Home IGIER  

The Econometrics of Asset Allocation and Risk Measurement

F.Corielli, C.Favero,M.Giacoletti, M.Guidolin

 This project is focussed on the main econometric ingredients for portfolio allocation and risk measurement: forecasting the distribution of returns of financial assets. The project is empirical and for each topic  MATLAB programmes are constructed to practically implement the tools introduced (we shall usethe Econometric Toolbox and codes  on the Lesage website and download and install the spatial econometrics toolbox).

 


Chapter 1. The Econometrics of Financial Returns: an introduction

1 Introduction
2 An Historical Perspective
2.1 The view from the 1960: Efficient Markets and CER
2.2 Empirical Challenges to the CER model: the DDG model
2.2.1 Predictability at different horizons
2.2.2 The Empirical Evidence .
2.3 A SimpleModel of Noise vs. Fundamentals
2.3.1 An interesting extension
3 Asset Pricing with Predictable Returns
4 Quantitative Risk Management and the behaviour of returns at high frequency
5 Afirst look at the data: Noise vs. Information in Asset Returns
6 The Plan of the book. Predictive Models in Finance
7 A first look at the data in MATLAB

DRAFT CHAPTER SLIDES,  MATLAB

Chapter 2. Returns

1 Returns
1.1 Simple and log Returns
1.2 Multi-period returns and annualized returns
1.3 Working with Returns
2 Stock and Bond Returns
2.1 Stock Returns and the dynamic dividend growthmodel
2.2 Bond Returns: Yields-to-Maturity, Duration and Holding Period Returns
2.2.1 Zero-Coupon Bonds
2.2.2 Coupon Bonds
2.3 A simple model of the term structure
3 Modeling Returns
3.1 Assessing Models by Simulation: Monte-Carlo and Bootstrap Methods
3.2 Stocks for the long run
4 Going to the data with MATLAB: the US Stock Market


Chapter 3  Linear Models of Financial Returns 

1 Econometric Modelling of Financial Returns: a general framework

1.1 The reduction process

1.2 Exogeneity and Identification

2 From theory to data: the CAPM

3 Graphical and Descriptive Data Analysis

4 Estimation Problem: Ordinary Least Squares

4.1 Properties of the OLS estimates

4.2 Residual Analysis

5 Interpreting Regression Results

5.1 The R2 as a measure of relevance of a regression

5.2 Inference in the Linear Regression Model

5.2.1 Elements of distribution theory

5.2.2 The conditional distribution  | X

5.2.3 Hypothesis Testing

5.2.4 The Partitioned Regression Model

5.2.5 The partial regression theorem

6 The effects of mis-specification

6.1 Misspecification in the Choice of Variables

6.1.1 Under-parameterization

6.1.2 Over-parameterization

6.2 Estimation under linear constraints

6.3 Heteroscedasticity, Autocorrelation, and the GLS estimator

6.3.1 Correction for Serial Correlation (Cochrane-Orcutt)

6.3.2 Correction for Heteroscedasticity (White)

6.3.3 Correction for heteroscedasticity and serial correlation (Newey-West)

7 Econometrics in action: From the CAPM to Fama and French Factors

7.1 Fama-French Factors and the Fama-MacBeth procedure

8 References

                         

DRAFT CHAPTER,SLIDES1SLIDES2SLIDES3  EVIEWSMATLAB


Chapter  4  The Constant Expected Return Model

1 The Constant Expected Returns Model
1.1 Regression Model Representation
2 A Static Asset Allocation Problem with Constant Expected Returns
3 What Happens in Practice ?
3.1 The resampled optimal mean-variance portfolio
3.2 Black and Litterman’s approach
4 Going to the Data: Asset Allocation and the CER model with MATLAB
4.1 Exploratory Data Analysis
4.2 Optimal Static Asset Allocation
4.3 Testing the model
4.4 The resampled optimal portfolio and efficient frontier
4.5 Black-Litterman

