Ameriprise Financial, Inc. (NYSE: AMP) — The Fee Engine Won; the Easy Re-Rating Did Too

Prepared as independent institutional-style research. Update report date: 2026-08-21. Price reference: $555.59 (2026-08-21 close). This report updates the 2026-06-21 work through Ameriprise’s Q2-2026 results and Form 10-Q. All figures are USD unless noted.

⚡ Analyst.s Take

This block is The following is the analyst.s subjective opinion. It is not investment advice. The institutional body below takes no position and carries no price target; that discipline applies everywhere except this fenced block.

Verdict: HOLD / accumulate only on material weakness. Not-a-short. Fair-value zone ~$540–650; preferred accumulation zone ~$480–515. Medium conviction. At $555.59, AMP trades at 12.9x trailing adjusted operating EPS of ~$43.01 and roughly 12.3–12.5x a ~$44.5–45.0 FY2026 run-rate. That remains cheaper than most scaled wealth peers, but it is no longer the unusually asymmetric ~$467 setup from June. The stock gained 18.9% while the operating evidence improved; much of the classification discount has narrowed before the hard part of the rate cycle is complete.

The prior thesis passed its first serious operating test. H1 adjusted operating EPS rose 20% to $22.33 even though AWM net investment income fell about 9%. In Q2, consolidated adjusted operating revenue rose 13%, adjusted operating earnings rose 14%, and adjusted operating EPS rose 22%; the six-point gap between earnings and per-share growth reflects a 6% lower diluted share count. More important, AWM advisory fees grew 23%, dwarfing an 8% decline in AWM NII. AWM pretax earnings rose 16% at a 28.9% margin, and total wrap flows accelerated to $6.9B. This is precisely the fee-engine-out-earns-spread-compression result the June report required. The bear test — EPS growth through easing while sweep economics stabilize — is now partly falsified in the bull’s favor, though not finally resolved because equity markets were strong and a simultaneous market drawdown plus further rate cuts remains untested.

The quality is real, but the headline flow number needs skepticism. Total AWM client flows slowed 29% year over year to $3.1B, while wrap flows rose 29% to $6.9B. The gap partly reflects Comerica advisor departures and tax-season cash activity; management will not disclose the precise organic/Comerica bridge, and Comerica’s roughly $19B is expected to be gone by the end of September. Huntington’s roughly $28B and 260 advisors should onboard in Q4 2026 or early Q1 2027, with full economics expected in Q4. That creates an attractive net channel swap, but it also means the clean organic growth rate is temporarily obscured. The previous bull falsification test therefore passes on fee growth but remains unscored on truly comparable net flows.

Three sources of earnings differ sharply in durability. AWM is now visibly a high-quality advice platform: $1.25T of client assets, $731.5B of wrap assets, $1.203M of trailing revenue per advisor, and 79 experienced-advisor additions in Q2. Columbia Threadneedle produced a strong 42.7% net margin and 23% pretax growth, but still lost $6.5B in the quarter and $12.7B in the half; favorable markets and Seligman technology strategies are masking a structurally negative flow franchise. RPS delivered $202M of quarterly pretax earnings, within its run-rate objective, while variable-annuity net flows stayed negative. I would underwrite AWM as the compounder, Columbia as a high-margin runoff-with-options, and RPS as a de-risked capital generator — not assign the same multiple to all three.

Capital allocation remains the strongest part of the case. AMP returned $932M in Q2, or 91% of adjusted operating earnings, including $774M of repurchases at an average price near $459. H1 repurchases were $1.65B at about $471 per share. The company therefore bought materially below today’s price and cut diluted shares 6% year over year. Excess capital was $2.1B and holding-company liquidity $2.8B. This is good execution, but it also means future repurchases at $556 create less value than the same dollars deployed near $460. The company had $1.1B left under the authorization at June 30, so price discipline now matters more than authorization size.

The rate bridge is better than the market feared, not solved. AWM NII was flat sequentially at $455M despite lower certificate balances; bank assets grew 6%, the portfolio yielded 4.67%, and new Q2 purchases yielded about 5.2% with roughly 4.5-year duration. Management has deliberately termed out the portfolio, with only about 6% floating-rate exposure. That reduces immediate Fed sensitivity. Yet total AWM cash balances fell 2% year over year and certificate balances fell 25%; off-balance-sheet brokerage cash revenue fell 16%. AMP is less dependent on sweep than LPLA, SCHW, or RJF, but cash sorting and rate repricing remain headwinds rather than catalysts.

The market framing has migrated from abandoned quality toward quality at a fair price. The stock sits above its 21-, 50-, and 200-day exponential moving averages ($549, $523, and $489) and only 1.5% below the January-2025 all-time high near $564. Unlike June, this is no longer a neglected value setup with an obvious re-rating runway. The clean current metric is earnings: market capitalization is approximately $49.1B using 88.33M shares outstanding, while enterprise value is analytically misleading for a regulated financial with client deposits and insurance liabilities. AZI’s directional screen places the composite at the 70.9th percentile and P/E at the 52.6th after peer control, but the underlying AMP per-share record conflicts with the filing. I treat those ranks as directional evidence of re-rating, not valuation facts.

What changes the call. A bullish flip requires two additional quarters of mid-teens AWM fee-led earnings growth, comparable organic wrap flows at least in the mid-single digits, and Huntington retention/economics landing without recruiting-cost leakage; at that point a mid-teens wealth multiple becomes more defensible. A bearish flip is two consecutive quarters in which advisory-fee growth falls below spread compression, total flows remain weak after Comerica has exited, or capital return continues at premium prices while organic growth slips. Tag: the fee engine won; the easy re-rating did too.

🔄 Changes Since the 2026-06-21 Report

Prior tests scored literally. The June bear thesis required adjusted operating EPS to grow through easing while sweep NII stabilized. H1 EPS +20% and Q2 AWM NII flat sequentially provide a strong first pass, so that test is partly falsified. The June bull thesis required fees to outgrow NII compression and flows to remain competitive. The first condition passed emphatically; the second is temporarily indeterminate because $6.9B of wrap inflows coexist with $3.1B of total flows and an undisclosed Comerica bridge.

📈 Stock Price Action — Five-Year Event Map

AMP rose from a five-year low near $210.6 in July 2022 to an all-time high near $564.1 in January 2025. It closed at $555.59 on 2026-08-21, only about 1.5% below that high and up 18.9% from the June report’s $467.43 reference. The current tape is constructive: spot is above the 21-day ($549.18), 50-day ($523.19), and 200-day ($489.13) EMAs; estimated market beta is 1.03. The price is fact; drivers are interpretation.

Cycle narrative. AMP first de-rated as a high-beta asset gatherer in 2022, then re-rated as Fed hikes monetized cash. The 2023 banking shock briefly challenged that narrative; the 2023–25 advance then stacked fee-AUM growth, spread income, and a shrinking denominator. The 2025 drawdown priced a peak in both NII and markets. The newest leg is different: Q2 2026 showed that fees can outgrow spread compression, validating a more durable earnings mix. However, the stock has returned almost to its prior peak before Columbia outflows, post-Comerica organic flows, and a full easing cycle have been resolved. (Prices from the five-year daily series; event attribution cross-checked against filing and earnings dates.)

1. Executive Summary

Ameriprise Financial is a wealth-management-led financial group whose legacy insurance and active-management sleeves still determine how the market prices the whole. The business has spent two decades shifting from manufactured insurance products toward recurring advice fees. FY2025 established the base: $18.48B of net revenue, $3.56B of GAAP net income, $3.86B of adjusted operating earnings, $39.29 of adjusted operating EPS, and 53.2% adjusted operating ROE excluding AOCI. Q2 2026 strengthened the mix argument. Adjusted operating net revenue rose 13% to $4.90B, earnings rose 14% to $1.03B, EPS rose 22% to $11.07, and adjusted operating ROE excluding AOCI reached 55%. Q2 earnings release, 2026-07-23.

The franchise has three unequal legs. Advice & Wealth Management is the core, with $1.25T of client assets, $731.5B of wrap assets, and $3.25B of quarterly net revenue. Q2 advisory fees rose 23% while AWM NII fell 8%, allowing pretax earnings to rise 16% at a 28.9% margin. Asset Management, principally Columbia Threadneedle, had $759B of AUM/A and a 42.7% net margin, but lost another $6.5B in Q2. Retirement & Protection Solutions produced $202M of pretax earnings, within its quarterly target, while its variable-annuity balances remain in net runoff. The earnings mix therefore combines a high-quality distribution franchise, a profitable but shrinking active manager, and a de-risked insurance cash generator.

