Expected Default Frequency (EDF) Calculator
Expected default frequency (EDF) is a credit-risk estimate: the probability that a firm defaults on its obligations over a set horizon, stated as a percentage. This calculator runs the Merton structural model, turning asset value, a default point and asset volatility into a distance to default and then into an EDF, and it can run that model backwards from a target EDF. It is a credit tool, not a counting tool: if you need the expected count of events for a chi-square test or a probability model, use the expected frequency calculator instead, because that page answers a different question.
Structural results are theoretical Merton probabilities: the normal curve maps distance to default straight into a percentage. Moody’s published EDF figures are calibrated against historical defaults instead and are capped, so the two can differ a lot for the same firm. The sections below explain the gap.
Quick Answer
With assets of $10,000,000, a default point of $9,000,000, asset volatility of 20%, zero drift and a 1-year horizon, the distance to default is 0.43 and the expected default frequency is 33.48%. That is the theoretical Merton reading: the normal curve converts a cushion of less than half a standard deviation into roughly a 1-in-3 default chance. A published Moody’s EDF for the same firm would be built from historical default data instead and would not be this number.
Key takeaways
- EDF here is a model output, not a rating: distance to default first, then a normal-curve conversion into a percentage.
- Leverage and volatility do the damage together. A firm can survive high debt at low volatility, or high volatility at low debt, but not both.
- The reverse solvers answer the questions analysts actually get asked: how much asset value, how little debt, or how much volatility sits behind a quoted EDF.
- Treat any single EDF as a ranking, not a verdict. Compare it with peers, other horizons and other models before pricing or declining credit.
- Quick Answer
- How to use the EDF calculator
- The Merton formulas behind EDF
- Worked example: $10M of assets against a $9M default point
- Reading an EDF: bands and the DD scale
- Theoretical EDF vs published EDF vs a Z-score
- Using EDF in loan pricing and limits
- What competing calculators miss
- Limits to keep in view
- Frequently asked questions
- Add This Calculator to Your Website
- References
How to use the EDF calculator
Pick a model tab, enter the firm’s figures, then click Calculate. Results stay hidden until you do.
- Structural model (default): leave Solve for set to EDF and enter V, D, volatility, drift and horizon. You get the EDF plus the distance to default behind it.
- Run it backwards: switch Solve for to V, D or volatility, enter a target EDF, and the tool returns the input that would produce exactly that EDF under the same assumptions.
- PD over horizon: if you already have a per-period default probability, this tab compounds it: EDF = 1 − (1 − PD)T. A 2.5% yearly PD over 3 years gives 7.31%.
Two input habits matter. The default point is not total debt: the convention used in the KMV literature, set out by Crosbie and Bohn in their Moody’s KMV methodology work, is short-term debt plus half of long-term debt. And clearing the drift or horizon fields is safe: the tool reads a blank drift as 0% and a blank horizon as 1 year, which the worked example below relies on.
The Merton formulas behind EDF
The structural model treats the firm’s assets as a moving target and default as the moment those assets fall to the default point.
Structural EDF = Φ(−DD) × 100, where Φ is the standard normal cumulative distribution function
Cumulative EDF = 1 − (1 − PD)T
The logic comes from Merton’s 1974 paper On the Pricing of Corporate Debt, which valued equity as a call option on the firm’s assets in the Black and Scholes framework. The first term inside DD, ln(V ÷ D), is the cushion: how far asset value sits above the default point, in log terms. The drift term adjusts that cushion for expected asset growth, and dividing by σ√T expresses the whole gap in standard deviations. A DD of 2 means assets could fall two standard deviations and still cover the default point. Converting DD with the normal CDF, the same area-under-the-curve calculation an area to z-score calculator performs, turns that distance into a probability.
One honesty note belongs with the formula. In the full KMV approach, V and σ are not typed in: they are solved iteratively from the firm’s equity value and equity volatility, and the resulting DD is then mapped to an EDF using a large historical default database. This tool takes V and σ as given and uses the normal-curve mapping directly. That makes it a transparent Merton probability calculator, ideal for learning, sensitivity work and quick screening, and not a replica of Moody’s production EDF.
