Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares

Detalhes bibliográficos
Autor(a) principal: Bohn, Liana
Data de Publicação: 2013
Tipo de documento: Dissertação
Idioma: por
Título da fonte: LOCUS Repositório Institucional da UFV
Texto Completo: http://locus.ufv.br/handle/123456789/3278
Resumo: The present economic structure is highly vulnerable to the crises stemming from extreme events, due to the poor understanding of the complex systems, especially in underdeveloped countries. The ability to predict the course of events in social systems is inherently limited, but there is a promising new approach to predict and understand the complex systems that have emerged through the integration of research in social sciences and mathematical prediction. This new approach, to be proposed in the present work, aims to develop a heuristic, from the recognition of regular patterns in macroeconomic indicators, to detect periods of large increases in unemployment rates in Brazil. For this, are used three methodological lines. The first is to evaluate the series of Brazilian unemployment (1985 to 2012), so as to recognize her regular patterns, which is made from the spline regression; the second seeks to identify, in the variable itself, signs of a change in long-term trend of unemployment, using the autocorrelation analysis and the fractal analysis; and, finally, to determine which macroeconomic variables present in advance a co-movement with the series of unemployment, use is made of discriminant analysis. The results reveal a robustness in the predictive ability of the suggested model. In the 27 analyzed years, there were six critical points that led to an accelerated increment in unemployment levels in the country. Before them, it is possible to observe an increment in the Hurst exponent, which indicates that the observations are more related and transmitting a performance occurred in the past, i.e., changes in the behavior of long-term unemployment are not merely randomized. This same result is not as noticeable from examination of the autocorrelation of the series that, despite being essentially high, is not increased in the vicinity of a change of state in unemployment. For purposes of forecasting, of the10 analyzed variables, five of them had significant correlations with unemployment and from these, three were chosen to discriminate the cases where there is, or not, an increment in their levels. At this juncture it is included the appearing of minimum points in the long-term trends of the series in terms of trade and imports, in consonance with negative and significant variations (from $ 10 - PPC) in the series of residuals of the minimum wage in PPC, which consists in the variations of the series observed in relation to the long-term trend. Thus, improvements in the terms of trade, increment of importations and falling real minimum wages in PPC precede periods of unemployment acceleration , together indicating the imminence of a long-term worsening of the labor market.
id UFV_c4547b13bbe3833d77ab76281c34a119
oai_identifier_str oai:locus.ufv.br:123456789/3278
network_acronym_str UFV
network_name_str LOCUS Repositório Institucional da UFV
repository_id_str 2145
spelling Bohn, Lianahttp://lattes.cnpq.br/3625641544852126Carvalho, Luciano Dias dehttp://lattes.cnpq.br/0061368522702958Silva, Evaldo Henrique dahttp://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4785521A7Bueno, Newton Paulohttp://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4788572J6Protil, Roberto Maxhttp://lattes.cnpq.br/27524351193659842015-03-26T13:18:06Z2013-09-132015-03-26T13:18:06Z2013-02-22BOHN, Liana. The gray swans in Brazil s unemployment: the prediction from regular patterns. 2013. 125 f. Dissertação (Mestrado em Desenvolvimento econômico e Políticas públicas) - Universidade Federal de Viçosa, Viçosa, 2013.http://locus.ufv.br/handle/123456789/3278The present economic structure is highly vulnerable to the crises stemming from extreme events, due to the poor understanding of the complex systems, especially in underdeveloped countries. The ability to predict the course of events in social systems is inherently limited, but there is a promising new approach to predict and understand the complex systems that have emerged through the integration of research in social sciences and mathematical prediction. This new approach, to be proposed in the present work, aims to develop a heuristic, from the recognition of regular patterns in macroeconomic indicators, to detect periods of large increases in unemployment rates in Brazil. For this, are used three methodological