In this section you find some Eviews exercises additional to the ones presented in the book. These exercises have been developed for the course of Financial Econometrics and Empirical Finance that prof. Guidolin teaches at Bocconi (you may also have a look here: Financial Econometrics ). However, we believe that these examples may be useful for all the Readers of the book (and not only), therefore we decided to make them avaliable through this page.
The magic world of cryptocurrencies. The first exercise that we propose is mostly related to the topics covered in Chapter 2 of the book. We analyze data concerning the Ethereum, one of the most famous cryptocurrencies (after the very popular and first-comer Bitcoin). Click on this link to download a workfile in which many tasks concerning preliminary data inspection and estimation of MA models have been performed. You also find a workfile that contains similar data for the Bitcoin itself: we leave to you as an exercise to replicate the analysis on this additional set of data. Here you find a presentation that discusses results on both Ethereum and Bitcoin.
Forecasting stock returns. The second exercise that we propose represents a transition from the topics covered in Chapter 2 of the book to VAR analysis (Chapter 3). Three series will be at work here: S&P Composite Index (that became S&P 500 in 1957), the rate of growth of the real (i.e., deflated using CPI inflation) dividends paid by the stocks in the index, and the cyclically adjusted price-earnings ratio over 10 years (CAPE-10). The series are collected for the (rather long!) period January 1881 – December 2018. The data are kindly offered by Prof. Robert Shiller. Click on this link to download a workfile where we go from an univariate analysis of each series to the estimation of a VAR model (that we then use for forecasting). In the folder you also find a workfile that contains the growth rate of real earnings: we leave it to you as a task to add it to the picture. Here you find a presentation that discusses results on both the dividend/CAPE and the real earnings exercises.
The effects of monetary policy. The third exercise that we propose focuses on VAR analysis, a topic that is covered in Chapter 3 of the book. The objective of this exercise is to understand the effects of conventional and uncoventional (i.e., quantitative easing) monetary policy in the US on a a set of time series that matter to fixed-income portfolio managers, namely 1- and 10-year Treasury bonds, Aaa long-term corporate bonds (10-year maturity) and Baa long-term corporate bonds (10-year maturity). The series are collected for a period spanning from January 1981 to March 2019. Click here to download a workfile where we study the effects of a shock to the Fed funds rate (that mimics a conventional monetary policy) and a shock to the 10-year Treasury yield (that represents a proxy for QE) using the variance decomposition and the impulse response function analysis. In the folder you also find a workfile that contains the series of the 30-year fixed mortagage rates: we leave to you as an exercise to study the effects of monetary policy on this additional series. Here you find a presentation that discusses results on both the corporate bond and the mortgage rate exercises. These exercises are inspired from our own research agenda; the interested Reader may take a look at: “Unconventional monetary policies and the corporate bond market” (Finance Research Letters, 2014) and “The impact of monetary policy on corporate bonds under regime shifts” (Journal of Banking and Finance, 2017).
The CDS-bond basis puzzle. The fourth exercise of this series focuses on unit root tests and cointegration (topics that are covered in Chapter 4 of the book). CDS premia (or spreads) and bond spreads (computed as the difference between bond yields and a risk-free rate) should be tied by a no-arbitrage relationship: they should indeed be equal. The goal of this exercise is to assess whether the 5-year spreads of a CDS written on the Italian Treasury bonds and the respective 5-year Treasury spreads (over the OIS swap rate, which is a proxy of the risk-free rate) do respect the theoretical no-arbitrage long run relationship. Click here to download a workfile where you will find the CDS and bond spread data (for the period January 2006 – September 2015, at weekly frequency). In this workfile you will learn how to test for the presence of unit root in the series, how to test for the presence of cointegration, how to specify a vector error correction model (VECM) and how to test restrictions on such a model. You will also find data of 5-year Spanish CDS and bond spreads: we leave to you to replicate the same exercise on these data. Here you find a presentation that discusses the results on the Italian and the Spanish data.
Volatility Modelling: A GARCH Analysis of the Euro/USD Exchange Rate. The fifth exercise brings us from mean to volatility modelling and, in particular, it focuses on GARCH models (a topic covered in Chapter 5 of the book). The example focuses on the daily time series of the EUR/USD exchange rate over the period January 4, 1999 – March 15, 2019. Once we have specified the appropriate model for the mean (that turns out to be an ARMAX(0,0) where the exogenous variable is the VXO), we turn to modelling the (rather rich) structure in the residuals. We use standard information criteria to select a GARCHX(1,2) with variance targeting and we estimate it jointly with the process for the mean. Finally, we test the residuals of our model. Click here to download the workfile for this exercise. We also leave you the data to replicate the same steps of the S&P500 time series. Here you find a presentation that discuss both the EUR/USD and the S&P500 cases.
Essentials of Time Series for Financial Applications 