Goldman Sachs (GS)
Banca / Servicios financieros
The world's leading investment bank (investment banking + trading + asset management), transformed after the consumer-banking retreat. But it trades at a double peak: peak earnings (ROTE 16%, 21% in Q1'26 — capital markets booming) × a demanding multiple (price-to-tangible-book 3×). At ~$894, normalizing to mid-cycle (~18× normalized earnings), the 5-year base yields ~$762 (-0%/year): Overvalued — a quality business priced for perfection; not even the bull scenario clears today's price by a comfortable margin.
Moat Compounder estimates the intrinsic value of Goldman Sachs (GS) at $762 per share on a five-year horizon. With the stock at $893.95 at 2026-10-12 close, the expected total return is -0.1% per year: overvalued. The analysis draws on 10-K FY2025 and 10-Q Q1 2026. Analysis dated 2026-08-03.
- Price
- $893.95
- Intrinsic value (5y, base)
- $762
- Total annual return (5y)
- -0.1%
- Status (nominal)
- Overvalued
- Margin of safety
- No margin
The essentials
- The world's leading investment bank: Global Banking & Markets is 71% of revenue (investment banking that ranks #1 in advisory, plus FICC and equities trading), Asset & Wealth Management is 29% (the durable fee stream — record AUS of $3.6bn, ROE 12.5%), and Platform Solutions is <1% (consumer, in wind-down: the Apple Card exiting, GreenSky/GM sold). The strategic transformation —the consumer-banking retreat to return to the core of banking/trading/management— de-risked the story and lifted ROTE to 16%.
- ⚠️ Double peak. Peak earnings: record Q2'26 (revenue $20.3bn, EPS $20.98, +78%), TTM ROTE 16.8% well above its through-cycle level (~14-15%), inflated by a one-off reserve release (−$1.1bn of provision) + a one-time tax benefit. And a demanding multiple: price-to-tangible-book 3×.
- An investment bank's EPS swings 2.6× across the cycle ($22.87 in 2023 to $59.45 in 2021) — it is the least predictable business, by GS's own admission ('by their nature, do not produce predictable earnings'). Normalizing earnings to mid-cycle (ROTCE ~15%) and with a disciplined exit multiple (~12× P/E, not today's peak multiple), the five-year base yields ~$762/share → a total return of -0%/year: Overvalued.
Intrinsic value — two valuation methods
Total return at 5 years: -0.1%/year = -3.1% appreciation + 3.0% dividend. The target price ($762) is ex-dividend; the $127 in dividends collected over 5 years are added separately.
The methods disagree: one places the value today above the price ($894) and the other below.
Pillars of the analysis
The verdict — today vs 5 years
Today — expensive, no margin of safety: at $894 trades ~23.8% above its value discounted to today (~$722); the expected return does not even reach the risk-free rate (4.5%).
At 5 years — Sobrevalorado: the expected total return is negative — the price already discounts a demanding scenario that, if not met, results in a loss.
The bridge: the return at 5 years falls below the risk-free rate (4.5%) — which is why there is not even a discount to today's value. To require a 15% annual return, it would need to be bought at ~$462.
Thesis
The business
Goldman Sachs is the world's leading investment bank: the #1 investment bank in advisory, one of the largest FICC and equities trading desks, and an asset management business ($3.6bn AUS) growing as the durable fee stream. The transformation —the consumer-banking retreat (GreenSky, cards) to return to the core— de-risked the story and lifted ROTE to 16%. But it is a cyclical business by nature: 69% of revenue depends on capital markets, and EPS swings 2.6× across the cycle. The moat is reputational and talent-based, narrower than a universal bank's (without the low-cost deposit franchise).
The valuation
A bank is valued on the equity —normalized P/E on net income to common plus price-to-tangible-book—, never on EV: deposits and debt are funding. The key for an investment bank is normalizing earnings to mid-cycle: today's (16.8% TTM ROTE, 21% in Q1'26) is at a capital-markets peak and inflated by one-offs (a −$1.1bn reserve release + a one-time tax benefit). Normalizing to a ~15% mid-cycle ROTCE (credit for the improved franchise, but without the peak), net income to common is ~$15.5bn.
