Net income recovered to $17.2 billion in 2025, the highest reported result since 2021. Return on tangible equity sits north of 14 percent, the buyback yield is approaching 4 percent, and the multiple has barely budged. The argument writes itself.
Goldman Sachs ended 2025 with $17.2 billion of net income, a 14 percent return on tangible equity, and a buyback authorisation that has retired roughly 13 percent of share count over the past three years. The market is paying 17.7 times trailing earnings for that performance. By comparison, JPMorgan trades at approximately 14 times, but with a lower ROTE and a more diversified, lower-multiple earnings stream.
The Capital Desk thesis is simple. Goldman has done the structural work. The trading business, after a multi-year platform investment, is generating record results at a more durable run-rate than the 2021 cohort. The investment banking franchise is rebuilding from the 2023 trough. Asset & Wealth Management revenue is now a meaningful diversifier. The private credit footprint is real and growing. And capital is being returned at a level the buyside has not yet fully discounted.
We are buyers at a $640 share price with a $700 target. The catalyst is a single quarter of buyback authorisation expansion combined with another print of 14 percent ROTE. We expect both within the next three quarterly reports.
The Capital Desk has covered Goldman across three full credit cycles. The pattern is consistent. The market mispriced the franchise at the 2009 trough, mispriced it again at the 2016 trough, and is mispricing it again here in 2026. Each time, the catalyst that closed the gap was a single quarter of better-than-feared trading revenue combined with a buyback authorisation expansion. The setup repeats with high precision.
The investment Goldman made between 2019 and 2023 in scaling the global markets platform is now generating results. The trading revenue line for 2025 sits at multi-year highs. Importantly, the mix has shifted. Equities now contributes a larger share of trading revenue than it did at the 2021 peak. The franchise is less FICC-dependent and therefore less rate-cycle-dependent.
The asset and wealth management business has scaled to a $3 trillion AUM run-rate. That fee revenue compounds steadily and is far less volatile than the trading book. Management fees from AWM now contribute approximately 30 percent of consolidated pre-tax income. That share was below 20 percent in 2019. The mix shift is doing real work on the consolidated return profile.
The private credit business, which is harder to see in the public disclosure, has grown to a meaningful balance sheet exposure. Goldman is one of the three or four largest non-bank lenders in the US middle market. The yield differential between bank lending and private credit is currently 200-300 basis points. Goldman is monetising both sides of that spread, originating in private credit and packaging in syndication.
The consumer-banking misadventure has been wound down. Apple Card runoff is complete. Marcus deposits have been migrated. The drag on consolidated ROE from consumer is now zero. That is a material positive that the market is still adjusting to.
TickerXray Report
Get the complete Goldman Sachs report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
Investment banking advisory and underwriting revenue compressed materially in 2022 and 2023 as deal volumes collapsed. Goldman's advisory revenue for 2023 was the lowest in a decade. The 2024 and 2025 results show a clear recovery profile, with advisory revenue rebuilding to approximately $4 billion per year and equity capital markets activity adding another $1.5-2 billion.
The Capital Desk view on the investment banking cycle is that the recovery is still in the early innings. Sponsor activity has not yet normalised. Strategic M&A volumes for 2025 sat well below the long-run average as measured by Refinitiv. The implied deal pipeline based on banker discussions and announced transactions suggests that 2026 advisory revenue should expand 15-25 percent from 2025 levels.
The investment banking franchise compounds in two ways. Direct fees on advisory and underwriting. And the financing follow-on that attaches to those mandates. A typical large M&A deal produces $30-50 million of advisory fees plus another $40-80 million of debt and equity financing revenue. The full mandate economics are roughly double the headline advisory fee. That stacking effect is what makes investment banking a structurally attractive franchise for the global money centre banks.
Goldman's competitive position in advisory remains industry-leading. The 2025 league tables show Goldman number one in global M&A advisory by deal value for the eighteenth consecutive year. That franchise durability is the kind of moat that justifies a premium multiple.
