MS 10-K Analysis: The $139B Sweep Deposit Paradox Behind Record Earnings

Morgan Stanley's record FY 2025 results rest on a single $139 billion asset that is simultaneously its strongest earnings engine and the target of class-action lawsuits. The $282 billion global financial services firm generated $70.6 billion in revenue across investment banking, wealth management, and trading, earning a 21.6% return on tangible equity — the highest among large bank peers. The market rewarded it with a 17.4x P/E, the richest multiple in the group.

The FY 2025 10-K, filed with the SEC on February 19, 2026, tells a story the earnings headlines cannot. Three operating segments delivered: Institutional Securities (IS) earned $33.1 billion in revenue, Wealth Management (WM) contributed $31.8 billion, and Investment Management added $6.5 billion. Revenue grew 14.4% firm-wide. EPS surged 28.4% to $10.21. Net income reached $16.2 billion. Management met its 20%+ ROTCE target and announced a $20 billion buyback authorization. On the surface, Morgan Stanley executed one of the strongest years in its history.

But beneath those numbers, the filing reveals dynamics that fundamentally reframe what investors are paying for. The $139 billion sweep deposit base driving record net interest income acceleration is the same asset class-action plaintiffs allege was unfairly compensated — Morgan Stanley's strongest growth driver and most dangerous legal liability are the same dollars. Institutional Securities overtook Wealth Management as the number-one revenue segment for the first time in the post-E*TRADE era, reintroducing cyclical risk the market hasn't priced. And the celebrated 300-basis-point efficiency improvement blends a structurally durable WM engine with a cyclical IS tailwind that may not repeat.

At $177.53 and 17.4x earnings, the market prices Morgan Stanley as a stable wealth management franchise with above-peer returns. The filing describes a firm whose revenue mix, earnings drivers, and risk profile have shifted meaningfully from that narrative — and the question is whether investors have noticed.

The Narrative Crack — IS Overtakes WM

Morgan Stanley's premium valuation rests on a story: the E*TRADE and Eaton Vance acquisitions transformed it from a cyclical investment bank into a stable, recurring-fee wealth management franchise. That story broke in FY 2025. Institutional Securities generated $33.08 billion in revenue — 46.8% of the firm total — overtaking Wealth Management's $31.75 billion (44.9%) for the first time since the acquisitions closed. The segment that investors pay a premium to avoid became the segment driving results.

The composition of IS dominance matters. Equity trading surged 32.6% to $11.94 billion, representing 67% of IS trading revenue — a concentration that makes IS performance highly correlated with equity market volumes. Interest rate trading declined 26.2%. Meanwhile, IB revenues from contracts with customers — the pure advisory and underwriting fees that meet ASC 606 definitions — fell from 91% of IB revenue (FY 2023) to 84% (FY 2025). The remaining 16% likely includes principal gains from bridge loans and warehousing, activities that carry balance sheet risk rather than earning fee income. IS is not just leading on revenue — it is doing so with a more volatile, more capital-intensive business mix than it carried two years ago.

IS pre-tax income doubled from $4.5 billion (FY 2023) to $11.2 billion (FY 2025), driven in part by compensation ratio compression of 670 basis points to 29.6%. That compression contributed approximately $2.2 billion in incremental pre-tax income — a figure that reflects revenue leverage on a surge year as much as structural discipline. The IS compensation ratio was 36.3% just two years ago; whether it sustains near 30% or reverts toward 33% will determine how much of the efficiency improvement is durable.

The peer context sharpens the concern. Goldman Sachs — structurally the most similar firm — trades at 16.9x earnings with a comparable IS-heavy revenue mix. Morgan Stanley's 0.5-turn P/E premium and 0.43x P/B premium over GS (2.52x vs 2.09x) have historically been justified by WM's recurring revenue stability. With IS now generating more revenue than WM, that premium justification weakens. Morgan Stanley's Institutional Securities segment generated $33.1 billion in FY 2025 revenue, overtaking Wealth Management for the first time since the E*TRADE acquisition, because equity trading surged 32.6% to represent 67% of IS trading revenue.

