| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 540.76 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Para que o normal funcionamento de um mercado económico se desenvolva de forma saudável um dos principais parâmetros a ter em conta é a informação contabilística que as diferentes empresas disponibilizam. No entanto, nem sempre essa informação é fidedigna. Por essa razão, a literatura de contabilidade tem procurado ativamente desenvolver modelos capazes de identificar manipulação contabilística por parte das empresas. Esta tese centra-se na identificação de um caso particular de manipulação contabilística, manipulação de resultados, em concreto por via de accruals, a partir do modelo de Dechow, Sloan & Sweeney (1995). Para testar a eficácia do modelo para identificar manipulação a partir de resultados reportados pelas empresas, estimei o modelo para uma amostra de resultados reportados por 23.606 empresas americanas no período de 1988 a 2015 (num total de 130.634 observações). Com base nos resultados de estimação, elaborei uma estimativa do valor dos accruals discricionários para cada empresa em cada ano. Por sua vez, como base nestas estimativas, desenvolvi e calculei um indicador para identificar a identidades das empresas que, em cada ano, de acordo com o modelo, manipularam os resultados reportados. Após esta identificação, comparei os resultados com a identidade das empresas que, em cada ano, segundo o regulador, efetivamente manipularam os resultados reportados. O indicador proposto, elaborado com base no modelo de Dechow, Sloan & Sweeney (1995), detetou 33,8% dos casos de manipulação identificados pelo regulador.
For the normal function of an economical market to develop in a healthy way, one of the main parameters to take into account is the information that different companies relay on their accounting’s standard. However, that information is not always trustworthy. For that reason, accounting literature has been searching models capable of identifying earnings management by companies. This thesis concentrates itself on identifying a particular case of earnings management, the manipulation of results, in concrete using accruals, considering the model by Dechow, Sloan & Sweeney (1995). To test the efficiency of the model on identifying results manipulation using results given by companies i estimated the model for a sample of results reported by 23.606 American companies in the period of 1988 to 2015 (in a total of 130.634 observations). With the results from the estimation as a base for the value of accruals discretionary for each company in each year. With these estimates in mind, I developed and calculated an indicator to identify the companies that, in each year, manipulated results reported. After the identification, I compared the results with the identity of the companies, in each year, according to the regulator had effectively manipulated results. The proposed indicator, elaborated with base on the model by Dechow, Sloan& Sweeney (1955), detected 33,8% of the cases identified by the regulator.
For the normal function of an economical market to develop in a healthy way, one of the main parameters to take into account is the information that different companies relay on their accounting’s standard. However, that information is not always trustworthy. For that reason, accounting literature has been searching models capable of identifying earnings management by companies. This thesis concentrates itself on identifying a particular case of earnings management, the manipulation of results, in concrete using accruals, considering the model by Dechow, Sloan & Sweeney (1995). To test the efficiency of the model on identifying results manipulation using results given by companies i estimated the model for a sample of results reported by 23.606 American companies in the period of 1988 to 2015 (in a total of 130.634 observations). With the results from the estimation as a base for the value of accruals discretionary for each company in each year. With these estimates in mind, I developed and calculated an indicator to identify the companies that, in each year, manipulated results reported. After the identification, I compared the results with the identity of the companies, in each year, according to the regulator had effectively manipulated results. The proposed indicator, elaborated with base on the model by Dechow, Sloan& Sweeney (1955), detected 33,8% of the cases identified by the regulator.
Description
Keywords
Manipulação de resultados Accruals Dechow, Sloan & Sweeney (1995) Earnings management