Goldman's capital return program is the cleanest in the large-cap banks. The 2025 program returned $13.4 billion of capital to shareholders through dividends and buybacks combined. Against a $285 billion market capitalisation, that produces a total shareholder yield of approximately 4.7 percent. The current dividend yield sits at 1.62 percent. The buyback yield therefore implies roughly 3.1 percent. Both are sustainable at current capital generation rates.
The Capital Desk framework values bank stocks on the product of ROTE and the multiple of book value the market is willing to pay. At 14 percent ROTE and 1.7 times tangible book, Goldman trades at a 13 percent discount to the implied fair multiple under a steady-state framework. That discount is the alpha.
The buyback authorisation is the lever the Capital Desk watches most carefully. Goldman repurchased $11 billion in 2025 at an average price below $590, a 9 percent discount to today's price. That is capital allocation competence. By comparison, the bank universe average buyback authorisation execution price was within 3 percent of the average market price across the year. Goldman delivered better than the peer average.
The Tier 1 capital ratio finished 2025 at approximately 14.6 percent, comfortably above the 10.9 percent regulatory minimum. The implied excess capital is approximately $20 billion. That capital is available for further buyback expansion or strategic deployment. In either case, it supports shareholder returns.
Investors who lived through the 2021 trading-revenue peak know what happens when extraordinary trading days normalise. The franchise that was earning $21.6 billion in 2021 fell to $8.5 billion of net income by 2023. The lesson was that the 2021 number was a function of unusual market activity, not a steady-state earnings profile.
The Capital Desk argument is that 2025 is structurally different from 2021. The composition of revenue has shifted. AWM is larger. Private credit is larger. Trading is more equities-weighted. Expenses have been disciplined. The 2025 efficiency ratio sits at approximately 60 percent, against the 65-70 percent range that defined the prior cycle peak. That 500-1,000 basis points of operating leverage flows directly to net income.
The scenario analysis works as follows. In a normal trading year, 2026 net income should land in the $16-18 billion range. In a strong trading year, $19-21 billion. In a weak trading year, $13-15 billion. Even the weak case produces a 12 percent ROTE, which keeps the buyback program running at the current pace and maintains capital return yield above 4 percent.
By comparison, JPMorgan's downside scenario produces a 14 percent ROTE because of the deposit funding advantage. Morgan Stanley's downside scenario produces a 12-13 percent ROTE because of the wealth management base. Goldman now sits inside that competitive band rather than being the outlier it was historically.
The most reliable long-term value driver in any bank stock is tangible book value per share growth. Goldman's tangible book value per share has compounded at approximately 11 percent annually over the past five years, reflecting the combination of retained earnings growth and steady share count reduction. That compounding alone produces a respectable equity return profile before any multiple expansion.
The Capital Desk model anchors the equity value to tangible book value times the through-cycle multiple. At today's price of $640, the price-to-tangible-book ratio sits at approximately 1.7 times. Through previous capital-return cycles, when ROTE has been at 13-15 percent for two consecutive years, the franchise has traded at 2.0-2.2 times tangible book. The implied valuation alone, holding tangible book flat, produces a $760-840 share price range.
Layer in the expected 8-10 percent tangible book value growth over the next 24 months and the framework produces a fair value range of $820-920 by mid-2027. That is 28-44 percent upside from the current price. The math is not the surprise. The surprise is that the market has not yet priced it.
This is one of those situations where the boring metric is the one that matters. Tangible book per share growth has been the most predictive single variable for bank equity returns since the post-2009 reset. Goldman is at the top of the bank universe on that metric. The price action will catch up.
A frequent point of investor scepticism on Goldman is that trading is too volatile to support a quality multiple. Across the 2018-2024 window, trading revenue ranged from $16.5 billion to $24.7 billion. That range covers roughly $8 billion of consolidated revenue volatility. Big numbers.
The Capital Desk look-through is different. The mix of trading revenue has shifted toward equities and away from FICC since 2021. Equities trading is more correlated with daily client volume and less correlated with the rate cycle. The smoothing effect is real. Across the past eight quarters, equities trading revenue has shown a standard deviation roughly 30 percent below the eight quarters prior. The franchise is producing more consistent results across product lines.
