Economic capital is the amount of money a financial institution estimates it needs to hold in order to absorb severe, unexpected losses and remain solvent. It is an internal figure, calculated by the firm itself using its own models and data, and it reflects the specific risks on that firm’s books rather than any standardized formula imposed from outside. Most large banks size it to a one-year horizon at a confidence level high enough to defend a target credit rating, and the resulting number shapes decisions on pricing, business-line investment, and how much capital the firm chooses to hold above its regulatory minimum.
What It Covers, and What It Doesn’t
Every lender expects to lose money. A bank with thousands of borrowers knows a predictable share will default each year, and those routine shortfalls get priced into products and funded through loan loss reserves. Economic capital is not about them.
It targets the tail of the loss distribution instead: the low-probability, high-severity events that blow past normal expectations. A credit portfolio might behave fine 99 years out of 100, but in that hundredth year, correlated defaults could produce losses far above the average. Economic capital quantifies the gap between what a firm expects to lose and what it could lose under extreme stress. Formally, it is the difference between a very high percentile of the total loss distribution and the expected loss.
Holding capital equal to that figure is what lets a firm survive a severe shock without becoming insolvent, meaning without reaching the point where losses exceed available resources and obligations can no longer be met.
How Economic Capital Is Calculated
Three inputs drive the number: the confidence level, the time horizon, and the risk types included. Change any one of them and the answer moves.
Confidence Level and Target Credit Rating
The confidence level sets the statistical bar for survival. A 99.95% confidence level means the firm is sizing capital so that, in any given year, the probability of insolvency is 5 in 10,000. Higher confidence means a bigger cushion and a lower chance of failure.
Firms pick the level to match a target credit rating. An institution aiming for AA needs a tighter standard than one content with A, because AA-rated firms are expected to default far less often. The mapping is imprecise. Default probabilities associated with an AA rating run from roughly two to seven basis points depending on the agency and methodology, corresponding to confidence levels anywhere from about 99.93% to 99.98%.1Bank for International Settlements. Range of Practices and Issues in Economic Capital Modelling One illustration submitted to the SEC: an AA-rated institution with a through-the-cycle default probability of three basis points would need capital sufficient to remain solvent at the 99.97% level.2U.S. Securities and Exchange Commission. Memorandum – Meeting With the Securities Industry and Financial Markets Association
The Basel Accord’s minimum regulatory capital implies a survival probability somewhere between 99.0% and 99.9%, but the economic capital that internationally active banks actually hold suggests a solvency standard “substantially higher than 99.9%.”3Bank for International Settlements. Regulatory and Economic Solvency Standards for Internationally Active Banks The regulatory floor sits well below where most large banks voluntarily set their internal bar.
Time Horizon
The horizon is the period over which the firm assumes it cannot raise new capital or meaningfully shed risk. About 80% of banks use one year for their overall economic capital calculation, which aligns with standard reporting cycles and gives management a reasonable window to react after a severe loss.1Bank for International Settlements. Range of Practices and Issues in Economic Capital Modelling
Individual risk types often use shorter measurement windows. Trading desks might calculate market risk over a single day or a two-week period, then scale those figures up to the one-year horizon using the square-root-of-time rule or more sophisticated statistical scaling methods that account for fat-tailed return distributions.4ResearchGate. Value-at-Risk Time Scaling for Long-Term Risk Estimation
Risk Types
Economic capital calculations are typically organized around three core categories: credit, market, and operational risk. Some firms also quantify capital for liquidity, reputational, and strategic risk, though those are harder to model and less consistently included.5Society of Actuaries. Practical Methods for Aggregating Banks Economic Capital
Credit risk is the potential for losses when borrowers fail to repay. Modeling it requires estimating default probabilities and loss severity for each exposure, then capturing how defaults across a portfolio might cluster during downturns.6Federal Reserve Bank of Chicago. Loss Given Default and Economic Capital Market risk is losses from adverse swings in interest rates, equity prices, foreign exchange, or commodities, usually measured first with a short-horizon Value-at-Risk and then scaled up. Operational risk covers losses from internal process failures, systems breakdowns, human error, and external events like fraud. Operational risk is the hardest category to model, because severe losses are rare enough that historical data is sparse and scenario analysis has to carry much of the weight.
Combining the Risks Into One Number
Once a firm calculates economic capital separately for each risk type, it has to combine them. Simply adding the numbers overstates total risk, because credit, market, and operational losses rarely hit their worst-case levels at the same moment. The reduction from the summed figure to the true combined number is called the diversification benefit, and capturing it is one of the main reasons firms adopt economic capital frameworks at all.5Society of Actuaries. Practical Methods for Aggregating Banks Economic Capital
Figuring out how much to reduce requires modeling how the risks move together: how likely is a credit shock to coincide with a trading loss and an operational failure? The most common technique uses a mathematical tool called a copula, which links the individual loss distributions and captures their tendency to become more correlated in a crisis, producing a combined loss distribution from which the diversified figure is drawn.5Society of Actuaries. Practical Methods for Aggregating Banks Economic Capital
In practice the diversification benefit typically runs about 8% to 12% of the undiversified total, depending on the confidence level and risk mix. At higher confidence levels the benefit shrinks, because extreme events are more likely to strike multiple risk categories together.5Society of Actuaries. Practical Methods for Aggregating Banks Economic Capital The correlation assumptions behind that benefit are among the most consequential and most uncertain inputs in the entire framework.
How It Differs From Regulatory and Accounting Capital
Economic capital is often confused with two other capital measures that serve very different purposes.
