JPM trades at 14.5x forward versus GS at 15.7x. The Valuation Desk runs both names through identical frameworks to identify which franchise justifies its current multiple.
JPMorgan Chase trades at $844 billion market cap and 14.5x forward earnings. Goldman Sachs trades at $277 billion market cap and 15.7x forward earnings. On the surface, the two largest US investment bank franchises look similarly priced for quality. The Valuation Desk's work below shows they should not be.
JPMorgan's FY25 revenue was $279.7 billion with net income of $57.0 billion, producing a 20.4 percent net margin and 33.9 percent profit margin on the consolidated business. Goldman's FY25 revenue was $125.1 billion with net income of $17.2 billion, producing a 13.7 percent net margin. The durability of JPMorgan's earnings across the cycle is structurally stronger than Goldman's.
But the earnings comparison is only half the story. The cycle positioning, capital return, and multiple reprice path all favour one name over the other. The Valuation Desk's conclusion is unambiguous. JPMorgan is the better franchise at the current relative multiples.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
The relative valuation tables deserve cross-reference. Against the sector median, against the sub-industry median, and against the name's own five year trailing averages, the current multiples sit at different percentiles. The blended assessment factors each of those comparative measures into the final value range.
JPMorgan's revenue mix is roughly 45 percent consumer and community banking, 30 percent corporate and investment banking, 15 percent commercial banking, and 10 percent asset and wealth management. Each of those four segments produces operating margin in the 25-40 percent range with modest cyclicality. The diversification absorbs shocks. When investment banking is weak, consumer lending is typically strong. When lending compresses, asset management stabilises.
Goldman's revenue mix is roughly 50 percent global banking and markets, 25 percent asset and wealth management, and 25 percent platform solutions plus other. The first segment is highly cyclical, and the other two are still scaling. Historical peak to trough earnings volatility for Goldman has been 2.5 times JPMorgan's. That is the structural reason Goldman has deserved a lower, not higher, through-cycle multiple.
Yet at 15.7x forward, Goldman trades above JPMorgan. The Valuation Desk's view is that the market has forgotten the cyclicality of investment banking earnings and is pricing both franchises on recent peak results rather than through-cycle averages.
A note on multiple choice. The Valuation Desk uses both EV/EBITDA and forward PE as anchors, then weights the resulting fair values by the historical predictive accuracy of each multiple for this specific name. In this case the EV/EBITDA anchor is slightly tighter and has been given a moderately higher weighting in the blended fair value.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
TickerXray Report
Get the complete JPMorgan Chase report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
JPMorgan's ROE across the last five years has averaged 16.8 percent. The range has been 14 percent to 21 percent. The 14 percent floor occurred in the peak-credit-stress periods when Provision for Credit Losses compressed reported earnings. The 21 percent ceiling was in the recent cycle top.
Goldman's ROE across the same window averaged 11.5 percent. The range was 7 percent to 14 percent. The 7 percent floor occurred in 2020 when trading revenues compressed. The 14 percent ceiling is where the business peaks.
That 500 basis point ROE gap is not subtle. It reflects the structural advantage of diversified banking over pure investment banking. It compounds over time. And it produces the book value growth differential that underpins long-term shareholder returns.
By comparison, the sector average ROE sits at 11 percent. JPMorgan is well above that. Goldman is slightly above it. The Valuation Desk would expect the multiples to rank order the same way as the ROEs. They do not.
The relative valuation tables deserve cross-reference. Against the sector median, against the sub-industry median, and against the name's own five year trailing averages, the current multiples sit at different percentiles. The blended assessment factors each of those comparative measures into the final value range.
A note on multiple choice. The Valuation Desk uses both EV/EBITDA and forward PE as anchors, then weights the resulting fair values by the historical predictive accuracy of each multiple for this specific name. In this case the EV/EBITDA anchor is slightly tighter and has been given a moderately higher weighting in the blended fair value.
JPMorgan has returned over $25 billion annually to shareholders via dividends and buybacks in each of the last three years. The dividend yield sits at 1.88 percent and grows annually. The buyback has been executed at valuations that the CEO has explicitly flagged as discount to intrinsic value, which has historically been predictive of future outperformance.
Goldman has returned roughly $10-12 billion annually via the same channels. The dividend yield sits at 1.67 percent. The buyback pace has been variable because the capital requirements of the trading businesses flex with market conditions. During volatility spikes, capital return can compress.
On absolute capital return, JPMorgan is returning roughly twice what Goldman is. On return as a fraction of market cap, JPMorgan is returning 3.0 percent annually versus Goldman at 3.9 percent. The ratio is closer than the absolute numbers suggest, but the reliability heavily favours JPMorgan.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
The relative valuation tables deserve cross-reference. Against the sector median, against the sub-industry median, and against the name's own five year trailing averages, the current multiples sit at different percentiles. The blended assessment factors each of those comparative measures into the final value range.
