Industry classification benchmark and industries’ predictive ability

Abstract

In this dissertation, we investigate Industry Classification Benchmark (ICB) to test if the industries are able to predict the market portfolio or not, and if so how many months ahead do they forecast? According to the final results, several conclusions have been made. Firstly, industries could forecast the aggregated market with information variables, such as Default Spread (BAA rate and AAA rate), Dividend Yield, Market Volatility and PER (price earnings ratio). Secondly, only five of thirty-two industries are able to predict the market portfolio in one month ahead, they are support services, food & drug retails, general retails, mobile telecommunications and nonlife insurance, respectively. Furthermore, those industries in the same supersectors of Industry Classification Benchmark possess the homological predictive power to the aggregated market portfolio.

Neste tese eu investigo a classificação de industrias:IndustryClassificationBenchmark (ICB) para testar se as industrias conseguem prever as variações da carteira de mercado, e se sim com quantos meses de avanço. Os resultados mostram em primeiro lugar que as industria podem prever para além das variáveis normalmente usadas tais como: oDefaultSpread (a diferença entre o spread das taxas BAA e AAA), a DividendYield, a volatilidade do mercado e opriceearnings ratio (PER). Em segundo lugar só 5 das 32 industrias são capazes de prever a carteira de mercado com um mês de avanço, elas são: supportservices, food&drugretails, general retails, mobile telecommunications enonlifeinsurance. Para além disso, estas industrias no mesmo supersector da Industry Classificatio nBenchmarktêm o mesmo poder preditivo homólogo para o mercado.