Nonlinear Seemingly Unrelated Regression Research Articles

Response of industrial customers to hourly pricing in Ontario’s deregulated electricity market

This paper uses hourly data from Ontario (Canada) between 2005 and 2008 to estimate the effects of real time wholesale electricity prices on demand by industrial customers. Nonlinear SUR estimates from Generalized Leontief (GL) specifications reveal elasticities of substitution from 0.02 to 0.07, confirming that industrial customers (connected to the transmission grid) shift consumption across peak and off-peak periods in order to reap benefits of lower prices. Estimates from FGLS and IV models suggest that this reduction in demand by industrial customers results in lower wholesale prices, which benefits all consumers. The policy lesson is that market based schemes that encourage Real Time Pricing (RTP) pricing should result in positive spillovers to all consumers.

Nonlinear SUR Models with Panel Data and Additive Auto-Correlated Errors

This paper is concerned with the estimation of nonlinear SUR models with additive AR(1) disturbances using panel data. We propose a transformation which eliminates auto-correlation for the whole system and yields a classical SUR-EC model. We present a general class of minimum distance estimators and derive the asymptotic properties. We also compare finite-sample properties of several possible estimators for our model, including some linear competitors commonly used in empirical research, and point out the cases when inference based on linear approximation and/or asymptotic distribution may be misleading with a small sample size in one or both dimensions.

Foreign exchange risk and risk exposure in the Japanese stock market

PurposeWhether stock returns are linked to exchange rate changes and whether foreign exchange risk is priced in a domestic context are less conclusive and thus still subject to a great debate. The purpose of this paper is to provide new empirical evidence on these two inter‐related issues, which are critical to investors and corporate risk management.Design/methodology/approachThis paper applies two different econometric approaches: Nonlinear Seemingly Unrelated Regression (NLSUR) via Hansen's Generalized Method of Moment (GMM) and multivariate GARCH in mean (MGARCH‐M) to examine the exchange rate exposure and its pricing.FindingsUsing industry data for Japan, similar to previous studies, foreign exchange risk is not priced based on the test of an unconditional two‐factor asset pricing model. However, strong evidence of time‐varying foreign exchange risk premium and significant exchange rate betas are obtained based on the tests of conditional asset pricing models using MGARCH‐M approach where both conditional first and second moments of industry returns and risk factors are estimated simultaneously.Research limitations/implicationsThe strong empirical evidence found in this study implies that corporate currency hedging not only results in more stable cash flows for a firm, but also reduces its cost of capital, and hence is justifiable.Originality/valueThis paper conducts an in‐depth investigation regarding the exchange rate exposure and its pricing by utilizing two different econometric approaches: NLSUR via Hansen's GMM and MGARCH‐M. In doing so, a more reliable conclusion about the exchange rate exposure and its pricing can be drawn.

Foreign exchange risk and risk exposure in the Japanese stock market

Purpose – Whether stock returns are linked to exchange rate changes and whether foreign exchange risk is priced in a domestic context are less conclusive and thus still subject to a great debate. This paper attempts to provide new empirical evidence on these two inter-related issues, which are critical to investors and corporate risk management. Design/methodology/approach – This paper applies two different econometric approaches: Nonlinear Seemingly Unrelated Regression (NLSUR) via Hansen’s (1982) Generalized Method of Moment (GMM) and multivariate GARCH in mean (MGARCH-M) to examine the exchange rate exposure and its pricing. Findings – Using industry data for Japan, similar to previous studies, foreign exchange risk is not priced based on the test of an unconditional two-factor asset pricing model. However, strong evidence of time-varying foreign exchange risk premium and significant exchange rate betas are obtained based on the tests of conditional asset pricing models using multivariate GARCH in mean (MGARCH-M) approach where both conditional first and second moments of industry returns and risk factors are estimated simultaneously. Research limitations/implications – The strong empirical evidence found in this study implies that corporate currency hedging not only results in more stable cash flows for a firm, but also reduces its cost of capital, and hence is justifiable. Originality/value – This paper conducts an in-depth investigation regarding the exchange rate exposure and its pricing by utilizing two different econometric approaches: Nonlinear Seemingly Unrelated Regression (NLSUR) via Hansen’s (1982) Generalized Method of Moment (GMM) and multivariate GARCH in mean (MGARCH-M). In doing so, a more reliable conclusion about the exchange rate exposure and its pricing can be drawn.

Estimating time-varying risk premia in UK long-term government bonds

Simple models of time-varying risk premia are used to measure the risk premia in long-term UK government bonds. The parameters of the models can be estimated using nonlinear seemingly unrelated regression (NL-SUR), which permits efficient use of information across the entire yield curve and facilitates the testing of various cross-sectional restrictions. The estimated time-varying premia are found to be substantially different to those estimated using models that assume constant risk premia. © 2004 Taylor and Francis Ltd.

Endogenous Parameter Time Series Estimation of the Ohlson Model: Linear and Nonlinear Analyses

This paper tests the empirical validity of the Ohlson (1995) model on a firm‐level time series basis. The coefficients of the earnings dynamic and valuation equations are first estimated by OLS. Next, recognizing the nonlinear relationships among the parameters, each equation is estimated by nonlinear Least Squares. Lastly, the model is estimated as a restricted system by nonlinear Least Squares and nonlinear SUR. In all cases, parameters are endogenously estimated. Irrespective of the estimation method, the Ohlson model often yields inconsistent or insignificant parameter estimates. Nevertheless, point estimates of equity risk premia are similar to those obtained from alternative methodologies.