The $139 Billion Paradox — NII Growth vs Sweep Litigation

Morgan Stanley's strongest growth driver and biggest legal risk derive from the same source: $139 billion in brokerage sweep deposits. These are client cash balances automatically swept from brokerage accounts into low-interest deposit accounts — balances that the Q3 2025 10-Q explicitly characterizes as having "low-cost funding characteristics."

"Deposits are primarily sourced from our Wealth Management clients and are considered to have stable, low-cost funding characteristics relative to other sources of funding."

That "low-cost funding" language is precisely what the class-action lawsuits challenge. Multiple suits allege that Morgan Stanley Securities and E*TRADE Securities failed to pay reasonable interest on these sweep products — keeping rates low while lending out the deposits at market rates. The economics confirm the plaintiffs' basic observation: WM net interest income grew 8.2% to $7.91 billion in FY 2025, even as WM interest expense declined 6% from $8.93 billion to $8.40 billion. MS earned wider spreads on a growing deposit base.

The net interest income acceleration was striking: firm-wide NII grew 21% from Q1 to Q4, from $2.35 billion to $2.86 billion, despite the Federal Reserve cutting rates during that period. This counter-intuitive trajectory reflects the structural advantage of a $139 billion low-cost deposit base — but also the precise economics that make the sweep lawsuits credible.

The asymmetry compounds the risk. The Q3 2025 10-Q disclosed WM NII sensitivity of +$188 million per 100bp rate increase versus -$228 million per 100bp rate decrease. Rate cuts cost WM NII more than equivalent rate hikes help, because deposit costs are already near floors in some sweep products. In the current rate-cutting environment, this creates a double headwind: falling rates reduce NII mechanically, while litigation could force higher sweep rates on the remaining spread.

This is the paradox investors must resolve: bulls projecting continued NII growth and bears pricing litigation risk are unwittingly debating the same variable — the sweep rate. Morgan Stanley holds $139 billion in brokerage sweep deposits that the filing calls "low-cost funding," generating $7.9 billion in wealth management NII while class-action lawsuits allege those same low rates are unfair — each 50bp forced rate increase would cost approximately $700 million per year.

The Efficiency Machine — Two Engines, Different Durability

Morgan Stanley's efficiency ratio improved 300 basis points in FY 2025, from 71.1% to 68.4% — the strongest improvement in the large bank peer group. But decomposing the improvement reveals two distinct engines with fundamentally different durability.

The structural engine lives in Wealth Management. WM non-compensation expenses stayed essentially flat at approximately $5.5 billion for three consecutive years — $5.64 billion in FY 2023, $5.41 billion in FY 2024, $5.46 billion in FY 2025 — while revenue grew from $26.3 billion to $31.8 billion. Non-comp as a percentage of WM revenue fell from 21.5% to 17.2% over the period. Meanwhile, the WM compensation ratio was rock-stable: 53.2%, 53.5%, 53.4% in consecutive years. The WM pre-tax margin improved from 24.9% to 27.2% to 29.3% entirely through non-comp operating leverage. Every incremental dollar of WM revenue drops through at approximately 47% operating margin — this is durable, structural cost absorption that has additional runway as long as WM revenue grows.

The cyclical engine lives in Institutional Securities. IS comp ratio compressed 670 basis points in two years — from 36.3% to 29.6% — contributing approximately $2.2 billion in incremental pre-tax income. Some of that compression reflects genuine discipline: the March 2025 reduction in force eliminated approximately 2% of the global workforce at a one-time severance cost of $144 million.

"During the first quarter of 2025 as a result of a March employee action, we recognized severance costs associated with a reduction in force ('RIF') of $144 million, included in Compensation and benefits expense. The RIF... impacted approximately 2% of our global workforce at that time."

