AMP 10-K Analysis: 53% ROE Built on Three Layers of Earnings Distortion

Ameriprise Financial reported a "record" 53.2% adjusted operating ROE in FY2025 — but its core pretax income actually declined 3.8%. The $46 billion wealth management and insurance company delivered 10.1% EPS growth, repurchased $2.9 billion of its own stock, and crossed $1.7 trillion in total client assets. Every headline metric says the compounding machine is accelerating.

The 10-K tells a different story. Strip away three layers of earnings distortion — a $1,004 million insurance fair value gain that inflated GAAP net income, a 5.1% share count reduction that manufactured more than half of EPS growth, and management's adjusted operating framework that excluded a worsening $366 million drag — and the underlying business contracted. Core pretax income fell from $3,639 million to $3,500 million, a decline masked by the rare alignment of three flattering forces moving in the same direction.

At $490 per share and a headline 13.5× P/E, Ameriprise looks like a cheap, high-returning capital allocator. Adjust for the insurance noise, and the true P/E is closer to 17.7× — on earnings that are shrinking, not growing. The question isn't whether AMP's buyback machine is impressive. It's whether the machine is buying overvalued stock.

The Three-Layer Earnings Gap

Every headline metric flatters Ameriprise's FY2025 performance — and all three distortions compound in the same direction. That rare alignment makes the stock harder to value than any single number suggests.

Start with the core operating layer. GAAP pretax income was $4,504 million, up from $4,267 million. But $1,004 million of the FY2025 figure came from a market risk benefit fair value gain — an insurance accounting line item driven by equity market movements and interest rate changes, not business operations. In FY2024, the equivalent MRB gain was $628 million. Stripping it from both years: core PTI fell from $3,639 million to $3,500 million, a 3.8% decline. The entire GAAP pretax improvement — and then some — was manufactured by insurance volatility.

Layer two: the insurance gain flowed through to net income. GAAP NI rose 4.8% to $3,563 million from $3,401 million — an increase of $162 million. The MRB gain alone was $1,004 million. Without it, net income would have declined. Management's adjusted operating earnings grew 10% to $4,877 million by excluding the "market impact on non-traditional long-duration products," which worsened from -$153 million to -$366 million. This created a widening gap between management's preferred metric and GAAP — and both were better than the core reality.

Layer three: financial engineering. Net income grew 4.8%, but EPS grew 10.1% — from $33.00 to $36.34. The 5.3 percentage point gap is entirely attributable to the 5.1% share count reduction, from 96.2 million to 91.3 million shares. Buybacks contributed more than half of per-share earnings growth. At $500.18 per share average, AMP spent $2.9 billion repurchasing stock to achieve this effect.

The three-layer decomposition is not an academic exercise — it changes the valuation math. At $490 per share, the headline P/E of 13.5× uses GAAP EPS of $36.34, which includes approximately $8.70 per share from the MRB insurance gain. Core EPS excluding MRB is roughly $27.64, putting the true P/E at approximately 17.7× — on declining core earnings. An investor relying on the headline multiple is pricing a growth stock. The filing reveals a shrinking earnings engine dressed in three layers of flattery.

"Most of our variable annuity products contain guaranteed minimum death benefits and a majority of our variable annuity products in force contain guaranteed minimum withdrawal and accumulation benefits. Decline or volatility in equity and/or bond markets could result in guaranteed minimum benefits being higher than what current account values would support."

The variable annuity guarantees remain in force despite AMP discontinuing new VA sales years ago. As long as the legacy book exists, MRB swings of $1 billion or more in either direction will continue distorting GAAP earnings, making year-over-year comparisons unreliable without the three-layer adjustment. Ameriprise Financial's core pretax income declined 3.8% to $3,500 million in FY2025, even as GAAP net income rose 4.8%, because a $1,004 million insurance fair value gain masked shrinking operating earnings.

The Advisor Cost Trap

Ameriprise's Advice & Wealth Management segment delivered 9% revenue growth in FY2025, reaching $11,741 million from $10,780 million. The headline looks strong. The cost structure underneath it does not.

AWM distribution expenses — advisor commissions, trail commissions, and platform costs — grew 12% to $6,513 million from $5,823 million, an increase of $690 million. Revenue grew $961 million. That means distribution expenses consumed 72% of the revenue gain, leaving only $271 million of the $961 million increase to flow through to margin improvement. In a year when equity markets rose 13% and drove most of the segment's asset growth, the advisor compensation model still ate nearly three-quarters of the upside.