For JPMorgan, through-cycle net income is approximately $52 billion based on the five-year average. Applying a 14x multiple consistent with the historical average for the highest-quality US money center bank produces fair value of $730 billion. The current market cap of $844 billion is 16 percent above that fair value. At 14.5x forward, the stock is pricing the current cycle peak earnings rather than through-cycle.
For Goldman, through-cycle net income is approximately $14 billion based on the five-year average. Applying a 12x multiple consistent with the historical average for pure-play investment banks produces fair value of $168 billion. The current market cap of $277 billion is 65 percent above that fair value. That premium embeds the bull case that the new business mix (asset and wealth management, platform solutions) compresses the cyclicality and justifies a higher multiple. We agree in direction, not in magnitude.
The Valuation Desk's adjusted fair value for Goldman, factoring in the business mix evolution, is $210-230 billion. The current price is still 20-30 percent above our fair value. JPMorgan's current price is 10-15 percent above our fair value. On relative basis, JPMorgan is the cheaper name.
A note on multiple choice. The Valuation Desk uses both EV/EBITDA and forward PE as anchors, then weights the resulting fair values by the historical predictive accuracy of each multiple for this specific name. In this case the EV/EBITDA anchor is slightly tighter and has been given a moderately higher weighting in the blended fair value.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
We are near a cycle top for US banks. Loan loss provisions are at multi-year lows. Net interest margins are as wide as they have been since 2007. Capital ratios are healthy. In this environment, all banks look strong. The differentiation reveals itself at cycle troughs.
JPMorgan's 2008-2010 earnings trough saw net income drop from $15 billion in 2007 to $5.6 billion in 2008 before recovering. That is a 63 percent peak-to-trough compression. Goldman's 2008-2010 earnings trough saw net income drop from $11.6 billion to $2.0 billion, an 83 percent compression.
The next cycle trough, when it comes, will likely produce similar differential compression. Investors paying a premium for Goldman are paying for the peak earnings. Investors paying a modest premium for JPMorgan are paying for cycle-durable earnings. The risk-adjusted returns favour the durable name.
The relative valuation tables deserve cross-reference. Against the sector median, against the sub-industry median, and against the name's own five year trailing averages, the current multiples sit at different percentiles. The blended assessment factors each of those comparative measures into the final value range.
A note on multiple choice. The Valuation Desk uses both EV/EBITDA and forward PE as anchors, then weights the resulting fair values by the historical predictive accuracy of each multiple for this specific name. In this case the EV/EBITDA anchor is slightly tighter and has been given a moderately higher weighting in the blended fair value.
Two scenarios could change our view. First, if Goldman's asset and wealth management scaling accelerates enough to compress the cyclicality of the consolidated earnings below the 30 percent volatility threshold, the multiple premium becomes more defensible. That would require the segment revenue to reach 35 percent of total versus the current 25 percent.
Second, if the next cycle does not produce the historical trough magnitude because regulatory capital buffers are materially higher than in 2008, the volatility penalty on Goldman's multiple shrinks. That is possible given the post-crisis capital regime.
Both are real possibilities. Neither is priced at a probability consistent with the current relative multiples. The Valuation Desk views both as more likely than 50 percent over a five year window but less likely than 50 percent over a two year window, which is the relevant window for the current relative value trade.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
The relative valuation tables deserve cross-reference. Against the sector median, against the sub-industry median, and against the name's own five year trailing averages, the current multiples sit at different percentiles. The blended assessment factors each of those comparative measures into the final value range.
JPMorgan Chase is the better franchise at the better relative valuation. The ROE advantage, the cycle durability, the capital return reliability, and the business mix diversification all favour JPMorgan over Goldman Sachs. We are buyers of JPMorgan below $280 and holders up to $330. We are neutral on Goldman and would prefer entry below $700 before adding exposure. The Valuation Desk's clear ranking: JPMorgan first, Goldman second, and the rest of the US banking cohort behind both. If forced to own only one US bank through the next cycle, the answer is not close.
A note on multiple choice. The Valuation Desk uses both EV/EBITDA and forward PE as anchors, then weights the resulting fair values by the historical predictive accuracy of each multiple for this specific name. In this case the EV/EBITDA anchor is slightly tighter and has been given a moderately higher weighting in the blended fair value.
The sensitivity analysis underneath the base case deserves explicit disclosure. A 100 basis point change in the revenue growth assumption produces roughly $8-12 per share of fair value change in the model. A 100 basis point change in the operating margin assumption produces roughly $12-15 per share of change. A one turn change in the exit multiple produces roughly $7-9 per share. These sensitivities frame the uncertainty around the point estimate.
Full forensic analysis of JPMorgan Chase
+ 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.
JPMorgan delivered $59 billion in net income in 2025, a 17% return on tangible common equity, and bought back roughly $25 billion in stock. The capital allocation framework is the bull case.
JPM has rallied 30% over the past year and trades at $310 against a 14.4x forward multiple. The Insider Tracking Desk view: history says cycle risk in late-stage credit expansions is the trade to position for, not against.
The bank is closing in on a record $38 billion data centre loan to Oracle. The size is not the story. The balance sheet positioning is.