GSH Financial mark GSH Financial
(431) 305-0352 Book a Consultation
GSH FinancialGSH Financial

Home / Financial Tools / Multi-Model Bankruptcy & Going-Concern Prediction Comparator

Graduate-Level Modeling · Credit Risk Research

Multi-Model Bankruptcy & Going-Concern Prediction Comparator

Every bankruptcy prediction model was built on a different sample, with different variables, in a different decade. The interesting question isn't which one is right, it's whether they agree.

How To Use This Model

Reading This Tool

Enter one set of financial statement figures and this tool scores the company against four independently developed bankruptcy prediction models at once.

Altman, Zmijewski, Springate and Ohlson were each estimated on different historical samples with different variables, seeing where they agree, and where they don't, tells you more than trusting any single score in isolation.

Financial Statement Inputs ($M)

Model Agreement

-

Altman Z-Score

-

Zmijewski Probability

-

Springate Score

-

Ohlson O-Score Probability

-

Distress Signal Across All Four Models (0 = Safe, 1 = Distress)

-

-

Why Four Models Can Legitimately Disagree

-

What Model Agreement Actually Buys You

No single model was estimated on a sample that looks like every company. When independently built models trained on different eras, industries and variable sets all point the same direction, that convergence is doing real work, it's much less likely to be an artifact of one model's specific blind spot.

A Word On Base Rates

All four of these models were built on samples with a much higher bankruptcy rate than the general population of companies, since researchers oversample failed firms to get enough distress cases to estimate from. Applied to an average, healthy firm, all four models tend to overstate the true probability of failure, treat the outputs as relative risk rankings, not calibrated real-world probabilities.

Models Flagging Distress

Raw Scores Against Each Model's Own Cutoff

ModelRaw ScoreDistress CutoffVerdict
Credit Risk & Bankruptcy Prediction Research

The Core Formulas

Altman Z = 1.2(WC/TA) + 1.4(RE/TA) + 3.3(EBIT/TA) + 0.6(MVE/TL) + 1.0(Sales/TA)
Zmijewski X = −4.3 − 4.5(NI/TA) + 5.7(TL/TA) − 0.004(CA/CL), P = 1/(1+e−X)
Springate Z = 1.03(WC/TA) + 3.07(EBIT/TA) + 0.66(EBT/CL) + 0.4(Sales/TA)
Ohlson O = −1.32 − 0.407·SIZE + 6.03(TL/TA) − 1.43(WC/TA) + 0.0757(CL/CA) − 1.72·OENEG − 2.37(NI/TA) − 1.83(FFO/TL) + 0.285·INTWO − 0.521·CHIN, P = 1/(1+e−O)

When To Actually Use This Model

  • Teaching comparative bankruptcy prediction research and model validation in an advanced credit risk or empirical accounting research course.
  • Cross-checking a single-model going-concern flag before it drives an audit opinion or lending decision.
  • Building intuition for how differently structured models (accounting ratios, market value inputs, dummy variables) can each miss different failure patterns.

Key Assumptions & Limitations

  • Ohlson's original SIZE variable scales total assets by a GNP price-level index that resets the model to a specific base year, simplified here to a direct log of total assets, treat the Ohlson output as directional, not literally reproducing the 1980 published model.
  • All four models were estimated decades ago on North American public company samples, industry, country, and era shifts can degrade any of them without warning.
  • None of these models incorporate qualitative factors, management quality, litigation risk, or industry-specific stress that a real going-concern assessment would weigh heavily.

Foundational References

Altman, E. I. (1968). Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. The Journal of Finance, 23(4), 589-609.

Zmijewski, M. E. (1984). Methodological Issues Related to the Estimation of Financial Distress Prediction Models. Journal of Accounting Research, 22, 59-82.

Ohlson, J. A. (1980). Financial Ratios and the Probabilistic Prediction of Bankruptcy. Journal of Accounting Research, 18(1), 109-131.

Springate, G. L. V. (1978). Predicting the Possibility of Failure in a Canadian Firm (Master's thesis). Simon Fraser University.

Want the market-implied credit view too?

Open The Merton Distance-To-Default Model →