Allometric relationships in Bactris gasipaes for heart-of-palm production agroecosystems in Costa Rica

Peach palm (Bactris gasipaes Kunth) agroecosystems for hearts-of-palm constitute a productive and sustainable land use for the humid tropics. Allometric models allow to predict biomass non-destructively at any time, and subsequently, to determine the span of growth phases, biomass and nutrient pools, and economic yields. The overall goals of this study were to obtain and validate predictive functions of above-ground dry biomass of peach palm shoots, and to relate standing biomass with heart-of-palm yields as well. Towards this purpose, peach palm shoots were harvested and separated into components (foliage, petiole and stem) in the Atlantic region of Costa Rica. Basal diameter (BD) was a more effective predictor of biomass than height to the fork between the spear leaf and the first fully expanded leaf, total height and number of leaves. Regression models explained 70–89% of the variance in component (foliage, petiole and stem) or total shoot biomass. Nonlinear regression, which independently calculates equation coefficients for biomass components and total shoot biomass, was compared with a nonlinear seemingly unrelated regression (NSUR) procedure, which simultaneously fits the component equations that predict leaf, petiole and stem in order to assure biomass additivity. Equation coefficients for NSUR fitted-regressions that also model unequal variances, were substantially different from those for individual regressions; e.g. Biomassleaf = 11·4739 BD1·8042, Residual mean square (RMS) = 69·9 for the individual equation, versus Biomassleaf = 6·841 BD2·086, RMS = 72·4 for the NSUR fitted-equation. NSUR equations had slightly less precision in estimating biomass than individual equations but consistently less bias. In separate harvests of peach palm plants within four stands ranging in age from 1·9 to 21 years, estimates of component and total above-ground shoot biomass were similar to observed values except for the youngest stand in which biomass was overestimated. In another harvest, yield of heart-of-palm per plant was linearly related to total above-ground biomass in two peach palm stands of age 5 and 9 years. The non-destructive estimation of above-ground biomass from easily measured plant dimensions will permit any-time, less expensive and reasonable precise biomass estimates in peach palm. Biomass data can be incorporated to decision support aids for nutrient management in heart-of-palm agroecosystems and serve other purposes such as for carbon sequestration calculations.

Time-varying market, interest rate, and exchange rate risk premia in the US commercial bank stock returns

This paper examines the role of market, interest rate, and exchange rate risks in pricing a sample of the US Commercial Bank stocks by developing and estimating a multi-factor model under both unconditional and conditional frameworks. Three different econometric methodologies are used to conduct the estimations and testing. Estimations based on nonlinear seemingly unrelated regression (NLSUR) via GMM approach indicate that interest rate risk is the only priced factor in the unconditional three-factor model. However, based on ‘pricing kernel’ approach by Dumas and Solnik [(1995). J. Finance 50, 445–479], strong evidence of exchange rate risk is found in both large bank and regional bank stocks in the conditional three-factor model with time-varying risk prices. Finally, estimations based on the multivariate GARCH in mean (MGARCH-M) approach where both conditional first and second moments of bank portfolio returns and risk factors are estimated simultaneously show strong evidence of time-varying interest rate and exchange rate risk premia and weak evidence of time-varying world market risk premium for all three bank portfolios, namely those of Money Center bank, Large bank, and Regional bank.

A Lintnerian Linear Accounting Valuation Model

This paper develops and tests a linear valuation accounting model based on Lintner's (1956) dividend model. Two test methodologies are employed on a firm-level time-series basis. First, abstracting from the nonlinear relationships among the parameters, the information dynamic and valuation equation are estimated by linear OLS and linear SUR. The estimated equations are evaluated by reference to the sign and value predictions of the model. Second, recognizing the underlying nonlinear relationships among the parameters, the Lintnerian system of equations is estimated by nonlinear OLS and nonlinear SUR. All parameters are estimated endogenously at the firm level, including each firm's cost of capital. The resulting parameter estimates are evaluated for statistical and, in the case of costs of capital, for economic significance. Results of the first (linear) methodology by and large confirm the validity of the Lintner model. The signs and values of the estimated coefficients are consistent with the predictions of the Lintner model except that the (mean) estimated book value coefficient in the price equation exceeds its theoretical upper bound. The results of the second (nonlinear) methodology are somewhat more problematic. Although the Lintner model yields statistically significant firm-level costs of capital, these estimates are not economically significant for the sample period.

Estimating the arbitrage pricing theory with observed macro factors

Analytical results are presented that drastically simplify joint estimation of the APT by Non-linear SUR and maximum likelihood when the factors are measured. Four alternative analytical expressions for calculating standard errors are also presented. © 1997 Elsevier Science S.A.

Integration or Segmentation of the Canadian Stock Market: Evidence Based on the APT

The domestic and international arbitrage pricing theories are modified to encom- pass the hypotheses that the Canadian and global North American equity markets are com- pletely or partly integrated (segmented). The exchange rate determination literature is used to identify potentially priced binational factors, and the sensitivities and factor prices are es- timated using non-linear SUR. The two equity markets are only partly integrated (segmented). Canadian stock returns are influenced by the pure domestic components of the term structure and lagged industrial production and by the pure international components of the differential

The budget deficit and the arbitrage pricing theory

Non-linear SUR techniques are used to demonstrate that the budget deficit is a priced factor in the Arbitrage Pricing Theory. The results also indicate that unexpected increases in the deficit lower stock prices.

Two estimators for the apt model when factors are measured

Non-linear SUR and ITSUR techniques are proposed for the estimation of the APT and the CAPM when the factors are observed. These techniques estimate all of the parameters of the model simultaneously and directly impose the model's non-linear parameter restrictions.