The contrast between distribution costs and overhead is striking. General and administrative expense actually declined 1%, reflecting what management calls "ongoing benefits from our initiatives to enhance operational efficiency." The cost discipline is real — but it's overwhelmed by the structural problem in advisor compensation.

"General and administrative expense decreased $28 million, or 1%, for 2025 compared to the prior year primarily reflecting ongoing benefits from our initiatives to enhance operational efficiency and effectiveness, as well as lower severance expenses, partially offset by higher performance fee compensation, higher volume-related expenses and investments for business growth."

The forward indicator is deteriorating. Wrap net inflows — the leading metric for future AWM revenue — decelerated 6.6% to $30.9 billion from $33.1 billion in FY2024. Meanwhile, market appreciation of $65.6 billion drove roughly 68% of total AUM growth. The wealth management segment is increasingly dependent on rising markets rather than organic client acquisition.

This creates a compounding vulnerability. In a flat equity market, AMP faces a double squeeze: fee revenue stalls because AUM stops appreciating, but advisor compensation doesn't flex down proportionally because it's contractually structured and tied to asset levels. The negative operating leverage that appeared even in a strong year would accelerate sharply in a weak one. Ameriprise's wealth management distribution expenses grew 12% in FY2025, consuming 72% of the segment's revenue increase, because advisor compensation scales with assets under management while wrap net inflows decelerated 6.6%.

Columbia Threadneedle: Award-Winning and Shrinking

Columbia Threadneedle Investments, Ameriprise's global asset management arm, earned a Barron's top-10 fund family ranking. Its segment assets shrank 8.5% anyway.

In FY2025, the Wells Fargo Equity Index — AMP's market proxy — rose 13%. Asset Management revenue grew only 3% to $3,621 million. AM segment assets fell from $7,350 million to $6,723 million, a decline of $627 million. Institutional net outflows consumed more than 100% of market-driven appreciation on the balance sheet — a striking underperformance in what should have been a tailwind year.

"Management and financial advice fees increased $966 million, or 10%, for 2025 compared to the prior year primarily reflecting market appreciation and continued wrap account net inflows as well as higher performance fees, partially offset by the cumulative impact of Asset Management and variable annuity net outflows."

The filing's own language confirms the structural problem: fee growth from market appreciation and AWM inflows was "partially offset" by AM outflows. This is not a one-quarter blip. Ameriprise Certificate Company deposits — another funding channel — declined $3.0 billion to $8.2 billion after eight consecutive quarters of net outflows. The certificate product, attractive in a rising-rate environment, is losing appeal as rates stabilize. Available-for-sale securities backing ACC certificates dropped 30% from $11,022 million to $7,725 million.

The paradox is that Columbia Threadneedle is simultaneously shrinking and still generating substantial cash for the parent. AM subsidiary dividends to Ameriprise rose 27% to $620 million — monetizing a declining asset base rather than reinvesting for growth. This is the behavior of a harvest-mode business, not a growth engine.

The "vertically integrated flywheel" thesis — that AWM advisors channel client assets into Columbia Threadneedle products, creating a closed-loop fee multiplier — is broken at the institutional layer. Retail flows through the advisor network may hold, but the institutional exodus is eroding the asset base from which AM earns management fees. Ameriprise's Columbia Threadneedle segment assets shrank 8.5% to $6,723 million in FY2025 despite a 13% equity market gain, because institutional net outflows more than consumed market-driven appreciation.

The $128 Billion Insurance Tail

Ameriprise's Retirement & Protection Solutions segment is often described as a legacy runoff business. The 10-K shows the opposite: it's growing, becoming less productive, and dragging the company's blended returns.

R&PS segment assets grew 9.6% to $127,778 million from $116,609 million — an increase of $11.2 billion. Revenue grew only 4.8% to $3,955 million. Revenue yield on those assets declined from 3.23% to 3.09%, a 14 basis point deterioration. The insurance book is absorbing more capital while generating proportionally less income.

The scale creates a structural issue for valuation. R&PS holds approximately 67% of Ameriprise's total segment assets — $128 billion out of roughly $191 billion — but generates only 21% of total revenue. For investors trying to value AMP as a wealth management company, two-thirds of the balance sheet belongs to an insurance operation with declining per-dollar returns.