lines. The first is to evaluate the series of Brazilian unemployment (1985 to 2012), so as to recognize her regular patterns, which is made from the spline regression; the second seeks to identify, in the variable itself, signs of a change in long-term trend of unemployment, using the autocorrelation analysis and the fractal analysis; and, finally, to determine which macroeconomic variables present in advance a co-movement with the series of unemployment, use is made of discriminant analysis. The results reveal a robustness in the predictive ability of the suggested model. In the 27 analyzed years, there were six critical points that led to an accelerated increment in unemployment levels in the country. Before them, it is possible to observe an increment in the Hurst exponent, which indicates that the observations are more related and transmitting a performance occurred in the past, i.e., changes in the behavior of long-term unemployment are not merely randomized. This same result is not as noticeable from examination of the autocorrelation of the series that, despite being essentially high, is not increased in the vicinity of a change of state in unemployment. For purposes of forecasting, of the10 analyzed variables, five of them had significant correlations with unemployment and from these, three were chosen to discriminate the cases where there is, or not, an increment in their levels. At this juncture it is included the appearing of minimum points in the long-term trends of the series in terms of trade and imports, in consonance with negative and significant variations (from $ 10 - PPC) in the series of residuals of the minimum wage in PPC, which consists in the variations of the series observed in relation to the long-term trend. Thus, improvements in the terms of trade, increment of importations and falling real minimum wages in PPC precede periods of unemployment acceleration , together indicating the imminence of a long-term worsening of the labor market.A atual estrutura econômica é altamente vulnerável às crises decorrentes de eventos extremos, vulnerabilidade esta que se dá pela pobre compreensão de sistemas complexos, especialmente nos países subdesenvolvidos. A capacidade de predizer o curso dos eventos em sistemas sociais é inerentemente limitada, mas existe uma nova e promissora abordagem para prever e entender os sistemas complexos que tem emergido através da integração de estudos em ciências sociais e da previsão matemática. Esta nova abordagem, ao ser proposta no presente trabalho, visa elaborar uma heurística, a partir do reconhecimento de padrões regulares nos indicadores macroeconômicos, para detectar períodos de grande aumento das taxas de desemprego no Brasil. Para tanto, utilizam-se três linhas metodológicas. A primeira tem por objetivo avaliar a série de desemprego brasileira (de 1985 a 2012), de modo a reconhecer nela padrões regulares, o que é feito a partir das regressões splines; a segunda busca identificar, na própria variável, indícios de uma mudança na tendência de longo prazo do desemprego, utilizando-se da análise de autocorrelação e da análise fractal; e, por fim, para determinar quais variáveis macroeconômicas apresentam com antecedência um comovimento com a série do desemprego, faz-se uso da análise discriminante. Os resultados encontrados revelam uma robusteza na capacidade preditiva do modelo sugerido. Nos 27 anos analisados, houve seis pontos críticos que conduziram a um aumento acelerado nos níveis de desemprego do país. Antes deles, é possível observar um aumento do expoente de Hurst, o que indica que as observações estão mais relacionadas e transmitindo um desempenho já ocorrido no passado, ou seja, as mudanças no comportamento de longo prazo do desemprego não são meramente aleatórias. Este mesmo resultado não é tão perceptível a partir do exame da autocorrelação da série que, apesar de ser essencialmente alta, não se amplia quando da proximidade de uma mudança de estado no desemprego. Para fins de previsão, das 10 variáveis analisadas, cinco delas obtiveram correlações significativas com o desemprego e destas, três foram escolhidas para discriminar os casos onde há, ou não, uma ampliação em seus níveis. Nessa conjuntura inclui-se o aparecimento de pontos mínimos nas tendências de longo prazo das séries dos termos de troca e da importação, em consonância com variações negativas e consideráveis (a partir de 10 dólares PPC) na série dos resíduos do salário mínimo em PPC, que consiste nas variações da série observada em relação à tendência de longo prazo. Deste modo, melhorias nas relações de