1,720,975 research outputs found
Forecasting realised volatility using regime-switching models
No full textThis paper extends standard AR and HAR models for realised volatility (RV) forecasting to include nonlinearity through two broad regime-switching approaches, the smooth transition and Markov-switching methods. Using daily data for eight international stock markets over the period 2007–2021, a comprehensive comparison is provided using a range of forecast tests that includes statistical and economic (risk management) based metrics. The results show that regime-switching models provide a better in-sample fit and out-of-sample forecasting, although this latter result is less clear-cut at the daily horizon. In comparing the two nonlinear approaches, we find that the abrupt transition technique of the Markov-switching model is preferred to the smooth transition one. It is believed that our results will be of interest to those especially engaged in risk management practice as well as for those modelling market behaviour.</p
Forecasting Realised Volatility: Does the LASSO approach outperform HAR?
Full text linkThe HAR model dominates current volatility forecasting. This model implies a restricted lag approach, with three parameters accounting for an AR(22) structure. This paper uses the Lasso method, which selects a parsimonious lag structure, while allowing both a flexible lag structure and lags greater than 22. In-sample results suggest that while significance is largely found among the first 22 lags, consistent with the HAR model, there is evidence that longer lags contain information, as Lasso models provide an improved fit. Out-of-sample forecasts for daily, weekly and monthly volatility, evaluated using MSE, QLIKE, MCS and VaR measures, suggest that the ordered Lasso model provides the preferred forecasts using an AR(100) at the daily level and an AR(22) for the weekly and monthly horizons. The results support the view that a more flexible lag structure is preferred over the HAR approach
Essays on financial volatility forecasting
Full text linkThe accurate estimation and forecasting of volatility is of utmost importance for anyone who
participates in the financial market as it affects the whole financial system and, consequently,
the whole economy. It has been a popular subject of research with no general conclusion as to
which model provides the most accurate forecasts. This thesis enters the ongoing debate by
assessing and comparing the forecasting performance of popular volatility models. Moreover,
the role of key parameters of volatility is evaluated in improving the forecast accuracy of
the models. For these purposes a number of US and European stock indices is used. The
main contributions are four. First, I find that implied volatility can be per se forecasted
and combining the information of implied volatility and GARCH models predict better the
future volatility. Second, the GARCH class of models are superior to the stochastic volatility
models in forecasting the one-, five- and twenty two-days ahead volatility. Third, when the
realised volatility is modelled and forecast directly using time series, I find that the HAR model
performs better than the ARFIMA. Finally, I find that the leverage effect and implied volatility
significantly improve the fit and forecasting performance of all the models
Essays in Financial Econometrics: Conditional Volatility, Realized Volatility and Volatility Spillovers
Full text linkThe accurate forecast of stock market volatility is of particular importance for policy makers, investors, and market participants who have certain levels of risk which they can bear. This thesis centres around the conditional volatility, realized volatility, and volatility spillovers in the context of their model extensions. In particular, we examine the behaviour of stock market volatility in a selection of international markets, the ability of different models to provide accurate volatility forecasts, and the nature of the interrelations between markets from the perspective of complex network theory. Focussing on the modelling and forecasting of volatility we compare some well-established conditional volatility models with realized volatility models and further investigate the use of a number of additional parameters in improving the forecast accuracy of the future realized volatility. In this regard, a wide range of additional parameters, from assets to commodities, extreme range estimators to overnight volatility, oil price to gold price, VIX to EPU, bond price to interest rate, are included. Moreover, those are classified as different information channels, namely local, regional, and global. In terms of volatility spillovers, a volatility spillover model is combined with complex network theory in order to construct a volatility network of international financial markets, consisting of nodes and edges. The main contributions of this thesis are four. First, using the thirty different stock market indices and more up-to-date data the realized volatility (HAR-RV) models outperform the conditional volatility (GARCHs) models and, moreover, decomposition of realized volatility into positive and negative realized semi-variances (HAR-PS) improve the forecast accuracy of HAR-RV model. Second, extreme range estimators such as Parkinson and Garman-Klass could contain additional information for forecasting the future realized volatility. Third, the role of global information at improving the forecasts of future realized volatility is more important than that of local and regional information. Lastly, the spillover networks of international financial markets are much denser in crisis periods compared to non- crisis periods and volatility spillovers in COVID-19 Crisis (2020) period are more transitive and intense than Global Financial Crisis (2008) period