"In 2009, RiverSource Life established an agreement to protect its exposure to Genworth Life Insurance Company ('GLIC') for its reinsured LTC. In 2016, substantial enhancements to this reinsurance protection agreement were finalized."

Beyond the GLIC tail risk, the MRB volatility is not a one-time event. Variable annuity guarantees remain in force — guaranteed minimum death, withdrawal, and accumulation benefits that fluctuate with equity and bond markets. The $1,004 million gain in FY2025 could just as easily be a $1 billion loss in a down market year, whipsawing GAAP earnings in either direction. RiverSource Life's dividend to the parent has been flat at $600 million for three consecutive years, suggesting the insurance subsidiary is at steady-state distribution capacity — generating cash but not growing it. Ameriprise's Retirement & Protection segment holds 67% of total assets but generates only 21% of revenue, with yield declining 14 basis points to 3.09% as the insurance book grows bigger and less productive.

The Buyback Endgame

Ameriprise has built the most aggressive buyback program in financial services. The question is whether it's creating value or destroying it at current prices.

The numbers are staggering. Cumulative treasury stock stands at $30.6 billion — equivalent to 66% of AMP's $46.4 billion market capitalization. The company has essentially bought itself 0.66 times over. In FY2025 alone, AMP repurchased 5.5 million shares at an average price of $500.18, spending $2,907 million. Buybacks represented 83% of total capital returns of $3,503 million (the remainder was $614 million in dividends, up from $593 million).

The buyback program's effectiveness depends on one condition: the stock must be undervalued for repurchases to create value. At 13.5× GAAP earnings, the case for undervaluation is plausible. At 17.7× core earnings — on a business where those core earnings declined 3.8% — the math is less forgiving. Each 1% reduction in share count now costs approximately $570 million. To maintain the current 5.1% annual shrink rate, AMP needs roughly $2.9 billion per year in buyback spending, which consumes the majority of free cash flow.

Meanwhile, capital demands elsewhere are growing. Off-balance-sheet commitments surged 32% to $3,322 million from $2,520 million, driven by pledged asset lines of credit growing 36% to $3,258 million. Two new commitment categories appeared for the first time: home equity lines of credit ($24 million) and private fund commitments ($7 million). The company is quietly expanding its advisor lending balance sheet — a capital-intensive pivot that competes with buybacks for cash.

The most telling signal is Ameriprise Advisor Capital, the advisor lending subsidiary. It swung from generating a $225 million dividend to consuming $320 million in capital contributions — a $545 million reversal buried in the liquidity section of the 10-K. Combined with debt refinancing at higher rates ($750 million of 5.20% notes replacing $500 million of 3.0% notes, adding $24 million in annualized interest expense), the capital return math is getting tighter.

"Short-term contractual obligations for the year 2026 include investment certificate maturities of $7.8 billion and estimated insurance and annuity benefits of $2.9 billion in addition to operating liquidity needs and maturing long-term debt in September 2026 of $500 million. We also hold banking and brokerage deposits of $25.6 billion that are payable on demand."

The $36.8 billion in short-term obligations — certificate maturities, insurance benefits, maturing debt, and demand deposits — provide context for the buyback debate. AMP's parent liquidity of approximately $2.2 billion plus a $3 billion credit facility provides adequate coverage, but the obligations underscore that the buyback program operates within a capital-constrained insurance holding company, not a pure-play asset manager. Ameriprise spent $2.9 billion on buybacks at an average price of $500.18 per share in FY2025, contributing more than half of its 10.1% EPS growth, because the share count declined 5.1% while net income grew only 4.8%.

What to Watch

Five metrics will determine whether AMP's "record results" narrative holds or collapses in FY2026:

1. Wrap net inflows (quarterly, annualized)

• FY2025 run rate: $30.9 billion (down from $33.1 billion)
• Bull case: Inflows reaccelerate above $34 billion annualized, suggesting the Q4 dip was seasonal and the AWM growth engine is intact
• Bear case: Inflows fall below $28 billion annualized, confirming structural deceleration and raising the market-dependence risk for fee revenue

2. AWM distribution expense ratio

• FY2025: 55.5% of AWM revenue
• Bull case: Ratio declines below 53%, indicating management has found structural cost efficiency in the advisor compensation model
• Bear case: Ratio rises above 57%, confirming the negative operating leverage thesis and signaling margin compression even if revenue grows