troca, aumento das importações e queda dos salários mínimos reais em PPC precedem períodos de aceleração do desemprego, conjuntamente sinalizando a eminência de uma piora de longo prazo no mercado de trabalho.Coordenação de Aperfeiçoamento de Pessoal de Nível Superiorapplication/pdfporUniversidade Federal de ViçosaMestrado em EconomiaUFVBRDesenvolvimento econômico e Políticas públicasComplexidadeInflaçãoIndicadores líderesSplineAnálise de discriminanteComplexityInflationLeaders indicatorsSplineDiscriminant analysisCNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::CRESCIMENTO, FLUTUACOES E PLANEJAMENTO ECONOMICO::FLUTUACOES CICLICAS E PROJECOES ECONOMICASOs cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regularesThe gray swans in Brazil s unemployment: the prediction from regular patternsinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisinfo:eu-repo/semantics/openAccessreponame:LOCUS Repositório Institucional da UFVinstname:Universidade Federal de Viçosa (UFV)instacron:UFVORIGINALtexto completo.pdfapplication/pdf1674238https://locus.ufv.br//bitstream/123456789/3278/1/texto%20completo.pdf50077c5ebeb1c9af3fa365ecc64688acMD51TEXTtexto completo.pdf.txttexto completo.pdf.txtExtracted texttext/plain263386https://locus.ufv.br//bitstream/123456789/3278/2/texto%20completo.pdf.txt22f1c8c01725a61672ce1a6e742619afMD52THUMBNAILtexto completo.pdf.jpgtexto completo.pdf.jpgIM Thumbnailimage/jpeg3481https://locus.ufv.br//bitstream/123456789/3278/3/texto%20completo.pdf.jpgb3d3fd2122928a402837e85c240a421aMD53123456789/32782016-04-09 23:07:07.637oai:locus.ufv.br:123456789/3278Repositório InstitucionalPUBhttps://www.locus.ufv.br/oai/requestfabiojreis@ufv.bropendoar:21452016-04-10T02:07:07LOCUS Repositório Institucional da UFV - Universidade Federal de Viçosa (UFV)false
dc.title.por.fl_str_mv Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
dc.title.alternative.eng.fl_str_mv The gray swans in Brazil s unemployment: the prediction from regular patterns
title Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
spellingShingle Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
Bohn, Liana
Complexidade
Inflação
Indicadores líderes
Spline
Análise de discriminante
Complexity
Inflation
Leaders indicators
Spline
Discriminant analysis
CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::CRESCIMENTO, FLUTUACOES E PLANEJAMENTO ECONOMICO::FLUTUACOES CICLICAS E PROJECOES ECONOMICAS
title_short Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
title_full Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
title_fullStr Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
title_full_unstemmed Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
title_sort Os cisnes cinzentos do desemprego no Brasil: a previsão a partir de padrões regulares
author Bohn, Liana
author_facet Bohn, Liana
author_role author
dc.contributor.authorLattes.por.fl_str_mv http://lattes.cnpq.br/3625641544852126
dc.contributor.author.fl_str_mv Bohn, Liana
dc.contributor.advisor-co1.fl_str_mv Carvalho, Luciano Dias de
dc.contributor.advisor-co1Lattes.fl_str_mv http://lattes.cnpq.br/0061368522702958
dc.contributor.advisor-co2.fl_str_mv Silva, Evaldo Henrique da
dc.contributor.advisor-co2Lattes.fl_str_mv http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4785521A7
dc.contributor.advisor1.fl_str_mv Bueno, Newton Paulo
dc.contributor.advisor1Lattes.fl_str_mv http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4788572J6
dc.contributor.referee1.fl_str_mv Protil, Roberto Max
dc.contributor.referee1Lattes.fl_str_mv http://lattes.cnpq.br/2752435119365984
contributor_str_mv Carvalho, Luciano Dias de
Silva, Evaldo Henrique da
Bueno, Newton Paulo
Protil, Roberto Max
dc.subject.por.fl_str_mv Complexidade
Inflação
Indicadores líderes
Spline
Análise de discriminante
topic Complexidade
Inflação
Indicadores líderes
Spline
Análise de discriminante
Complexity
Inflation
Leaders indicators
Spline
Discriminant analysis
CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::CRESCIMENTO, FLUTUACOES E PLANEJAMENTO ECONOMICO::FLUTUACOES CICLICAS E PROJECOES ECONOMICAS
dc.subject.eng.fl_str_mv Complexity
Inflation
Leaders indicators
Spline
Discriminant analysis
dc.subject.cnpq.fl_str_mv CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::CRESCIMENTO, FLUTUACOES E PLANEJAMENTO ECONOMICO::FLUTUACOES CICLICAS E PROJECOES ECONOMICAS
