RDP 2016-09: Why Do Companies Fail? 1. Introduction

We explore the determinants of corporate failure in Australia. Our paper is motivated by two observations: corporate failure is closely linked to financial stability risks, and it is also associated with fluctuations in the business cycle. Despite these close links, there has been little research to date on why Australian companies fail.

The link between aggregate corporate failures and the business cycle is shown in Figure 1. The aggregate failure rate typically rises during economic downturns. This is most evident in the early 1990s recession. But there also seem to have been structural breaks in the level of the aggregate failure rate; for instance, the failure ate increased sharply around the start of the 2000s. This is due, at least in part, to changes in corporate insolvency law that reduced the cost of entering insolvency for companies.^[1] This suggests a range of factors are important in determining corporate failure.

Financial stability risks are closely related to failure among non-financial companies because failing companies often do not repay creditors. History shows that bank credit losses in Australia are closely tied to conditions in the non-financial corporate sector (Rodgers 2015). Non-financial companies account for a disproportionate share of bank non-performing loans; 35 per cent of outstanding non-performing assets but 23 per cent of banks' outstanding credit at the end of June 2016. More notably, non-performing loans to non-financial companies explained more than half of the total stock of non-performing loans during the global financial crisis of 2008–09.

It is important to be clear at the outset on the meaning of ‘company failure’. Company failure is defined as an entry into external administration – the process by which a company's creditors take control of the company's assets to recover funds owed to them. Failure can include voluntary administration, liquidation and receivership. By this definition, a company must be indebted for it to fail.

A company failure is not the same as a company ‘exit’. Fewer than 10 per cent of business exits are estimated to be due to failure (Productivity Commission 2015). Failures do not always result in companies ceasing their activities – some companies in external administration will be sold as a going concern or will satisfy their creditors and regain control from external administrators. Similarly, a company may exit the business population for reasons other than failure. For example, a company may exit voluntarily if it does not earn a sufficient return or because the owners retire.

So what causes companies to fail? A regular survey of external administrators by the Australian Securities and Investments Commission (ASIC) points to a wide range of possibilities, including: weak organisational structure, poor business strategy, technological change and economic conditions. For instance, more than 4 in 10 respondents cite ‘poor strategic management’ as a cause of company failure, suggesting internal ‘structural’ causes at the company level. But a similar share of respondents also indicate that ‘cyclical’ factors specific to the company, such as ‘inadequate cash flow’, are important. Furthermore, around 1 in 5 respondents attribute failure to ‘general economic conditions’, suggesting that the state of the economy might play a role over and above company-specific factors.

The challenges in identifying the causes of corporate failure are highlighted by a well-publicised Australian case – the collapse of the electronics retailer, Dick Smith Electronics. In its own company reports, Dick Smith Electronics attributed failure to cyclical factors specific to the company, namely unexpected weakness in sales and constraints on its ability to finance inventory investment. But some equity market analysts have suggested that structural company-specific factors were important too; in particular, poor corporate governance arrangements and low transparency in financial reporting (Knapp 2016).

In this paper we attempt to disentangle the relative importance of the factors that cause company failure. To do so, the causes of failure are grouped into three broad categories: 1) company-specific factors that vary with time, which we label ‘cyclical’ factors, such as profitability and leverage; 2) ‘structural’ company-specific factors that do not necessarily vary with time, such as whether the company is listed on the stock exchange or is a subsidiary of a parent company; and 3) external macroeconomic conditions, such as the state of the real economy.

Identifying the causes of company failure should help policymakers to better understand the underlying drivers of the business cycle. For instance, corporate failure could be a key channel through which weaker demand for goods and services translates into lower output and employment during a recession, particularly if there are costs involved in shifting labour and capital from failed firms to surviving firms. There is evidence that business exits amplify and propagate the effects of aggregate shocks (Clementi and Palazzo 2016). Corporate failures can also have ‘ripple’ effects if businesses that are suppliers to or customers of the failing companies also subsequently fail (Bams, Pisa and Wolff 2015). In the extreme, a wave of company defaults may set off a downward spiral of fire sales, falling asset prices and reduced credit supply, which exacerbates the economic downturn (Acharya, Bharath and Srinivasan 2007).

