Rolling Regression Research Articles

The non-linear link between electricity consumption and temperature in Europe: A threshold panel approach

This paper investigates the relationship between electricity demand and temperature in the European Union. We address this issue by means of a panel threshold regression model on 15 European countries over the last two decades. Our results confirm the non-linearity of the link between electricity consumption and temperature found in more limited geographical areas in previous studies. By distinguishing between North and South countries, we also find that this non-linear pattern is more pronounced in the warm countries. Finally, rolling regressions show that the sensitivity of electricity consumption to temperature in summer has increased in the recent period.

The Narrow Money Demand Behaviour in Indonesia, 1970– 2005

This paper specifies a simple partial adjustment model of narrow money demand for Indonesia in an open economy context and estimates it with annual data for the period 1970-2005. Empirical results suggest that real income, inflation (a proxy for expected inflation), and the return on foreign financial assets are the major determinants of narrow money demand in this country. The deposit rate of interest (a proxy for the interest rate on alternative domestic financial assets) is statistically insignificant. The CUSUM and CUSUMSQ of residuals' tests suggest that the narrow money demand function remained stable during the sample period of study. Having selected an empirically parsimonious model, both the recursive and rolling regression techniques are applied to investigate the possibility of structural change in the regression coefficients possibly due to financial reforms and innovations since the early 1990s. Empirical results suggest that since the late 1990s the long-run income elasticity of the demand for narrow money has remained relatively stable around a value not different from one while the narrow money demand elasticities with respect to the rates of inflation and the return on foreign financial assets have declined markedly. The paper interprets the overall empirical results in a historical context and draws implications on the strategy of monetary policy in Indonesia.

Central Bank Forex Interventions Assessed Using Realized Moments

This paper studies and assesses the impact of G3 Central Bank interventions on the DEM/USD exchange rate properties using daily realized moments of exchange rate returns (obtained from intraday data) for the period 1989-2001. Event studies in terms of the realized moments for the intervention day, the days preceding and following the intervention day illustrate the shape of this impact. Rolling regressions results for an ARFIMA model for realized moments are used to measure the intervention impact and characterize its significance. The analysis confirms previous findings of an increase of volatility after a coordinated Central Bank intervention. It highlights new findings on the timing and the persistence of coordinated interventions on exchange rate volatility, on important volatility spillovers, on the impact on exchange rate covariances and correlations and on skewness coefficients.

Central bank FOREX interventions assessed using realized moments

This paper assesses the impact of G3 official central bank interventions on daily realized moments of DEM/USD exchange rate returns obtained from intraday data, 1989–2001. Event studies of the realized moments for the intervention day, the days preceding and following the intervention illustrate the shape of this impact. Rolling regressions results for an AR(FI)MA model for realized moments are used to measure the impact and its significance. The analysis confirms previous empirical findings of a temporary increase of volatility after a coordinated central bank intervention. It highlights new findings on the timing and the temporary nature of the impact of coordinated interventions on exchange rate volatility and on cross-moments between foreign exchange markets.

Time-series and cross-sectional excess comovement in stock indexes

Abstract This paper is an empirical investigation of the excess comovement among 82 industry indexes in the U.S. stock market between January 5, 1976 and December 31, 2001. We define excess comovement as the covariation between two assets beyond what can be explained by fundamental factors. In our analysis, the fundamental factors are sector groupings and the three Fama-French factors. We then estimate residuals of joint (FGLS) rolling regressions of these fundamentals on industry returns. Finally, we compute excess comovement as the mean of square unconditional, statistically significant correlations of these residuals. We show that excess comovement is high (about 0.07, i.e., equivalent to an average absolute correlation of 0.26), statistically significant, and represents an economically significant portion (almost 30%) of the average gross square return correlation. Excess comovement is also uniformly significant across industries and over time and only weakly asymmetric, i.e., not significantly different in rising or falling markets. We explain more than 23% of this market-wide (and up to 73% of sector-wide) excess square correlation by its positive relation to proxies for information heterogeneity and U.S. monetary and real conditions, and its negative relation to market volatility and the level of the short-term interest rate. This evidence is consistent with the implications of portfolio rebalancing and product market theories of financial contagion, but offers little or no support for the correlated liquidity shock channel.