But $144 million in severance savings is marginal against $33 billion in IS revenue. The primary driver of IS comp compression is revenue leverage — more revenue per banker and trader on a surge year. If IS revenue reverts even partially toward FY 2024 levels ($28.1 billion vs $33.1 billion), the comp ratio rises and the efficiency ratio worsens. The peer comparison underscores how far MS still has to go: at 68.4%, Morgan Stanley's efficiency ratio remains the worst among large bank peers — above Goldman Sachs (64.4%), Bank of America (65.6%), and Wells Fargo (65.5%). Morgan Stanley's wealth management non-compensation expenses stayed flat at $5.5 billion for three years while revenue grew from $26.3 billion to $31.8 billion, reducing non-comp as a share of revenue from 21.5% to 17.2% — a structural efficiency improvement independent of the cyclical IS comp ratio compression.

The Capital Return Inflection — SCB Cut Changes the Math

The Federal Reserve's decision to cut Morgan Stanley's stress capital buffer from 5.1% to 4.3% in September 2025 was the quiet inflection point of the year. It lowered the required CET1 ratio from 13.5% to 11.8%, expanding the capital buffer from 1.5 percentage points to 3.2 percentage points — approximately $17.9 billion in excess CET1 capital on estimated risk-weighted assets of $560 billion.

"On September 30, 2025, the Federal Reserve announced that it had reduced Morgan Stanley's SCB from 5.1% to 4.3%, effective on October 1, 2025."

This enabled the $20 billion buyback authorization and the Q4 acceleration: MS repurchased approximately $1.5 billion in Q4 alone, up from the ~$1.15 billion quarterly average. But the headline buyback figures overstate the per-share value creation. MS spent $4.585 billion on share repurchases in FY 2025, retiring approximately 25.8 million shares at $177.53 average price. The net share count reduction, however, was only 19.2 million shares — a 1.2% reduction in the float. The difference of approximately 6.6 million shares represents stock-based compensation dilution that the buyback merely offset. At $177.53 per share, roughly $1.17 billion — 26% of the total buyback spend — went to neutralizing SBC dilution rather than creating incremental per-share value. SBC expense grew 18.7% to $1.93 billion, outpacing revenue growth of 14.4%.

The shareholder yield gap is stark. At 3.81%, MS returns barely half what Goldman Sachs (6.1%) and Wells Fargo (7.7%) provide, despite having the highest CET1 ratio in the peer group. The SCB cut provides the regulatory room to close this gap — the capacity is there. Whether management deploys it is the test. An additional complication is the $640 million in annual preferred stock dividends — a permanent capital claim invisible in most earnings summaries that reduces the effective return to common shareholders by 3.8%.

If Q1 2026 buybacks exceed $1.75 billion, the annualized total shareholder yield reaches approximately 5.2% — closing half the gap with Goldman Sachs. If buybacks decelerate despite excess capital, management may be conserving for litigation or balance sheet needs rather than returning capital. Morgan Stanley's stress capital buffer cut from 5.1% to 4.3% tripled the firm's excess CET1 capital to approximately $17.9 billion, enabling buyback acceleration — but at the current 3.81% total shareholder yield, MS still returns barely half what Goldman Sachs (6.1%) and Wells Fargo (7.7%) provide.

What to Watch: Three Metrics That Test the Thesis

Morgan Stanley's FY 2025 results were unambiguously strong — record revenue, record earnings, highest ROTCE among large bank peers. The question is not whether the year was good, but whether its drivers are durable or vulnerable. Three metrics over the next two quarters will provide the answer.

1. WM net interest income (Q1 2026). WM NII accelerated through FY 2025 quarters, reaching approximately $2.1 billion in Q4. If Q1 2026 WM NII exceeds $2.0 billion, the NII acceleration trajectory is intact despite Fed rate cuts — confirming the structural advantage of the $139 billion sweep deposit base. If WM NII falls below $1.9 billion, the rate-cut sensitivity is reasserting despite the sweep advantage, the margin path to 30% stalls, and the sweep deposit feedback loop tilts toward the litigation risk side.