The moat is concentrated rather than company-wide. AWM has a Greenwald-style combination of customer captivity and platform scale: long-lived advisor relationships, tax and transfer friction, integrated planning and banking, and fixed technology/compliance costs spread across more than 10,000 advisors. The evidence is economic rather than promotional: trailing revenue per advisor reached $1.203M, up 12%; wrap assets rose 19%; and the company added 79 experienced advisors in Q2. The limiting factor is that advisors themselves are mobile. LPL, Raymond James, Morgan Stanley, and others can bid for books with rich transition packages, so AMP must keep reinvesting to preserve the relationship asset.

The principal update is that the fee engine has begun to prove it can offset the rate cycle. H1 adjusted operating EPS rose 20% while AWM NII declined about 9%. In Q2, AWM NII was flat sequentially at $455M, as bank assets grew and the portfolio’s 4.67% gross yield benefited from termed-out reinvestment. This does not make earnings rate-proof: certificate balances fell 25% year over year, off-balance-sheet cash revenue fell 16%, and only a strong equity market has tested the fee hedge so far. It does show that AMP is less cash-dependent than sweep-heavy peers.

Organic growth requires a cleaner read. Q2 wrap flows rose 29% to $6.9B, but total client flows fell 29% to $3.1B because of advisor departures, including Comerica. Management expects the approximately $19B Comerica relationship to be fully out by the end of Q3, while roughly $28B and 260 Huntington advisors should arrive in Q4 2026 or early Q1 2027. Because the company did not quantify the exact bridge, neither the headline flow deceleration nor the wrap acceleration is a clean measure of steady-state organic growth.

Financial quality remains high. The Q2 adjusted-to-GAAP difference was driven primarily by market effects on long-duration products, a recurring source of accounting noise rather than a cash impairment. Diluted shares fell 6% year over year; AMP returned $932M, or 91% of operating earnings, in Q2 and repurchased $774M of stock at an average around $459. Excess capital was $2.1B and holding-company liquidity $2.8B. The balance sheet and RiverSource capital buffer make permanent impairment unlikely; the relevant risk is a large cyclical decline in fees, spread, and multiple at the same time.

At the August 21 close, AMP’s approximately $49.1B equity value equals about 12.9x trailing adjusted operating EPS of $43.01. That is still a discount to scaled wealth platforms, but it is materially less compelling than at the June report’s $467.43 reference. The institutional question is no longer whether the fee engine exists; Q2 answered that. It is whether comparable organic flows and fee-led earnings remain durable after Comerica, through Huntington onboarding, and through a less favorable market. This institutional memo states no investment conclusion.

2. Business Overview

Q2-2026 operating snapshot. AWM produced $3.246B of net revenue and $939M of pretax adjusted operating earnings; Asset Management produced $947M and $274M; RPS produced $975M and $202M; Corporate and closed blocks lost $81M. AWM therefore contributed about two-thirds of positive operating-segment pretax earnings in the quarter, consistent with the FY2025 mix. Management’s description of “core distribution earnings” growing in the low-30% range is directionally helpful but not auditable: on the earnings call the CEO suggested core could be about 70% of AWM earnings while the CFO resisted a precise split. The report therefore uses the disclosed fee, spread, expense, and segment lines rather than treating “core” as a measured KPI. Q2 earnings release, 2026-07-23.

What AMP is. Ameriprise Financial provides financial planning, advice, brokerage, banking, asset management, annuities and insurance to ~3.5M+ retail clients and institutional clients, primarily in the US with a meaningful EMEA/Asia asset-management presence. It is organized into four segments: Advice & Wealth Management, Asset Management, Retirement & Protection Solutions, and Corporate & Other. The firm is a savings-and-loan holding company (and financial holding company) regulated by the Federal Reserve, owns a federal savings bank (Ameriprise Bank, FSB) and the Ameriprise Certificate Company, and manufactures annuities/insurance through RiverSource Life. (Fact — FY2025 10-K, Item 1.)

Consolidated revenue mix (FY2025, GAAP). Total net revenues $18,480M (+7%). The build: Management & financial advice fees $11,109M (≈60% of total revenue, +10%); Distribution fees $2,117M (+3%); Net investment income $3,570M (−2%); Premiums/policy/contract charges $1,587M (+2%); Other $528M. (Fact — 10-K MD&A.) The single most important structural fact: ~60% of revenue is asset-/advice-based management fees that scale with AUM — recurring, market- and flow-sensitive — with net investment income (~19% of revenue) the rate-sensitive leg. This is a fee-gathering machine wrapped around a shrinking insurance/spread book.

Segment pretax adjusted operating earnings (FY2025). AWM $3,411M (+6%), Asset Management $1,016M (+10%), RPS $846M (+17%); Corporate & Other a loss. The three operating segments sum to $5,273M, mixing AWM 64.7% / AM 19.3% / RPS 16.0%. (Fact — 10-K segment tables.) AWM is now roughly two-thirds of operating earnings and the sole growth engine — the visible result of AMP’s two-decade shift away from its annuity-company roots.

How AWM actually makes money (the crown jewel, quantified). AWM total net revenues $11,741M (+9%); pretax adjusted operating margin 29.1% (down 90bps from 30.0%). Three income streams:

• (1) Advisory wrap fees — the core. Ending advisory wrap assets $664.4B; total wrap $670.4B (+17%) on net inflows $30.9B plus $65.6B market appreciation; average advisory wrap +15% to $606.2B. This drives the $7,371M management/advice-fee line (+14%). (Fact — 10-K wrap table.)
• (2) Distribution fees $2,520M (+3%) — point-of-sale, asset-based 12b-1, brokerage transactional, and third-party sweep-placement fees. Note +$119M from transactional activity but −$52M brokerage cash revenue — the cash-cow erosion is already visible at the margin.
• (3) Net investment income / cash spread $1,956M (−11%) — the rate-sensitive piece, earned on Ameriprise Bank and Certificate Company assets. Banking/deposit interest expense ($431M, −35%) is netted here.

So AWM revenue is roughly 63% recurring advice/management fees + 21% distribution + 16% NII — the spread leg is the swing factor on rates but is a minority of AWM revenue, and (per management) a smaller share of earnings than at cash-dependent peers like Schwab, where net interest is ~49% of total revenue.

The cash / bank engine. Total AWM client cash $87.0B: Ameriprise Bank deposits $23.7B (+6%), Certificate balances $8.2B (−$3.0B, eight straight quarters of outflow), broker-dealer on-balance-sheet $1.9B, off-balance-sheet sweep $5.1B, and $48.1B in third-party cash products (MMFs/brokered CDs). Cash-sweep balances ~$29.4B (Q1-26). The economics: AMP sweeps client cash into its own bank, paying clients ~0.28% on bank deposits while investing the $20.5B AFS portfolio at a ~4.6% yield (≈4yr duration, mostly AA+ structured, only 9% floating-rate after deliberate term-out). (Fact — 10-K client-cash table + Q1-26 call.) This is the same spread economics as Schwab/Morgan Stanley but far smaller as a share of earnings, and deliberately stabilized.

The advisor force. >10,000 financial advisors across four affiliation channels: Ameriprise Franchise Group (independent franchisees — the largest), Ameriprise Advisor Group (W-2 employees), Ameriprise Personal Wealth Group (centralized virtual advice / succession vehicle), and Ameriprise Financial Institutions Group (AFIG — advisors based in third-party banks/credit unions). Adjusted operating net revenue per advisor reached $1,122K in FY2025 (+8%) and a record $1.2M run-rate in Q1-26 (+10%). Client satisfaction 4.9/5; J.D. Power 2026 ranked 3rd of 23. In Q1-26 AMP signed Huntington Bank to AFIG (~260 advisors, ~$28B, onboarding Q4-26) while losing Comerica (early-terminated after Fifth Third’s acquisition; $25M one-time make-whole; ~$18B leaving through Q3-26). (Fact — Q1-26 call.)

Asset Management (Columbia Threadneedle). Total managed assets $678.1B; segment net revenues $3,621M (+3%); pretax $1,016M (+10%) on equity appreciation, +$55M performance fees, and cost cuts; margin reached 44% in Q1-26 (above the 35–39% target). The structural problem: net outflows of −$31.7B in FY2025 (Global Retail −$25.3B, Global Institutional −$22.1B), worse than −$20.4B in 2024 — only market beta keeps total AUM rising. Mitigants (ETF platform >$10B, top-10 SMA provider, alternatives) are right-direction but sub-scale.

Retirement & Protection Solutions (RiverSource). Pretax $846M (+17%; ~$800M/yr run-rate); VA account balances $91.3B; living-benefit rider mix down to 46% from 50% (no new living benefits sold since mid-2022). RiverSource is the exclusive new-annuity manufacturer to AMP advisors; legacy blocks are being reinsured (LTC ~50% ceded to Genworth; fixed annuities ~89% ceded). It is a de-risking, cash-generative captive manufacturer, not a growth story.

