Evaluating multivariate volatility forecasts : how effective are statistical and economic loss functions?

Multivariate volatility forecasts are an important input in many financial applications, in particular portfolio optimisation problems. Given the number of models available and the range of loss functions to discriminate between them, it is obvious that selecting the optimal forecasting model is challenging. The aim of this thesis is to thoroughly investigate how effective many commonly used statistical (MSE and QLIKE) and economic (portfolio variance and portfolio utility) loss functions are at discriminating between competing multivariate volatility forecasts. An analytical investigation of the loss functions is performed to determine whether they identify the correct forecast as the best forecast. This is followed by an extensive simulation study examines the ability of the loss functions to consistently rank forecasts, and their statistical power within tests of predictive ability. For the tests of predictive ability, the model confidence set (MCS) approach of Hansen, Lunde and Nason (2003, 2011) is employed. As well, an empirical study investigates whether simulation findings hold in a realistic setting. In light of these earlier studies, a major empirical study seeks to identify the set of superior multivariate volatility forecasting models from 43 models that use either daily squared returns or realised volatility to generate forecasts. This study also assesses how the choice of volatility proxy affects the ability of the statistical loss functions to discriminate between forecasts. Analysis of the loss functions shows that QLIKE, MSE and portfolio variance can discriminate between multivariate volatility forecasts, while portfolio utility cannot. An examination of the effective loss functions shows that they all can identify the correct forecast at a point in time, however, their ability to discriminate between competing forecasts does vary. That is, QLIKE is identified as the most effective loss function, followed by portfolio variance which is then followed by MSE. The major empirical analysis reports that the optimal set of multivariate volatility forecasting models includes forecasts generated from daily squared returns and realised volatility. Furthermore, it finds that the volatility proxy affects the statistical loss functions’ ability to discriminate between forecasts in tests of predictive ability. These findings deepen our understanding of how to choose between competing multivariate volatility forecasts.

Semi-parametric forecasting of spikes in electricity prices

The occurrence of extreme movements in the spot price of electricity represents a significant source of risk to retailers. A range of approaches have been considered with respect to modelling electricity prices; these models, however, have relied on time-series approaches, which typically use restrictive decay schemes placing greater weight on more recent observations. This study develops an alternative, semi-parametric method for forecasting, which uses state-dependent weights derived from a kernel function. The forecasts that are obtained using this method are accurate and therefore potentially useful to electricity retailers in terms of risk management.

Forecasting the future risk of Barmah Forest Virus disease under climate change scenarios in Queensland, Australia

BACKGROUND Mosquito-borne diseases are climate sensitive and there has been increasing concern over the impact of climate change on future disease risk. This paper projected the potential future risk of Barmah Forest virus (BFV) disease under climate change scenarios in Queensland, Australia. METHODS/PRINCIPAL FINDINGS We obtained data on notified BFV cases, climate (maximum and minimum temperature and rainfall), socio-economic and tidal conditions for current period 2000-2008 for coastal regions in Queensland. Grid-data on future climate projections for 2025, 2050 and 2100 were also obtained. Logistic regression models were built to forecast the otential risk of BFV disease distribution under existing climatic, socio-economic and tidal conditions. The model was applied to estimate the potential geographic distribution of BFV outbreaks under climate change scenarios. The predictive model had good model accuracy, sensitivity and specificity. Maps on potential risk of future BFV disease indicated that disease would vary significantly across coastal regions in Queensland by 2100 due to marked differences in future rainfall and temperature projections. CONCLUSIONS/SIGNIFICANCE We conclude that the results of this study demonstrate that the future risk of BFV disease would vary across coastal regions in Queensland. These results may be helpful for public health decision making towards developing effective risk management strategies for BFV disease control and prevention programs in Queensland.

Beyond point forecasting : evaluation of alternative prediction intervals for tourist arrivals

This paper evaluates the performances of prediction intervals generated from alternative time series models, in the context of tourism forecasting. The forecasting methods considered include the autoregressive (AR) model, the AR model using the bias-corrected bootstrap, seasonal ARIMA models, innovations state space models for exponential smoothing, and Harvey’s structural time series models. We use thirteen monthly time series for the number of tourist arrivals to Hong Kong and Australia. The mean coverage rates and widths of the alternative prediction intervals are evaluated in an empirical setting. It is found that all models produce satisfactory prediction intervals, except for the autoregressive model. In particular, those based on the biascorrected bootstrap perform best in general, providing tight intervals with accurate coverage rates, especially when the forecast horizon is long.

