Earnings management among bankrupt non-listed firms: evidence from Spain

Abstract

We analyse whether Spanish non-listed bankrupt firms are more inclined to manage earnings in comparison with their non-bankrupt pairs during the years preceding a legal procedure for bankruptcy. We also investigate the techniques these companies employ to manage earnings and when they start using earnings manipulation practices. Analysing a matched sample of bankrupt and healthy companies, we find that bankrupt firms manage earnings upwards more than their healthy pairs. They achieve that by employing both accrual and real activity manipulation. These two practices start at least three years before the beginning of the bankruptcy procedure, but real activity manipulation stops the year immediately before filing for bankruptcy. Findings also indicate that earnings management tools change based on the industry in which firms operate and the number of years preceding the bankruptcy. This evidence is relevant to governments, monitoring bodies and all those involved in an insolvency procedure.

Analizamos si las empresas españolas no cotizadas en concurso de acreedores manipulan más los estados contables en comparación con sus pares no concursadas durante los años anteriores a su entrada en el proceso legal del concurso. También investigamos las técnicas que esas empresas emplean para manipular sus resultados y cuándo empiezan a utilizar dichas prácticas. Utilizando una muestra pareada de empresas en concurso y empresas “sanas”, encontramos que aquellas en concurso manipulan al alza sus resultados más que sus pares “sanas”. Para manipular el resultado, nuestro estudio demuestra que las empresas utilizan tanto alteraciones en la contabilidad como decisiones operativas, con la finalidad de maximizar el resultado real. Esas dos prácticas comienzan al menos tres años antes de que empiece el concurso de acreedores, mientras que la manipulación de actividades reales finaliza el año inmediatamente anterior a su declaración de concurso. Los resultados también indican que las técnicas de manipulación del resultado dependen del sector en el que operan las empresas y del número de años anteriores al concurso. Esta evidencia es relevante para gobiernos, organismos supervisores y todos aquellos interesados en un proceso de insolvencia.

Keywords:

• earnings management
• bankruptcy
• code law setting
• non-listed companies
• real activity manipulation
• accrual manipulation

Palabras clave:

• Manipulación del resultado
• Concurso de acreedores
• Marco institucional del derecho romano
• Empresas no cotizadas
• manipulación de actividades reales
• manipulación de devengos

Acknowledgements

We acknowledge the support of Complutense University of Madrid and BSCH project No. 931559 of INIFCO Research Troup. We thank Dr Javier Martinez for his help in relation to the legal topics, Dr Manuel Nuñez for his useful suggestions and the participants to the workshop held at the Complutense University of Madrid for their helpful insights. We also appreciated the support of the two editors in charge for this manuscript, Dr. Juan Manuel García Lara and Dr. Beatriz García Osma and one anonymous reviewer, since their comments and suggestions significantly improved the quality of our work. Last, but not least, we express our gratitude to the executive editor Dr. Pablo de Andrés Alonso for all his assistance.

Notes

^1. In general, we use the term “bankruptcy” to refer to firm failure in a legal process. “Failure”, “Insolvency” and “Bankruptcy” are used with the same meaning throughout this paper.

^2. This is according to article 198 of the Spanish Act on Insolvency (www.mjusticia.gob.es); this website is “accessible free of charge on the Internet and shall publish all the insolvency resolutions that require publication pursuant to the provisions of this Act”.

^3. To match a company on the basis of the industry, we use the first two digits of the NACE Rev. 1 sector classification.

^4. We match companies on the basis of two identified legal forms: “corporation” and “limited companies”.

^5. If more than one healthy firm met our matching criteria (industry, legal form and years), the company included in the control sample is one randomly selected from the pool of available matches.

^6. We also decided to match by size using legal form because, among unlisted companies, it implies different requirements and limitations that are relevant during situations of financial distress that most of the times lead to bankruptcy. For example, in the presence of financial needs, the issue of financial instruments might differ among different legal forms. This situation would not happen with a sample of listed companies. In any case, to be on the safe side, we control for firm size, measured as the natural logarithm of total assets, in all our regression models.

^7. Abnormal discount and extended credit term might also reflect firms’ strategies as if a firm is performing worse than its competitors because of lower quality goods, they are forced to sell their product using more favourable terms. In this case, our proxy would capture firms’ underlying economic fundamentals rather than opportunistic behaviours.