Why Debt Is Raw Material, Not Financing: The FIG Paradigm

Introduction

Every coverage group in investment banking has its own sector-specific knowledge: Healthcare bankers learn FDA approval pathways, TMT bankers learn subscriber metrics, Energy bankers learn reserve economics. But FIG is the only coverage group where the foundational framework of corporate finance itself needs to be rebuilt from the ground up. The reason comes down to a single conceptual shift: for financial institutions, debt is not a financing choice. It is the raw material of the business.

This distinction is not academic. It is the reason that enterprise value, EBITDA, unlevered free cash flow, and WACC-based discounting all produce meaningless or misleading results when applied to banks, insurance companies, and other financial institutions. Understanding why these standard tools break, and what replaces them, is the single most important conceptual leap for anyone entering FIG. Interviewers test this concept relentlessly because it separates candidates who have genuinely internalized FIG thinking from those who have simply memorized a list of differences.

In every non-financial industry, the textbook framework treats a company as having two separable components: the operating business (which generates revenue, incurs costs, and produces cash flow) and the capital structure (the mix of debt and equity used to finance the operating business). This separation is the foundation of modern valuation.

When you calculate enterprise value, you add equity value and net debt together to arrive at the total value of the operating business, independent of how it is financed. When you compute EBITDA, you measure operating profitability before interest expense specifically because interest is treated as a financing cost, not an operating one. When you build an unlevered DCF, you project free cash flow to the firm (stripping out interest and debt payments) and discount at WACC precisely because you want to value the business separately from its financing decisions.

Enterprise Value (EV)

The total value of a company's operating business, calculated as equity value (market capitalization) plus total debt minus cash and equivalents. EV represents what a buyer would pay for the entire business: taking on both the equity claim and the debt obligations while receiving the cash. The concept relies on the assumption that debt is separable from operations, meaning you could theoretically pay off all the debt and the underlying business would continue functioning. This assumption holds for manufacturers, retailers, technology companies, and nearly every other sector, but it fails completely for financial institutions.

This framework works beautifully for an industrial company. Take a manufacturer with $500 million in equity value and $200 million in net debt. Enterprise value is $700 million. If you paid off the debt entirely, the factories would still run, the customers would still buy, and the operating cash flows would remain largely unchanged (minus the tax shield). The debt is genuinely separable from the business.

Now apply that same logic to a commercial bank. JPMorgan Chase has roughly $4.2 trillion in total assets, funded primarily by approximately $2.4 trillion in customer deposits, plus hundreds of billions in wholesale borrowings, federal funds, and other liabilities. Its equity base is approximately $340 billion.

If you tried to "remove the debt" from JPMorgan the way you would for an industrial company, you would be removing the deposits and borrowings that fund every loan, every mortgage, every securities portfolio, and every trading position. There would be no bank left. The deposits and borrowed funds are not financing the business. They are the business, in the same way that steel is the raw material for an automaker or crude oil is the raw material for a refinery.

This is not a minor technical nuance. It invalidates the entire standard valuation toolkit in three specific ways.

Enterprise Value Becomes Meaningless

Enterprise value equals equity value plus net debt. For JPMorgan, that would mean adding $340 billion in equity to roughly $3.8 trillion in "net debt" (total liabilities minus cash and equivalents). The resulting enterprise value of over $4 trillion tells you nothing useful about the operating value of the business. It is simply restating the balance sheet size, which reflects how much raw material (deposits and borrowings) the bank has gathered, not how profitably it deploys that raw material.

The deeper problem is conceptual: enterprise value assumes you can toggle the debt level independently from the business. For a bank, the debt level is the business scale. A bank with $100 billion in deposits operates a fundamentally different (and larger) business than one with $10 billion in deposits, not because it chose a more aggressive capital structure, but because it has gathered more raw material to deploy.

EBITDA Loses All Meaning

EBITDA strips out interest expense, depreciation, and amortization to isolate operating profitability. For a financial institution, this produces a nonsensical result.

Interest expense is an operating cost. When a bank pays 2.5% on deposits to fund loans yielding 6.5%, the deposit interest is the cost of raw material, not a capital structure decision. Stripping it out would be like calculating EBITDA for a manufacturer but adding back the cost of goods sold. The resulting number would tell you nothing about profitability.

