What it means
Banks and bondholders need to know whether a borrower is likely to repay. Reading a full set of accounts takes time, so researchers looked for a way to condense the information into one number.
The Z-score used five ratios, and the Zeta model extended the idea with more ratios and updated methods. The model examines a company's profitability, earnings stability, ability to cover its debts, accumulated profits, liquidity, capital structure and size.
Each factor is given a weight based on how well it separated failed companies from survivors in historical data. The weighted result is a score, and companies below a cut-off are treated as higher risk.
A key advantage is consistency. Two analysts applying the same model to the same accounts get the same score, which removes a lot of subjectivity.
The score can also be tracked over time, so a falling score warns that a borrower is weakening before a default occurs. The exact weights of the Zeta model were kept proprietary and sold commercially, so the public discussions of it focus on the variables and the approach.
Anyone building a similar model today would estimate their own weights from their own data. Statistical models like this also depend on the period and industry from which the data was taken.
The nuance is that a score is a probability guide, not a prophecy. It uses past accounts, which may be out of date, and it can be thrown off by accounting differences or unusual events.
Credit professionals therefore use it alongside cash flow analysis, industry knowledge and conversations with management. For a non-specialist, the useful lesson is that credit risk can be described with a handful of ratios that anyone can calculate.
Rising debt, falling profit, thin liquidity and small size all push a score the wrong way. A manager who watches these figures in their own company can see trouble before a lender does.
In practice
Real-world examples.
Example
A bank credit officer scores a manufacturing borrower before approving a $5,000,000 loan. The score is well above the bank's cut-off, so the loan goes to committee with a favourable note. The officer still reads the cash flow forecast.
Example
A bond investor monitors the scores of 40 companies in a portfolio each quarter. When one company's score falls for three quarters in a row, she sells the bonds. The company's rating is cut six months later.
Example
A supplier considers giving a new customer 60 days' credit. The credit controller calculates a score from the customer's published accounts and sets a $50,000 limit instead of $150,000. She reviews the limit every six months and updates it when new accounts are published, so that the credit control team always works from the latest figures.
Formula
Calculation
Score = Sum of (weight x ratio) across the model's variables
The real Zeta weights are proprietary, so this is a simplified, purely illustrative example with invented weights. Take three ratios: EBIT (operating profit) to total assets of 0.10, retained earnings to total assets of 0.20, and working capital to total assets of 0.15, with illustrative weights of 2.0, 1.5 and 1.0. The score is 2.0 x 0.10 + 1.5 x 0.20 + 1.0 x 0.15 = 0.20 + 0.30 + 0.15 = 0.65. If the lender's illustrative cut-off is 0.50, a score of 0.65 is above it and the company passes this screen.Case study
Seen in the real world.
Hartwell Credit Partners is an illustrative, fictional lender that uses a Zeta-style model to screen applicants for working capital loans. Applicants with scores above a set line go to a fast-track team, while those below are reviewed in depth. In one quarter, 120 applications arrive.
The model sends 85 to fast-track and 35 to the full review. Of the 35, the analysts find that 9 have temporary problems that explain the low score, and they approve those loans with conditions. The other 26 are declined or reduced.
The illustrative lesson is that the model saves time by sorting the pile, but humans still make the final decision on borderline cases. Hartwell reviews its cut-off each year using its own repayment data, since a model should be recalibrated as conditions change.
Watch out
Common mistakes.
- Treating the score as a certain prediction, when it only estimates the likelihood of failure.
- Using old weights on a new industry or period, when the relationships between ratios and failure can change.
- Relying on a score without reading the accounts, when one-off items and accounting choices can distort the ratios.
Questions
People also ask.
Who created the Zeta model?
It was developed in the 1970s by Edward Altman with colleagues as a more detailed successor to the Z-score.
How is it different from the Z-score?
It uses more variables and a different statistical set-up, though both combine ratios into one risk score.
Can I build my own version?
Yes, you can estimate weights from your own historical data, which is how many banks build internal credit models.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related