Verdict. AMP is a wealth-management-led franchise (~65% of earnings, ~60% of revenue recurring advice/management fees) with two supporting legs — a melting active manager and a de-risked annuity book — that the market weights more heavily than their earnings contribution warrants. The business has been deliberately and successfully repositioned toward the highest-quality fee segment.

3. Industry Dynamics

Current capital-cycle evidence. The scarce input is the productive advisor, and competing platforms are increasingly capitalizing the cost of winning one. AMP’s adviser loans rose about 15% in H1, while LPLA’s advisor-loan balance rose more than 50% year over year in its latest update. AMP management said some industry packages imply cash paybacks of up to eight years and argued that matching them would add little incremental profit. That discipline protects return on capital but can sacrifice gross recruiting. The correct industry verdict is therefore structurally attractive on the demand side but worsening on the supply side: household wealth and fee-based advice grow, while increasingly expensive advisor mobility transfers part of the value pool to labor.

AMP straddles three industries of sharply different structural quality, and its earnings are concentrated in the best one.

US wealth management / advice (~65% of earnings) — structurally GOOD. Secular demand is rising on both sides of the demographic ledger: baby-boomer retirement decumulation and younger-generation wealth accumulation. The multi-decade migration from commission/transactional to fee-based advisory continues to expand the recurring-revenue profit pool. The economics favor scale players with a fixed advice/tech platform and high client switching costs. AMP targets the mass-affluent ($500K–$5M investable) plus HNW. (Fact — 10-K Item 1; Interpretation.)

Active asset management (~19% of earnings) — structurally BAD. Active mutual funds saw ~$640B of outflows in 2025 (~$4T cumulative over the decade); passively-managed assets (~$19.1T) now exceed active (~$16.2T); fees are compressing faster than in prior cycles. Survivors must reach BlackRock-scale, pivot hard to active ETFs/SMAs/private markets, or be acquired. (Fact — Morningstar/McKinsey 2025.) Columbia Threadneedle is squarely caught: net outflows every year, sub-scale globally, pivoting late. This is a melting (slowly) ice cube whose value to AMP is mostly captive manufacturing + cash harvesting, not a standalone franchise.

Insurance/annuity (~16% of earnings) — mature, capital-intensive, but a de-risked slice. AMP runs a shrinking, reinsured, cash-generative book rather than chasing new long-tail liabilities. Industry pressures (rates, capital intensity, longevity assumptions) apply, but AMP’s deliberate de-risking neutralizes most of the tail.

Verdict. Mixed-to-good, bifurcated by segment, but favorable on a weighted basis because AMP’s earnings are concentrated in the structurally good wealth/advice segment. The active-AM industry is structurally bad (a multiple cap on ~19% of earnings); the cash-cow/NII risk that haunts the discount brokers is materially less acute for AMP. The principal industry headwind is the rising cost of advisor talent.

4. Competitive Position

Q2 moat scorecard: passed, with a market-assistance caveat. Management and advice fees increased $353M year over year while gross AWM NII fell $41M — an 8.6x dollar offset. Yet average advisory wrap assets rose 21% and AMP’s average weighted equity index rose about 27%; roughly 90% of sequential wrap-asset growth came from market appreciation and other effects rather than net flows. AWM margin was flat at 28.9%, not expanding. The evidence supports a narrow, maintenance-intensive moat capable of monetizing rising assets, but it does not yet prove cycle-independent operating leverage.

The moat lives in AWM, and it is real. In Greenwald’s taxonomy, AWM possesses a customer-captivity (switching-cost/habit) advantage layered with modest economies of scale on the advice/tech platform — the most durable combination type. It is not a network effect and not a low-cost-production moat. The captivity shows up in financial outcomes, which is the only test that matters:

• Rising productivity: adjusted operating net revenue per advisor +8–10%/yr to $1.2M — clients consolidate assets with their advisor over time (wrap +17%, net inflows $30.9B even in a cautious year).
• Sticky multi-product relationships: planning fees, wrap, banking (HELOCs/checking/pledged lending), and annuities/insurance bundled around one advisor relationship raise the cost of leaving.
• High satisfaction/retention: 4.9/5 client satisfaction, J.D. Power 3rd of 23 — the advisor is the switching-cost asset, and AMP’s integrated tech/CRM/advice platform retains advisors and their books.
• Returns far above cost of capital: even on the conservative ~13–14% reported GAAP ROE, returns clear the hurdle comfortably; on the operating-economics basis (adjusted operating ROE-ex-AOCI 53.2%) AWM is an extraordinarily high-return, capital-light fee business.

Pressure test — what limits the moat. AWM is partly a distribution business, not a pure franchise. The advisor — not the Ameriprise brand — owns much of the client relationship, and advisors are mobile: LPL and RJF recruit them aggressively, and AMP’s own Q1-26 flows were dented by departures. Captive manufacturing (Columbia funds + RiverSource annuities pushed through AMP advisors) adds economics but also conflict-of-interest/fiduciary risk. The captivity is real at both the client-advisor and the platform/scale level, but the firm must keep paying (comp, tech) to retain advisors — this is a moat that requires continuous reinvestment, not a toll bridge.

Columbia Threadneedle — no durable moat. A melting active manager: persistent net outflows, fee compression, sub-scale globally. Decent fund performance (>70% of funds above peer median on 1/3/5yr) and a late ETF/SMA/alts pivot are mitigants, not a moat. It has neither scale (vs BlackRock/Vanguard) nor genuine captivity (institutional money repositions to passive at will).

RPS/RiverSource — narrow captive-channel advantage. Exclusive new-annuity manufacturing to AMP’s own advisors gives a guaranteed, low-acquisition-cost distribution channel — a real edge versus insurers fighting for shelf space — but the underlying annuity/life business is a commoditizing, capital-intensive industry. The moat is the channel, not the product.

Head-to-head. Versus LPL (~32,000 advisors, pure custodial/platform, recruiting machine): AMP has far fewer advisors but higher productivity and a more integrated, branded, higher-margin advice model plus captive manufacturing and a bank — LPL wins the headcount war, AMP wins the per-advisor-economics war. Versus Raymond James (the closest analog — branded, advice-led, multi-channel, a bank/NII leg): AMP’s productivity and capital-return discipline differentiate. Versus Morgan Stanley Wealth ($9.3T assets, HNW/UHNW + workplace funnel + investment bank): the closest moat analog but at far larger scale and a richer client tier — AMP plays mass-affluent, MS plays up-market. Versus Schwab (scale/cost moat, RIA custody + discount brokerage, NII ~49% of revenue): a different, far more cash-cow-dependent model — AMP’s advice-led, lower-cash-dependence mix is a relative defensive advantage in the AI-cash-cow debate.

Verdict. A durable advantage in the segment that matters (AWM, ~65% of earnings) — a real customer-captivity-plus-scale moat evidenced by rising productivity, sticky relationships, high retention, and through-cycle returns well above cost of capital. But it is maintenance-intensive (mobile advisors in an inflating recruiting war), partly distribution rather than pure franchise, and diluted by a no-moat Columbia Threadneedle and a narrow-captive RPS. Superior, more durable, and less cash-cyclical than the discount brokers — but not an unassailable toll bridge.

5. Growth History and Forward Opportunities

Q2-2026 growth decomposition. Wrap net flows of $6.935B on beginning wrap assets of $664.2B equal roughly 4.18% annualized, just inside the prior report’s 4–5% hurdle. Total client flows of $3.053B on $1.149T of beginning client assets equal only 1.06% annualized, which fails it. Both numbers are valid but answer different questions: wrap flow measures movement into fee-bearing advice accounts; total client flow includes brokerage and cash movement and was depressed by Comerica-linked departures. Management did not provide an ex-Comerica bridge. Huntington is a channel replacement — approximately $28B arriving against $19B leaving — rather than evidence of purely organic share gain.

The strongest growth proof is in revenue, not ending assets. Q2 advisory fees rose 23%, planning fees 5%, distribution fees 13%, and AWM net revenue 16%; adjusted operating revenue per advisor reached $1.203M, up 12%. But the average equity index and wrap asset base rose faster than the fee line, and AWM margin stayed flat. The quarter therefore showed powerful market-assisted monetization plus positive organic wrap growth, not a clean new structural growth rate.