At ~$894, that implies ~18× normalized earnings (and 3× tangible book value). The base scenario projects earnings growing modestly (~3%/year) with a buyback that reduces shares ~2%/year, and a disciplined exit multiple of ~12× P/E (equivalent to ~1.8× tangible book, well below today's 3×). That yields ~$762/share → a total return of -0%/year. Not even the bull scenario (sustained 16% ROTE + 14× premium multiple) clears today's price by a comfortable margin.
The margin of safety
No margin of safety: the price already discounts a demanding scenario. At ~$894, Goldman Sachs is a quality business priced for perfection — the textbook case of stacked optimism in a financial: peak earnings (capital markets booming + one-offs) multiplied by a demanding multiple (3× tangible book). The verdict is Overvalued: the franchise has genuinely improved (exit from consumer, growing AWM, mid-teens ROTE), but the price discounts both the earnings peak AND the multiple peak holding at once. Normalizing earnings to mid-cycle with a disciplined multiple, the base ($762) sits well below the price, and even the bull scenario barely reaches it. There is no discount: no missing buyers or motivated sellers are identifiable. The bear scenario (a capital-markets cycle downturn, with ROTE falling to ~11% and the multiple compressing to ~10× — as happened in 2022-2023) has a large downside. Value discipline says to wait for a cycle trough — not to buy an investment bank at the peak of both earnings and multiple at once.
What to watch
Three things. The capital-markets cycle (the central disconfirmation test) — today's earnings are at an M&A/issuance/trading boom; an investment bank's EPS fell 60% from peak to trough in the 2021-2023 cycle, so the key metric is whether the cycle holds or reverses. The multiple — at 3× tangible book (vs a 1.2× historical level), any earnings or cycle disappointment re-rates sharply; paying a premium multiple on peak earnings is a double risk. And governance — the $80M retention awards to Solomon and Waldron (with a strong shareholder backlash, say-on-pay at 66%), succession, and the litigation legacy (1MDB). At this price, the best entry point comes with a capital-markets cycle trough — not with earnings at the peak and a demanding multiple.
Educational / informational. Does not constitute investment advice.
Valuation by multiples
The multiple is applied to the metric per share (EPS / Core FFO): that metric already reflects the evolution of the share count (buybacks or issuance), so the share count does not enter as a separate step. The implied equity (~$232.3 bn) is the metric carried to the equivalent of today's share count — the detail is in the piece's (i).
Discounted cash flow to present value (DCF)
Net income to common normalized to mid-cycle (≈ distributable; the reported TTM ~$17.1bn is inflated by a reserve release + the capital-markets peak) as the base. Move the assumptions: the value recalculates live. The verdict remains anchored by multiples; the DCF contrasts it at present value.
Risk does not inflate the rate: protection is required separately, as a margin of safety over the value. The floor avoids discounting at the pace of a depressed market rate.
| Year | Projected earnings | Discount factor | Present value |
|---|---|---|---|
| 1 | $16 bn | 0.957 | $15.4 bn |
| 2 | $16.6 bn | 0.916 | $15.2 bn |
| 3 | $17.2 bn | 0.876 | $15.1 bn |
| 4 | $17.8 bn | 0.839 | $14.9 bn |
| 5 | $18.4 bn | 0.802 | $14.8 bn |
Reverse DCF — what growth the price discounts
The inverse approach: instead of projecting growth to obtain the value, the market price ($894) is taken as given and it solves for what annual owner-earnings growth would need to hold for 5 years for the present value —at the method's rate (4.5%, no-growth terminal)— to equal that price. It is the disconfirmation test: the expectations the price already pays for, contrasted against the method's projection.
The market discounts less growth (-5.0%/year) than we project (3.5%/year) → if the base case holds, there is margin: perception is more pessimistic than the estimated reality.
That growth implies ~$12 bn of owner earnings in year 5 (vs ~$18.4 bn of our base case). It recalculates if the DCF assumptions are edited.
