Introduction
The excess return model (also called the residual income model) is the second core intrinsic valuation methodology for financial institutions, alongside the dividend discount model. While the DDM values a bank as the present value of future dividends, the excess return model values a bank as its current book value plus the present value of all future returns above the cost of equity. The insight is powerful: a bank's P/TBV multiple is not arbitrary. It is book value (which accounts for the "1.0x") plus a premium or discount that reflects whether the bank creates or destroys value relative to what shareholders require. This decomposition is the theoretical engine behind the entire P/TBV valuation framework and explains why JPMorgan trades above 2.5x book while some banks trade below 0.7x.
The Formula
The excess return model starts from a simple premise: a dollar of book value is worth exactly a dollar if the bank earns its cost of equity on it (no value created, no value destroyed). But if the bank earns above its cost of equity, that dollar of book value is worth more than a dollar because it generates returns in excess of what investors require. The excess return for any period is:
Where \( ROE_t \) is return on equity, \( K_e \) is cost of equity, and \( BV_{t-1} \) is beginning book value. The equity value is then:
In the single-stage version (constant ROE, constant growth), this simplifies to:
The "1" is book value itself. The second term is the present value of excess returns, which determines the premium or discount to book.
- Excess Return (Residual Income)
The excess return is the dollar amount of earnings left over after charging equity capital its required return. If a bank has $100 billion in book equity, earns 15% ROE ($15 billion in net income), and its cost of equity is 10% ($10 billion equity charge), the excess return is $5 billion per year. This $5 billion represents genuine value creation: the bank is generating returns above what shareholders could earn elsewhere at equivalent risk. Capitalizing this excess return as a perpetuity at the cost of equity minus growth rate produces the premium above book value. Conversely, a bank with 8% ROE and 11% cost of equity generates negative excess returns of $3 billion per year on the same $100 billion equity base, explaining why it trades below book value.
How Excess Returns Drive P/TBV
The excess return model provides the theoretical foundation for the entire spectrum of bank P/TBV multiples. The connection is direct and mathematical:
| ROE vs. Cost of Equity | Excess Return | P/TBV | Interpretation |
|---|---|---|---|
| ROE significantly > Ke | Large positive | 2.0x+ | Strong value creation (JPMorgan, Morgan Stanley) |
| ROE moderately > Ke | Moderate positive | 1.0-2.0x | Modest value creation (well-run regionals) |
| ROE = Ke | Zero | 1.0x | Breaking even on capital (industry median) |
| ROE moderately < Ke | Moderate negative | 0.7-1.0x | Mild value destruction (underperforming banks) |
| ROE significantly < Ke | Large negative | < 0.7x | Severe value destruction (distressed banks) |
JPMorgan illustrates the model's explanatory power. With ROTCE of approximately 20% in 2024-2025 and an estimated cost of equity of approximately 11%, JPMorgan's excess return spread is roughly 9 percentage points. This wide positive spread, sustained over many years through scale advantages, a dominant deposit franchise, and diversified fee revenue, produces a large present value of excess returns that justifies the bank's 2.6x+ P/TBV premium.
Citigroup provides the opposite case. For over a decade, Citigroup's ROTCE lingered around 5-8%, below its estimated cost of equity of 11-13%. The negative excess return spread meant the bank was destroying value: shareholders would have been better off if the bank had simply returned all of their equity. The stock traded below 1.0x P/TBV for years, only recently crossing above 1.0x as management's restructuring pushed ROTCE toward 10-11%.
When to Use Excess Return vs. DDM
The DDM and excess return model produce identical results under consistent assumptions. The choice between them depends on practical considerations, not theoretical superiority.
The DDM is preferred when the bank has a long, consistent dividend history with a predictable payout ratio (most large US commercial banks), when the analyst wants to model the explicit link between regulatory capital constraints and distributable cash flow, and when the audience (investors, clients) expects a DDM-based valuation (which remains the industry standard for bank pitch books).
The excess return model is preferred when the bank pays no dividends or has an irregular payout (restructuring situations, banks under ECB dividend restrictions during 2020-2021), when the analyst wants to anchor the valuation on book value (a known, audited number that reduces dependence on long-term projections), or when the objective is to decompose P/TBV into its component parts (book value versus the excess return premium). In practice, many FIG analysts use both models as cross-checks: if the DDM and excess return model produce significantly different values under the same assumptions, there is likely an internal inconsistency in the inputs.
European banks present a particularly instructive case for the excess return model. The entire European banking sector traded below 1.0x P/TBV from 2010 through early 2025 (the EURO STOXX Banks Index only broke above 1.0x P/TBV for the first time in over 15 years in early 2025). The excess return model explains this clearly: European banks earned average ROE of approximately 10% against an estimated cost of equity of 12-17% (per Bloomberg and ECB research), producing negative excess return spreads across the sector. The ECB's negative interest rate policy (2014-2022) compressed net interest margins, and fragmented markets limited scale efficiencies. As European bank profitability improved (consensus ROE reaching approximately 13% for 2026), the excess return spread narrowed, driving the P/TBV recovery. The model's prediction and the market's behavior aligned precisely.
The ROE-P/TBV regression that FIG analysts run across bank peer groups is an empirical application of the excess return model. Banks plotting above the regression line trade at a premium to what their ROE-based excess return would justify; banks below the line trade at a discount. Understanding the excess return model transforms this regression from a statistical exercise into an economically grounded valuation tool.
The excess return model completes the intrinsic valuation toolkit for financial institutions, providing both a standalone valuation methodology and the theoretical explanation for why relative valuation multiples vary across banks. Together with the DDM and the justified P/BV framework, it forms a complete, internally consistent system for valuing any financial institution from first principles.
Free Interview PDF Guide