Historical decomposition (5yr). Total net revenues grew $11.9B (2020) → $18.48B (2025), a ~9.2% CAGR — matching management’s stated “9% compounded annual revenue growth.” GAAP net income $1.53B → $3.56B. Adjusted operating EPS reached $39.29 in FY2025, a ~20% five-year CAGR (“20% compounded annual EPS growth,” per the CFO). (Fact — 10-K MD&A; Q1-26 call.) The ~11-point wedge between ~9% revenue growth and ~20% EPS growth is the buyback engine: diluted average shares fell 123.8M (2020) → 98.2M (2025), ~21%. Roughly half of per-share earnings growth is buyback-driven, half operating — AMP is a capital-return compounder as much as an organic-growth story.

Organic vs acquired. The only material acquisition in the window is BMO’s EMEA asset-management business (closed Nov-2021, ~$845M, +~$124B AUM to Columbia Threadneedle). Everything else is organic plus bolt-on/reinsurance — management self-describes the model as “organic growth, built not bought.” (Fact — 10-K; Q1-26 call.) Tellingly, the one acquisition reinforced the weakest segment (active AM, now in outflows). AWM growth is essentially 100% organic.

The AWM growth math (the engine). Management & financial advice fees $6,492M → $7,371M, +14%, on average wrap assets +15%. Net flows $30.9B on a ~$570B starting wrap base ≈ ~5.4% organic wrap growth; the rest of the +17% wrap jump is market beta. Crucially, the advice-fee flywheel (+$879M) outran the NII decline (−$239M) by ~3.7:1 in 2025 — the central proof point that fee growth offsets cash-spread roll-off. Productivity reached a record $1.2M/advisor (+10%). (Fact — 10-K AWM table; Q1-26 call.)

The NII headwind. AWM NII fell 11% (−$239M) in 2025; the average rate paid on bank deposits dropped to 0.28% (from 0.44%); certificate balances fell $3.0B (eight straight quarters). This is the principal cyclical headwind — the 2022–24 high-rate vintage inflated AWM earnings via the cash spread, and Fed cuts now compress it. Mitigants AMP has deliberately built: a termed-out bank book (9% floating, ~4yr duration locking ~4.6–5.0% yields), a full bank product suite (HELOCs/checking/pledged lending) growing NII via volume not just spread, and ~$48B of third-party MMF cash management frames as redeployment optionality into fee-earning wrap as rates fall.

The AM drag. Total AUM net outflows −$31.7B in FY2025 (worse than −$20.4B in 2024), improving to −$5.9B in Q1-26 on +26% gross retail sales. “Growth” here is margin/efficiency-driven (44% Q1 margin on cost-out and back-office outsourcing), not flow-driven — a managed decline, not a growth contributor.

Forward drivers (ranked). (1) AWM wrap/advice-fee growth — highest-quality, durable; plus the $48B third-party cash redeployment optionality. (2) AFIG bank/credit-union channel — Huntington signed (~260 advisors/$28B, 10-yr, “IRRs very good”); the thesis that banks/credit unions increasingly want wealth management but won’t make the platform investment, so they outsource to AMP — a structural pipeline, albeit lumpy and contestable (partners can be acquired, à la Comerica). (3) Advisor productivity via AI/tech — embedded AI “takes away hours of work per week,” framed as capacity not headcount cuts. (4) Signature Wealth UMA — a fast-ramping mid-2025 wrap launch bringing new money and a fee-rate uplift. (5) Bank product suite — pledged lending/HELOC/checking growing NII via volume. (6) Buyback-driven EPS — ~5%/yr share shrink at ~11–13x continues mechanically under the $4.5B authorization.

Forward headwinds. AM secular outflows; NII/cash-spread compression on cuts; the inflating advisor-recruiting war suppressing net headcount; and a 2026 flow air-pocket between Comerica outflows (Q2–Q3) and Huntington inflows (Q4).

Verdict. High-quality engine, medium-durability growth rate. The durable core (AWM advice-fee growth +14%, rising productivity) is genuinely high quality — recurring, fee-on-AUM, capital-light, organic. But the headline ~20% EPS CAGR is flattered on three counts that will not all repeat: ~half is buyback mechanics; the 2022–24 NII vintage is rolling over; and market appreciation, not flows, supplied the bulk of AUM growth. Net organic flow growth is ~5% in AWM and negative in AM. Forward growth is real but lower and lumpier than the headline implies — normalizing toward low-double-digit.

6. Financial Quality

Current quality bridge. H1-2026 GAAP net income was $2.028B versus adjusted operating earnings of $2.092B. The $64M difference was modest relative to earnings and mainly reflected market impacts on long-duration products, realized investment effects, and tax. H1 diluted shares averaged 93.7M versus 99.4M, so the 13% increase in adjusted operating earnings became 20% EPS growth. Q2 diluted shares were 92.9M; 88.33M common shares were outstanding on July 24. Earnings quality is high, but per-share growth must always be separated into operating growth and denominator shrink.

The balance-sheet read remains conservative. At June 30 management reported $2.1B of excess capital and $2.8B of holding-company liquidity. The AWM bank portfolio’s gross yield was 4.67%; Q2 purchases of roughly $1.1B reportedly yielded around 5.2% with 4.5-year duration, while only about 6% of the book was floating rate. That extends income but adds duration; it reduces near-term cut sensitivity without eliminating mark-to-market or reinvestment risk.

The headline correction: the real ROE, and why equity is tiny. FY2025 GAAP net income $3,563M sits on year-end common equity of only $6,549M (common $3M + APIC $10,377M + retained earnings $27,662M − treasury stock $30,601M + AOCI −$1,305M). Real GAAP ROE on year-end equity ≈ 54%; AMP’s own 10-K reports return on equity excluding AOCI of 49.1% and adjusted operating ROE-ex-AOCI of 53.2% (target: >30%; “best-in-class,” up 60bps, per the proxy). (Fact — 10-K consolidated balance sheet + MD&A ROE reconciliation.)

A critical data note: all current ROIC.ai AMP ratios and multiples are excluded, not only ROE and P/B. The underlying AMP record contains a corrupted per-share denominator and produces mutually inconsistent P/E, P/S, and enterprise-value outputs. Real common equity and filing-derived adjusted earnings are used instead. (Interpretation, reconciled to the filings.)

Equity is small for two structural reasons. (1) Relentless capital return: treasury stock has built to $30,601M (2025) from $18,879M (2020); of 338.1M shares issued, only 91.3M are outstanding (~247M bought back and parked in treasury), so retained earnings ($27.7B) are almost entirely offset by treasury stock ($30.6B). (2) AOCI drag: AOCI swung from +$629M (2020) to −$2,546M at the 2022 rate-hike trough, recovering to −$1,305M (2025) as the AFS bond book’s unrealized losses bleed back. AMP deliberately runs a thin-equity, high-ROE model funded by insurance/bank float and recurring fee cash flow — the thin equity is a choice, and the engine of the per-share compounding.

GAAP vs adjusted operating — a small, legitimate gap. FY2025 GAAP NI $3,563M ($36.28) → adjustments of just −$295M (almost entirely a −$366M non-cash market-impact mark on the runoff variable-annuity/IUL book, partly tax-offset) → adjusted operating earnings $3,858M ($39.29) on 98.2M diluted shares. The 2023 GAAP dip ($25.37/sh; adj op $28.86) was driven by VA hedge/market marks (−$191M), the SEC recordkeeping accrual, severance, and $62M integration — precisely the items the adjusted measure backs out. (Fact — 10-K reconciliation.) This is the rare case where the non-GAAP measure is the more economically representative number, not a flattering one. Cross-check: operating cash flow / net income ran 2.34x in 2025 (1.94x in 2024) — cash generation massively exceeds GAAP NI (insurance/working-capital float), confirming the earnings are real cash, not accrual fiction.

Rate/NII share of earnings (the cyclical kicker, now reversing). AWM net investment income $1,956M (2025) vs $2,195M (2024), −11%; net AWM spread (NII − deposit interest) ≈ $1,525M, roughly flat YoY because deposit costs fell too. That is ~$1.5B of net spread inside the $3.4B AWM segment — a material slug of rate-driven income that ballooned 2022–24 and is now compressing. The bull case requires advice-fee growth (+14%) to keep outrunning NII decay; in 2025 it did (AWM still +6% with NII down). This is the single biggest earnings-quality swing factor: the record ROE is partly a high-rate vintage, partially offsettable by the asset-gathering flywheel.

Does it improve with scale? Yes — in AWM (asset-light, fee-on-AUM, 30%+ incremental margins) and RPS (spread/fee on a de-risked book). Asset Management is the structurally weaker, flow-challenged unit that masks franchise quality. Consolidated adjusted operating margin 26.9% (+40bps).