2. Firm-wide efficiency ratio (Q1 2026). The 68.4% FY 2025 figure benefited from IS comp at a cyclical low, a $144 million severance event, and 14.4% revenue growth — conditions that may not repeat. If Q1 efficiency falls below 68%, the structural WM non-comp leverage engine is overpowering any IS comp reversion — thesis strengthened. If Q1 efficiency rises above 70%, IS comp discipline is unwinding and the 300bp improvement was partially a one-year event.

3. Buyback pace (Q1 2026). The SCB cut unlocked the capacity. Q4 showed early acceleration. If Q1 buybacks exceed $1.75 billion, the annualized total shareholder yield rises to approximately 5.2%, closing half the gap with Goldman Sachs and signaling management confidence. If buybacks fall below $1.0 billion despite $17.9 billion in excess capital, management is sending a signal of its own — possibly conserving capital in anticipation of adverse litigation outcomes.

At $177.53, the market prices Morgan Stanley at 17.4x earnings — a 29% premium above the Bank of America/Wells Fargo average of ~14.1x and a 3% premium above Goldman Sachs. For that multiple to be fair at a 10% cost of equity, MS needs approximately 8-10% annual EPS growth sustained over five years, reaching $15-16 by 2030 with a terminal P/E of 14x. The filing supports this trajectory through structural WM non-comp leverage, $356 billion in WM net new assets, and $17.9 billion in deployable excess capital. But it complicates it through IS cyclicality, sweep deposit litigation exposure, and a 68.4% efficiency ratio that is still the worst among peers. The investor question is not whether Morgan Stanley had a great year — it is whether the sources of that greatness are durable enough to sustain a 17.4x multiple through the inevitable normalization.

Frequently Asked Questions

What does Morgan Stanley's FY 2025 10-K filing reveal about its business mix?

The FY 2025 10-K shows that Institutional Securities (IS) overtook Wealth Management (WM) as Morgan Stanley's number-one revenue segment for the first time in the post-E*TRADE era. IS generated $33.08 billion (46.8% of revenue) versus WM's $31.75 billion (44.9%). This matters because MS's premium valuation (17.4x P/E versus 13-15x for bank peers) is traditionally attributed to the stability of recurring WM fee income. With IS — which is driven by equity trading (67% of IS trading) and cyclical IB advisory — now leading, the risk profile of MS's earnings has shifted. IS grew 18% year-over-year while WM grew 12%, and IS pre-tax income doubled from $4.5 billion (FY 2023) to $11.2 billion (FY 2025).

How large is Morgan Stanley's sweep deposit exposure, and why does it matter?

Morgan Stanley held $139.1 billion in brokerage sweep deposits as of Q3 2025. These are client cash balances automatically swept from brokerage accounts into low-interest deposit accounts. The filing explicitly characterizes them as having "low-cost funding characteristics." This matters for two reasons: first, the low sweep rate enables MS to earn wide NII spreads — WM NII grew 8.2% to $7.9 billion in FY 2025 while WM interest expense declined 6%. Second, multiple class-action lawsuits allege that Morgan Stanley Securities and E*TRADE Securities failed to pay reasonable interest on these sweep products. At $139 billion, even a 50bp forced rate increase would cost approximately $700 million per year in NII, compressing WM pre-tax margin from 29.3% to approximately 27.1%.

Is Morgan Stanley's efficiency ratio improvement sustainable?