Essays on Realised Volatility Forecasting for International Stock Markets
Full text linkModelling and forecasting market volatility is an important topic within finance research, with the aim of producing accurate forecasts, as confirmed by the plethora of academic papers written over the past few decades. Understanding volatility is crucial for market participants such as investors, policymakers, and academics. The linear Heterogeneous Autoregressive (HAR) model currently dominates the volatility models for forecasting Realised Volatility (RV). This thesis enters the ongoing volatility forecasting debate by developing further the HAR model. First, within the HAR setting volatility jumps, realised semi-variance and the leverage effect are added. With the use of a selection of loss functions and forecasting comparisons it is found that adding the leverage effect into the HAR model can produce the most accurate forecasts over daily, weekly, and monthly horizons. Second, this thesis compares the foresting ability of the Autoregressive (AR) model with flexible lags, generated by the Least Absolute Shrinkage & Selection Operator (Lasso) approach (es), to the HAR model with a fixed lag structure. In-sample results show the Lasso approach to improve the model fitness, and the out-of-sample results indicate a more flexible lag structure is preferred, especially the ordered Lasso performs the best. Third, this thesis incorporates the Smooth Transition and Markov-switching approaches with the linear HAR model in a further forecasting exercise. In-sample results show that the regime-switching models provide better estimation accuracy than the linear HAR model. For the out-of-sample results, although the regime-switching models have limited forecasting ability over the daily horizon, these do outperform the linear HAR model over weekly and monthly horizons. The Markov-switching model is found to be the best, by consistently exhibiting the most accurate forecasts over time. All the above findings have been evaluated within a risk management setting (Value at Risk & Expected Shortfall)
Uncovering hidden information and relations in time series data with wavelet analysis: three case studies in finance
Full text linkThis thesis aims to provide new insights into the importance of decomposing aggregate time series data using the Maximum Overlap Discrete Wavelet Transform. In particular, the analysis throughout this thesis involves decomposing aggregate financial time series data at hand into approximation (low-frequency) and detail (high-frequency) components. Following this, information and hidden relations can be extracted for different investment horizons, as matched with the detail components. The first study examines the ability of different GARCH models to forecast stock return volatility in eight international stock markets. The results demonstrate that de-noising the returns improves the accuracy of volatility forecasts regardless of the statistical test employed. After de-noising, the asymmetric GARCH approach tends to be preferred, although that result is not universal. Furthermore, wavelet de-noising is found to be more important at the key 99% Value-at-Risk level compared to the 95% level. The second study examines the impact of fourteen macroeconomic news announcements on the stock and bond return dynamic correlation in the U.S. from the day of the announcement up to sixteen days afterwards. Results conducted over the full sample offer very little evidence that macroeconomic news announcements affect the stock-bond return dynamic correlation. However, after controlling for the financial crisis of 2007-2008 several announcements become significant both on the announcement day and afterwards. Furthermore, the study observes that news released early in the day, i.e. before 12 pm, and in the first half of the month, exhibit a slower effect on the dynamic correlation than those released later in the month or later in the day. While several announcements exhibit significance in the 2008 crisis period, only CPI and Housing Starts show significant and consistent effects on the correlation outside the 2001, 2008 and 2011 crises periods. The final study investigates whether recent returns and the time-scaled return can predict the subsequent trading in ten stock markets. The study finds little evidence that recent returns do predict the subsequent trading, though this predictability is observed more over the long-run horizon. The study also finds a statistical relation between trading and return over the long-time investment horizons of [8-16] and [16-32] day periods. Yet, this relation is mostly a negative one, only being positive for developing countries. It also tends to be economically stronger during bull-periods