Year-by-year model — bank (equity)
A bank is valued on equity (P/E × common earnings + price/tangible book), not EV: deposits and debt are funding. It returns capital via dividend + buyback —the buyback reduces the share count, so earnings per share grow faster than aggregate earnings—. Total return adds the dividend collected along the way. In edit mode, the metric, shares, dividend, and exit multiple can be adjusted.
| US$ bn / per share | TTM | +1a | +2a | +3a | +4a | +5a |
|---|---|---|---|---|---|---|
| Operation (editable: metric, shares) | ||||||
| Utilidad a común norm. ($bn) | 15.07 | 15.459 | 15.945 | 16.431 | 16.917 | 17.5 |
| growth | — | +3% | +3% | +3% | +3% | +3% |
| ROTCE | 15.0% | 15.5% | 15.2% | 15.0% | 15.0% | 15.0% |
| Utilidad / acción (EPS) | $49.41 | $51.71 | $54.43 | $57.23 | $60.13 | $63.46 |
| Shareholder return (dividend + buyback) | ||||||
| Dividend / share | 20 | 21.6 | 23.3 | 25.2 | 27.2 | 29.4 |
| Payout (div / metric) | 40% | 42% | 43% | 44% | 45% | 46% |
| Shares (M · buyback) | 305 | 298.932 | 292.963 | 287.094 | 281.324 | 275.753 |
| Valor tangible / acción | 336.61 | 354.347 | 373.366 | 393.386 | 413.406 | 434.427 |
| Multiples at today's price (equity, no EV) — how they compress | ||||||
| P/E (price / per share) | 18.1x | 17.3x | 16.4x | 15.6x | 14.9x | 14.1x |
| P/tangible book | 2.7x | 2.5x | 2.4x | 2.3x | 2.2x | 2.1x |
| Value curve (per share × P/E exit multiple by year) | ||||||
| Value / share | — | $646 | $680 | $715 | $737 | $762 |
| Total return vs price | — | (-25%) | (-10%) | (-4%) | (-2%) | (-0%) |
An investment bank is valued on equity (P/E × earnings to common + P/tangible book), not on EV. The key is normalizing earnings to mid-cycle: today's (ROTE TTM 16.8%, Q1'26 21%) are at a capital-markets peak and inflated by non-recurring effects (a reserve release of −$1.1bn + a one-off tax benefit). Rolling grid: the base year is the TTM with earnings to common normalized to ~$15.5bn (mid-cycle ROTCE ~15%) and the projection is 12-month windows (+1y…+5y). It returns capital via dividends (forward $20/year, +8%) + buybacks (which reduce the share count ~2%/year — modest and price-aware, because buying at today's multiple of tangible book is expensive and destroys some tangible book value per share). Base case: earnings to common +3%/year, exit P/E 12× (~1.8× tangible book, disciplined from today's multiple). A double peak: peak earnings × a peak multiple → a quality business priced for perfection.
Today's multiple compresses on its own going forward as the metric per share grows (accelerated by the buyback, which reduces the share count). The exit multiple at 3 years is higher than the terminal at 5 years —at 3 years there is more growth still ahead—. Total return adds the dividend collected; the required return is applied to the base scenario.
Scenarios (bear / base / bull) — at 5 years
Value sensitivity
Value per share by growth scenario (rows) and the compression or expansion of the exit multiple (columns). The color shows whether it beats the required return.
| Growth ↓ / Multiple → | Compression−15% | Base multiple | Expansion+15% |
|---|---|---|---|
| BearCapital-markets cycle downturn (M&A/issuance/trading reverse from the boom) · base ~10× P/E (cycle downturn) | $450 -9.0% | $529 -6.4% | $608 -4.1% |
| BaseEarnings normalized to mid-cycle (ROTCE ~15%) growing ~3%/year · base ~12× P/E (mid-cycle, disciplined) | $648 -3.0% | $762 -0.1% · base case | $876 2.4% |
| BullThe improved franchise sustains ROTE ~16% + AWM compounds the durable fee stream + the capital-markets cycle stays strong; the market sustains a 14× P/E premium multiple (~2.2× tangible book) · base ~14× P/E (franchise + premium multiple) | $898 2.9% | $1,056 6.0% | $1,214 8.8% |
Multiples — today
High today = growth is being paid for; they cheapen toward 3 and 5 years (see Projections).
