The importance of common cyclical features in VAR analysis: a Monte-Carlo study

Despite the commonly held belief that aggregate data display short-run comovement, there has been little discussion about the econometric consequences of this feature of the data. We use exhaustive Monte-Carlo simulations to investigate the importance of restrictions implied by common-cyclical features for estimates and forecasts based on vector autoregressive models. First, we show that the ìbestî empirical model developed without common cycle restrictions need not nest the ìbestî model developed with those restrictions. This is due to possible differences in the lag-lengths chosen by model selection criteria for the two alternative models. Second, we show that the costs of ignoring common cyclical features in vector autoregressive modelling can be high, both in terms of forecast accuracy and efficient estimation of variance decomposition coefficients. Third, we find that the Hannan-Quinn criterion performs best among model selection criteria in simultaneously selecting the lag-length and rank of vector autoregressions.

The importance of Common-Cyclical Features in VAR analysis: a Monte-Carlo study (Preliminary Version)

Despite the belief, supported byrecentapplied research, thataggregate datadisplay short-run comovement, there has been little discussion about the econometric consequences ofthese data “features.” W e use exhaustive M onte-Carlo simulations toinvestigate theimportance ofrestrictions implied by common-cyclicalfeatures for estimates and forecasts based on vectorautoregressive and errorcorrection models. First, weshowthatthe“best” empiricalmodeldevelopedwithoutcommoncycles restrictions neednotnestthe“best” modeldevelopedwiththoserestrictions, duetothe use ofinformation criteria forchoosingthe lagorderofthe twoalternative models. Second, weshowthatthecosts ofignoringcommon-cyclicalfeatures inV A R analysis may be high in terms offorecastingaccuracy and e¢ciency ofestimates ofvariance decomposition coe¢cients. A lthough these costs are more pronounced when the lag orderofV A R modelsareknown, theyarealsonon-trivialwhenitis selectedusingthe conventionaltoolsavailabletoappliedresearchers. T hird, we…ndthatifthedatahave common-cyclicalfeatures andtheresearcherwants touseaninformationcriterium to selectthelaglength, theH annan-Q uinn criterium is themostappropriate, sincethe A kaike and theSchwarz criteriahave atendency toover- and under-predictthe lag lengthrespectivelyinoursimulations.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We consider model selection criteria which have data-dependent penalties as well as the traditional ones. We suggest a new two-step model selection procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties and we prove its consistency. Our Monte Carlo simulations measure the improvements in forecasting accuracy that can arise from the joint determination of lag-length and rank using our proposed procedure, relative to an unrestricted VAR or a cointegrated VAR estimated by the commonly used procedure of selecting the lag-length only and then testing for cointegration. Two empirical applications forecasting Brazilian inflation and U.S. macroeconomic aggregates growth rates respectively show the usefulness of the model-selection strategy proposed here. The gains in different measures of forecasting accuracy are substantial, especially for short horizons.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We consider model selection criteria which have data-dependent penalties as well as the traditional ones. We suggest a new two-step model selection procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties and we prove its consistency. Our Monte Carlo simulations measure the improvements in forecasting accuracy that can arise from the joint determination of lag-length and rank using our proposed procedure, relative to an unrestricted VAR or a cointegrated VAR estimated by the commonly used procedure of selecting the lag-length only and then testing for cointegration. Two empirical applications forecasting Brazilian in ation and U.S. macroeconomic aggregates growth rates respectively show the usefulness of the model-selection strategy proposed here. The gains in di¤erent measures of forecasting accuracy are substantial, especially for short horizons.

Liberalização da conta de capital e fluxos de portfólio para o Brasil

The present work seeks to investigate the dynamics of capital account liberalization and its impact on short run capital flows to Brazil in the period of 1995-2002, considering different segments such as the monetary, derivative and equity markets. This task is pursued by developing a comparative study of financial flows and examining how it is affected by the uncovered interest parity, country risk and the legislation on portfolio capital flows. The empirical framework is based on a vector autoregressive (VAR) analysis using impulse-response functions, variance decomposition and Granger causality tests. In general terms the results indicate a crucial role played by the uncovered interest parity and the country risk to explain portfolio flows, and a less restrictive (more liberalized) legislation is not significant to attract such flows.

O fator comum associado à dinâmica de preços das commodities : a relação de cointegração e o fator dinâmico

Este trabalho analisa a importância dos fatores comuns na evolução recente dos preços dos metais no período entre 1995 e 2013. Para isso, estimam-se modelos cointegrados de VAR e também um modelo de fator dinâmico bayesiano. Dado o efeito da financeirização das commodities, DFM pode capturar efeitos dinâmicos comuns a todas as commodities. Além disso, os dados em painel são aplicados para usar toda a heterogeneidade entre as commodities durante o período de análise. Nossos resultados mostram que a taxa de juros, taxa efetiva do dólar americano e também os dados de consumo têm efeito permanente nos preços das commodities. Observa-se ainda a existência de um fator dinâmico comum significativo para a maioria dos preços das commodities metálicas, que tornou-se recentemente mais importante na evolução dos preços das commodities.