Partially. The 300bp improvement (71.1% to 68.4%) comes from two distinct engines with different durability. The WM engine improved through non-comp operating leverage: WM non-comp expenses were flat at approximately $5.5 billion while revenue grew from $26.3 billion to $31.8 billion, reducing non-comp as a percentage of revenue from 21.5% to 17.2% over three years. WM comp ratio was rock-stable at 53.2-53.5%. This is structural and has runway. The IS engine improved through comp ratio compression: IS comp/revenue fell from 36.3% (FY 2023) to 29.6% (FY 2025), contributing approximately $2.2 billion in incremental pre-tax income. But this reflects revenue leverage on a cyclical surge year — if IS revenue normalizes, the comp ratio will rise. MS's 68.4% efficiency ratio is still the worst among large bank peers (GS 64.4%, BAC 65.6%, WFC 65.5%).

How does Morgan Stanley's valuation compare to Goldman Sachs?

MS trades at 17.4x P/E and 2.52x P/B versus GS at 16.9x and 2.09x. The approximately 0.5-turn P/E premium and 0.43x P/B premium are traditionally attributed to MS's WM stability. However, in FY 2025: IS overtook WM as the number-one segment, making MS's revenue mix more similar to GS's; GS returned 108% of net income to shareholders ($17.6 billion) versus MS's 69% ($11.2 billion); and GS's total shareholder yield is 6.1% versus MS's 3.81%. The gap suggests that either MS's premium will compress as its revenue mix converges toward GS's profile, or MS must demonstrate that its WM engine produces superior returns even when IS leads on revenue.

What is the ROTCE discrepancy between different sources?

The MetricDuck pipeline reports ROTCE at 19.1% while the filing reports 21.6% — a 250bp gap. This is fully explained by two factors: the pipeline's tangible common equity denominator includes approximately $10 billion in preferred equity (the pipeline maps preferred_stock as zero due to an XBRL mapping limitation), and management calculates ROTCE using net income to Morgan Stanley minus preferred dividends (approximately $640 million per year) in the numerator. Correcting both: $16.2 billion divided by approximately $75.2 billion average true TCE equals 21.6%. The filing's 21.6% is the economically accurate number — MS generates a 21.6% return on tangible common equity, making it one of the highest-returning G-SIBs.

What did the SCB reduction mean for Morgan Stanley's capital position?

The Federal Reserve reduced MS's stress capital buffer from 5.1% to 4.3% effective October 1, 2025. This lowered the required CET1 ratio from 13.5% to 11.8%. At 15.0% actual CET1, MS's capital buffer expanded from 1.5 percentage points to 3.2 percentage points. On estimated risk-weighted assets of approximately $560 billion, the excess CET1 capital is approximately $17.9 billion. This is the enabling event for the $20 billion buyback authorization and explains the Q4 buyback acceleration — approximately $1.5 billion in Q4 alone versus the approximately $1.15 billion per quarter FY average. The SCB cut effectively tells the market: MS is now viewed by the Fed as less systemically risky, and has the capital room to be significantly more aggressive on shareholder returns.

How much of Morgan Stanley's buyback actually shrinks the share count?

Not as much as the headline suggests. MS spent $4.585 billion on buybacks in FY 2025, which at $177.53 per share retires approximately 25.8 million shares. But the net share count reduction was only approximately 19.2 million shares (1.2% of the float). The difference of approximately 6.6 million shares represents SBC-related dilution that the buybacks merely offset. At approximately $177.53 per share, roughly $1.17 billion of the $4.585 billion buyback (26%) went to neutralizing SBC dilution rather than creating per-share value. SBC expense grew 18.7% to $1.926 billion, outpacing revenue growth of 14.4%.

What is the IB contract revenue decline signaling?

Investment banking revenues from contracts with customers declined from 91% (FY 2023) to 90% (FY 2024) to 84% (FY 2025). The remaining 16% in FY 2025 likely includes principal gains and losses from bridge loan commitments, warehousing positions, or other arrangements that do not meet the ASC 606 contract revenue definition. This seven-percentage-point decline over three years signals that IB is increasingly taking principal positions rather than earning pure advisory and underwriting fees. The practical impact: MS's IB revenue is becoming more volatile, less predictable, and more dependent on mark-to-market outcomes — characteristics that contradict the "fee-based, stable" narrative that supports the premium multiple.