Forward multiples
With today's price fixed and the metric growing, what multiple is being paid at 3 and 5 years. Today's high multiple is the price of growth: if the business grows, the entry multiple cheapens on its own.
Optionalities
They are valued separately, with their own rationale, and are not incorporated into the base or the verdict (they are excess return). When assigning them value — in Editmode —, the total with optionalities updates live, without moving the base.
The verdict, the base CAGR, and the margin of safety are always calculated on the base; optionalities do not alter them (with optionalities at $0 they do not move).
Maximum price to pay today — by required return
Each card fixes a required annual return and answers: if the business is worth $762 in 5 years, what is the maximum that can be paid today to obtain that return? Since it now trades at $894, the margin of safety is how much cheaper the market is than that maximum. The three thresholds: 4% covers inflation (the floor), 10% is the long-term average return, and 15% is the level of a great investment.
Return and margin of safety calculator
The maximum price to pay today to earn the required return, with the dividend collected as a separate flow. Both controls are editable.
With a target price of $762 in 5 years plus $127 of dividends collected (the dividend adds to the return, not to the price) and a required return of 4.5% annually, the maximum to pay today is $722. Against the current market price ($894), the margin of safety is -23.8% (trades above the maximum → a premium is paid) and the total return at that price would be -0.1% annually.
Valuation quality
- Multiple vs the band. ~18× normalized earnings (3× tangible book) — well above an investment bank's [10,14] P/E band. The exit multiple (12×) is disciplined; today's is at the peak.
- Double peak. Peak earnings (inflated ROTE + one-offs) × peak multiple (3× tangible book) — the textbook case of stacked optimism in a financial. There is no discount.
- Margin of safety. 5-year base ~$779 vs price ~$1,086 → total return -0%/year, negative. No margin; a quality business priced for perfection. Wait for a cycle trough.
ROTCE vs the 10% bar — the bank's engine
The quality bar — return bands
The return on capital is judged against absolute bands; the value-creation floor is the market's opportunity cost (~10%). A stock's volatility does not measure business risk.
ROTCE 16% → excellent (15-20%). The bar is a measure of business quality, not the method's discount rate: value is discounted to today at the risk-free rate, and protection is required separately, as a margin of safety.
Quality — profitability and capital
- ROTCE vs the 10% hurdle. ROTE 16.0% (FY2025) above the 10% hurdle — good, but volatile (8-24% across the cycle). The 16.8% TTM is at the peak.
- Earnings quality / volatility. Earnings are the least predictable of any bank's (GS says so itself: 'by their nature, do not produce predictable earnings'). EPS swings 2.6×. Business quality is high, but predictability is low.
- Capital. CET1 12.5% (requirement 11.4%, ~110bp of buffer as of Mar-2026 — the capital buffer compressed from 14.3% at Dec-2025 on asset growth + the buyback), passed CCAR. Well capitalized, but the buffer is tighter than a universal bank's.
Revenue trajectory
Values in US$ bn. The % over each bar is the year-over-year (YoY) growth — each year, historical and projected, vs the prior one (the TTM vs the TTM from a year ago). The path comes from the same source as the table; years without their own series in the model are interpolated between the anchors. Historical solid, projection in a lighter shade.
Where the growth comes from · by segment
Weight in revenue and year-over-year (YoY) growth, in reported USD.
Revenue is dominated by Global Banking & Markets (71%, +15% in 2025 — the #1 investment bank in advisory + FICC + equities), the engine but also the source of volatility. Asset & Wealth Management (29%, +8%) is the durable fee stream GS is looking to grow (AUS $3.6bn, alternatives targets). Platform Solutions (<1%) is in wind-down (the Apple Card exiting). GBM's growth is cyclical (M&A/trading boom); AWM's is the most stable.
Growth engine — operating drivers
Annual levels from the official filing (10-K); the % over each bar is the year-over-year (YoY) growth vs the prior year.