Depreciation and amortization are trivial. Banks and insurance companies are asset-light businesses in the physical sense. They do not have factories, heavy equipment, or massive tangible asset bases requiring depreciation. Adding back D&A (which is already minimal) while also adding back the primary operating cost (interest) produces a metric that bears no relationship to actual operating performance.

Net Interest Margin (NIM)

The difference between a bank's interest income (earned on loans, securities, and other earning assets) and its interest expense (paid on deposits, borrowings, and other funding), expressed as a percentage of average earning assets. NIM is the banking equivalent of gross margin for a manufacturer. The industry-wide NIM was approximately 3.28% in Q4 2024. A bank with a higher NIM either earns more on its assets, pays less for its funding, or both. NIM is the single most important profitability metric for spread-based financial institutions and the starting point for any bank income statement analysis.

Unlevered DCF and WACC Break Down

The standard DCF projects unlevered free cash flow (FCFF), which removes debt-related cash flows to isolate cash generated by the operating business. You then discount at WACC, which blends the cost of debt and equity to reflect the overall cost of capital. Both steps fail for banks.

You cannot calculate FCFF because there is no meaningful "unlevered" version of a bank. Removing interest expense and debt repayments from a bank's cash flows removes the core business. And WACC is problematic because the cost of debt for a bank is not a financing cost to be weighted and blended; it is the cost of raw material. Using WACC to discount bank cash flows would be like using a weighted average of equity cost and steel cost to value an automaker.

Because the standard tools fail, FIG has developed an entirely separate valuation methodology built on one principle: value equity directly, rather than trying to value the total enterprise and subtract debt.

This means two things in practice. First, all FIG valuation multiples are equity-based. Second, the intrinsic valuation model (equivalent to the DCF in other sectors) discounts cash flows to equity holders at the cost of equity, not cash flows to the firm at WACC.

Price / Tangible Book Value (P/TBV)

P/TBV is the foundational FIG multiple. It measures what the market pays for each dollar of a bank's hard equity after stripping out goodwill and intangible assets from prior acquisitions. The logic: a bank's balance sheet is marked much closer to fair value than an industrial company's (financial assets have observable market prices, loans are subject to impairment testing under CECL), so tangible book value is a more meaningful reference point than it would be for a semiconductor company or an airline.

Current P/TBV multiples illustrate the range. JPMorgan Chase trades at approximately 2.6x tangible book value, reflecting its industry-leading ROTCE of roughly 20-22%. Bank of America trades at approximately 1.9x TBV with an ROTCE around 15%. Wells Fargo targets 17-18% ROTCE and trades around 1.8x TBV. Regional and community banks span a wider range: from 0.8x TBV for underperformers to 2.1x for high-quality franchises.

The relationship holds globally. European banks spent much of 2015-2022 trading below 0.7x TBV as negative interest rates compressed profitability to well below cost of equity. The re-rating since then has been dramatic: UniCredit, which traded below 0.4x TBV in 2020, reached approximately 1.9x TBV by early 2026 after delivering a 19.2% ROTE in 2025. Deutsche Bank, despite years of restructuring, still trades around 0.8x book value, reflecting lower returns. The persistent gap between US and European multiples (and within Europe itself) reinforces the same principle at work.

The key driver of P/TBV is ROTCE relative to the cost of equity. A bank earning 18% ROTCE against a 10% cost of equity is creating significant excess value on each dollar of equity and deserves a premium to tangible book. A bank earning 8% ROTCE against a 10% cost of equity is destroying value and will trade at a discount. This relationship can be expressed formally through the justified P/BV ratio: P/BV = (ROE - g) / (CoE - g), where banks with higher ROE (or ROTCE) relative to cost of equity command higher multiples.

The Dividend Discount Model (DDM)

The DDM is the FIG equivalent of the unlevered DCF. Instead of projecting free cash flow to the firm and discounting at WACC, you project the total cash that can be distributed to equity holders (dividends plus share buybacks, collectively called total shareholder payout) and discount at the cost of equity.

The DDM captures a critical regulatory reality that has no parallel in corporate finance: total payout is constrained by regulatory capital requirements. In each projection year, the bank earns net income, which increases retained earnings and capital ratios. But it can only distribute the excess above regulatory minimums (plus any management buffer). If the bank needs to retain capital to fund loan growth or absorb higher risk weights, the payout shrinks. If loan growth slows or capital ratios are comfortable, payout expands.