Irving Fisher, the UIP Puzzle and the 'Peso Problem'

In this paper, we first review Irving Fisher's seminal work on UIP and on the closely related equation linking interest rates and inflation relation. We go on to re-examine the performance of UIP since the advent of floating exchange rates in the 1970s. Like Fisher a century ago, we find that the failures of UIP are tied in with individual episodes in which errors surrounding exchange-rate expectations have been persistent but in the end transitory. We see evidence of this behavior both in the changed coefficients estimates from rolling regressions. We also find considerable commonality in deviations from UIP and PPP suggesting that these deviations are both driven by a common factor as the forecasting errors in exchange rates. Using a dynamic latent factor model we find that deviations from UIP are almost completely due to forecasting errors in exchange rates. Once the variation in the size of the forecasting errors is taken into account we find empirical support of a unitary value for the implied beta in the Fama regression. Using a number of countries we find unanimous support that deviations from UIP are driven by errors made in forecasting exchange rates, a result which we attribute to Irving Fisher who first mentioned this a century to date.

Testing for convergence of the Okun’s Law coefficient in Europe

Without convergence of macroeconomic parameters representative of the nature of adjustment mechanisms, even a common shock to members of a monetary union can lead to different macroeconomic consequences across the union and, eventually, to the need for more or less co-ordinated specific policies. In this paper, we test for the presence of convergence of the Okun’s Law coefficient (OLC) among several alternative groupings of European economies. The empirical strategy adopted is based on the evaluation of the time path of rolling regression estimates of the OLC for European countries. We then use a testing procedure suggested by Evans [Evans P (1996) J Econ Dynam Control 20:1027–1049] to investigate the convergence, or non-convergence, of the OLC in several groups of European countries by examining how the cross-country variance of the OLC evolves over time in these groups. A hypothesis of medium-term convergence of the OLC is rejected for most of the European country groups examined.

Is There a Risk-Return Tradeoff? Evidence from High-Frequency Data

This paper examines the intertemporal relation between risk and return for the aggregate stock market using high-frequency data. We use daily realized, GARCH, implied, and range-based volatility estimators to determine the existence and significance of a risk-return tradeoff for several stock market indices. We find a positive and statistically significant relation between the conditional mean and conditional volatility of market returns at the daily level. This result is robust to alternative specifications of the volatility process, across different measures of market return and sample periods, and after controlling for macroeconomic variables associated with business cycle fluctuations. We also analyze the risk-return relationship over time using rolling regressions, and find that the strong positive relation persists throughout our sample period. The market risk measures adopted in the paper add power to the analysis by incorporating valuable information, either by taking advantage of high frequency intraday data (in the case of realized, GARCH, and range volatility) or by utilizing the market's expectation of future volatility (in the case of implied volatility index).

Was the Introduction of IFRS in Europe Value-Relevant?

The introduction of IFRS in Europe was intended to improve accounting quality and facilitate cross-border financial transactions. This study examines whether IFRS increases value-relevance of corporate disclosure using traders' response in financial markets of 15 EU countries. If the regulatory reform improves informativeness of financial reporting, there will be a significant decrease in the bid-ask spread, and more particularly in its information asymmetry component (AIc). Using AIc of 3,907 companies in the period 2003-2005, we investigate how information asymmetry changes when legal origin, country earnings management and corporate disclosure scores differ. On average, the spread decreases over the period, whereas AIc does not. We suggest that IFRS adoption is positively associated with a significant increase in both size and volatility of AIc for UK firms. The increased and stable proportion of free float shares for UK and Scandinavian firms we also attribute to the regulatory change. It seems that AIc does not become more volatile under IFRS for French and Scandinavian origin groups, whereas it changes greatly for the UK and German groups. Our results show that firms with a higher country score for earnings management have a lower AIc. Additionally, higher disclosure is positively correlated with larger information asymmetry. Using aggregate AIc obtained with rolling regressions, we explain variance in information asymmetry with market value, spread size, volume, free float and returns. Our empirical results confirm that, when spread size increases, AIc decreases across the four country groups. The marginal effect of the spread size on information asymmetry is stronger for UK firms. We argue that although average bid-ask spread decreases over the period 2003-2005, it does not necessarily lead to a lower AIc in 2005.