Worked example: $10M of assets against a $9M default point
A firm holds $10,000,000 of assets against a $9,000,000 default point, with 20% asset volatility, no expected drift and a 1-year horizon. The cushion is thin, and the model says so bluntly.
| Step | Arithmetic | Value |
|---|---|---|
| Log cushion | ln(10,000,000 ÷ 9,000,000) | 0.1054 |
| Drift adjustment | (0 − 0.20² ÷ 2) × 1 | −0.0200 |
| Distance to default | (0.1054 − 0.0200) ÷ (0.20 × 1) | 0.43 |
| EDF | Φ(−0.43) × 100 | 33.48% |
The reverse solvers round-trip the same example. Set a target EDF of 33.48% with D at $9,000,000 and volatility at 20%, and solving for V returns $9,999,789.15. Solving for D from $10,000,000 of assets returns $9,000,189.77, and solving for volatility returns 20.00%. The small gaps against the round inputs come from the target being rounded to two decimals, not from solver error.
| Check (real calculator runs) | DD | EDF |
|---|---|---|
| Safer firm: V $15,000,000, D $5,000,000, σ 25%, μ 5%, T 1 | 4.47 | 0.00% |
| Worked example stretched to T 5 | 0.01 | 49.52% |
| Underwater firm: V $8,000,000, D $9,000,000, σ 20%, μ 0%, T 1 | −0.69 | 75.46% |
| Same worked example at σ 40% | 0.06 | 47.47% |
Read the T 5 row carefully. Stretching the horizon on a thin cushion pushes EDF toward 50%, because the drift adjustment turns negative and time gives volatility more room to work. Horizon is an assumption, not a detail.
Reading an EDF: bands and the DD scale
An EDF means little until it is placed on a scale. The badge under your result uses these bands, and they are deliberately strict, because they describe a theoretical probability rather than a calibrated one:
| EDF | Band | Typical reading |
|---|---|---|
| Up to 0.10% | Minimal | Very strong credits on a one-year view. |
| 0.10% to 1% | Low | Solid investment-grade territory. |
| 1% to 5% | Elevated | Speculative-grade risk: price it, monitor it. |
| 5% to 20% | High | Distress is a realistic scenario inside the horizon. |
| Over 20% | Distress zone | The cushion is under about one standard deviation. |
The DD scale underneath those bands is fixed by the normal curve, and it is worth memorising a few points:
| Distance to default | −1 | 0 | 0.5 | 1 | 1.5 | 2 | 2.5 | 3 |
|---|---|---|---|---|---|---|---|---|
| Theoretical EDF | 84.13% | 50.00% | 30.85% | 15.87% | 6.68% | 2.28% | 0.62% | 0.13% |
Two properties follow. First, EDF moves fastest exactly where weak firms live: between DD 0 and DD 2, each half-step of DD cuts the EDF roughly in half. Second, a theoretical EDF can pass 50% once DD turns negative, which simply means assets already sit below the default point. Moody’s published EDF measures behave differently: as their methodology paper EDF Measures as Overlays to Credit Ratings sets out, published corporate EDFs are calibrated to realised default rates and capped at 50% (35% for financial firms), so a 75.46% reading like the underwater check above has no published equivalent.
Theoretical EDF vs published EDF vs a Z-score
Three different credit answers are often confused with each other. They are not interchangeable:
This tool: theoretical Merton EDF
Inputs V, D, σ, μ, T go in directly and the normal curve does the mapping. Strength: fully transparent, instant sensitivity and reverse solves. Limit: no historical calibration, so levels run hot for weak firms.
Moody’s KMV EDF (CreditEdge)
V and σ are inferred from equity prices, then DD is mapped through a proprietary default database covering decades of firms. Strength: levels track realised defaults. Limit: subscription product, method only partly visible, EDF capped.