DRAFT CHAPTER PART1 PART2SLIDESMATLAB


Chapter 5  Univariate Time-Series

1 Introduction
2 Time-Series
3 Analysing Time-Series: Fundamentals
3.1 Conditional and Unconditional Densities
3.2 Stationarity
3.3 ARMA Processes
4 Persistence: Monte-Carlo Experiment
5 AsymptoticTheory
5.1 Basic elements of asymptotic theory
5.1.1 Convergence in distribution
5.1.2 Convergence in probability
5.1.3 Central limit theorem (Lindeberg−Levy)
5.1.4 Slutsky’s Theorem
5.1.5 Cramer’s Theorem
5.1.6 Mann−Wald Theorem
5.2 Application to models for stationary time-series
6 Estimation of ARMA models. The Maximum Likelihood Method
6.1 MLE of an MA process
6.2 MLE of an AR process
7 Putting ARMA models at work
7.1 An Illustration
8 Trends
8.1 Univariate decompositions of time-series
8.1.1 Beveridge−Nelson decomposition of an IMA(1,1) process
8.1.2 Beveridge−Nelson decomposition of an ARIMA(1,1) process
8.1.3 Deriving the Beveridge−Nelson decomposition in practice
8.1.4 Assessing the Beveridge−Nelson decomposition
9 Asset Allocation with a simple TVER model: the SOP method
10 Univariate time-series analysis and portfolio allocation with TVER in

DRAFT CHAPTERSLIDES,MATLABDATASET COCHRANE(94)

 


Chapter 6  Multivariate Time Series Analysis

 1 Stochastic trends and spurious regressions

2 Dynamic Models and Spurious Regressions

2.1 Non Stationary Time-Series, Cointegration and Error Correction Models
2.2 Static Regressions and Dynamic Models .
3 Spurious Regressions and the predictability of returns at  different frequencies 
4 Cointegration with Multiple Cointegrating Vectors 
4.1 The Johansen procedure
4.2 Identification of multiple cointegrating vectors
4.3 Hypothesis testing with multiple cointegrating vectors
5 Using VAR Models 
6 Identification of VAR 
7 Description of VAR models
8 From VAR innovations to structural shocks
8.1 Choleski decomposition
8.2 CVAR and Identification of shocks
8.3 Sign Restrictions
8.4 GIRF 
9 Cointegration and Present Value Models
9.1 Cointegration and multivariate trend-shocks decompositions 
9.2 Forecasting froma Cointegrating VAR
9.3 VECM and common trends representations
10 An Application: Cointegration, Noise and Information
in Stock Market Returns
11 Risk, Returns and Portfolio Allocation with Cointegrated VARs 
11.1 Inspecting the mechanism: a bivariate case 
11.2 A VAR with many assets and predictors
11.3 Mean-Variance Analysis with a VAR model
12 Multivariate Time-Series with Matlab
 
 

Chapter 7 The Term Structure of Risk:  Bayesian Methods


Chapter 8 Risk Measurement with High Frequency Data

1. Introduction

2. The Evidence form High Frequency Data

2.1 Heteroscedsticity

2.2 Testing and Measuring deviations from Normality

3. GARCH variance models

3.1 a formal GARCH test

3.2 Forecasting with GARCH models

4. Maximum Likelihood Estimation of GARCH models

5. GARCH specification estimation and forecasting in MATLAB

6. From GARCH to VaR

7. Backtesting VaR

 

LECTURE NOTESDRAFT CHAPTER,MATLAB


Chapter 9 Modelling Heteroscedasticity and non-normality

Contents

1 Introduction

2 Computing Measures of Risk without simulation

3 Simple Models for Volatility

3.1 Rolling window variance model 

3.2 Exponential variance smoothing: the RiskMetrics model . .

3.3 Are GARCH(1,1) and RiskMetrics different?

4 BeyondGARCH

4.1 Asymmetric GARCH Models (with Leverage) and Predetermined Variance Factors

4.2 Exponential GARCH .

4.3 Threshold (GJR) GARCHmodel .

4.4 NAGARCHmodel 

4.5 GARCH with exogenous (predetermined) factors

4.5.1 One example with VIX predicting variances

4.6 Component GARCH Models: Short- vs. Long Run Variance Dynamics

5 Modelling Non-Normality

5.1 t-Student Distributions for Asset Returns .

5.2 Estimation: method ofmoments vs. (Q)MLE

5.3 ML vs. QML estimation of models with Student t innovations

5.4 A simple numerical example.

5.5 A generalized, asymmetric version of the Student t 

5.6 Cornish-Fisher Approximations to Non-Normal Distributions

5.7 A numerical example 

6 Direct Estimation of Tail Risk: A Quick Introduction to Extreme Value Theory

LECTURE NOTESDRAFT CHAPTERMATLAB 


Chapter 10: Correlation Modeling

LECTURE NOTES


Chapter 11: Markov and Regime Switching Models 

LECTURE NOTES


Chapter 12: Multivariate GARCH Modeling

LECTURE NOTES

 

 



Last updated October 29, 2014