Better Macadamia crop forecasting

Better Macadamia crop forecasting.

Better Macadamia Crop Forecasting - Part 2

The continually expanding macadamia industry needs an accurate crop forecasting system to allow it to develop effective crop handling and marketing strategies, particularly when the industry faces recurring cycles of unsustainably high and low commodity prices. This project aims to provide the AMS with a robust, reliable predictive model of national crop volume within 10% of the actual crop by 1 April each year by factoring known seasonal, environmental, cultural, climatic, management and biological constraints, together with the existing AMS database which includes data on tree numbers, tree age, variety, location and previous season's production.

Understanding urban rail travel for improved patronage forecasting

Improved forecasting of urban rail patronage is essential for effective policy development and efficient planning for new rail infrastructure. Past modelling and forecasting of urban rail patronage has been based on legacy modelling approaches and often conducted at the general level of public transport demand, rather than being specific to urban rail. This project canvassed current Australian practice and international best practice to develop and estimate time series and cross-sectional models of rail patronage for Australian mainland state capital cities. This involved the implementation of a large online survey of rail riders and non-riders for each of the state capital cities, thereby resulting in a comprehensive database of respondent socio-economic profiles, travel experience, attitudes to rail and other modes of travel, together with stated preference responses to a wide range of urban travel scenarios. Estimation of the models provided a demonstration of their ability to provide information on the major influences on the urban rail travel decision. Rail fares, congestion and rail service supply all have a strong influence on rail patronage, while a number of less significant factors such as fuel price and access to a motor vehicle are also influential. Of note, too, is the relative homogeneity of rail user profiles across the state capitals. Rail users tended to have higher incomes and education levels. They are also younger and more likely to be in full-time employment than non-rail users. The project analysis reported here represents only a small proportion of what could be accomplished utilising the survey database. More comprehensive investigation was beyond the scope of the project and has been left for future work.

Dynamic Factor Models and Forecasting Finnish Macroeconomic Variables

In recent years, thanks to developments in information technology, large-dimensional datasets have been increasingly available. Researchers now have access to thousands of economic series and the information contained in them can be used to create accurate forecasts and to test economic theories. To exploit this large amount of information, researchers and policymakers need an appropriate econometric model.Usual time series models, vector autoregression for example, cannot incorporate more than a few variables. There are two ways to solve this problem: use variable selection procedures or gather the information contained in the series to create an index model. This thesis focuses on one of the most widespread index model, the dynamic factor model (the theory behind this model, based on previous literature, is the core of the first part of this study), and its use in forecasting Finnish macroeconomic indicators (which is the focus of the second part of the thesis). In particular, I forecast economic activity indicators (e.g. GDP) and price indicators (e.g. consumer price index), from 3 large Finnish datasets. The first dataset contains a large series of aggregated data obtained from the Statistics Finland database. The second dataset is composed by economic indicators from Bank of Finland. The last dataset is formed by disaggregated data from Statistic Finland, which I call micro dataset. The forecasts are computed following a two steps procedure: in the first step I estimate a set of common factors from the original dataset. The second step consists in formulating forecasting equations including the factors extracted previously. The predictions are evaluated using relative mean squared forecast error, where the benchmark model is a univariate autoregressive model. The results are dataset-dependent. The forecasts based on factor models are very accurate for the first dataset (the Statistics Finland one), while they are considerably worse for the Bank of Finland dataset. The forecasts derived from the micro dataset are still good, but less accurate than the ones obtained in the first case. This work leads to multiple research developments. The results here obtained can be replicated for longer datasets. The non-aggregated data can be represented in an even more disaggregated form (firm level). Finally, the use of the micro data, one of the major contributions of this thesis, can be useful in the imputation of missing values and the creation of flash estimates of macroeconomic indicator (nowcasting).

Testing the Forecasting Performance of Ibex 35 Option-implied Risk-neutral Densities

Published also as: Documento de Trabajo Banco de España 0504/2005.

Volatility and VaR Forecasting in the Madrid Stock Exchange

This paper provides an empirical study to assess the forecasting performance of a wide range of models for predicting volatility and VaR in the Madrid Stock Exchange. The models performance was measured by using different loss functions and criteria. The results show that FIAPARCH processes capture and forecast more accurately the dynamics of IBEX-35 returns volatility. It is also observed that assuming a heavy-tailed distribution does not improve models ability for predicting volatility. However, when the aim is forecasting VaR, we find evidence of that the Student’s t FIAPARCH outperforms the models it nests the lower the target quantile.