Essays on volatility forecasting
No full textStock market volatility has been an important subject in the finance literature for which now an enormous body of research exists. Volatility modelling and forecasting have been in the epicentre of this line of research and although more than a few models have been proposed and key parameters on improving volatility forecasts have been considered, finance research has still to reach a consensus on this topic. This thesis enters the ongoing debate by carrying out empirical investigations by comparing models from the current pool of models as well as exploring and proposing the use of further key parameters in improving the accuracy of volatility modelling and forecasting. The importance of accurately forecasting volatility is paramount for the functioning of the economy and everyone involved in finance activities. For governments, the banking system, institutional and individual investors, researchers and academics, knowledge, understanding and the ability to forecast and proxy volatility accurately is a determining factor for making sound economic decisions. Four are the main contributions of this thesis. First, the findings of a volatility forecasting model comparison reveal that the GARCH genre of models are superior compared to the more ‘simple’ models and models preferred by practitioners. Second, with the use of backward recursion forecasts we identify the appropriate in-sample length for producing accurate volatility forecasts, a parameter considered for the first time in the finance literature. Third, further model comparisons are conducted within a Value-at-Risk setting between the RiskMetrics model preferred by practitioners, and the more complex GARCH type models, arriving to the conclusion that GARCH type models are dominant. Finally, two further parameters, the Volatility Index (VIX) and Trading Volume, are considered and their contribution is assessed in the modelling and forecasting process of a selection of GARCH type models. We discover that although accuracy is improved upon, GARCH type forecasts are still superior
Essays on Financial Market Volatility
Full text linkVolatility is an important component of market risk analysis and it plays a key role in many financial activities, such as risk management, asset pricing, hedging, and diversification strategies. This thesis consists of four empirical essays that evaluate the utility of a wide range of econometric models as well as explore and propose the use of further novel methods to enhance the understanding of volatility mechanisms across emerging and developed financial markets of Asia. Specifically, the first empirical essay provides an in-depth analysis on the characteristics of volatility phenomenon by comparing various GARCH models using three different frequencies with 24 years of data. The findings reveal robust empirical evidence that asymmetric GARCH models outperform in daily and weekly return series, while symmetric GARCH models outperform in monthly return series, indicating that different frequencies have their own structure and characteristics. The second empirical chapter investigates the forecast ability of a number of representative econometric models belonging to two main model groups based on recursive and rolling window methods. The obtained results report that frequency of the data and choice of forecast method have strong effects on performance of the models. Furthermore, existence of strong volatility asymmetry has been found in the higher frequencies of data which is also systematically confirmed by the superiority of the asymmetric models in daily and weekly series. On the other hand, it is found that the monthly series of Asian stock markets are less sensitive to the leverage effects, thus the predictive capability of symmetric GARCH genre of models are more superior in lower frequencies. The third empirical chapter extended the volatility forecasting exercise by evaluating the utility of advanced Machine Learning models in comparison to traditional forecasting models. The findings indicate that the neural network prediction models exhibit improved forecasting accuracy across both statistical and economic based metrics, offering new insights for market participants, academics, and policymakers. The obtained results are further evaluated by the risk management settings of Value at Risk (VaR) and Expected Shortfall (ES). The final empirical essay introduced an Early Warning System (EWS) by integrating DCC correlations with state-of-the-art Deep Learning (DL) model. The novel results demonstrate that the bursts in volatility spillovers are successfully verified by the proposed model and EWS signals are generated with high accuracy before the 12-month period of crises, providing supplementary information that contributes to the decision-making process of practitioners, as well as offering indicative evidence that facilitate the assessment of market vulnerability to policymakers
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