Balance sheet / liquidity / capital. Recourse corporate (senior-note) debt is ~$3.1–3.3B (the “$5.9B” some screens show includes non-recourse/consolidated-entity debt); net debt is negative at the consolidated level (more cash/investments than debt). RiverSource Life RBC is 523% of Company Action Level ($2,731M total adjusted capital vs $522M CAL) — very strong; excess capital ~$2.1B. Cash & equivalents $10.1B. SBC is modest at $206M (~5% of NI) — low dilution for a financial.

Free cash flow / capital generation. OCF $8,323M (2025) on tiny $162M capex; the headline OCF is inflated by insurance/working-capital float, so the durable distributable figure is the ~$2.5–3.5B/yr of holdco cash, equivalently framed as “~88% of adjusted operating earnings returned.”

Verdict. Elite economics that genuinely improve with scale in the segment that matters. The ~50%+ ROE is real (not a data artifact), the GAAP-to-adjusted gap is small and legitimate, cash generation exceeds GAAP earnings, the balance sheet is a fortress, and dilution is minimal. The one earnings-quality watch item is the rate/NII vintage rolling over — durable, but partly offsettable.

7. Capital Allocation

Q2 update. AMP returned $932M, or 91% of adjusted operating earnings, including $774M of repurchases and $158M of dividends. It repurchased 1.7M shares at an average near $459 in Q2 and 3.3M shares for $1.653B at an average $470.84 in H1. Approximately $1.1B remained under the authorization at June 30, expiring June 30, 2027. This was accretive execution relative to the $555.59 report-date price. The forward question is not whether management will return capital — the record is clear — but whether it will retain the same price discipline after the re-rating.

Capital allocation is, alongside the AWM moat, the strongest part of the AMP story — and the clearest evidence management thinks like owners.

The buyback machine. Shares outstanding fell 116.8M (2020) → 91.3M (2025), −22% in five years (~−26% from 2019). Buybacks (cash-flow statement): $2,030M (2021), $1,978M (2022), $2,127M (2023), $2,448M (2024), $2,907M (2025) — cumulatively ~$12B+ over 2021–25. The Jul-2023 $3.5B authorization was exhausted by Q2-2025; a new $4.5B authorization (Apr-2025) runs through Jun-30-2027. In Q1-26 management explicitly “took advantage of the decline in our PE multiple” to repurchase 1.6M shares. (Fact — 10-K; 8-Ks; Q1-26 call.) Buying back ~5% of shares/yr at ~11–13x earnings while ROE stays ~50% is mechanically accretive — this is the textbook high-ROE capital-return compounder.

The dividend. DPS grew $4.01 (2020) → $6.53 (2025), a ~10% CAGR, raised 8% in 2025 and a further 6% in Q1-26 — yet the payout is only ~16–17% of net income, leaving a long runway. Total capital return is ~$3.4B/yr, ~85–90% of adjusted operating earnings, the 15th consecutive year of returning capital at a “differentiated rate.” (Fact — proxy; 10-K.)

M&A discipline. The one material deal (BMO EMEA, ~$845M, 2021) reinforced the weakest segment; otherwise growth is organic plus small bolt-on/reinsurance (cash-flow “acquisitions of subsidiaries” of $427M/$1,125M/$1,431M in 2023–25 are largely reinsurance/block transactions, not transformative M&A). AMP has not chased dilutive equity-funded deals at high multiples — capital goes to buybacks instead. Positive.

The June-2026 debt raise. An 8-K (filed 2026-06-09; priced 2026-06-04) issued $300M 4.800% senior notes due 2031 + $450M 5.350% notes due 2036 (~$742.5M net, off the Feb-2024 shelf), most likely to refinance the $500M 2.875% notes maturing September-2026. (Fact — 8-K + 424B5.) Routine terming-out at a modest rate step-up; recourse debt stays ~$3.1–3.3B; net debt remains negative. Not balance-sheet stress, not new leverage for buybacks — disciplined refinancing.

Incentive alignment — genuinely strong (a rarity for a large financial). The 2026 proxy’s Annual Incentive scorecard is Financial 70% (Net Revenues 15%, Earnings 25%, Adjusted Operating EPS 20%, Return on Equity-ex-AOCI 20%, Balance Sheet 20%) + Strategic 30%; ROE-ex-AOCI scored 5.0 “Distinguished” in 2025. Long-term PSUs are earned on EPS growth + average ROE + a relative-TSR modifier (0–150% with ±25pp TSR overlay). CEO Cracchiolo’s FY2025 total comp was ~$30.8M; CFO Berman ~$9.4M; say-on-pay support ~89%. (Fact — 2026 DEF 14A.) Tying ~40% of variable pay to ROE and the rest to per-share metrics — with relative-TSR discipline — is materially better aligned than the absolute-revenue/AUM frameworks most banks and asset managers use.

The governance demerit. Cracchiolo is long-tenured Chairman + CEO (CEO since the 2005 spinoff, ~21 years; combined roles) with no publicly named successor, and 2026 brought senior-officer retirements (CIO Davies 6/30/26; AWM’s Sweeney 4/3/26) that thin the bench. Pay quantum is high but performance (53% ROE, record EPS) supports it; the concentration/succession issue is the real flaw.

The insider read. Across the full 60-month corpus of 311 Form 4s, there are zero code-P open-market purchases and 674,490 code-S shares sold for roughly $269M. Recent CEO and CFO sales were option monetization accompanied by essentially unchanged direct stakes, so they are not evidence of a fresh fundamental view. The pattern is mildly negative/neutral: equity compensation creates selling supply, but the data show no conviction buying.

Verdict. An intelligent, shareholder-aligned capital allocator: relentless accretive buybacks, a fast-growing well-covered dividend, M&A discipline, a fortress balance sheet, and a genuinely ROE/per-share-aligned comp structure. The blemishes are governance (combined Chair/CEO, no named successor) and the absence of insider conviction buying — neither thesis-breaking.

8. Changes and Headwinds — Last Two Years

Latest-quarter additions. Q2 resolved one old concern and sharpened three others. Fee-led earnings growth decisively offset spread pressure; however, the total-flow headline weakened, Columbia remained in outflow, and the stock returned almost to its all-time high. Comerica’s accelerated exit and Huntington’s later onboarding create an unusual two-quarter comparison window. Asset Management’s back-office transformation is expected to finish in Q3, which should make the underlying expense base easier to judge. Management’s AI claims — more than 6,000 advisors using tools and potentially large weekly time savings — remain a hypothesis because no audited conversion to revenue, client retention, or cost savings was disclosed.

AFIG channel build / partner churn (the biggest strategic change). Q1-26 signed Huntington Bank (~260 advisors, ~$28B, 10-yr deal, conversion Q4-26) as a takeaway after a >1-year competitive process, while losing Comerica (early-terminated after Fifth Third’s acquisition; $25M one-time make-whole; ~$18B leaving, conversion ~end-Q3-26). (Fact — Q1-26 call.) Net of the two, AMP adds ~$10B+ and 260 advisors — structurally strengthening (AFIG is a validated, growing outsourced-wealth channel) but exposing the channel’s weakness: bank partners get acquired and walk, making AFIG inherently lumpier and lower-switching-cost than the core channels. Net: mild positive with a 2026 flow air-pocket.

RPS de-risking (multi-year, ongoing). No new VA living benefits sold since mid-2022; rider mix down to 46%; LTC ~50% ceded to Genworth, fixed annuities ~89% ceded; RiverSource RBC 523%. (Fact — 10-K.) This strengthens the thesis — AMP is systematically shrinking the insurance tail risk that scares investors about owning a fee-based wealth manager with a legacy annuity book, converting RPS into a de-risked ~$800M/yr cash spigot. (New, modest residual: reinsurer credit exposure.)

Capital structure / buyback continuity. The June-2026 $750M refinancing (above); the Apr-2025 $4.5B buyback authorization; the 6% Q1-26 dividend raise. All strengthen the capital-return continuity.

Regulatory — SEC off-channel-communications settlement (resolved). August-2024: AMP among 26 firms settling SEC recordkeeping charges over unapproved messaging apps; AMP paid $50M (censure + remediation). One-time, industry-wide, ~1.4% of one year’s NI — a governance/compliance demerit, not thesis-changing; this was the “regulatory matter” behind part of the 2023 GAAP dip.

Regulatory/litigation — cash-sweep class actions (AMP is a named defendant). In 2024 AMP was hit with putative class actions including Mehlman & Hultman v. Ameriprise alleging that low-yield sweep economics harmed clients. The August 19, 2025 ruling compelled arbitration for advisory accounts and dismissed the fiduciary-duty count; the surviving brokerage theories are contract, implied covenant, and unjust enrichment. The exposure remains unquantified in the 10-Q. This is a real but contained tail risk because AMP is less sweep-dependent than SCHW/MS and arbitration narrows the class mechanism; it should not be described as a surviving brokerage-fiduciary claim.