This makes the DDM a balance sheet-driven model. You start by projecting balance sheet growth (loan growth, deposit growth, securities portfolio changes), derive the income statement from the balance sheet (NII is a function of earning asset yields and funding costs), calculate regulatory capital ratios, and then determine how much can be paid out. The terminal value uses a Gordon Growth Model: terminal payout divided by (cost of equity minus long-term growth rate).

The debt-as-raw-material concept is most intuitive for commercial banks, but it applies in modified forms across every FIG sub-sector. Understanding these extensions demonstrates the depth of FIG thinking that interviewers reward.

Insurance companies operate a parallel model through float. When policyholders pay premiums, the insurer holds those funds until claims are paid, which may be years or decades later (especially for long-tail liability lines like workers' compensation or medical malpractice). During that holding period, the insurer invests the float and earns investment income. The premiums are the insurance equivalent of deposits: they are the raw material of the investing business, not a financing mechanism. This is why Warren Buffett has described Berkshire Hathaway's insurance float (over $173 billion as of year-end 2024) as "better than free" funding, because policyholders effectively pay the insurer to hold their money through the underwriting profit on top of investment income.

Specialty finance companies like BDCs and mortgage REITs depend on borrowed funds as their primary business input. A mortgage REIT borrows through repurchase agreements (repo) at short-term rates and invests in mortgage-backed securities earning higher long-term yields. The leverage is not a financing decision in the traditional sense; it is the business model. Removing the borrowings eliminates the entity's ability to operate.

Asset managers present a middle case. Traditional and alternative asset managers earn fee-based revenue from AUM, not spread income. Their business model does not depend on debt as raw material, and they can be valued on AUM-based multiples or EV/EBITDA. However, some asset managers (particularly those with significant balance sheet co-investment or seed capital) have meaningful leverage embedded in their operations. And wealth management platforms that custody client assets face regulatory capital requirements that echo banking constraints. The result is a hybrid: pure fee-based asset managers can use standard valuation, while balance-sheet-intensive financial firms within the asset management umbrella require equity-based methods.

The debt-as-raw-material paradigm does not just change how you value financial institutions in isolation. It reshapes how you analyze every M&A transaction in FIG.

In standard M&A, accretion/dilution analysis measures whether the deal increases or decreases EPS for the acquirer, and the financing mix (cash, debt, stock) is the primary lever. In bank M&A, the standard accretion/dilution framework is modified to focus on tangible book value dilution and the earn-back period, as discussed in the FIG deal flow article. Because the acquirer is paying a premium to TBV (and because goodwill is deducted from regulatory capital), the central question becomes: how long until the combined entity's TBV per share recovers to its pre-deal level through retained earnings and synergies?

The capital impact assessment, unique to FIG M&A, directly follows from the debt-as-raw-material paradigm. Since the acquirer cannot "deleverage" the target post-acquisition (you cannot remove a bank's deposits without removing the business), the combined entity must maintain regulatory capital ratios against the full combined balance sheet. Pro forma CET1, Tier 1, and Total Capital ratios determine deal feasibility. A deal that pushes the combined entity below regulatory minimums is either restructured (reducing the premium, changing the consideration mix, or divesting assets) or abandoned.

The deposit premium analysis is another FIG-specific M&A tool that flows directly from the paradigm. When an acquirer pays a premium for a target bank's deposits, it is paying for the raw material acquisition cost advantage that those deposits represent. A bank with $10 billion in core deposits at an average cost of 1.5% has a structural funding advantage over one relying on wholesale funding at 4.5%. The deposit premium (typically 2-8% of core deposits) capitalizes that cost advantage into the purchase price, creating a core deposit intangible asset that is amortized over 7-10 years. This entire analytical layer exists because deposits are business inputs, not balance sheet liabilities to be minimized.

The debt-as-raw-material concept is not merely theoretical. It is the lens through which every piece of FIG analysis is conducted: from the bank income statement (where NII, the spread on raw material, is the first line) to the balance sheet (where liabilities are the funding base, not a burden to be reduced) to M&A (where capital impact and deposit premiums replace leverage analysis and synergy-driven EPS accretion as the primary deal metrics).

Internalizing this paradigm is what allows a FIG banker to look at a bank's $50 billion deposit base and see an asset, not a liability. It is what makes the DDM feel intuitive rather than arbitrary. And it is what enables you to answer the most common FIG interview question with the conceptual depth that separates prepared candidates from those who have only skimmed the surface.