Regulatory capital is the minimum a bank must hold to comply with rules set by external authorities, principally the Basel Committee on Banking Supervision. Basel III requires Common Equity Tier 1 of at least 4.5% of risk-weighted assets, Tier 1 of at least 6%, and total capital of at least 8%.7Bank for International Settlements. Definition of Capital in Basel III In the United States, large banks face additional requirements including a stress capital buffer of at least 2.5% and, for globally systemic institutions, a surcharge of at least 1.0%.8Board of Governors of the Federal Reserve System. Annual Large Bank Capital Requirements The formulas are standardized, designed to protect the system as a whole. Economic capital uses the firm’s own models to reflect its specific portfolio. A bank concentrated in low-risk loans may calculate an economic capital well below regulatory minimums; a bank concentrated in risky assets may land far above. The gap tells management where regulation is over- or under-charging relative to the firm’s own view.
Accounting capital is shareholders’ equity on the balance sheet under GAAP or IFRS. It looks backward, using historical cost and accrual rules to show what the book value of assets minus liabilities is today. Economic capital looks forward, estimating what could happen under severe stress using probabilistic models and current market values. Accounting capital satisfies reporting requirements; economic capital informs management’s internal view of whether the firm is genuinely solvent on a risk-adjusted basis.
What Firms Do With the Number
The usefulness of economic capital lies in the decisions it drives. Three applications dominate.
The first is capital allocation across business lines. Economic capital provides a way to distribute the firm’s capital budget based on each unit’s contribution to overall risk. A trading desk with volatile returns gets charged more capital than a retail lending division with stable, predictable losses, even if both produce the same revenue. The allocation is based on the marginal risk each unit adds to the corporate portfolio, so it reflects diversification effects a cruder method would miss. Without it, a firm can inadvertently subsidize risky businesses with capital that should support more productive activities.
The second is risk-adjusted performance measurement, most commonly through Risk-Adjusted Return on Capital, or RAROC. A business unit’s risk-adjusted profit (revenue minus operating costs and expected losses) is divided by its allocated economic capital.9BBVA. Economic Profit and Risk-Adjusted Return on Economic Capital A unit creates shareholder value only when its RAROC exceeds the firm’s cost of equity, which acts as the hurdle rate. A loan portfolio returning 15% on its economic capital looks attractive when the cost of equity is 10%; a trading desk returning 8% is destroying value under the same benchmark. RAROC makes that comparison possible across businesses with very different risk profiles.
The third is risk-based pricing. For a commercial loan, the interest rate must cover funding costs, expected losses, the cost of holding economic capital against unexpected losses, and a return sufficient for shareholders. If expected revenue cannot produce a RAROC above the hurdle rate, the product is either repriced or discontinued. Every transaction gets priced to compensate for its actual risk, not just its expected outcome.
Where the Models Fall Short
Economic capital frameworks have real blind spots, and the 2007–2009 financial crisis exposed several of them. Anyone relying on an economic capital figure should understand where it bends.
Tail data is scarce. Economic capital is fundamentally about rare events, and rare events by definition leave little evidence. Validating a 99.97% confidence level would require thousands of years of clean data, which no institution has. Tail estimates end up depending heavily on assumptions about the shape of loss distributions rather than on observed losses. The Basel Committee has acknowledged that validating these models is “intrinsically difficult” given the need to evaluate extreme quantiles with limited data.1Bank for International Settlements. Range of Practices and Issues in Economic Capital Modelling
Correlations are unstable. The diversification benefit that makes economic capital attractive is also one of its biggest vulnerabilities. Correlations between risk types tend to spike in crises: credit losses surge while market values collapse and operational failures multiply. Models calibrated on normal-period data can dramatically overstate diversification precisely when the benefit is most needed. The Basel Committee has noted that correlation estimates “depend heavily on explicit or implicit model assumptions” and that “due consideration for conservatism” is warranted.1Bank for International Settlements. Range of Practices and Issues in Economic Capital Modelling
The models are procyclical. During booms, risk measures look benign and the models produce reassuringly small capital requirements. During downturns, risk is reassessed as high, capital requirements spike, and the increased demands can push firms to cut lending at the moment credit is most needed. The Bank for International Settlements has described this pattern as risk being “underestimated in booms and overestimated in recessions,” creating a feedback loop that amplifies real economic swings.10Bank for International Settlements. Procyclicality of the Financial System and Financial Stability – Issues and Policy Options
Many assumptions cannot be tested. Building the model requires choices about how losses behave, how risks interact, and how markets function under stress. Many of those choices are, as the Basel Committee has put it, “no more than untested hypotheses” that “may appear coherent and plausible” but cannot be verified empirically.1Bank for International Settlements. Range of Practices and Issues in Economic Capital Modelling Two banks with identical portfolios can produce materially different economic capital estimates simply by making different modeling choices. The output is an informed estimate, not a precise measurement.
A Parallel in Insurance
Economic capital is not confined to banking. The European Union’s Solvency II framework imposes an analogous requirement on insurers through the Solvency Capital Requirement, defined as a Value-at-Risk calculation at a 99.5% confidence level over one year, meaning insurers must hold enough capital to survive a 1-in-200-year loss event.11PMC. Solvency II Solvency Capital Requirement for Life Insurance The 99.5% threshold is lower than what large banks target internally, which typically runs between 99.9% and 99.98%. The difference reflects the distinct risk profiles of insurance versus banking: insurance liabilities are generally longer-dated and less exposed to sudden run dynamics. The underlying logic is the same. Translate all material risks into a single capital figure calibrated to a chosen survival probability over a chosen horizon.