Altman Z-Score: Formula, Zones and How to Read It

What the Altman Z-score is

The Altman Z-score is a measure of financial strain that rolls five ratios into a single number. Edward Altman, a finance professor, built it in 1968 by comparing public manufacturers that had gone bankrupt with similar ones that had not, and asking which combination of accounting ratios best told the two groups apart. The answer was a weighted sum of five ratios, and the result is the score.

A higher Altman Z-score describes a company whose balance sheet and earnings look less like those of the failed firms in the original sample. A lower one looks more like them. That is a description of resemblance, and it is worth keeping in mind through everything below: the score summarises the company's position on its latest statements, and it says nothing certain about what happens next.

The appeal is that it reads several parts of the statements at once. A debt-to-equity ratio looks at leverage and an interest coverage ratio looks at whether profits carry the interest bill. The Z-score folds liquidity, accumulated profit, earning power, market value and asset use into one figure, so a weakness in one area can be offset by strength in another.

The five ratios and their weights

Each of the five ratios divides a line from the statements by a measure of the company's size, so companies of different sizes can be compared.

Working capital over total assets. Working capital is current assets minus current liabilities, the cushion for bills due within a year. A company that has been losing money often sees this shrink first. The current ratio reads the same two lines another way.

Retained earnings over total assets. Retained earnings are the profits a company has kept over its whole life rather than paid out. A young company or one with years of losses scores low here, which is part of why the model tends to rate young firms as riskier.

Operating earnings over total assets. Earnings before interest and tax, divided by assets, measures how much the assets produce before financing costs. Altman's original model gives this ratio the largest weight of the five, because earning power proved the strongest sign separating the two groups.

Market value of equity over total liabilities. This compares what the market values the shares at with everything the company owes. It is the one ratio that uses a share price, so it moves with the market, and it carries the smallest weight.

Sales over total assets. Revenue divided by assets measures how hard the assets work. It varies a great deal between industries, which is one reason the original model suits manufacturers better than other businesses.

The weights multiply each ratio before they are added. The reading at the end of this page prints the formula sentence Tenet's report serves for that company, which names the version and the period and, when the score is assembled from the filed statements rather than taken from the data vendor, the weights, so they are stated beside the score rather than restated here.

The zones and what they describe

The model sorts scores into three zones. At the top is a safe zone, where companies look little like the failed firms in Altman's sample. At the bottom is a distress zone, where they look much like them. Between the two is a grey zone, where the score does not separate them clearly.

The lines between the zones belong to the version of the model in use. Altman's original model for public manufacturers set one pair of lines, and the variants he fitted later, for private companies and for non-manufacturers, use different ratios and different lines. A score read against the wrong version's lines can land in the wrong zone, so the first thing to check is which version produced it.

The two best-known variants change the inputs as well as the lines. The version for private companies replaces the market value of equity with its book value, because a private company has no share price. The version for non-manufacturers drops sales over total assets altogether, because that ratio swings so widely between industries that it distorts the comparison. Each has its own weights and its own zone boundaries, which is why a score quoted without its version is hard to read.

A zone describes resemblance at one point in time. A company in the distress zone has a balance sheet and earnings that look like those of companies that later failed; many such companies recover, and some companies in the safe zone run into trouble. The zones are a reason to read the statements more closely, not a verdict.

Which companies it fits and which it does not

The original model was fitted on public manufacturers, and that is where it works best: companies with physical assets, inventory, receivables and a mix of debt and equity that looks like the firms in the sample.

It fits badly in several places. Banks and insurers fund themselves largely with deposits and policy liabilities, so the working-capital and liabilities ratios mean something else for them, and the model is not used there. Asset-light businesses such as software companies can score very high simply because their total assets are small next to their sales and market value. Utilities and real estate companies carry heavy long-term debt by design, which pulls their scores down without signalling the same strain. Young companies score low on retained earnings almost by definition.

For these groups, analysts use other measures, or one of the later variants that drops the ratios that do not fit. Tenet's Munger Compounders screen uses a minimum Altman Z-score as one of its gates, and the methodology page linked in the reading below states the basis it uses.

Where the Altman Z-score misleads

The market value term moves with the share price. A falling share price lowers the score even when nothing in the statements has changed, and a rising one lifts it. A Z-score computed at a market peak can flatter a company that the statements alone would rate lower.

The statements are a snapshot. The balance-sheet ratios use one date's figures. A company that has just raised cash, sold a division or paid down debt can look very different a quarter later, and the reverse is true of one about to make a large payment.

Accounting choices feed straight in. Retained earnings are reduced by buybacks in some presentations and by large write-downs, and operating earnings include one-off items unless they are taken out. Reading the accounting red flags guide alongside a surprising score is a good habit.

It is not a forecast. The original paper reported how well the score separated its sample, and later studies reported different results in different periods and markets. None of that turns a score into a statement about a particular company's future.

Where to go from here

The Altman Z-score is a compact summary of a company's balance sheet and earnings, best read against its own history, against similar companies and with the version of the model in mind. The guide to reading a balance sheet covers the lines that feed four of the five ratios.

To see the Altman Z-score on real companies, open Apple's Altman Z-score and Coca-Cola's Altman Z-score on the Statistics tab of each company's Tenet report. The reading below shows one manufacturer's score with its period and the formula that built it.