Leadership / governance. 2026 senior-officer retirements (CIO Davies, AWM’s Sweeney); Cracchiolo remains Chair+CEO with no named successor. Weakens at the margin — orderly retirements that nonetheless thin the bench amid an unaddressed CEO-succession question.

Verdict. Net mildly strengthens, with two new overhangs. Strengtheners: AFIG validation (Huntington > Comerica net), continued RPS de-risking, disciplined refinancing, sustained buyback/dividend growth. Weakeners: AMP is now confirmed a named cash-sweep defendant (contained but real), and CEO/senior-officer succession is unaddressed. The strategic direction — organic, de-risked, capital-returning, channel-expanding — is reinforced.

9. Risk Analysis

Risk migration since June. Fundamental risk fell modestly because the fee engine passed the first NII-offset test; valuation and market risk rose because the shares gained 18.9% and sit near the all-time high. Flow opacity also rose: the disclosed total and wrap metrics diverge, and management did not quantify Comerica’s exact effect. The most informative near-term risk monitor is therefore post-Comerica total client flow, not consolidated EPS alone.

The dominant risks are cyclical/beta risks stacked on a structurally good franchise — not idiosyncratic-blowup risks (no leverage cliff, no credit cliff, no going-concern issue) — plus two structural/governance overhangs (the recruiting war and succession). The asymmetry: limited downside protection if the market/rate cycle turns (earnings are near a cyclical high), offset by a fortress balance sheet and a relentless buyback that cushions per-share results.

Catastrophic-loss / total-loss assessment. The risk of a permanent total loss is low: no near-term recourse-debt cliff, RiverSource RBC of 523% at year-end, substantial holding-company liquidity, modest SBC, and a fee-based revenue model. Gross debt was about $3.85B after June’s $750M prefunding, and cash is distributed across regulated entities, so “negative net debt” is not an appropriate claim. The realistic downside is a cyclical earnings/multiple drawdown, not impairment.

10. Valuation Discussion — What the Price Now Requires

Use equity-value lenses, not enterprise value. AMP is a regulated savings-and-loan holding company with a bank, certificate funding, separate accounts, and insurance liabilities. Netting those financial assets and liabilities into a conventional industrial-company enterprise value produces nonsense. The useful lenses are price to normalized adjusted operating EPS, price to book interpreted alongside ROE excluding AOCI, segment earnings power, and capital-return yield. The July 24 share count of 88.33M and the August 21 close of $555.59 imply equity value of roughly $49.1B. Q2 Form 10-Q, filed 2026-08-04.

The clean earnings bridge. FY2025 adjusted operating EPS was $39.29. Replacing H1 2025’s $18.61 with H1 2026’s $22.33 produces trailing adjusted operating EPS of about $43.01, so spot is 12.9x. Annualizing H1 mechanically produces $44.66 and a 12.4x multiple, but that is not a forecast: second-half market levels, performance fees, insurance seasonality, and interest-rate reinvestment can move the result. The proper conclusion is a current multiple in the low-13x area on normalized near-term earnings, not the June report’s low-11x observation.

Peer comparison needs like-for-like denominators. The latest in-repository updates put LPLA near 15.7x FY2026 adjusted EPS and 13.4x a FY2027 build, RJF near 14.5x forward earnings, and Schwab at a higher multiple consistent with its custody scale and stronger near-term growth. Morgan Stanley also trades at a premium because its wealth platform is combined with a scaled institutional franchise. AMP remains cheaper, but the discount is now measured in a few turns rather than the extreme spread implied by the June screen. A direct comparison also overstates AMP’s cheapness unless one discounts Columbia’s outflows and RPS’s capital/actuarial complexity.

Embedded expectations at 12.9x. A simple payout-growth identity should not be used literally for a financial company whose repurchases regularly exceed accounting retention, but it is directionally useful. With a low cash-dividend payout and total capital returns around 90% of adjusted earnings, the multiple does not require high retained-growth economics. It does require that normalized EPS avoid a large decline. The price is broadly consistent with mid-single-digit organic earnings growth plus continued share retirement, or low-single-digit operating growth combined with a gradual reduction in the conglomerate discount.

Three expectations are now in the price:

1. Fees keep outrunning spread pressure. Q2 passed this test with advisory fees +23% versus NII -8%. The market cannot assume that pace forever, but it now has evidence that a modest version is durable.
2. The Comerica/Huntington transition is economically positive. Approximately $19B leaves and $28B arrives, but the timing, retention, and exact revenue contribution remain undisclosed. A clean net win is implied; meaningful leakage would disappoint.
3. Capital return stays accretive. The stock is almost $100 above AMP’s Q2 average repurchase price. The same payout rate produces less share shrink at today’s valuation, so the next phase needs more operating growth and less denominator help.

Reverse-engineering the growth burden. Starting from a normalized adjusted operating EPS base in the mid-$40s, a five-year outcome can be decomposed into AWM fee growth, spread normalization, Columbia flows/margin, RPS stability, and share count. If AWM operating earnings compound high single digits, the other operating sleeves are flat to slightly down, and shares fall roughly 3% annually, consolidated EPS can plausibly compound high single to low double digits. If equity markets are flat, AWM organic growth slips to low single digits, NII continues to fall, and buybacks slow because the stock is dearer, EPS growth can fall toward low single digits. The current multiple discounts the latter more than the former, but no longer offers a large cushion against it.

These are explicit assumption sets, not targets. The current 12.9x trailing adjusted multiple corresponds to roughly 3–4% perpetual EPS growth under an 85–90% total capital-return ratio and a 10–10.5% cost of equity. A third-party forward multiple around 10.8x implies approximately $51.7 of forward EPS — about 20% above trailing adjusted EPS — which shows that near-term consensus expectations are materially stronger than the terminal growth embedded in the current price.

Capital-cycle read. Wealth-management economics attract capital in the form of richer advisor transition packages, better technology, and higher cash yields. LPLA’s advisor loans and transition assistance have expanded materially, while RJF has been winning displaced advisors. AMP’s refusal to match deals with cash paybacks as long as eight years protects returns, but it can reduce headline recruiting. That is exactly the Marathon tension: a good industry can become a poor investment if competitive capital floods the scarce input. AMP’s superior per-advisor productivity is useful only if it can preserve flows without paying away the economics.

Segment-sum logic without false precision. The prior report’s illustrative sum of parts used a mid-teens multiple for AWM, a low-double-digit multiple for Asset Management, and a high-single-digit multiple for RPS. Q2 raises the earnings base for AWM and AM, but the stock has moved close to the prior arithmetic result. The useful conclusion is not a single calculated number; it is that the market now gives partial credit to the wealth sleeve while still discounting the other two. Further re-rating requires evidence that AWM is both organically growing and less cyclical than the consolidated label suggests.

Sensitivity hierarchy. The largest valuation variable is not one turn of P/E; it is the interaction of market levels and AWM organic growth. The current 10-K estimates that a 10% equity decline would reduce adjusted pretax fees by about $410M over the next twelve months; the broader GAAP sensitivity is about $788M before management action. The second variable is the cash-book reinvestment rate and balance mix. The third is the rate of share retirement. Columbia flow improvement matters, but at approximately one-fifth of operating-segment earnings it cannot independently rescue a weak AWM outcome.

Data-quality note. Current third-party AMP valuation outputs contained internally inconsistent EPS, P/E, P/S, and enterprise-value records. AZI’s peer-controlled percentile screen — composite 70.9th, P/E 52.6th, P/B 74.6th, P/S 85.6th — is useful only directionally because the raw AMP per-share denominator does not reconcile to filings. Filing-derived adjusted EPS, the actual share count, and current price take precedence. That produces a more defensible, if less cosmetically precise, valuation read.

11. Variant Perception

Consensus belief. AMP is a well-run diversified financial whose exceptional per-share growth is partly a high-rate and bull-market artifact. Cash spread rolls down as rates fall, Columbia Threadneedle remains in secular outflow, and repurchases create a flattering gap between earnings and EPS. A low-teens multiple is therefore appropriate even if AWM is a good franchise.

Strongest bull case. The company is increasingly a high-return wealth platform whose minority active-management and insurance sleeves obscure the quality of the whole. Q2 offers the cleanest recent evidence: advisory fees rose $353M while gross AWM NII fell only $41M; AWM pretax earnings rose 16%; and H1 adjusted operating EPS rose 20%. AWM needs neither the scale of Schwab nor LPL’s recruiting aggression to create value if it keeps increasing revenue per advisor and returning most earnings at sensible prices. The bull is right if post-Comerica organic wrap growth remains at least mid-single digit, Huntington retains its asset base, and AWM margins hold around the high-20s while rates normalize.