The FICC business has also shifted toward more financing revenue and less pure flow trading. Financing revenue is closer to a spread business than a trading business. It produces a more predictable run-rate. Goldman's financing revenue for 2025 finished at multi-year highs and has been a positive contributor across both 2024 and 2025.
The point is that the trading book today is a structurally different earnings stream than the trading book of 2017 or 2021. The market multiple has not yet adjusted to that shift. That gap is the trade.
The growth of the private credit franchise inside Goldman is the most underappreciated story in the equity. The business has been built quietly through Goldman Sachs Alternatives, the asset management arm that now manages approximately $310 billion in private credit, private equity, real estate, and infrastructure assets. The blended fee rate sits above 80 basis points, and the carry component adds another 30-50 basis points across the cycle.
The scale of the private credit book means Goldman is now competing directly with Apollo, Blackstone Credit, and Ares for sponsor-led financings. The fee economics on the private credit business are structurally more durable than the fee economics on the public market trading business. A $300 billion AUM franchise growing at 15 percent a year, generating 80-100 basis points of fees, produces roughly $4 billion of recurring revenue. That is a material positive on the consolidated valuation framework.
The consensus model still treats Goldman as a trading-heavy investment bank. The current revenue mix says otherwise. Approximately 35 percent of pre-tax income now comes from fee-based, recurring sources. That share was below 25 percent five years ago. The valuation multiple should reflect the higher-quality revenue mix. It does not yet.
Historically, when bank multiples have failed to reflect mix-driven margin and ROE improvements, the catalyst has been a single capital markets day or strategic update from management. The next such update is the 2026 investor day, which the Capital Desk expects in the second half of the year.
Three risks deserve acknowledgement. First, a deep credit cycle would expose the private credit book. The asset class has not yet seen a full default cycle. Loss content could prove higher than current marks suggest. Goldman's private credit underwriting has been conservative, but the asset class is untested.
Second, regulatory capital requirements could increase further. The Basel III endgame proposals continue to evolve. A more punitive final rule would reduce the buyback program by 10-20 percent. The Capital Desk is monitoring the timeline.
Third, trading revenue is inherently volatile. A 2018-style market environment would compress trading revenue meaningfully. Our weak-case scenario already prices that risk, but the downside in any single quarter could be sharp.
We accept these risks. The Capital Desk view is that the capital return yield at current levels compensates for them with room to spare. The Goldman equity offers the most attractive combination of return on capital and shareholder yield in the large-cap bank universe.
The Capital Desk target on Goldman sits at $700-720, implying 10-13 percent upside before considering the dividend and buyback yield, which adds another 4-5 percent annually. The thesis runs on three legs: durable ROTE in the 13-15 percent range, ongoing share count reduction, and multiple expansion as the market re-rates the higher-quality revenue mix.
Across the past four bank capital return cycles (2011-2014, 2014-2017, 2017-2020, 2021-2024), the pattern is consistent. Banks that combine 13 percent or higher ROTE with a 4 percent or higher total shareholder yield have produced 15-22 percent annualised total returns over the subsequent three-year window. Goldman currently checks both boxes.
We are buyers at $640 with confidence. The catalyst to add aggressively is a Q3 2026 buyback authorisation expansion above $15 billion. That would signal management's confidence in the through-cycle earnings profile and would accelerate the share count reduction faster than the model currently assumes.
Full forensic analysis of Goldman Sachs
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
GS just printed a 14.5% revenue growth quarter, the highest in three years. The composition is the issue, not the headline. Trading is doing the work that capital markets cannot.
Across the last three earnings seasons, the gap between Goldman Sachs's operating execution and the broader investment-banking peer set has widened materially. The Capital Desk reads the data and concludes the multiple has not yet caught up.
Goldman's $17.2 billion in net income and 38.3% operating margins are driven by a structural role as the financial intermediary for the AI infrastructure buildout.