description The present economic structure is highly vulnerable to the crises stemming from extreme events, due to the poor understanding of the complex systems, especially in underdeveloped countries. The ability to predict the course of events in social systems is inherently limited, but there is a promising new approach to predict and understand the complex systems that have emerged through the integration of research in social sciences and mathematical prediction. This new approach, to be proposed in the present work, aims to develop a heuristic, from the recognition of regular patterns in macroeconomic indicators, to detect periods of large increases in unemployment rates in Brazil. For this, are used three methodological lines. The first is to evaluate the series of Brazilian unemployment (1985 to 2012), so as to recognize her regular patterns, which is made from the spline regression; the second seeks to identify, in the variable itself, signs of a change in long-term trend of unemployment, using the autocorrelation analysis and the fractal analysis; and, finally, to determine which macroeconomic variables present in advance a co-movement with the series of unemployment, use is made of discriminant analysis. The results reveal a robustness in the predictive ability of the suggested model. In the 27 analyzed years, there were six critical points that led to an accelerated increment in unemployment levels in the country. Before them, it is possible to observe an increment in the Hurst exponent, which indicates that the observations are more related and transmitting a performance occurred in the past, i.e., changes in the behavior of long-term unemployment are not merely randomized. This same result is not as noticeable from examination of the autocorrelation of the series that, despite being essentially high, is not increased in the vicinity of a change of state in unemployment. For purposes of forecasting, of the10 analyzed variables, five of them had significant correlations with unemployment and from these, three were chosen to discriminate the cases where there is, or not, an increment in their levels. At this juncture it is included the appearing of minimum points in the long-term trends of the series in terms of trade and imports, in consonance with negative and significant variations (from $ 10 - PPC) in the series of residuals of the minimum wage in PPC, which consists in the variations of the series observed in relation to the long-term trend. Thus, improvements in the terms of trade, increment of importations and falling real minimum wages in PPC precede periods of unemployment acceleration , together indicating the imminence of a long-term worsening of the labor market.
publishDate 2013
dc.date.available.fl_str_mv 2013-09-13
2015-03-26T13:18:06Z
dc.date.issued.fl_str_mv 2013-02-22
dc.date.accessioned.fl_str_mv 2015-03-26T13:18:06Z
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
status_str publishedVersion
dc.identifier.citation.fl_str_mv BOHN, Liana. The gray swans in Brazil s unemployment: the prediction from regular patterns. 2013. 125 f. Dissertação (Mestrado em Desenvolvimento econômico e Políticas públicas) - Universidade Federal de Viçosa, Viçosa, 2013.
dc.identifier.uri.fl_str_mv http://locus.ufv.br/handle/123456789/3278
identifier_str_mv BOHN, Liana. The gray swans in Brazil s unemployment: the prediction from regular patterns. 2013. 125 f. Dissertação (Mestrado em Desenvolvimento econômico e Políticas públicas) - Universidade Federal de Viçosa, Viçosa, 2013.
url http://locus.ufv.br/handle/123456789/3278
dc.language.iso.fl_str_mv por
language por
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universidade Federal de Viçosa
dc.publisher.program.fl_str_mv Mestrado em Economia
dc.publisher.initials.fl_str_mv UFV
dc.publisher.country.fl_str_mv BR
dc.publisher.department.fl_str_mv Desenvolvimento econômico e Políticas públicas
publisher.none.fl_str_mv Universidade Federal de Viçosa
dc.source.none.fl_str_mv reponame:LOCUS Repositório Institucional da UFV
instname:Universidade Federal de Viçosa (UFV)
instacron:UFV
instname_str Universidade Federal de Viçosa (UFV)
instacron_str UFV
institution UFV
reponame_str LOCUS Repositório Institucional da UFV
collection LOCUS Repositório Institucional da UFV
bitstream.url.fl_str_mv https://locus.ufv.br//bitstream/123456789/3278/1/texto%20completo.pdf
https://locus.ufv.br//bitstream/123456789/3278/2/texto%20completo.pdf.txt
https://locus.ufv.br//bitstream/123456789/3278/3/texto%20completo.pdf.jpg
bitstream.checksum.fl_str_mv 50077c5ebeb1c9af3fa365ecc64688ac
22f1c8c01725a61672ce1a6e742619af
b3d3fd2122928a402837e85c240a421a
bitstream.checksumAlgorithm.fl_str_mv MD5
MD5
MD5
repository.name.fl_str_mv LOCUS Repositório Institucional da UFV - Universidade Federal de Viçosa (UFV)
repository.mail.fl_str_mv fabiojreis@ufv.br
_version_ 1801212847758245888