Accounting scores (Altman Z-score family)
Financial ratios are weighted into a score with cut-off zones. Strength: works for private firms with no traded equity. Limit: book values move slowly, so the signal lags market stress by quarters.
For portfolio work, the honest use of this page is ranking and stress testing: sort borrowers by EDF, shock volatility or asset value, and watch who migrates bands. For a portfolio-level quality snapshot, the average credit quality calculator answers the adjacent question, and for tail losses beyond a threshold, the expected shortfall calculator in this batch covers the loss side rather than the default side.
Using EDF in loan pricing and limits
The standard pricing identity is expected loss = EDF × loss given default × exposure at default. On a $1,000,000 loan with an EDF of 2% and a 45% loss given default, expected loss is $9,000, and that $9,000 is the floor the interest margin has to clear before capital costs and profit.
That identity is why the reverse solvers earn their place. A credit committee that caps EDF at 2% can ask this tool what default point a borrower would need to stay inside the cap, or what asset value a restructuring has to restore. The same expected-value logic, probability times payoff summed across outcomes, is what a conditional expected value calculator formalises for general uncertain payoffs. Keep one discipline in all of it: an EDF built on a guessed volatility is a guessed EDF, so run the volatility solver and the sensitivity checks before quoting a price.
What competing calculators miss
Free Merton tools are scarce, and the ones that exist aim at a different job:
- Full option-pricing labs (xdage Merton lab): rich balance-sheet modelling with equity value, debt value and Greeks. Powerful, but the default probability is one line in a large dashboard, with no reverse solve from a target EDF.
- Excel templates (Ryan O’Connell Merton template): nine credit metrics including risk-neutral and physical default probabilities, but locked in a spreadsheet: no instant web run, no banding, no worked web example to check against.
- Software documentation (MathWorks Merton functions): exact method references for MATLAB users, including the time-series calibration. There is no calculator a non-programmer can use.
- Explainers (university notes, FasterCapital, Bradford Lynch): the DD formula and the iterative calibration are explained well, and one widely read page prints the DD numerator with a plus sign on the volatility term where the physical-measure formula needs (μ − σ²/2). None of them lets you test the numbers.
- The gap this page fills: forward EDF with DD shown, three reverse solvers, a cumulative-PD tab, explicit bands, and an on-page statement of where the theoretical number stops and Moody’s calibrated EDF begins.
Limits to keep in view
The model assumes asset values follow a lognormal path with constant volatility, a single default point, and default only at the horizon. Real firms jump, refinance, and default early. Asset value and volatility are also estimates in any real use, and small errors in volatility move weak-firm EDFs a long way, as the 20% versus 40% check above shows (33.48% against 47.47% on identical leverage). The Basel framework’s internal-ratings guidance, summarised in the Basel Framework, exists precisely because regulators expect PD estimates to be validated against realised defaults rather than trusted from a formula alone. Use this tool the same way: as a sharp first estimate that earns a second look, not as the second look itself. In Legacy mode (the PD over horizon tab), the per-period default probability must sit between 0% and 100%: entries above 100% are rejected as invalid, and the cumulative EDF is capped at 100%, because a probability cannot exceed 100%.
Frequently asked questions
Expected default frequency is the estimated probability that a firm defaults on its obligations within a set horizon, usually one year, expressed as a percentage. An EDF of 2% means the model expects about 2 defaults per 100 similar firms over that horizon. The term is associated with the Moody’s KMV approach, which estimates it from the firm’s asset value, debt and asset volatility.
First compute the distance to default: DD = [ln(V / D) + (mu – sigma squared / 2) x T] / (sigma x square root of T). Then EDF = Phi(-DD) x 100, where Phi is the standard normal cumulative distribution function. For example, $10,000,000 of assets against a $9,000,000 default point with 20% volatility, zero drift and a 1-year horizon gives DD = 0.43 and EDF = 33.48%.
Distance to default is the number of standard deviations by which the firm’s expected asset value exceeds its default point at the horizon. A DD of 0.43 means the cushion is less than half a standard deviation, which the normal curve maps to a 33.48% default probability. Higher DD means a safer firm, and a negative DD means assets already sit below the default point.