Forecasting accuracy and estimation uncertainty using VAR models with short- and long-term economic restrictions: a Monte-Carlo study

Using vector autoregressive (VAR) models and Monte-Carlo simulation methods we investigate the potential gains for forecasting accuracy and estimation uncertainty of two commonly used restrictions arising from economic relationships. The Örst reduces parameter space by imposing long-term restrictions on the behavior of economic variables as discussed by the literature on cointegration, and the second reduces parameter space by imposing short-term restrictions as discussed by the literature on serial-correlation common features (SCCF). Our simulations cover three important issues on model building, estimation, and forecasting. First, we examine the performance of standard and modiÖed information criteria in choosing lag length for cointegrated VARs with SCCF restrictions. Second, we provide a comparison of forecasting accuracy of Ötted VARs when only cointegration restrictions are imposed and when cointegration and SCCF restrictions are jointly imposed. Third, we propose a new estimation algorithm where short- and long-term restrictions interact to estimate the cointegrating and the cofeature spaces respectively. We have three basic results. First, ignoring SCCF restrictions has a high cost in terms of model selection, because standard information criteria chooses too frequently inconsistent models, with too small a lag length. Criteria selecting lag and rank simultaneously have a superior performance in this case. Second, this translates into a superior forecasting performance of the restricted VECM over the VECM, with important improvements in forecasting accuracy ñreaching more than 100% in extreme cases. Third, the new algorithm proposed here fares very well in terms of parameter estimation, even when we consider the estimation of long-term parameters, opening up the discussion of joint estimation of short- and long-term parameters in VAR models.

Social welfare analysis in a simple financial economy with risk regulation

In the last years, regulating agencies of rnany countries in the world, following recommendations of the Basel Committee, have compelled financiaI institutions to maintain minimum capital requirements to cover market risk. This paper investigates the consequences of such kind of regulation to social welfare and soundness of financiaI institutions through an equilibrium model. We show that the optimum level of regulation for each financiaI institution (the level that maximizes its utility) depends on its appetite for risk and some of them can perform better in a regulated economy. In addition, another important result asserts that under certain market conditions the financiaI fragility of an institution can be greater in a regulated econolny than in an unregulated one

Comparativo de metodologias de mensuração de VAR para o mercado financeiro brasileiro

Várias metodologias de mensuração de risco de mercado foram desenvolvidas e aprimoradas ao longo das últimas décadas. Enquanto algumas metodologias usam abordagens não-paramétricas, outras usam paramétricas. Algumas metodologias são mais teóricas, enquanto outras são mais práticas, usando recursos computacionais através de simulações. Enquanto algumas metodologias preservam sua originalidade, outras metodologias têm abordagens híbridas, juntando características de 2 ou mais metodologias. Neste trabalho, fizemos uma comparação de metodologias de mensuração de risco de mercado para o mercado financeiro brasileiro. Avaliamos os resultados das metodologias não-paramétricas e paramétricas de mensuração de VaR aplicados em uma carteira de renda fixa, renda variável e renda mista durante o período de 2000 a 2006. As metodologias não-paramétricas avaliadas foram: Simulação Histórica pesos fixos, Simulação Histórica Antitética pesos fixos, Simulação Histórica exponencial e Análise de Cenário. E as metodologias paramétricas avaliadas foram: VaR Delta-Normal pesos fixos, VaR Delta-Normal exponencial (EWMA), Simulação de Monte Carlo pesos fixos e Simulação de Monte Carlo exponencial. A comparação destas metodologias foi feita com base em medidas estatísticas de conservadorismo, precisão e eficiência.

Mercado brasileiro: aplicação de análise de componentes principais no cálculo de VAR para carteiras de renda fixa

A abordagem do Value at Risk (VAR) neste trabalho será feita a partir da análise da curva de juros por componentes principais (Principal Component Analysis – PCA). Com essa técnica, os movimentos da curva de juros são decompostos em um pequeno número de fatores básicos independentes um do outro. Entre eles, um fator de deslocamento (shift), que faz com que as taxas da curva se movam na mesma direção, todas para cima ou para baixo; de inclinação (twist) que rotaciona a curva fazendo com que as taxas curtas se movam em uma direção e as longas em outra; e finalmente movimento de torção, que afeta vencimentos curtos e longos no mesmo sentido e vencimentos intermediários em sentido oposto. A combinação destes fatores produz cenários hipotéticos de curva de juros que podem ser utilizados para estimar lucros e perdas de portfolios. A maior perda entre os cenários gerados é uma maneira intuitiva e rápida de estimar o VAR. Este, tende a ser, conforme verificaremos, uma estimativa conservadora do respectivo percentual de perda utilizado. Existem artigos sobre aplicações de PCA para a curva de juros brasileira, mas desconhecemos algum que utilize PCA para construção de cenários e cálculo de VAR, como é feito no presente trabalho.Nesse trabalho, verificaremos que a primeira componente principal produz na curva um movimento de inclinação conjugado com uma ligeira inclinação, ao contrário dos resultados obtidos em curvas de juros de outros países, que apresentam deslocamentos praticamente paralelos.