Strongest bear case. Q2 was a market-assisted peak disguised as structural growth. Average wrap assets rose 21% and the weighted equity index about 27%, yet AWM margin did not expand. Approximately 90% of the sequential increase in wrap assets came from markets and other effects. Total-client flow annualized near 1%, Columbia lost another $6.5B, certificate balances fell 25%, and the stock moved back near its high. The bear is right if flat or lower markets expose weak organic gathering, falling cash balances continue to pressure spread, and repurchases slow as valuation rises.

What is genuinely variant now. The market no longer ignores the fee engine; the re-rating since June shows that. The remaining variant is more specific: AMP may be able to compound EPS at a high-single-digit rate with only modest market appreciation because advice growth, termed-out bank income, and share shrink can work together. That claim has not been proven through a flat-market easing cycle. It should be judged against flows and revenue per advisor, not management’s broad “core” characterization.

Positioning read. At $555.59, AMP is above its 21-, 50-, and 200-day EMAs and only about 1.5% below its historical high. Price-series beta is approximately 1.03; the broader factor regression carries Market 1.13, Dividend 0.70, Financials 0.65, Growth -0.49, and zero Momentum loading. This is neither a falling knife nor the abandoned relative-value setup described in June. It is a recovered quality name benefiting from a favorable Financials/Value/Dividend regime, not generic momentum. AZI’s directional valuation screen also moved higher, but its raw AMP per-share record conflicts with the filings, so absolute percentiles are downgraded.

The five assumptions that matter most:

1. Post-Comerica comparable total and wrap flows.
2. Huntington advisor and asset retention, plus the timing of full economics.
3. AWM advisory-fee growth versus net spread compression under the actual rate path.
4. The amount of share retirement achievable at a higher stock price.
5. Columbia’s institutional outflows and whether its high margin survives weaker markets.

12. Fact vs. Interpretation Table

13. Open Questions

1. What were Q2 total-client and wrap flows excluding all Comerica-linked departures, and what is the steady-state rate after September?
2. What proportion of Huntington’s approximately $28B and 260 advisors will be retained, when will transfers complete, and what incremental revenue and expense will be recognized?
3. Can management provide a stable, reconcilable definition of AWM “core distribution” earnings versus cash earnings?
4. How quickly do bank reinvestment yields and certificate runoff change AWM NII under the actual easing path?
5. Does Columbia’s 42.7% margin hold if equity markets are flat and institutional redemptions remain around current levels?
6. How much capital will be returned after the remaining authorization is used, and how price-sensitive will repurchases be near historical highs?
7. What is the quantified reserve or reasonably possible loss range for the surviving cash-sweep litigation theories?
8. What is the succession plan for a combined Chair/CEO who has led AMP since the 2005 separation?
9. Do the advertised AI time savings convert to higher client capacity, lower support cost, better retention, or merely convenience?

14. What Must Be True

For the quality-compounder interpretation:

• AWM advisory and planning fees must continue to outgrow net spread compression, preferably through at least two flat- or down-market quarters.
• Comparable wrap flows should remain at least in the mid-single digits after Comerica, without economically irrational recruiting packages.
• Huntington should retain most assets and advisors and produce positive incremental economics after transition costs.
• Columbia’s outflow should moderate enough that market appreciation is not the sole support for AUM and margin.
• Repurchases should continue to reduce shares without consuming balance-sheet flexibility or becoming valuation-insensitive.

Falsification test: two consecutive post-Comerica quarters with comparable wrap flow below roughly 3%, advisory-fee growth below the decline in cash earnings, and no offsetting productivity gain would refute the claim that the wealth engine is carrying the group.

For the peak-earnings / deserved-discount interpretation:

• Rate cuts and cash sorting must pull AWM NII materially lower even with the termed-out bank book.
• Equity-market normalization must expose weak organic asset gathering and limit fee growth.
• Columbia’s persistent active outflow must keep the blended business from earning a wealth-platform multiple.
• Higher stock prices must reduce buyback accretion enough that EPS growth converges toward modest operating growth.

Falsification test: adjusted operating EPS continues to grow at a double-digit rate through a full easing year with flat markets, post-Comerica wrap flows remain at least mid-single digit, and bank NII stabilizes. That outcome would invalidate the core peak-earnings argument.

The highest-information scorecard for the next two quarters is therefore: comparable wrap flow, AWM advisory-fee growth, AWM NII, revenue per advisor, diluted shares, and Huntington retention. Consolidated EPS by itself is insufficient because market effects and repurchases can mask changes in franchise health.

15. Source Appendix

• Ameriprise Q2-2026 earnings release, furnished 2026-07-23 — consolidated and segment adjusted results, AWM assets/flows/cash, AM flows/performance, RPS results, earnings reconciliation, capital return.
• Ameriprise Q2-2026 Form 10-Q, filed 2026-08-04 — GAAP statements, share count, capital and liquidity, repurchase details, legal contingencies, Item 5©.
• Ameriprise FY2025 Form 10-K, filed 2026-02-19 — business model, five-year base, segment economics, risk sensitivities, capital, AOCI, reinsurance and reserves.
• Ameriprise 2026 proxy statement, filed 2026-03-20 — compensation design, governance, ownership, succession-process disclosure, pay-versus-performance.
• Ameriprise quarterly-results hub — Q1/Q2 earnings calls and presentations; Q2 call dated 2026-07-23.
• Ameriprise Q2 results announcement, 2026-07-23 — release timing and company context.
• EDGAR Form 4 corpus, 2021-08-21 through 2026-08-21 — insider transaction sweep; transaction codes reconciled rather than treating grants or tax withholding as open-market activity.
• Local five-year daily price series through 2026-08-21 — close, moving averages, beta, and event map. AZI percentile outputs are used only directionally because the raw AMP per-share record does not reconcile to filings.
• FactorsToday local outputs through 2026-08-21 — market/factor exposure and related-stock checks.

The institutional body contains no investment recommendation or price framing. Facts are distinguished from interpretation and assumptions; management claims are treated as hypotheses where no filing bridge exists.

APPENDIX A — Standard Diligence Questionnaire

Supplemental to the research memo. Update report date 2026-08-21. Answers are grounded in the disclosed sources; Fact / Interpretation / Assumption is labeled where it matters. Where a question does not map to a wealth-management/insurance hybrid, the correct sector analog is given.

General

What thoughtful questions have other investors asked about this company? The recurring institutional debates: (1) Is AMP “really” a wealth manager or a rate-levered diversified financial? — the classification that drives the multiple. (2) How much of the record ~53% ROE / ~$39 EPS is a high-rate cash-sweep vintage that reverses on Fed cuts? (3) Can AWM advice-fee growth keep outrunning NII compression through a cut cycle? (4) Is the cheap multiple a value opportunity or a correct discount for the annuity tail + active-AM melt + equity beta? (5) What is the CEO-succession plan after 21 years of Cracchiolo? (6) How exposed is AMP to the cash-sweep litigation wave vs Schwab/Morgan Stanley? (Interpretation, synthesized from the call Q&A and peer coverage.)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? A high (Interpretation). Trailing adjusted operating EPS is about $43.01 and adjusted operating ROE ex AOCI is 55%. Q2 proved fees can offset spread compression, but average wrap assets rose 21% in a strong equity market. A simultaneous rate-cut-plus-equity-drawdown would still compress fees and spread together.

Driven by the external environment or internal actions? Both. Internal: the deliberate shift to fee-based advice, relentless buyback (shares −22% in 5yr), RPS de-risking, productivity gains ($1.2M/advisor). External: the rate cycle (sweep NII) and equity-market levels (fee-on-AUM). The durable internal actions are why earnings held up even as NII rolled over.

How stable are revenues? ~60% of revenue is recurring management/advice fees on AUM — stable but market-linked. The rate-sensitive NII leg (~19% of revenue) is the swing factor. Through-cycle, AMP grew revenue every year 2020–25 ($11.9B → $18.48B). (Fact.)

Outlook for products/services? Wealth advice is a secular grower; the active-AM product set (Columbia) is in structural decline (mitigated by ETF/SMA/alts pivot); RPS is a deliberate managed runoff. (Interpretation.)

How big will this market be? US wealth/advice is a large, growing, domestic-led market (boomer decumulation + next-gen accumulation), with EMEA/Asia asset-management exposure via Columbia Threadneedle. AMP targets the mass-affluent ($500K–$5M) plus HNW. (Fact/Interpretation.)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Wealth/advice: more competitive for advisor talent (inflating LPL/RJF recruiting war) but structurally favorable for asset gatherers. Active AM: brutally competitive and consolidating. (Interpretation.)