An investment bank's engine is cyclical: investment banking fees (M&A advisory + underwriting) rise and fall with the deal cycle, and trading (FICC + equities) with market volatility — together 69% of revenue. The durable fee stream is asset management (AUS $3.6bn of assets under supervision), which GS is looking to grow to smooth cyclicality. The quality metric is ROTE (16%, but it swings 8-24% across the cycle). It returns almost everything via dividend + buyback (though the buyback at {{mult2_hoy}} tangible book destroys some tangible book value per share).
Projections
| Metric | FY23 | FY24 | FY25 | TTM | +1A | +2A | +3a | +4A | +5a |
|---|---|---|---|---|---|---|---|---|---|
Net revenue | $46.3 bn | $53.5 bn (+16%) | $58.3 bn (+9%) | $66.2 bn (+18%) | $66.8 bn (+1%) | $67.3 bn (+1%) | $67.9 bn (+1%) | $70.1 bn (+3%) | $72.3 bn (+3%) |
Net income | $8.5 bn | $14.3 bn (+68%) | $17.2 bn (+20%) | $21 bn (+35%) | $19.5 bn (-7%) | $18.2 bn (-7%) | $16.9 bn (-7%) | $17.4 bn (+3%) | $18 bn (+3%) |
Diluted EPS | 22.87 | 40.54 (+77%) | 51.32 (+27%) | 54 | 55.492 (+3%) | 57.025 (+3%) | 58.6 (+3%) | 61.67 (+5%) | 64.9 (+5%) |
ROTCE | 8.1% | 13.5% (+5pp) | 16.0% (+2pp) | 16.8% | 16.2% (-4%) | 15.6% (-4%) | 15.0% (-4%) | 15.0% (+0%) | 15.0% (+0%) |
The % are the annual (year-over-year) growth: each year —historical and projected— vs the prior one; the TTM (trailing 12m) vs the TTM of a year ago, to avoid overlapping windows. The historicals are exact figures from the official filings; the projected years come from the year-by-year model (the intermediate years without their own series are interpolated between the anchors). The projected columns (+1y…+5y) are 12-month windows counted from the TTM close (30-jun-2026): the projection starts from the most recently reported data, not the fiscal year. The projected base is realistic and unbiased — the risk discount is applied at the end, via the required return. The rationale for each metric is in the (i).
Growth quality
- Real, not nominal. Growth (revenue +9% in 2025) is real, driven by the capital-markets boom + AWM. No monetary inflation distorting it.
- Growth engine. The cyclical investment banking/trading boom (not extrapolable) + AWM's durable fee stream (the most stable) + the buyback. Cyclical, not secular.
- Cyclicality. Capital-markets revenue is cyclical; EPS swings 2.6×. Today in an M&A/trading boom — the key metric is whether it holds or reverses.
Moat strength
The business and its moat
The business: what it does and how it makes money
Goldman Sachs is, first and foremost, an investment bank: it advises on mergers and acquisitions, underwrites equity and debt issuance, and acts as an intermediary (trading) in fixed income, currencies, commodities (FICC), and equities markets. That business —Global Banking & Markets, 71% of revenue— is the engine, but also the source of the volatility: investment banking fees ($9.3bn in 2025) rise and fall with the M&A and issuance cycle, and trading ($31bn combined) with market volatility and flows. The second business is Asset & Wealth Management (29%): it earns management fees on $3.6bn of assets under supervision (a record, +15%), plus private banking and alternatives — a much more durable, capital-light fee stream, which GS is looking to grow (ROE target ~17-19%).
The third segment, Platform Solutions, is consumer — and it is in retreat. After years of losses in Marcus/cards, GS sold GreenSky and the GM card, and is exiting the Apple Card. That transformation —returning to the core of investment banking + management and retreating from consumer banking— is the operating thesis that de-risked the story and lifted profitability. GS returns almost everything it earns: in 2025 it returned $16.8bn (buyback $12.4bn + dividend $4.4bn), roughly 100% of net income to common.
Scale and competitive position
Goldman is the most prestigious investment banking franchise in the world: #1 or top-2 in M&A advisory, a leader in underwriting, and one of the largest FICC and equities trading desks. In investment banking it competes with Morgan Stanley (its closest rival), JPMorgan, and the advisory boutiques (Evercore, Centerview); in trading, with JPMorgan, Morgan Stanley, and the large houses. Its edge is the brand, decades of corporate relationships, talent (hence the retention awards), and the scale of the trading desk, which gives it flow information and the capacity to intermediate large transactions.