To the best of our knowledge, there has been limited research to date that examines the causes of corporate failure in Australia.^[2] Our paper therefore makes several key contributions to the corporate failure literature.

First, we estimate the model based on a large panel of companies, including: 1) publicly listed; 2) publicly unlisted; and 3) private companies. Most existing research on corporate failure relies on samples of publicly listed companies.^[3] The additional cross-sectional variation means our results are more applicable to the broader population of companies than previous research; private companies typically make up 99 per cent of companies and nearly 50 per cent of both total sales and assets.

The additional cross-sectional variation also allows us to determine the relative importance of ‘structural’ company-specific factors, such as the type of ownership. Previous research has studied whether corporate survival is affected by the legal structure of a company and, in particular, whether it has limited liability or not (e.g. Harhoff, Stahl and Woywode 1998; Mata and Portugal 2002; Esteve-Pérez, Sanchis-Llopis and Sanchis-Llopis 2010). But, to the best of our knowledge, few studies have looked at how the risk of failure varies between private and public companies.

The broad-based nature of the company sample also allows us to better identify the underlying cyclical drivers of failure. There is extensive evidence that cyclical company-specific factors, such as profitability and leverage, are important determinants of failure (Bhattacharjee et al 2009; Black et al 2012). We examine the extent to which this is true for both public and private companies.

Second, we explore the extent to which macroeconomic conditions matter to corporate failure. There has been extensive research into the links between firm dynamics and the business cycle (see Siegfried and Evans (1994), Caves (1998), and Gil (2010) for literature reviews). And the literature points to a clear role for the business cycle in determining business exits (Gil 2010). We examine whether this is true even after controlling for changes over time in company-specific factors, such as leverage and profitability.

Third, we quantify the potential risks to financial stability via defaults on non-financial corporate debt. To do this, we combine our company-level estimates of the probability of default with balance sheet information on the outstanding stock of debt to construct estimates of ‘debt at risk’ (DAR) for each company in our sample. The outstanding stock of debt includes estimates of ‘net trade credit’ (or business-to-business lending), which makes up a significant share of debt for the average Australian company (Fitzpatrick and Lien 2013). We then use this framework to analyse both aggregate risks and the distribution of those risks.

Our key findings include:

• Structural characteristics are important determinants of failure. Public companies, and in particular listed companies, are more likely to fail than comparable private companies. This is consistent with these companies having a greater separation of ownership and control, which may encourage managers to take more risk. As far as we know, this is a novel finding in the corporate failure literature.
• Cyclical company-level characteristics are important too. Corporate failure is more likely when companies have high leverage, low liquidity and low profitability. Structural and cyclical characteristics together appear to determine the relative riskiness of companies.
• In contrast, aggregate conditions, such as the macroeconomic environment, appear to determine the annual level of the corporate failure rate.
• Debt at risk is estimated to comprise a small amount of total corporate debt (currently less than 1 per cent). However, DAR is highly concentrated among some large companies; it is possible for a company to be identified as a very low risk of default and yet have such large holdings of debt that they pose a significant risk to financial stability in aggregate.

Footnotes

Several events occurred around 2000/01 which may have led to a permanent rise in the average company failure rate. These include: the introduction of employee entitlement schemes which effectively reduced the cost of companies entering insolvency the introduction of the goods and services tax (GST), which may have increased the cost of being a registered company; and legislation introducing harsher penalties for trading while insolvent (Connolly et al 2015). The structural break also appears to reflect measurement issues and, in particular, an increase in the share of registered companies that are inactive'. We discuss this in more detail later. [1]

The Australian research that does exist has typically taken an accounting approach to develop credit risk models rather than focus on the links between corporate failures and the business cycle. These papers include Castagna and Matolscy (1981), Jones and Hensher (2004) and Gharghori, Chan and Faff (2006). Black et al (2012) focus on the factors that influence the survival of small businesses (rather than all businesses). [2]

Kim (2011) estimates a survival model for financial distress based on a sample of Australian listed companies. [3]