3. Core pretax income ex-MRB

• FY2025: $3,500 million (-3.8% YoY)
• Bull case: Core PTI inflects positive — any growth in this metric invalidates the "declining earnings engine" thesis
• Bear case: Core PTI declines again, proving the "record results" narrative is entirely dependent on insurance fair value gains and buyback-driven share shrinkage

4. Columbia Threadneedle net flows

• FY2025: Net outflows consuming >100% of market gains on AM segment balance sheet
• Bull case: Two consecutive quarters of positive net flows would signal the institutional exodus has reversed
• Bear case: AM segment assets drop below $6 billion, indicating the harvest-mode trajectory is accelerating

5. MRB fair value direction

• FY2025: $1,004 million gain
• The MRB is mechanistically linked to equity markets and interest rates. In a down market, a $1 billion loss is possible — which would reverse the entire GAAP earnings picture and expose the 3.8% core decline to investors who haven't been watching

At $490 per share, the market prices Ameriprise at 13.5× GAAP earnings — a value stock multiple on a business that returned 7.4% to shareholders through buybacks and dividends. The filing supports the capital return story: AMP generates exceptional free cash flow, has near-zero capital intensity, and maintains extraordinary ROE. But the 10-K complicates the growth story: core earnings are declining, the cost structure creates negative operating leverage, the asset management arm is shrinking, the insurance balance sheet is growing less productive, and the buyback program's value-creation math depends on a stock price that may already reflect optimistic assumptions. The true P/E of 17.7× on declining core earnings suggests investors are paying a growth multiple for a business that isn't growing — yet.

Frequently Asked Questions

What is Ameriprise Financial's core earnings growth when you strip insurance volatility?

Ameriprise reported GAAP net income growth of 4.8% in FY2025 ($3,563M vs $3,401M). However, this includes a $1,004M market risk benefit fair value gain. Stripping MRB from pretax income reveals core earnings declined 3.8% ($3,639M to $3,500M). The entire reported earnings increase was manufactured by insurance accounting, not operational improvement.

Why does Ameriprise's GAAP ROE (59.8%) exceed its adjusted operating ROE (53.2%)?

AMP has -$892M in accumulated other comprehensive income (AOCI) losses, primarily unrealized bond losses, which reduces GAAP equity and inflates GAAP ROE. Management's adjusted ROE excludes AOCI from equity, using a larger denominator — so the adjusted number is actually lower. Both numbers are extraordinarily high, but the GAAP figure is partially artificial.

How much of Ameriprise's EPS growth comes from buybacks vs. earnings growth?

In FY2025, EPS grew 10.1% while net income grew only 4.8%. The 5.3 percentage point gap is attributable to the 5.1% share count reduction (96.2M to 91.3M shares). Buybacks contributed more than half of EPS growth. AMP spent $2.9B on repurchases at $500.18/share average, with $30.6B in cumulative treasury stock.

Is Columbia Threadneedle a growth engine or a drag on Ameriprise?

Columbia Threadneedle is a decelerating drag. In FY2025, Asset Management revenue grew only 3% despite equity markets rising 13%. AM segment assets shrank 8.5% ($7,350M to $6,723M) as institutional net outflows more than consumed market-driven gains. The filing confirms "cumulative impact of Asset Management net outflows" partially offset fee growth.

What drives Ameriprise's negative operating leverage?

AWM distribution expenses — advisor commissions, trail commissions, and platform costs — grew 12% versus revenue growth of 9%, consuming 72% of the segment's revenue increase. G&A actually declined 1%. The negative leverage is structural: advisor compensation scales with AUM, so costs rise faster than revenue when markets drive most of the growth.

What is the Genworth LTC reinsurance risk in Ameriprise's filing?

RiverSource Life has a reinsurance agreement with Genworth Life Insurance Company for long-term care insurance, established in 2009 with enhancements in 2016. The provisions are described as confidential. Genworth has been under financial stress with multiple states monitoring its solvency. The total exposure is not disclosed, making it an unquantifiable tail risk.

How does Ameriprise's valuation look on a core earnings basis?

At $490/share, AMP trades at 13.5× trailing GAAP EPS ($36.34). But GAAP EPS includes approximately $8.70/share from the MRB insurance gain. Core EPS ex-MRB is approximately $27.64, putting the true P/E at roughly 17.7× — significantly more expensive on a business with declining core profitability.