Against the universal banks (JPMorgan, Bank of America), GS lacks a low-cost deposit franchise — it funds itself more in the market, a more expensive and more cyclical model. Its 16% ROTE is achieved with a mix heavily dependent on capital markets. The bet on growing Asset & Wealth Management is, in part, to add the durable fee stream that smooths that cyclicality — but AWM is still only 29% of revenue.
The moat: why it is hard to compete
Goldman's moat is reputational and talent-based, more than structural. The 'Goldman Sachs' brand opens boardroom doors: when a company sells, goes public, or makes a large acquisition, GS is on the short list on the strength of its prestige and relationships. That relational capital, plus talent (a machine for recruiting and retaining top bankers and traders) and the scale of the trading desk (which gives it flow, information, and the capacity to take the other side of large transactions), are real barriers for an entrant. Systemic-bank (GSIB) regulation also protects incumbents.
But it is a narrower, more fragile moat than a universal bank's. There is no low-cost deposit franchise anchoring funding; there are no strong switching costs (a corporate client engages several banks); and talent is mobile (bankers leave with their relationships — hence the $80M retention awards). ROTE, though good, is volatile. The moat's direction is favorable in one respect (AWM is growing, adding a durable fee stream), but the core remains a capital-markets business, cyclical by nature.
Moat direction and threats
The strategic direction is favorable: the consumer-banking retreat (which was destroying capital) + growth in Asset & Wealth Management (the durable fee stream, AUS at $3.6bn en route to $750bn alternatives targets by 2030) improve the mix and lift ROTE toward a more stable mid-teens. The bull case is that GS consolidates as a top-quality investment bank + asset manager with a sustained 15-16% ROTE and a permanent premium multiple.
The threats are cyclicality and valuation. Cyclicality: 69% of revenue (investment banking + trading) depends on the M&A/issuance cycle and market volatility — today in a boom (record Q2'26), but EPS already fell from $59 (2021) to $23 (2023) in a prior cycle; extrapolating the peak is the error. Valuation: the stock trades at 3× tangible book value on peak earnings. And there are idiosyncratic risks: governance (the $80M retention awards to Solomon and Waldron drew a strong shareholder backlash), litigation legacy (1MDB), and regulatory capital risk (the Basel III Endgame, though re-proposed on more benign terms).
Business / sector quality
- Relevant in 10 years (recurrence). Investment banking + trading + asset management are essential and GS is the leader; it will still be in use in 10 years. But the revenue is transactional/cyclical, not recurring like a subscription.
- Pricing power. Moderate: the #1 brand gives it pricing power in advisory fees, but trading is price-competitive and management fees are under pressure from indexing.
- Moat and its direction. Narrow moat (reputational + talent + trading scale), favorable direction (AWM is growing, adding a durable fee stream), but more fragile than a universal bank's (no low-cost deposits, mobile talent).
- Cyclicality. The disconfirmation test: 69% of revenue tracks the capital-markets cycle; EPS swings 2.6× ($23 to $59). Today at a peak — extrapolating it is the cyclical's classic error (§2).
Capital adequacy (CET1)
CET1 of 12.5% vs the regulatory requirement with buffers of 11.4% → +1.1pp of excess: room to absorb stress and return capital (dividend + buybacks).
Credit quality
Under CECL the reserve anticipates the expected losses over the entire life of the loan; the charge-offs are the losses already realized. A low charge-off rate (~0.7%) versus the through-cycle rate (~1%) indicates a healthy loan book — but it is a point in the cycle, not a permanent floor.
Company health / solvency
- !Capital adequacy (CET1)CET1 12.5% (+1.1pp above the minimum)
- !Reserve coverage (allowance / loan book)1.00% of the loan book
- ✓Credit quality (net charge-off rate)0.60% annually
- ✓Value creation (ROTCE − 10% bar)+6pp
- ✓Funding (deposit base)Deposits $561 bn
A traffic-light interpreted by the method (not generic): float (negative WC) adds up, capex is judged by incremental ROIC (malinvestment test), and a lender is not subjected to corporate solvency. The (i) shows the derivation of each number.