Is there a risk–return trade‐off? Evidence from high‐frequency data

AbstractThis paper examines the intertemporal relation between risk and return for the aggregate stock market using high‐frequency data. We use daily realized, GARCH, implied, and range‐based volatility estimators to determine the existence and significance of a risk–return trade‐off for several stock market indices. We find a positive and statistically significant relation between the conditional mean and conditional volatility of market returns at the daily level. This result is robust to alternative specifications of the volatility process, across different measures of market return and sample periods, and after controlling for macro‐economic variables associated with business cycle fluctuations. We also analyze the risk–return relationship over time using rolling regressions, and find that the strong positive relation persists throughout our sample period. The market risk measures adopted in the paper add power to the analysis by incorporating valuable information, either by taking advantage of high‐frequency intraday data (in the case of realized, GARCH, and range volatility) or by utilizing the market's expectation of future volatility (in the case of implied volatility index). Copyright © 2006 John Wiley & Sons, Ltd.

M1 demand and volatility

The reasons for the instability within the M1 market during the 1970s and 1980s have been attributed to an actual breakdown in the M1 market, multicollinearity, or an inappropriate specification. The present paper offers a more general reason: increased volatility within the money demand determinants produced an econometric environment which made estimation of the relationship problematic. In order to investigate this possibility, the traditional demand vector is re-estimated with additional generalized autoregressive conditional heteroskedasticity representation for income, inflation, and interest rate volatility as conditioning variables. The rolling regression results highlight a significant stability within the M1 demand vector and its long- and short-run parameters.

Portfolio optimisation under changing risk via time‐varying beta

PurposeThe paper is aimed at modelling time varying betas via a state space representation in order to decompose the marginal contribution to risk of downside and upside deviations of asset returns in portfolio optimisation.Design/methodology/approachThe approach enables to take into account the relationship between risk and excess returns in up‐side and down‐side markets and to arrange a flexible asset allocation model which directly incorporates the investor risk tolerance to positive or negative expected market moves. The model volatility through state space models and the Kalman filter, widely used to recursively and optimally estimate time varying betas.FindingsThe study shows that the application of an asset allocation model which splits beta in two parts, one related to Bear and the other to Bull markets, and reconciles them with a non negative risk aversion parameter may produce interesting financial results if compared with typical passive portfolios. The proposed model was tested by conducting extensive empirical evaluations on a set securities belonging to eight different markets. The outcomes show that active strategies can be developed and can lead to better performances.Research implicationsThe research affects optimisation models in particular considering the volatility indicators usually estimated not only by researchers but also by practitioners.Originality/valueIn financial literature we find empirical evidence that the constant beta model may be inaccurate and hazardous to use in asset allocation decisions and many statistical techniques have been developed to estimate time dependent betas. Rolling regression procedures allow to capture beta dynamics but require the definition of the estimation period. The paper provides an empirical analysis referred both to European and American market data which let us to allocate assets avoiding the usual limits of standard volatility indicators.

Insight and strategy in multiple-cue learning.

In multiple-cue learning (also known as probabilistic category learning) people acquire information about cue-outcome relations and combine these into predictions or judgments. Previous researchers claimed that people can achieve high levels of performance without explicit knowledge of the task structure or insight into their own judgment policies. It has also been argued that people use a variety of suboptimal strategies to solve such tasks. In three experiments the authors reexamined these conclusions by introducing novel measures of task knowledge and self-insight and using "rolling regression" methods to analyze individual learning. Participants successfully learned a four-cue probabilistic environment and showed accurate knowledge of both the task structure and their own judgment processes. Learning analyses suggested that the apparent use of suboptimal strategies emerges from the incremental tracking of statistical contingencies in the environment.

Pricing and Hedging Illiquid Energy Derivatives: An Application to the JCC Index

In this paper we discuss a simple strategy for pricing and hedging illiquid financial products, such as the Japanese crude oil cocktail (JCC) index, the most popular OTC energy derivative in Japan. First, we review the existing literature for computing optimal hedge ratios (OHR) and we propose a critical classification of the existing approaches. Second, we compare the empirical performance of different econometric models (namely, regression models for price levels, price first differences, and price returns, as well as error correction and autoregressive distributed lag models) in terms of their computed OHR, using monthly data on the JCC over the period Janury 2000-January 2006. Third, we illustrate and implement a reliable procedure to price and hedge a swap contract on the JCC with a variable oil volume. We explain how to compute a bid/ask spread and to construct the hedging position for the JCC swap. Fourth, we evaluate our swap pricing scheme with backtesting and rolling regression techniques. Our empirical findings show that it is not necessary to use complicated econometric techniques, since the price-level regression model permits to compute more precise optimal hedge ratios relative to its competing alternatives.