What should investors watch in Q1 2026?

Three metrics determine whether the thesis holds. First, WM NII — if Q1 WM NII exceeds $2.0 billion, the NII acceleration is intact despite Fed rate cuts; below $1.9 billion signals the sweep deposit advantage is fading. Second, the efficiency ratio — below 68% confirms structural improvement continues; above 70% signals IS comp discipline is unwinding as revenue normalizes. Third, buyback pace — above $1.75 billion per quarter signals management is deploying the SCB capital buffer; below $1.0 billion despite capacity suggests management is conserving capital, possibly in anticipation of adverse sweep litigation outcomes.

What is the asymmetric rate sensitivity in Wealth Management?

The Q3 2025 10-Q disclosed WM NII sensitivity: a 100bp rate increase adds $188 million, but a 100bp rate decrease costs $228 million. This asymmetry means rate cuts hurt WM NII more than equivalent rate hikes help it. The $40 million gap per 100bp reflects the convexity of the deposit pricing structure — in a falling rate environment, asset yields compress faster than MS can reduce deposit costs (which are already near floors in some sweep products). Combined with the sweep deposit litigation, this creates a double headwind: rates falling reduces NII, and lawsuits could force higher deposit costs on the remaining spread.

Methodology

Data Sources

• Primary: Morgan Stanley FY 2025 10-K filing (filed 2026-02-19 with the SEC), Q3 2025 10-Q (filed 2025-11-03), and Q4 2025 8-K earnings release. All figures tagged [FILING] sourced from MD&A, segment footnotes, deposit footnotes, and capital management sections via MetricDuck's filing text extraction pipeline.
• Secondary: MetricDuck metrics pipeline for XBRL-sourced financial data. Pipeline data was cross-verified against filing text; critical corrections applied for ROTCE (preferred stock mapping), provision for loan losses (NII concept mismatch), and segment-level breakdowns where pipeline aggregation diverged from filing presentation.
• Peer data: Goldman Sachs (FY 2025 10-K), Bank of America (FY 2024/TTM Q3 2025), and Wells Fargo (FY 2025) data from MetricDuck pipeline extractions and filing research. BAC data is approximate due to different reporting period. Peer metrics are directionally correct but have not undergone the same filing-level verification applied to Morgan Stanley.
• Derived calculations: All derived figures computed from filing-sourced inputs with formulas documented inline as MDX comments. The sweep deposit scenario analysis uses the $139.1 billion Q3 2025 deposit balance, not the FY year-end figure (which was not separately disclosed in the 10-K at this granularity).

Limitations

1. BAC data is FY 2024 / TTM Q3 2025, not FY 2025. Direct peer comparison is approximate due to different reporting periods.
2. Sweep deposit balance ($139.1B) is from Q3 2025 10-Q, not the FY 2025 10-K year-end figure. The year-end balance may differ.
3. CET1 ratios for BAC and WFC are estimated (~11-12%) from pipeline data and public disclosures, not verified against their respective 10-K filings.
4. Excess CET1 capital (~$17.9B) uses estimated RWA (~$560B). The actual RWA figure from the 10-K may differ, affecting the precision of the excess capital estimate.
5. Private credit exposure is not separately disclosed in the filing, limiting assessment of the IS balance sheet expansion's risk composition.
6. No forward guidance on NII or sweep rates. Morgan Stanley provides no quantitative forward guidance on NII trajectory, sweep deposit pricing, or litigation reserves, limiting the precision of scenario analysis.

Disclaimer:

This analysis is for informational purposes only and does not constitute investment advice. The author does not hold positions in MS, GS, BAC, or WFC. Past performance and current metrics do not guarantee future results. All data is derived from public SEC filings and may contain errors or omissions from the automated extraction process.