They measure the same event, and EDF is best read as one specific way of estimating a probability of default: the market-based, structural way that runs through distance to default. Probability of default is the general term and can also come from credit ratings, accounting models or scorecards. This calculator’s PD over horizon tab shows the related compounding step, turning a per-period PD into a cumulative figure.
No. Moody’s KMV infers asset value and volatility from equity prices, then maps distance to default to an EDF using a large historical database of actual defaults, and caps published corporate EDFs at 50%. This calculator takes asset value and volatility as inputs you supply and maps distance to default with the normal distribution directly. It is the theoretical Merton probability: transparent and useful for ranking and sensitivity work, but not a calibrated Moody’s figure.
On the bands this page uses, up to 0.10% is minimal, 0.10% to 1% is low, 1% to 5% is elevated, 5% to 20% is high, and over 20% is a distress-zone reading. Strong investment-grade firms usually sit in the first two bands on a one-year view. Any band only means something against peers in the same industry and cycle, so compare before you judge.
Volatility is the yardstick distance to default is measured in, so doubling it roughly halves the distance. In the worked example, holding assets at $10,000,000 and the default point at $9,000,000, raising volatility from 20% to 40% moves EDF from 33.48% to 47.47%. The effect is strongest for firms near their default point and flattens for very safe firms, which is why two firms with identical leverage can carry very different EDFs.
Through expected loss: expected loss = EDF x loss given default x exposure at default. A $1,000,000 loan with an EDF of 2% and a loss given default of 45% carries an expected loss of $9,000, and the interest margin has to clear that figure before capital costs and profit. Lenders also use EDF bands to set approval limits and monitoring intensity.
They are unrelated measures that share two words. Expected default frequency is a credit probability: the percentage chance a firm defaults over a horizon. Expected frequency is a statistics count: how many observations you would expect in a category under a hypothesis, used in chi-square tests and probability models. This page covers the credit measure only.
Add This Calculator to Your Website
Embed this calculator on your own site with the code below. Pick a size to preview it live, then copy the code.
Powered by Calculator Zilo
References
- Merton, R. C. On the Pricing of Corporate Debt: The Risk Structure of Interest Rates, Journal of Finance, 1974. Accessed October 21, 2026.
- Black, F., and Scholes, M. The Pricing of Options and Corporate Liabilities, Journal of Political Economy, 1973. Accessed October 21, 2026.
- Dwyer, D. W., Kocagil, A. E., and Stein, R. M. The Moody’s KMV EDF RiskCalc v3.1 Model: Next-Generation Technology for Predicting Private Firm Credit Risk (Moody’s KMV modeling methodology). Accessed October 21, 2026.
- Campbell, J. Y., Hilscher, J., and Szilagyi, J. In Search of Distress Risk, NBER Working Paper No. 12362 (published in the Journal of Finance, 2008). Accessed October 21, 2026.
- Bharath, S. T., and Shumway, T. Forecasting Default with the Merton Distance to Default Model (method summary and estimation notes), Review of Financial Studies, 2008. Accessed October 21, 2026.
- Vassalou, M., and Xing, Y. Default Risk in Equity Returns, Journal of Finance, 2004. Accessed October 21, 2026.
- Altman, E. I. Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy, Journal of Finance, 1968. Accessed October 21, 2026.
- Bank for International Settlements. The Basel Framework (internal ratings and probability of default standards). Accessed October 21, 2026.
- Board of Governors of the Federal Reserve System. SR 11-7: Guidance on Model Risk Management. Accessed October 21, 2026.
- National Institute of Standards and Technology. NIST Digital Library of Mathematical Functions, Chapter 7: Error Functions, Dawson’s and Fresnel Integrals (normal CDF approximations). Accessed October 21, 2026.
- MathWorks. Estimate default probability using time-series version of Merton model (mertonByTimeSeries documentation). Accessed October 21, 2026.