Argenti’s A score

Argenti’s A-score is not a reliable model. To get the results, you need to send questionnaires to firms; and if one likely firm is predicted to fail as per Argenti’s Z-score model, chances are that firm will answer the questionnaire with bias to hide its true operational status, or worst of all, will never answer the questionnaire.

The observer also needs to visit the company and its factories, meet its directors and get to know them well, in order to make objective assessments. That is not easy and takes a lot of time – it’s not practical in a day-to-day application but just a one-off analysis. That’s a major disadvantage of this model.

Very easily, if a company is failing, by theory the incompetent CEO has to be replaced… and that’s it!

The case scenario will be very much simplified. It will give you a few hints on where things have gone wrong with the company, and you label them accordingly to certain defects, mistakes or symptoms. Alex Watt is a ‘think-out-of-the-box man’ and so may test you by linking the two models A and Z together… think about it. 🙂

The Altman Z-score is a good predictive model and in the past have proven to be 80%-90% accurate in predicting corporate failure within three years before the company’s financial distress. Scores below 1.8 would show greater certainty in a company failing (but not definitely), and above 3.0 will be financially sound. The grey area between 1.8 and 3.0 shows the company being reasonably managed well.

While being a model aging 40+ years since the 1960s, its predictive accuracy is still as relevant as today. In 2009, Morgan Stanley used the Z-score model on a group of European companies, and found that companies with weaker balance sheets underperformed the market more than two-thirds of the time, and this commonly attributes to smaller private companies which had difficulty in obtaining financial loans and are usually at the introduction or growing phase of its product-life-cycle. Moreover, it was proven that a company with a Z-score of less than 1 underperformed the market by 4%.

A little bit beyond text book knowledge:

1. During an economic recession, investors would usually invest in defensive share portfolios – those shares that give consistent dividends and low volatility in the share pricing. More importantly, the Z-score model helps investors to narrow down even further which companies are still financially viable during the recession, hence having better ability to choose the right shares.

2. Shareholders use the Z-score as a potential warning if the company may go bankrupt. Bankruptcy costs and the creditor protection system is very costly and time-consuming, if the company does go bust. The Z-score is also used by banks to evaluate the credit rating of the company in servicing a loan.

3. The Z-score uses five determinants, four of them using total assets as a denominator, since a company must have good asset backing to be financially healthy.

Some clear disadvantages of the Z-score model:

1. The management will have experienced the symptoms of financial distress even though they don’t use the Z-score model. In an attempt to hide their distressed condition, they may resort to window dressing – distorting accounting data, hence ultimately affect Z-score results which is wholly based on historical accounting data – NOT cash flows. A company could be labelled to fail within 3 years, but it instead does well, and the opposite goes true.

2. Numerical data cannot tell you the whole story. You need to know what goes behind closed doors. This is what Z-score can’t tell you. Deeper analysis on the management structure can be investigated using the qualitative Argenti’s A-Score model.

3. The Z-score model will decrease in accuracy if compared against other firms, because they may run on different accounting policies (ie. difference in inventory valuation). While it can be compared between two companies in the same industry, it cannot be compared across many other industries. The Z-score model basically is a unique indicator of corporate failure to the specific company alone.

The below points are just dessert for your brain – these will clearly not be in the exam:

1. Z-score model was only tested on manufacturing companies 40 years ago, and although it is still quite accurate right now, there are better models that can predict corporate failure like Shumway’s Hazard Model. However, Altman’s Z-score model is still preferable because it is simple to calculate despite the risk of manipulation in accounting data, while Shumway’s Hazard model is complex.

2. The definition of ‘failure’ differs across international boundaries. It could mean liquidation as ordered by the court, while in other countries, failure may also be defined if the auditor issues a qualified opinion on its financial statements. The exact definition of failure is not strict – a company may also enter into a reorganisation or restructuring plan as an outcome.

3. The Z-score cannot be applied on financial industries as they run different accounting standards and a possible high amount of window dressing involved, and this is crucial as it is the financial industries that are the hardest hit during a recession. However, banks ‘never fail’ – they just simply get merged or being acquired by other banks.

angryhamtaro thank you so much for the insite. if this model was tested 40years ago for manufacturing companies why is it still being used to date 2012 and why has ACCA not introduced another model? I am actually learning of the Shumway’s Model now. So will it be a crime if in the exam u mentioned Shumway’s model if the question is open to assess corporate failure?