Assessment of Energy Conservation in Indian Cement Industry and Forecasting of Co2 Emissions

Cement industry ranks 2nd in energy consumption among the industries in India. It is one of the major emitter of CO2, due to combustion of fossil fuel and calcination process. As the huge amount of CO2 emissions cause severe environment problems, the efficient and effective utilization of energy is a major concern in Indian cement industry. The main objective of the research work is to assess the energy cosumption and energy conservation of the Indian cement industry and to predict future trends in cement production and reduction of CO2 emissions. In order to achieve this objective, a detailed energy and exergy analysis of a typical cement plant in Kerala was carried out. The data on fuel usage, electricity consumption, amount of clinker and cement production were also collected from a few selected cement industries in India for the period 2001 - 2010 and the CO2 emissions were estimated. A complete decomposition method was used for the analysis of change in CO2 emissions during the period 2001 - 2010 by categorising the cement industries according to the specific thermal energy consumption. A basic forecasting model for the cement production trend was developed by using the system dynamic approach and the model was validated with the data collected from the selected cement industries. The cement production and CO2 emissions from the industries were also predicted with the base year as 2010. The sensitivity analysis of the forecasting model was conducted and found satisfactory. The model was then modified for the total cement production in India to predict the cement production and CO2 emissions for the next 21 years under three different scenarios. The parmeters that influence CO2 emissions like population and GDP growth rate, demand of cement and its production, clinker consumption and energy utilization are incorporated in these scenarios. The existing growth rate of the population and cement production in the year 2010 were used in the baseline scenario. In the scenario-1 (S1) the growth rate of population was assumed to be gradually decreasing and finally reach zero by the year 2030, while in scenario-2 (S2) a faster decline in the growth rate was assumed such that zero growth rate is achieved in the year 2020. The mitigation strategiesfor the reduction of CO2 emissions from the cement production were identified and analyzed in the energy management scenarioThe energy and exergy analysis of the raw mill of the cement plant revealed that the exergy utilization was worse than energy utilization. The energy analysis of the kiln system showed that around 38% of heat energy is wasted through exhaust gases of the preheater and cooler of the kiln sysetm. This could be recovered by the waste heat recovery system. A secondary insulation shell was also recommended for the kiln in the plant in order to prevent heat loss and enhance the efficiency of the plant. The decomposition analysis of the change in CO2 emissions during 2001- 2010 showed that the activity effect was the main factor for CO2 emissions for the cement industries since it is directly dependent on economic growth of the country. The forecasting model showed that 15.22% and 29.44% of CO2 emissions reduction can be achieved by the year 2030 in scenario- (S1) and scenario-2 (S2) respectively. In analysing the energy management scenario, it was assumed that 25% of electrical energy supply to the cement plants is replaced by renewable energy. The analysis revealed that the recovery of waste heat and the use of renewable energy could lead to decline in CO2 emissions 7.1% for baseline scenario, 10.9 % in scenario-1 (S1) and 11.16% in scenario-2 (S2) in 2030. The combined scenario considering population stabilization by the year 2020, 25% of contribution from renewable energy sources of the cement industry and 38% thermal energy from the waste heat streams shows that CO2 emissions from Indian cement industry could be reduced by nearly 37% in the year 2030. This would reduce a substantial level of greenhouse gas load to the environment. The cement industry will remain one of the critical sectors for India to meet its CO2 emissions reduction target. India’s cement production will continue to grow in the near future due to its GDP growth. The control of population, improvement in plant efficiency and use of renewable energy are the important options for the mitigation of CO2 emissions from Indian cement industries

Forecasting the spot prices of various coffee types using linear and non-linear error correction models

Este documento estima modelos lineales y no-lineales de corrección de errores para los precios spot de cuatro tipos de café. En concordancia con las leyes económicas, se encuentra evidencia que cuando los precios están por encima de su nivel de equilibrio, retornan a éste mas lentamente que cuando están por debajo. Esto puede reflejar el hecho que, en el corto plazo, para los países productores de café es mas fácil restringir la oferta para incrementar precios, que incrementarla para reducirlos. Además, se encuentra evidencia que el ajuste es más rápido cuando las desviaciones del equilibrio son mayores. Los pronósticos que se obtienen a partir de los modelos de corrección de errores no lineales y asimétricos considerados en el trabajo, ofrecen una leve mejoría cuando se comparan con los pronósticos que resultan de un modelo de paseo aleatorio.