ⓘ Has a loan book: credit health (delinquency, coverage, normalized CECL, RoA/RoE) is evaluated separately, not with corporate solvency.
Health — balance sheet risks
- Capital. CET1 12.5% (~110bp above the 11.4% requirement as of Mar-2026 — the capital buffer compressed from 14.3% at Dec-2025 on asset growth + the buyback), passed 2026 CCAR. Well capitalized, but the buffer is tighter than it appears.
- Funding. Deposits $561bn, but funding is more market-based than a universal bank's (without the low-cost deposit franchise) — more expensive and more cyclical.
- Credit. Credit is benign and wholesale (the 2025 provision was a net release of −$1.1bn); GS's real risk is market/counterparty, not consumer charge-offs. Litigation legacy (1MDB among them; ~$1.4bn of aggregate firm-wide exposure).
Who runs it
- The $80M retention awards to Solomon and Waldron (Jan-2025, vesting through 2030) took Solomon's 2025 pay to $118.9M (740:1 vs the median employee) — a governance flashpoint that drew a strong backlash (say-on-pay at 66%).
- The strategic transformation: GS's consumer-banking retreat (sold GreenSky and the GM card, is exiting the Apple Card) to refocus capital on investment banking + management — de-risking the story and lifting ROTE.
- Capital allocation: returned $16.8bn in 2025 (buyback $12.4bn + dividend $4.4bn), ~100% of net income to common; the nuance is that buying back at {{mult2_hoy}} tangible book destroys some tangible book value per share.
Capital allocation — indicators
The quality test: it reinvests at 16% vs the 10% bar (market opportunity cost) → each $1 retained creates value.
Sources and uses of cash
How cash comes in and how it is deployed. In green, the business's own cash (the owner-FCF it generates and reinvests); in gray, the float and credit — customer and funding money, which is not the shareholder's.
GS returns almost everything it earns: in 2025 it returned $16.8bn (buyback $12.4bn + dividend $4.4bn), roughly 100% of net income to common, with a $40bn buyback authorization outstanding. The key nuance: buying back at {{mult2_hoy}} tangible book value is expensive and destroys some tangible book value per share — at this price, the dividend (growing aggressively from a low payout) is a better use than the buyback. Reinvestment goes toward growing risk-weighted assets (RWA) and AWM.
Shares — ownership and dilution
Who owns the shares — the alignment and whether there is a controlling shareholder.
Management / capital allocation
- Skin in the game. Insiders 0.55% (dispersed ownership, no founder). ⚠️ The $80M retention awards (Solomon + Waldron) drew a strong shareholder backlash; pay tied to TSR, not ROTCE.
- Capital allocation. The transformation (exit from consumer) was well-executed; it returns ~100% of earnings. But buying back at 3× tangible book is expensive and destroys some tangible book value per share.
- Execution / integrity. Executed the strategic pivot (the consumer-banking retreat). ⚠️ The governance flashpoint ($80M awards, say-on-pay 66%) + the litigation legacy (1MDB) weigh on shareholder-facing integrity.
Why it is not cheap
- It is not cheap: it trades at 3× tangible book value on peak earnings.
- The market has already priced in the thesis: the transformation (exit from consumer) + the capital-markets boom + a sustained premium multiple are all in the price.
- No missing buyers or motivated sellers are identifiable — the characteristic discount of an opportunity does not exist; the expected return (-0%/year) is negative because it is a double peak (earnings × multiple).
Goldman Sachs is not a case of a cheap stock — it is the opposite: a quality business priced for perfection, and the textbook example of stacked optimism in a financial. The why-cheap taxonomy does not apply: there are no missing buyers (it is an ultra-covered blue chip) and no motivated sellers (there was no one-off that sank the result — on the contrary, the 2025 one-offs inflated it). The price reflects that the market embraced the transformation (exit from consumer) + the capital-markets boom. The perception-reality gap runs the opposite way from a bargain: the market is paying today a peak multiple (3× tangible book) on peak earnings (ROTE inflated by the cycle + one-offs), and normalizing both to mid-cycle, the expected return is negative. With no identifiable source of discount, the analytical edge is missing: value discipline says to wait for a capital-markets cycle trough, not to buy at the peak.