How profitable is the business (ROIC, ROE)? Elite. Adjusted operating ROE-ex-AOCI was 55.0% in Q2 on a trailing basis. All current ROIC.ai AMP ratios are excluded because the per-share/equity record is corrupted. Filing-derived returns far exceed the cost of capital. (Fact/Interpretation.)

How profitable is the industry — competitors, barriers to entry? Wealth/advice earns high, recurring fee-on-AUM margins with real barriers (client switching costs, advisor relationships, platform scale, regulatory licensing). Active AM earns compressing fees with low barriers against passive. (Interpretation.)

Can the business be easily understood? Mostly — a fee-gathering wealth manager with a bank spread leg and a runoff insurance book. The complications are the GAAP-vs-adjusted insurance marks and the AOCI/thin-equity ROE optics. (Interpretation.)

Can it be undermined by foreign low-cost labor? No — advice is relationship- and regulation-bound and domestic. The relevant disruption vector is technology/AI + fee transparency on the cash-sweep spread, not offshoring. (Interpretation.)

Do brands matter? Moderately. The Ameriprise/Columbia Threadneedle/RiverSource brands matter, but the advisor owns much of the client relationship — the brand supports retention, it is not the primary moat. (Interpretation.)

Nature of competition / customers’ switching costs? Competition is on advisor productivity/economics, platform/tech, and client experience (AMP wins on per-advisor economics, not headcount). Client switching costs are real (multi-product relationships, tax/transfer friction, advisor loyalty) but advisor switching costs are lower (advisors are recruitable). (Interpretation.)

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The AWM franchise value — advisor force, client relationships, productivity — is largely unrecognized; conversely, separate-account and policyholder assets appear on the balance sheet but are not common equity. This accounting structure is one reason aggregator ratios require filing reconciliation. (Interpretation.)

Off-balance-sheet liabilities? Reinsured legacy blocks (LTC ~50% ceded to Genworth; fixed annuities ~89% ceded) leave RiverSource primarily liable as direct insurer — a residual reinsurer-credit exposure. Off-balance-sheet client sweep (~$5.1B) and third-party cash (~$48B) are placement relationships, not liabilities. (Fact/Interpretation.)

How conservative is the accounting? Reasonably. The adjusted-operating measure backs out non-cash insurance marks (legitimate, not flattering); OCF/NI 2.34x corroborates real cash earnings; SBC is modest. The deliberate de-risking of the VA book reduces actuarial-assumption risk. (Interpretation.)

How CapEx-hungry? Very light — capex ~$162M on $18.5B revenue; the business is human-capital and technology-intensive, not physical-asset-intensive. (Fact.)

Capital Allocation & Management

How much FCF does the business generate, and how is it used? Headline OCF ~$8.3B is float-inflated; the durable holdco-distributable figure is ~$2.5–3.5B/yr. Capital is returned at ~85–90% of adjusted operating earnings (~$3.4B in 2025, 15th straight year), split buybacks-heavy plus a growing dividend. (Fact.)

Significant acquisitions recently? No transformative M&A; the one material deal (BMO EMEA AM, 2021, ~$845M) reinforced the weakest segment. Recent “acquisitions of subsidiaries” line items are largely reinsurance/block transactions. Discipline is a positive. (Fact/Interpretation.)

Buying back shares? Yes. AMP repurchased $774M in Q2 at an average near $459 and $1.653B in H1 at $470.84; $1.1B remained authorized at June 30. Diluted Q2 shares fell 6% year over year. (Fact.)

Issuing large amounts of new shares to insiders? No — SBC is modest (~$206M, ~5% of NI); net dilution is minimal. (Fact.)

Compensation policy of directors/management? Genuinely shareholder-aligned: annual scorecard weights ROE-ex-AOCI 20% + adj op EPS 20% + earnings 25% + revenue 15%; PSUs on EPS growth + average ROE + relative-TSR. CEO ~$30.8M, CFO ~$9.4M; say-on-pay ~89%. (Fact.) Demerit: combined Chair/CEO, no named successor.

Motivations of management? Per-share value creation and capital-return discipline are deeply embedded (“built not bought,” 15 years of differentiated return, multiple comp metrics aligned). The risk is key-person concentration in Cracchiolo. (Interpretation.)

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — a US C-corporation common stock (NYSE: AMP), standard 1099 dividends. (Fact.)

Dividend policy? Quarterly cash dividend, DPS $4.01 (2020) → $6.53 (2025), ~10% CAGR, raised 8% in 2025 and 6% in Q1-26; payout only ~16–17% of NI (long runway). Trailing yield ~1.4%. The primary return vehicle is buybacks, not the dividend. (Fact.)

How profitable is the business? Among the most profitable large financials in the US: 55% adjusted operating ROE ex AOCI, 27.2% consolidated pretax adjusted margin, and 28.9% Q2 AWM margin. (Fact.)

Is net income diverging from cash from operations? OCF exceeds NI (2.34x in 2025) due to insurance/working-capital float — cash generation is stronger than GAAP NI, a positive divergence. Distributable cash is the ~$2.5–3.5B holdco figure, not the float-inflated OCF. (Fact/Interpretation.)

Risks & Downside

What factors would cause the stock to decline? (1) A simultaneous Fed-cut + equity-drawdown compressing both NII and fee AUM; (2) NII compressing faster than fees grow; (3) AWM net-flow deceleration / advisor attrition in the recruiting war; (4) an adverse cash-sweep litigation outcome forcing higher sweep payouts; (5) a disorderly CEO succession; (6) accelerating Columbia Threadneedle outflows. (Interpretation, from the risk matrix.)

Risk of a catastrophic loss? Low. No near-term recourse-debt cliff, RiverSource RBC 523% at year-end, $2.8B of holding-company liquidity at Q2, modest SBC, and a fee-based revenue model. Cash is regulated/nonfungible and gross debt was about $3.85B after prefunding, so a simple net-debt claim is inappropriate. The realistic downside is a cyclical earnings/multiple drawdown, not impairment. (Interpretation.)

Chance of a total loss? Remote — a diversified, profitable, well-capitalized franchise with a fortress balance sheet. (Interpretation.)

Recent News & Events

Has the business environment changed recently? Yes. Q2 fee growth outweighed spread compression, but the advisor-recruiting war intensified. Comerica’s approximately $19B is expected out by end-Q3 and Huntington’s approximately $28B/260 advisors begins onboarding in Q4 or early 2027. (Fact.)

Significant acquisitions? None recently of scale; Huntington is a 10-year channel win (~$28B), not an acquisition. (Fact.)

Change in accounting policies? No material change; the long-duration-insurance (LDTI) framework drives the non-cash VA marks that the adjusted measure backs out. (Fact.)

Recent changes — new markets, facilities, management? AFIG bank/credit-union channel expansion (Huntington); Signature Wealth UMA launch; 2026 senior-officer retirements (CIO Davies, AWM’s Sweeney); $750M June-2026 senior-notes refinancing; new $4.5B buyback authorization. (Fact.)

APPENDIX B — Source Appendix

Update report date 2026-08-21. Primary sources prioritized. Facts cite the underlying filing/data; interpretations are labeled in the memo. Sources were re-checked through 2026-08-22.

Primary — SEC filings (EDGAR CIK 0000820027)

Primary — Earnings call

Litigation & regulatory (public)

• Mehlman & Hultman v. Ameriprise Financial, D. Minn. No. 0:24-cv-03018 — Memorandum Opinion & Order, 2025-08-19: advisory arbitration compelled; fiduciary count dismissed; contract, implied-covenant, and unjust-enrichment theories survived at that stage.
• SEC off-channel-communications settlement, Aug-2024 — AMP $50M (of ~$393M / 26 firms). SEC press release; Compliance Week; Star Tribune.

Quantitative data services

• ROIC.ai MCP — AMP ratios and multiples excluded because the current AMP per-share record is corrupted; filing values take precedence.
• AZI — five-year daily price CSV through 2026-08-21; directional valuation screen composite 70.9th percentile, P/E 52.6th, P/B 74.6th, P/S 85.6th after peer control. Absolute ranks are downgraded because the underlying raw AMP per-share record is inconsistent with filings.
• FactorsToday — all-factors loadings: Market 1.13, Dividend 0.70, Financials 0.65, Growth -0.49, Momentum zeroed; recent strength aligns with a favorable Financials/Value/Dividend regime rather than a generic momentum factor.

Industry & trade data (public)

• Morningstar / McKinsey 2025 asset-management trends — active-MF outflows ~$640B (2025), ~$4T cumulative; passive AUM ~$19.1T > active ~$16.2T; fee compression; active ETFs ~$1.2T cumulative inflows.
• InvestmentNews / Financial Planning 2025 — advisor recruiting (LPL ~32,000 advisors +601, RJF record headcount +313, LPL/Commonwealth ~77.5% retention).