Unlocking RAROC: How Risk-Adjusted Return on Capital Works
By Markets & Investing DeskUpdated Nov 10, 2026
Definition
Risk-adjusted return on capital (RAROC) measures profitability by deducting expected losses from income and dividing the result by the capital held against unexpected losses.
Key takeaways
- The formula is revenue minus expenses minus expected loss plus income from capital, divided by capital.
- In credit, expected loss equals probability of default times loss given default times exposure at default.
- Raising default probability from 1% to 3% cuts the example loan's RAROC from 19.6% to 8.4%.
- A loan below the hurdle rate, often tied to the cost of equity, is rejected or repriced.
Risk-adjusted return on capital (RAROC) is a profitability measure that subtracts the losses an activity is expected to generate from its income, then divides the result by the capital needed to absorb losses beyond that expectation. A loan, trading desk or business line is judged by the return it earns on the capital its risk consumes, not by its gross revenue.
The RAROC Formula
RAROC = (r − e − el + ifc) ÷ c
- r (revenue): interest income, fees and other income the activity generates.
- e (expenses): operating costs plus funding costs. For a bank, interest paid on deposits and other borrowings belongs here, because it is the cost of funding the loan.
- el (expected loss): the average loss anticipated over the measurement period, usually one year.
- ifc (income from capital): the return earned by investing the capital held against the position, calculated as capital × risk-free rate. It is added back because that capital is not idle.
- c (capital): the amount set aside to cover unexpected losses. Institutions often derive it from a value-at-risk or similar model at a chosen confidence level.
The numerator is therefore a risk-adjusted profit. The denominator is the capital that risk requires.
Estimating Expected Loss
Expected loss is the probability of a loss multiplied by its size. In credit, this is commonly broken into three parts:
Expected loss = probability of default (PD) × loss given default (LGD) × exposure at default (EAD)
PD is the chance the borrower defaults within the period. LGD is the share of the exposure lost after recoveries from collateral. EAD is the amount owed at the time of default. Expected loss covers average outcomes, and capital covers the gap between that average and a severe but plausible loss.
Worked Example: RAROC for a Loan
Hypothetical figures for illustration only. A bank considers a one-year, $10 million corporate loan.
| Item | Assumption | Amount |
|---|---|---|
| Interest income | 6% on $10,000,000 | $600,000 |
| Fees | Flat | $20,000 |
| Funding cost | 4% cost of funds | $400,000 |
| Operating expenses | Flat | $50,000 |
| Expected loss | PD 1% × LGD 45% × EAD $10,000,000 | $45,000 |
| Economic capital | Set by the bank’s model | $800,000 |
| Income from capital | $800,000 × 4% risk-free rate | $32,000 |
Numerator: $620,000 − $400,000 − $50,000 − $45,000 + $32,000 = $157,000
RAROC: $157,000 ÷ $800,000 = 19.6%
Now raise the probability of default to 3% and keep everything else the same. Expected loss becomes $135,000, the numerator falls to $67,000, and RAROC drops to 8.4%. In practice the capital requirement would usually rise with the higher default risk too, pushing RAROC lower still.
When to Reject: The Hurdle Rate
A RAROC figure has meaning only against a benchmark, typically the institution’s hurdle rate, often tied to its cost of equity, the return shareholders require. Suppose the hurdle here is 12%. The first loan (19.6%) clears it and adds value on a risk-adjusted basis. The second (8.4%) does not, so the bank would reject it or reprice it with a higher rate, larger fees or more collateral.
How RAROC Is Used
- Loan pricing: Banks work backward from a target RAROC to find the minimum rate a borrower must pay. This is common in corporate lending, SME loans and trade finance.
- Capital allocation: Capital is limited. RAROC shows which business lines earn the most per dollar of risk capital, so management can direct capital toward them.
- Performance measurement: Comparing desks or units on raw profit rewards whoever takes the most risk. RAROC puts a low-risk custody business and a high-risk trading desk on the same scale.
- Product analysis: For a credit card portfolio, high interest income may be offset by high expected losses. RAROC shows whether the margin survives once those losses and the capital they require are counted.
Non-bank companies also apply the framework to gauge how credit, market and operational risk affect their returns.
Origins at Bankers Trust
Bankers Trust built RAROC in the late 1970s under a team led by Dan Borge. As the bank shifted toward securities and derivatives, it needed to compare the profitability of trading and wholesale activities whose risks differed widely, which simple return on capital could not do. Other banks built their own versions through the 1980s, often under different names, but RAROC remains the best-known label.
RAROC vs. RORAC vs. RARORAC
The three acronyms describe where the risk adjustment is applied:
- RAROC adjusts the return for risk by deducting expected loss, and divides by economic capital.
- RORAC (return on risk-adjusted capital) uses an unadjusted return over risk-adjusted capital. The risk sits in the denominator only.
- RARORAC (risk-adjusted return on risk-adjusted capital) adjusts both the numerator and the denominator.
Institutions do not use these labels consistently. Some call a RORAC-style calculation RAROC, so check the formula before comparing figures from different sources.
How RAROC Differs From Other Return Measures
Return on investment (ROI) divides profit by the amount invested and ignores how likely losses are. Two loans with the same interest margin show the same ROI even if one borrower is far more likely to default. RAROC charges the riskier loan its higher expected loss and a larger capital requirement.
The Sharpe ratio also adjusts for risk, but differently. It measures excess return over the risk-free rate per unit of standard deviation, which suits portfolios with observable return histories. RAROC is built for positions such as loans, where risk shows up as default and recovery rather than daily price swings. It also links the result to the capital a firm must hold.
Limitations
- Model dependence: PD, LGD and economic capital are estimates. Optimistic inputs produce a flattering RAROC, and small changes in assumptions can move the result a lot.
- Data demands: Reliable loss estimates require extensive default and recovery history. That history may be thin for new products, small borrowers or unusual markets.
- Tail risk: A high RAROC does not mean an activity is safe. A position with a small chance of a very large loss can score well if the capital model understates that tail.
- Single-period view: The standard calculation covers one period and can miss risks that build over a multi-year relationship.
- Not a full decision: Strategy, concentration, liquidity and regulatory capital rules can override what the ratio suggests.
FAQs
Can individual investors use RAROC?
Rarely. The formula needs default probabilities, loss estimates and an economic capital figure, inputs that banks model internally. Individual investors usually rely on measures such as the Sharpe ratio, which work with return and volatility data that is easier to obtain.
The Bottom Line
RAROC matters to banks and other firms that need to price loans, allocate limited capital, and compare business lines with very different risks on one scale. The ratio is useful only against a hurdle rate and only as reliable as its loss and capital estimates. Treat it as one input alongside strategy, concentration, liquidity, and regulatory rules, and check how a figure was calculated before comparing it across sources.