Why did off-balance-sheet commitments surge 32% in FY2025?

Total funding commitments grew from $2,520M to $3,322M, driven by pledged asset lines of credit surging 36%. Two new categories appeared: home equity lines ($24M) and private fund commitments ($7M). Ameriprise Advisor Capital swung from generating a $225M dividend to consuming $320M in capital — a $545M swing buried in footnotes.

Is Ameriprise's 43% FCF margin sustainable?

AMP's 43.2% FCF margin is partially inflated by derivative counterparty cash collateral inflows, a non-recurring working capital item. Per Ameriprise's FY2025 10-K MD&A — Liquidity, OCF growth reflects both fee-based earnings and increased cash collateral from derivative counterparties. A normalized FCF margin is likely closer to 35-38% — still exceptional, but less than the headline.

How aggressive is Ameriprise's buyback program compared to peers?

AMP's program is the most aggressive among financial services peers. Cumulative treasury stock of $30.6B equals 66% of market cap. Buybacks represent 83% of total capital returns versus 71% for Chubb, roughly 2% for Blackstone, and 0% for Progressive. Share count is declining 5.1% annually, costing $570M per 1% reduction at current prices.

What are the key risks to Ameriprise's investment thesis?

The primary risk is market dependence: 68% of AUM growth came from market appreciation, not organic flows. In a flat market, AMP faces a triple headwind — declining fee revenue, sticky advisor costs creating negative operating leverage, and reduced FCF for buybacks. Additional risks include Columbia Threadneedle outflows accelerating, MRB losses in a down market, and unquantified Genworth LTC exposure.

What should investors watch in Ameriprise's next filing?

Three key metrics: (1) wrap net inflows — if they fall below $28B annualized, the AWM deceleration is structural; (2) AWM distribution expense ratio — if it stays above 55%, negative operating leverage is baked in; (3) PTI ex-MRB — if core earnings decline again, the "record results" narrative collapses regardless of what markets do.

Methodology

Data Sources

This analysis is based on Ameriprise Financial's FY2025 Annual Report (10-K) filed with the SEC, accessed via the MetricDuck filing text extraction pipeline. Financial metrics were extracted using the MetricDuck XBRL metrics processor and cross-verified against filing text in the following sections: MD&A — Results of Operations, MD&A — Liquidity, footnote on commitments and contingencies, segment footnote, risk factors, and footnote on debt.

Peer data for HOOD, CB, BX, and PGR is sourced from MetricDuck pipeline data and prior filing research iterations. All peer metrics are from FY2025 annual filings except where noted.

Key Derived Calculations

• Core PTI ex-MRB: GAAP pretax income minus change in fair value of market risk benefits
• MRB per-share contribution: MRB gain × (1 - effective tax rate of 20.9%) / diluted shares outstanding
• Core P/E: Market price / (GAAP EPS - MRB per-share contribution)
• Distribution expense consumption: AWM distribution expense increase / AWM revenue increase
• R&PS revenue yield: Segment revenue / segment assets
• Buyback cost per 1% reduction: Total repurchases / percentage share count decline

Limitations

1. Derivative collateral quantification: The filing confirms derivative counterparty cash collateral as a material driver of OCF growth but does not isolate the exact dollar amount. Our normalized FCF estimate (35-38% margin) is approximate.
2. Columbia Threadneedle flow granularity: Net flow data by channel (institutional vs retail) and geography is not available in the 10-K. The "net outflows" conclusion is confirmed qualitatively by the filing but not quantified at the sub-segment level.
3. Genworth exposure: The GLIC reinsurance tail risk is identified as material but cannot be sized due to confidential provisions. We flag it as a known unknown rather than a quantified risk.
4. Peer comparison limitations: PGR data is from pipeline metrics only (no filing-level research completed). Metrics definitions may vary across financial services sub-sectors.
5. MRB tax treatment: We apply the effective tax rate (20.9%) to the MRB gain for the per-share calculation. The actual tax treatment of MRB fair value changes may differ due to temporary versus permanent differences.
6. Forward guidance: The 12-15% adjusted EPS growth target appears in the 10-K's capital allocation discussion as an ongoing management target; the exact origin document (10-K vs investor day) is not pinpointed.

Disclaimer: This analysis is for informational purposes only and does not constitute investment advice. The author does not hold positions in AMP, HOOD, CB, BX, or PGR. 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.