BNY — Q3 2026 Financial Report Analysis

BNY (The Bank of New York Mellon) earns most of its money from fees for holding, servicing and moving assets for institutions, plus net interest income (NII) — the gap between what it earns investing the cash clients leave with it and what it pays them on those deposits. In the third quarter of 2026 (quarter ended Sept. 30), total revenue rose 12% from a year earlier to $5.68 billion, fees and NII both grew double digits, and noninterest expense (the cost of running the bank) rose 8%. Net income applicable to common shareholders rose 20% to $1.61 billion and diluted earnings per share (EPS) rose 24% to $2.34, the extra EPS growth coming from a 3.6% lower share count. Against the previous quarter, though, the picture is flat-to-down: revenue slipped $22 million, expenses rose $50 million, and EPS fell from $2.45.

This report is based on BNY's earnings release (Exhibit 99.1 to its Oct. 15, 2026 8-K) and the financial supplement filed with it. The full Q3 10-Q isn't filed yet, so some detail (cash flow statement, full MD&A) isn't available.

Key metrics

Pre-tax operating margin is income before taxes as a share of total revenue. Net interest margin is NII as a share of average interest-earning assets — low at a custody bank because most of those assets are central-bank deposits and high-quality bonds rather than loans. Return on tangible common equity (ROTCE) is annualized profit over shareholders' equity after removing goodwill and other acquired intangibles; it's a non-GAAP (company-adjusted) measure. The CET1 ratio is common equity tier 1 capital as a share of risk-weighted assets, regulators' core measure of a bank's loss-absorbing cushion. BNY doesn't state an efficiency ratio or a net charge-off ratio in the release, so neither is shown.

Year to date: revenue $16,783M (+12.6% from $14,901M), net income applicable to common shareholders $4,866M (+25.4% from $3,879M), diluted EPS $7.02 (+30.2% from $5.39).

Takeaway: The year-over-year numbers are strong, but the quarter-over-quarter turn matters more for what comes next: revenue stalled at $5.7 billion while costs kept rising, so operating leverage (revenue growth minus expense growth) was −184 basis points versus Q2 after a run of positive quarters. The main reason is that NII, the fastest-growing part of revenue in the first half, fell as average deposits dropped 1% and deposit costs rose. If deposits don't come back, fee growth alone has to carry earnings.

Where the revenue came from

Fees (+12%, $4,078M). BNY attributes the increase "primarily" to "net new business, higher client activity and higher market values, partially offset by the mix of AUM flows." Investment services fees — what BNY charges for custody, fund administration, corporate trust, clearing and similar work — rose 15% to $2,973M. Foreign exchange revenue rose 19% to $198M, though it fell from $229M in Q2. Investment management and performance fees grew only 3% to $803M.

Net interest income (+15%, $1,423M). Year over year, the increase "primarily" reflects "the reinvestment of investment securities at higher yields and balance sheet growth, partially offset by deposit margin compression" — maturing low-yield bonds being rolled into higher-yield ones. Average loans rose 19% to $86.8 billion and average interest-earning assets 6% to $396.5 billion. The average rate on interest-bearing deposits fell to 2.43% from 2.90% a year earlier. Sequentially, NII fell 2% "primarily reflecting lower deposit balances": average deposits were $311.1 billion, down from $314.0 billion in Q2.

Investment and other revenue fell to $175M from $208M a year earlier, reflecting weaker fixed income and equity trading, seed capital results (gains or losses on BNY's own money in its funds) and the absence of a $12M disposal gain in Q3 2025, partly offset by a strategic equity investment gain. It includes $25M of net losses on investment securities sales, against $30M a year earlier.

Segment results

Securities Services (custody, fund services, corporate trust, depositary receipts) grew fastest. Issuer Services fees rose 43% to $449M on higher Corporate Trust revenue, and Asset Servicing fees rose 12% to $1,264M. Securities lending revenue rose 48% to $92M. But the segment's margin fell to 35.8% from 39.3% in Q2: revenue was flat sequentially while expenses rose 5% ($1,808M vs. $1,722M), which BNY attributes to "higher revenue-related expenses and investments." Depositary Receipts revenue dropped sequentially, the seasonal decline the Q2 filing had warned about, and the segment's FX revenue fell 16% from Q2.

Market and Wealth Services (Pershing clearing for brokers and advisers, payments, tri-party collateral management) remains the most profitable segment at a 51.4% margin. Clearance and Collateral Management fees rose 17% to $464M on "higher collateral balances and clearance volumes." Payments and Trade revenue rose 14% to $579M, and Pershing brought in $48 billion of net new assets on its U.S. platform, versus $22 billion a year earlier.