A cross-country analysis of the Okun's Law coefficient convergence in Europe

The paper examines whether or not evidence is consistent with convergence of the Okun's Law coefficient (OLC) among several alternative groupings of European economies. A two-step empirical strategy is employed. The first step obtains rolling regression estimates of the OLC for individual European countries. The second step examines how the cross-country variance of the OLC evolves over the decade until 2002 in the selected country groupings. Evidence is found consistent with convergence of the OLC among northern European countries, and among countries with centralized wage bargaining, but an absence of convergence in other country groups.

Bayesian Analysis of Stochastic Betas

AbstractWe propose a mean-reverting stochastic process for the market beta. In a simulation study, the proposed model generates significantly more precise beta estimates than GARCH betas, betas conditioned on aggregate or firm-level variables, and rolling regression betas, even when the true betas are generated based on these competing specifications. Our model significantly improves out-of-sample hedging effectiveness. In asset pricing tests, our model provides substantially stronger support for the conditional CAPM relative to competing beta models and helps resolve asset pricing anomalies such as the size, book-to-market, and idiosyncratic volatility effects in the cross section of stock returns.

가구주 학력이 외식수요에 미치는 효과

본 연구는 총 가구주의 교육정도가 외식수요에 미치는 효과를 밝히기 위하여 식료품에 대한 교육계층별 지출패턴을 조사하여 외식에 대한 지출구조와 비교하였다. 그 결과 소득에서 식료품에 대한 지출이 차지하는 비율이 가구주가 무학일 경우 가장 높고 교육정도가 높을수록 낮아져 대학원 졸업에서 가장 낮게 나타났다. 모든 교육계층에서 외식비가 소득에서 차지하는 비율이 시간의 흐름과 더불어 꾸준히 상승하는 것으로 나타났다. 이것은 외식문화가 보편화ㆍ대중화되어 낮은 교육계층에는 외식비가 가계에 부담이 될 가능성이 크다는 것을 의미하였다. 외식수요함수를 추정한 결과 외식가격상승은 무학과 대학원졸을 제외한 교육계층에서 외식수요를 감소시키고, 소득증가는 모든 교육계층에서 외식수요를 증가시키는 요인으로 작용하였다. 그리고 모든 교육계층에서 소득계수가 가격계수보다 커서 가격보다 소득이 외식수요에 큰 영향을 미쳤다. 시간에 따른 외식 수요 패턴을 보기 위한 전향적 이동회귀는 소득과 가격의 영향력이 감소함을 알 수 있었다. 외식의 단기적 특성을 보기 위해 충격반응함수를 도입한 결과는 모든 교육계층에서 가격충격은 외 식수요를 감소시키며, 소득충격은 외식수요를 증가시키나, 가격충격은 소득충격에 비해 외식수요에 미치는 충격의 크기도 대단히 작았다.

Instability and predictability of factor betas of industrial stocks: The Flexible Least Squares solutions

By using the Flexible Least Squares (FLS) method, this study traces out month-to-month alterations in factor betas. The time variation paths of factor betas reveal time-varying correlations between different factor betas. Moreover, the FLS method is found to be able to produce more accurate and stable forecasts of industry costs of equity than rolling regressions and other methods, in terms of smaller forecasting errors and standard deviation of forecasting errors, thanks to the better estimates of factor betas.

Asymmetric volatility of stock returns during the Asian crisis: Evidence from Indonesia

This paper investigates the conditional volatility in stock returns in Indonesia over the period covered by the Asian crisis. Rolling regression parameter estimates from three asymmetric volatility models suggested that all parameters, including those capturing asymmetric response, were time-varying. The precise pattern of adjustment was sensitive to model selection. Nevertheless, increases in asymmetric response patterns appear to coincide with the very large exchange rate devaluations in the rupiah over this period and were followed by more general symmetric short-term volatility in the post crisis period. Estimates from a smooth transition volatility model indicated both sign and size asymmetries during the crisis period.

Uncovered interest rate parity and analysis of monetary convergence of potential EMU accession countries

This paper analyzes deviations from uncovered interest rate parity which are interpreted as indicator of the substitutability of currencies. Backward recursive statistical tests and error correction models are applied to study the co-movement of interest rates, and rolling regressions are used to study size and volatility of country specific risk premiums. In accordance to their degree of monetary integration with the Euro area, new EU members and accession countries are divided into three groups. Estonia and Lithuania seem to exhibit the highest degree of monetary integration with the Euro area.