Search

Electricity market time series include several systematic components describing the long-term dynamics, the annual, weekly and daily periodicities, calendar effects, jumps, etc. As a result, modelling electricity variables requires the estimation of these components. For this purpose two main approaches have been proposed in the literature: the deterministic and the stochastic. Although an inappropriate modelling of systematic components could have important consequences on the prediction of loads and prices, in the literature it has not yet been assessed, which approach is more appropriate for electricity markets time series. This work aims at filling this gap by comparing the deterministic and the stochastic approach in a systematic way and in a homogeneous framework, both for loads and prices. In the deterministic case, components are represented by smoothing splines and dummy variables, while in the stochastic case they are described by stochastic processes common to the unobserved component modelling literature. As systematic components are not observable, the comparison is based on the prediction implications of the two procedures. This allows us to account for possible compensations among estimated components on the final result. Predictive performance is mainly assessed with respect to the one-day-ahead horizon, but also seven-day-ahead predictions are considered. The two approaches are evaluated on loads and prices of four important wholesale electricity markets: the Italian IPEX, the Scandinavian Nord Pool, the British EPEX SPOT UK and North American PJM.

This paper considers how well the approach of combining forecasts extends to the context of electricity prices. With the increasing popularity of regime switching and time-varying parameter models for predicting power prices, the multi model and evolutionary considerations that usually support the combining of simpler time series methods may be less applicable when the individual models incorporate these features. We address this question with a backtesting analysis on British day-ahead prices. Furthermore, given the volatility of power prices and concerns about accurate forecasting under extreme price excursions, we evaluate the results using various error metrics including expected shortfall. The comparisons are furthermore carefully simulated to consider model selection uncertainty in order to realistically test the value of combining as an ex ante policy. Overall, our results support combining for both accurate operational planning and risk management

Performance persistence is a relevant issue when evaluating the predictability of future results of managed portfolios. A related crucial aspect is the stability over time of the measure used to assess the performance, defined as the degree of association between the rankings of financial assets induced by the performance measure throughout subsequents periods. In this work a general class of possible criteria to measure stability is proposed. Then, the attention is focused on a specific index, whose asymptotic expected value and variance are derived under the null hypothesis of absence of stability. Furthermore, two statistical tests for evaluating the significance of stability are discussed. An application to a large set of US equity mutual funds shows that stability may remarkably vary, as the performance measure or the time widow width where it is computed change.

In this paper we present an extensive comparison of four different classes of models for daily forecasting of spot electricity prices, including ARMAX, constant and time-varying parameter regression models as well as non linear Markov regime-switching regressions. They are selected for particular reasons related to the emerging body of research on the price formation processes observed in electricity markets. The analyses are conducted for representative trading periods of the day in the UK Power Exchange prompt market, with the price series adjusted for their deterministic components and spikes. They show that relative out-of-sample forecasting performances are distinctly different for each trading period, season and across the actual performance metrics. No model consistently outperforms the others, but the ARMAX approach performs well in most cases and the Diebold and Mariano test indicates that, when it is not the best, the ARMAX model is not statistically different from the best. Nevertheless, we suggest that subtle differences in performance between different methods under different conditions are consistent with the apparent variations in the price formation processes by time of day and by season. We conclude with some observations on the disparities between the model specifications appropriate for understanding in-sample price formation and those for accurate out-of-sample predictions.

In the mutual funds industry the rating process is very important, and Morningstar is surely the most influential international rating agency. In this work we consider the problem of evaluating if the risk component is adequately accounted for in the Morningstar rating. To face this problem we compare the ratings produced giving different weights to the risk component. The focus of the analysis is on testing the hypothesis that two similar rating procedures with different risk parameters (or, in statistical terms, two raters) are equivalent. To that end, first the notion of −equivalence is introduced and then a Monte Carlo test for the hypothesis of −equivalence is described. Finally, to answer the question on the role of risk in the Morningstar rating, we analyze 1763 monthly return time series of US mutual funds. Results show that the current Morningstar classification, based on a risk-adjusted measure, only marginally accounts for risk and that if we want that risk really matters, the risk parameter should be increased.

A statistically correct procedure for evaluating management skill in the mutual fund industry is described. The approach is based on random and semi-random portfolios and is completely data-driven, allowing reference to a specific benchmark or to peer-groups to be avoided. Rather, it produces a new benchmark, which represents the minimum level of performance that skilled managament should reach. An application to some Italian equity funds is also given.

In this paper the prediction of Value-at-Risk by means of models accounting for higher moment dynamics is studied. We consider the GARCHDSK model, which allows for dynamic skewness and kurtosis, and compare its performance with that of several widely adopted models. The analysis is based on the study of sequences of (long and short) VaR violations, for which the hypotheses of absence of autocorrelation and of correct coverage rates are assessed. Both in-sample and out-of-sample results are investigated.

Distributional theory for Quasi-Maximum Likelihood estimators in long memory conditional heteroskedastic models is not formally defined, even asympotically. Because of that, this paper analyses the performance of the Likelihood Ratio and the Lagrange Multiplier misspecification tests for Periodic Long Memory GARCH models. The real size and power of these tests are studied by means of Monte Carlo simulations with respect to the class of Generalized Long Memory GARCH models. An application to the USD/JPY exchange rate is also provided.

In this paper marginal and conditional skewness of financial return time series is studied, by means of testing procedures and suitable models, for nine major international stock indexes. To analyze time-varying conditional skewness a new GARCH-type model with dynamic skewness and kurtosis is proposed. Results indicate that there are no evidences of marginal asymmetry in the nine series, but there are clear findings of significant time-varying conditional skewness. The economic significance of conditional skewness is analyzed and compared by considering Value-at-Risk, Expected Shortfall and Market Risk Capital Requirements set by the Basel Accord.

A distinguishing feature of the intra-day time-varying volatility of financial time series is given by the presence of long-range dependence of periodic type due mainly to time-of-the-day phenomena. In this work we introduce a model able to describe the empirical evidence given by this periodic longmemory behaviour. The model, named PLM-GARCH (Periodic Long Memory GARCH), represents a natural extension of the FIGARCH model proposed for modelling long-range persistence of the volatility of financial time series. Periodic long memory versions of EGARCH (PLM-EGARCH) models are also considered. Some properties and characteristics of the models are given and an estimation procedure based on quasi maximum likelihood is established. Further possible extensions of the model to take into account multiple sources of periodic long-memory behaviour are suggested. Some empirical applications on intra-day financial time series are also provided.

In this work we introduce an extension of the FIGARCH model, called SFIGARCH, to account for periodic and long memory components in intraday volatility of financial returns. An application on a real time series is provided.

This paper examines the problem of evaluating the presence of asymmetry in the marginal distribution of financial returns, by means of a suitable statistical test. After a brief description of existing tests, a bootstrap procedure is proposed which leads to an approximation of the Bai and Ng test under weaker assumptions. A Monte Carlo study showed that our test works properly and that, in terms of power, it is competitive with existing tests. An application to real financial time series is also presented.

One of the main concerns in air pollution is excessive tropospheric ozone concentration. The aim of this work is to develop statistical models giving short-term prediction of future ground-level ozone concentrations. Since there are few physical insights about the dynamic relationship between ozone, precursor emissions and/or meteorological factors, a nonparametric and nonlinear approach sems promising in order to specify the prediction models. First, we apply four nonparametric procedures to forecast daily maximum 1-hour and maximum 8-hours averages of ozone concentrations in an urban area. Then, in order to improve the prediction performances, we combine the time series of the forecasts. This idea seems to give promising results.