Investment and Wealth Management (asset management and private wealth) remains the laggard. Revenue grew 4% to $857M; Wealth Management rose 11% to $295M while Investment Management was flat at $562M. Assets under management fell 2% in the quarter to $2,187 billion on $27 billion of negative market impact and $12 billion of net outflows. The mix flipped from Q2: long-term strategies took in $10 billion (fixed income +$9 billion, index +$4 billion, liability-driven investments +$3 billion, equity −$5 billion), while cash (money-market) strategies lost $22 billion after gaining $24 billion in Q2.

Costs and capital

Noninterest expense rose 8% to $3,489M on "higher revenue-related expenses and investments, partially offset by efficiency savings." Staff costs, the largest line, rose only 3% to $1,805M, and full-time employees were 46,600, 5% fewer than a year ago (49,200). Non-staff lines grew faster: professional, legal and purchased services +17% to $474M, software and equipment +9% to $593M, sub-custodian and clearing +19% to $168M. The provision for credit losses was a $1M charge against a $7M release a year ago — negligible either way.

BNY returned $1.4 billion to common shareholders: $436M of dividends and $937M of buybacks. It repurchased 5.8 million shares at an average $160.51, versus 8.0 million at $137.62 in Q2, so a higher share price bought fewer shares. The CET1 ratio was 10.9% at Sept. 30, down from 11.0% at June 30 and 11.9% at Dec. 31, 2025; BNY says CET1 capital fell mainly on buybacks, dividends and a decline in accumulated other comprehensive income (unrealized losses on securities), partly offset by earnings and lower risk-weighted assets. The Tier 1 leverage ratio fell to 5.7% from 6.1% a year ago. The average liquidity coverage ratio was 111%.

What the headline numbers hide

• Buybacks and the tax rate both move EPS. Average diluted shares fell 3.6% (686.9 million vs. 712.9 million), which is most of the gap between 20% net income growth and 24% EPS growth. Lower preferred dividends ($82M vs. $106M) also added to common earnings. The effective tax rate rose to 22.8% from about 21.3% a year ago (computed from $499M of tax on $2,186M of pre-tax income vs. $395M on $1,852M), which held back net income — so pre-tax income (+18%) is the cleaner measure of operating growth.
• GAAP vs. adjusted is a small gap. BNY's adjusted figures exclude "notable items": $23M of severance and a $1M litigation reserve release this quarter; a year ago, $50M of severance, $3M of litigation reserves, a $14M FDIC special-assessment credit and a $12M disposal gain. Adjusted EPS was $2.37 vs. $1.91, the same +24% as GAAP, so one-offs don't flatter the comparison.
• Sequential momentum turned negative. Revenue fell 0.4% from Q2 while expenses rose 1.5%; BNY's own operating-leverage figure is −184 bps sequentially (−144 bps adjusted), against +389 bps year over year. Part is seasonal (Q2 depositary receipts and FX activity), but the NII drop on lower deposits is not obviously seasonal.
• Cash conversion can't be checked yet. The earnings release has no cash flow statement; that waits for the 10-Q. For a custody bank, operating cash flow swings with client balances and is a weak guide to earnings quality anyway.
• Period-end balances are noisy. Period-end deposits were $335.4 billion, well above the $311.1 billion quarterly average, and the supplement notes the March 31 spot balance sheet was "temporarily elevated" by delayed payment processing. NII is earned on average balances, so the average deposit figure (down 1% from Q2, up 4% from a year ago) is the one that matters for the trend.

Did last time's read hold up?

Mostly yes, with one miss. Our Q2 report said NII growth from reinvesting maturing bonds should continue and named "deposit margin compression" as the offset to watch. Year over year the reinvestment tailwind held (+15%), but sequentially it was outweighed by lower deposit balances and a higher deposit rate, so NII fell from the prior quarter. The Q2 10-Q warned that third-quarter volume-related fees "may decline due to reduced client activity" and that Q2 benefits from depositary receipts seasonality; both showed up (Depositary Receipts and FX revenue fell from Q2). We also flagged that buybacks drove EPS ahead of profit and depended on BNY running CET1 near 11%: CET1 edged to 10.9% and buybacks slowed to $937M from $1.1B, partly because the stock was about 17% more expensive.

Outlook

The earnings release gives no numeric guidance for revenue, NII or expenses. Our read: fee growth looks durable — Issuer Services, collateral management and Pershing net new assets all grew double digits, and those reflect client activity and new business rather than only market levels. The weak spot is NII, now about a quarter of revenue: with deposits slipping and deposit costs edging up, the reinvestment benefit that drove the first half is no longer enough on its own to grow NII sequentially. Watch Q4 average deposits and the interest-bearing deposit rate first, then whether expense growth (+8%) stays below fee growth (+12%). With CET1 at 10.9% and an 86% year-to-date payout ratio, further buybacks are likely to track earnings rather than exceed them, so EPS growth from share count reduction should be roughly what it is now, not more.