Return asymmetry — risk/reward
The annual return (CAGR at 5 years) in each scenario, with the total period return below — the margin of safety made visual: upside range wide, downside range narrow.
Even in the bear scenario, the return holds at -6%/year (-28% total): the margin of safety protects the downside. The bull (+6%/year, +34% total) exceeds it comfortably — a favorable asymmetry, with a narrow downside range and a wide upside range.
Bear case — disconfirmation
- Capital-markets cycle downturn: the M&A/issuance/trading boom reverses (as in 2022-2023, when EPS fell from $59 to $23); ROTE falls to ~11% and net income to common declines sharply.
- The multiple compresses (~1.3× tangible book / ~10× P/E in the bear case) from today's 3× — compression alone, without still-higher earnings, is a large headwind; paying a peak multiple on peak earnings is the double risk.
- The narrower moat bites: without a low-cost deposit franchise, the more expensive, cyclical funding pressures results in a high-rate or stress environment; talent (bankers) leaves with its relationships.
- Idiosyncratic risks: the governance backlash ($80M awards), the litigation legacy (1MDB among them; firm-wide reasonably possible exposure is ~$1.4bn), and a regulatory capital tightening that reduces ROTCE.
Bull case — the thesis for
- The improved franchise consolidates: the consumer-banking retreat + growth in Asset & Wealth Management (AUS at $3.6bn, alternatives to $750bn by 2030) lift ROTE to a stable mid-teens, meriting a permanent premium multiple.
- The capital-markets cycle stays strong: the M&A recovery (the 'conviction cycle') + IPOs + strong trading hold longer than normalized, and today's earnings turn out not to be as peak as they appear.
- Basel III Endgame relief (re-proposed on more benign terms) frees up capital → more buybacks + a higher ROTCE.
- The market sustains a premium multiple (2.2× tangible book or more) for the improved franchise + the mid-teens ROTE + the brand — justifying a valuation close to today's price.
Risks — what breaks the base case
- Capital-markets cycle. The central disconfirmation test: earnings swing 2.6× ($23 to $59); today at a peak. A cycle downturn + multiple reversion is the largest risk to the return.
- Double-peak valuation. 3× tangible book (vs a 1.19× historical level) × peak earnings. Any disappointment re-rates sharply. Paying a peak multiple on peak earnings is stacked optimism (§2).
- Governance. The $80M retention awards (Solomon + Waldron) + say-on-pay at 66% + pay tied to TSR more than ROTCE + succession are governance points to watch.
- Narrow moat + litigation. No low-cost deposit franchise (more expensive/cyclical funding), mobile talent, and the litigation legacy (1MDB among them; ~$1.4bn of aggregate firm-wide exposure).
Lenses — the value investing thinkers
Each thinker's analytical framework applied to our data.
The disagreement starts with the business, not just the price.
- Buffett / Graham Quality + margin of safety
A narrow moat and ROTCE 16% above the 10% bar, but the price sits above value (no margin) → great business, expensive.
- Peter Lynch Growth at a reasonable price (GARP)
A stalwart growing 4% at a PEG of 5.2 → expensive for its growth.
- Joel Greenblatt Cheap and high-return (Magic Formula)
Not applicable — the Magic Formula excludes financials and regulated businesses (EBIT/EV does not capture the operating leverage).
- Howard Marks Perception vs reality + cycle
The price discounts -5% vs our 4%: perception is more pessimistic than reality, with an identified source of the discount.
- Seth Klarman Capital protection (bear scenario)
Bear-scenario floor -10%/yr over 5y (material loss) → risk of permanent capital loss.
- Pat Dorsey Moat strength (Five Rules)
A narrow moat, stable; sources: intangibles, efficient scale → partially passes the Five Rules.
- Aswath Damodaran Expectations implied by the price
Justifying the price requires discounting -5%, within what we project (4%) — the story squares with the numbers.
