher YoY. Debt capital markets activity fell 18% YoY in July as the 10-year yield climbed to its highest level since 2002, and Bank of America has guided 3Q26 investment-banking fees down at least 10% YoY, with trading flat.
Figure 8: Capital-markets issuance in 2026 is on record pace

We expect capital-markets growth to slow. Moving into 2027, the high base and elevated yields point to much slower revenue growth, with debt underwriting most exposed if companies keep delaying refinancing. AI-related bond issuance, a healthy IPO pipeline and elevated volatility should still provide support. Investment banks therefore remain the clearest earnings driver within financials, but a more cyclical one that relies on yields stabilising and equity markets holding up.
Life insurers continue to benefit from higher reinvestment yields, while P&C pricing softens
Life insurers remain one of the clearer beneficiaries of higher-for-longer rates. Prudential’s fixed-maturity investment yield rose to 4.67% in 2Q26 from 4.39% a year earlier, while MetLife’s net investment income grew 18% YoY. Because insurers reinvest long-duration portfolios gradually, portfolio yields can keep rising for several quarters even without another sharp increase in policy rates.
P&C insurers face the opposite cycle. Global commercial insurance pricing fell -6% YoY in 2Q26, including a -12% decline in property pricing, with rates falling in every major region. Softer renewals reduce the scope for further underwriting-margin expansion, especially if claims inflation stays elevated or catastrophe losses return to more normal levels. This matters for the index, as Berkshire Hathaway, its second-largest constituent, is heavily exposed to P&C and reinsurance. We therefore prefer life insurers over P&C insurers.
Figure 9: Commercial insurance pricing is falling across all major regions

Muse is not yet an earnings threat
Meta’s AI agent, Muse, added to September’s pressure on global financials, although its contribution is difficult to separate from other factors. The MSCI World Financials Index gave back some of its gains from early September. On 22 September, US wealth platforms were among the sharpest decliners as investors assessed the potential threat from AI agents. The sell-off also coincided with a flatter US yield curve and concerns about softer third-quarter investment-banking fees.
Muse has sharpened concerns that AI agents could eventually direct customers’ spare cash towards higher-yielding options, putting pressure on banks’ low-cost deposits and brokers’ cash-sweep income. That is a plausible longer-term risk, but there is no evidence yet that Muse has caused meaningful customer outflows or earnings losses. We see AI agents as a long-term risk concentrated in retail banks that rely on low-cost deposits and wealth platforms that earn cash-sweep or advisory fees. For now, this is not a demonstrated near-term threat to the broader financial sector’s earnings.
Selective opportunities across global banking markets
Europe: Banks continue to report stronger-than-expected results, with STOXX Europe 600 Banks Index revenue and earnings up about 9% and 12% year on year in 2Q26. However, further NIM gains may be limited, and higher consensus forecasts leave less room for earnings surprises. M&A provides selective support, though regulatory and cross-border constraints remain. Hence, we are Neutral on European banks.
Japan: Banks are benefiting from the BOJ’s rate hikes, which have widened their domestic deposit-lending spread and supported NII. Fee and trading income also grew strongly, broadening the earnings recovery. Further rate increases could extend the NIM tailwind, although rising credit costs are a risk.
Read more: MUFG: Banking on Japan’s rate hikes for more upside
Singapore: Earnings remain resilient, with firmer SORA helping to stabilise NII and wealth management and trading supporting fee-led growth. Valuations are above historical averages, so further share-price gains will depend more on earnings delivery and dividends. The 7 October 2026 pullback following Citi’s OCBC downgrade appears to reflect profit-taking after a strong YTD rally, rather than a broad deterioration in bank fundamentals. While valuations limit price upside, attractive dividend income supports a compelling total-return profile. Within the sector, we prefer DBS for its earnings profile and dividend yield.
Read more: Singapore banks: Sell-off tests the rally, but fundamentals hold
China: Deposit-rate controls are helping contain banks’ funding costs, while corporate and consumer lending support loan growth. But weak mortgage demand and pressure on loan yields continue to constrain NII. The Big Four offer stable dividends and may provide portfolio ballast, but after their share-price rally, potential upside is now in the single digits.
Read more: Four State-owned Banks 1H26 Release: How Should Investors Position in Stocks After Price Surge?
Although we remain cautious on global financials overall, we see selective opportunities in Japanese banks, where rising domestic rates are widening lending spreads and fee and trading income are broadening earnings growth. We are Neutral on European, Singapore and Chinese banks as their income is resilient, but valuations limit share-price upside.
Valuation leaves little room for disappointment
Valuation remains the main obstacle to a more constructive view. The S&P Global 1200 Financial Sector Index trades at around 1.93x book value, about 42% above its 10-year average of 1.36x.
Earnings have clearly beaten our expectations. Our FY2026 EPS estimate of 202.2 is 11% above our January forecast of 182, and we raise our fair P/B from 1.3x to 1.5x to reflect a structurally higher return on equity. Even so, EPS growth is expected to slow from 14.4% in FY2026 to around 7-8% in FY2027 and FY2028. Applying 1.5x to FY2028 book value per share gives a fair value of 2,586, implying downside of around -5%.
Figure 10: The SGFS Index P/B ratio remains around two standard deviations above its 10-year average

Table 1: Earnings estimates of the SGFS Index
|
|
FY25 |
FY26E |
FY27E |
FY28E |
|
PE Ratio |
15.18 |
13.46 |
12.57 |
11.68 |
|
Earnings Per Share (EPS) |
176.81 |
202.2 |
216.53 |
233.09 |
|
Earnings Growth (YoY%) |
|
14.36% |
7.09% |
7.65% |
|
Dividend Yield (%) |
2.50% |
2.93% |
3.21% |
3.53% |
|
P/B Ratio |
1.94 |
1.85 |
1.72 |
1.58 |
|
Book value per share |
1,350 |
1,486 |
1,587 |
1,724 |
|
Projected Fair Price |
|
|
|
2586 |
|
Potential Upside (Based on 1.5x fair P/B ratio) |
|
|
|
-5% |
|
Source: Bloomberg Finance L.P., iFAST Estimates. Data as of 5 October 2026. |
||||
Key Takeaway
The broad financial sector still lacks a common catalyst strong enough to support a sustained rerating. Large US banks are fundamentally healthy, but matured NIM, increasingly volume-dependent NII growth and already-benign credit leave limited room for positive earnings surprises, while P&C pricing is softening and capital-markets growth faces a high base. With the index trading around two standard deviations above its 10-year average P/B, we maintain our 2.0 Stars "Unattractive" rating.
Investors holding broad financial-sector exposure may consider taking profits at current levels. Those who prefer to stay invested should be selective, as investment banks retain a visible capital-markets catalyst, while life insurers continue to benefit from rising reinvestment yields.
For redeployed capital, we see better opportunities in Asia semiconductors, emerging markets, Asia ex-Japan technology and Singapore. These markets offer stronger earnings visibility, less-stretched valuations, AI-driven growth, multi-year memory and hardware demand, supportive reforms and buybacks.

