FMSN60/MASM18 Financial Statistics Lecture 1, Introduction and stylized facts Magnus Wiktorsson
People and homepage 222 86 25, MH:130 (Lecturer) samuel.wiqvist@matstat.lu.se , 222 79 83, MH:326 (Computer exercises) 222 04 85 , MH:223 (Computer exercises) susann.nordqvist@matstat.lu.se , 222 85 50, MH:221 (Course secretary) //www.maths.lth.se/matstat/kurser/fmsn60masm18/ ▶ Magnus Wiktorsson: magnusw@maths.lth.se , ▶ Samuel Wiqvist: ▶ Carl Åkerlindh: carl.akerlindh@matstat.lu.se , ▶ Susann Nordqvist: ▶ http:
Purpose: data, and use these tools in combination with economic theory. The main applications are valuation and risk management. statistical tools supporting courses like ’EXTQ35 Financial Valuation and Risk Management’ or ’FMSN25/MASM24 Valuation of Derivative Assets’. ▶ The course should provide tools for analyzing ▶ The course is intended to provide necessary
Inference problems? such as stochastic volatility or credit default intensity) the model? How do we estimate parameters in general models? Cross covariance and auto covariance. Often results in Non-linear, Non-Gaussian, Non-stationary models... ▶ Forecast prices, interest rates, volatilities (under the P and Q measures) ▶ Filtering of data (e.g. estimating hidden states ▶ Distribution of prediction errors; can we improve ▶ What about extreme events?
Inference problems? such as stochastic volatility or credit default intensity) the model? models? Non-stationary models... ▶ Forecast prices, interest rates, volatilities (under the P and Q measures) ▶ Filtering of data (e.g. estimating hidden states ▶ Distribution of prediction errors; can we improve ▶ What about extreme events? ▶ How do we estimate parameters in general ▶ Cross covariance and auto covariance. ▶ Often results in Non-linear, Non-Gaussian,
Example I – Daily interest data - big crisis in Sweden during the early 1990s happen again? Covariation with of market factors? - Can this Forecasts - 0.5 % or 500 %? Models and distributions. See Section 2.4 in the book for more information. STIBOR and REPO Yields 1992 REPO STIBOR 1W STIBOR 1M 500 STIBOR 3M STIBOR 6M 100 10 Q1−92 Q2−92 Q3−92 Q4−92 Q1−93
Example I – Daily interest data - big crisis in Sweden during the early 1990s See Section 2.4 in the book for more information. happen again? STIBOR and REPO Yields 1992 REPO STIBOR 1W STIBOR 1M 500 STIBOR 3M STIBOR 6M 100 10 Q1−92 Q2−92 Q3−92 Q4−92 Q1−93 ▶ Forecasts - 0.5 % or 500 %? ▶ Covariation with of market factors? - Can this ▶ Models and distributions.
Electricity spot price and Hydrological situation
Example II – Forward prices on Nordpool movements on the futures on Nordpool on yearly contracts. (days or weeks). prizes. 1. Hydrological situation is the energy stored as snow, ground water or in reservoirs 2. Time to maturity. 3. Perfect or imperfect markets. ▶ Traders are interested in predicting price ▶ Or predicting the movements on short horizons ▶ Expected movement and/or prob. of declining ▶ What about fundamental factors? ▶ Other factors – suggestions?
Ex – Forwards on Nordpool, contd. There is a strong dependence between the hydrological situation and the price. model should we use? Is the relation linear? ▶ How do we model this dependence, e.g. what ▶ How do we fit the chosen model? ▶ How do we know if the model is good enough? ▶ One supermodel or several models? ▶ Adaptive models?
Contents The course treats estimation, identification and validation in non-linear dynamical stochastic models for financial applications based on data and prior knowledge. There are rarely any absolutely correct answers in this course, but there are often answers that are absolutely wrong. This was expressed by George Box as All models are wrong - but some are useful! Think for yourself, and question the course material!
Contents The course treats estimation, identification and validation in non-linear dynamical stochastic models for financial applications based on data and prior knowledge. There are rarely any absolutely correct answers in this course, but there are often answers that are absolutely wrong. This was expressed by George Box as All models are wrong - but some are useful! Think for yourself, and question the course material!
Contents, 2 Discrete and continuous time. identification and model validation. other approaches. Kalman filters (and versions thereof) and particle filters ▶ Parameter estimation (LS, ML, GMM, EF), model ▶ Modelling of variance, ARCH, GARCH, ..., and ▶ Stochastic calculus and SDEs. ▶ State space models and filters
Course goals -Knowledge and Understanding For a passing grade the student must: family, stochastic volatility, and models use for high-frequency data, formula, Girsanov transformation, martingales, Markov processes, filtering, Kalman filters and particle filters, above model families. ▶ handle variance models such as the GARCH ▶ use basic tool from stochastic calculus: Ito’s ▶ use tools for filtering of latent processes, such as ▶ statistically validate models from some of the
Course goals -Skills and Abilities For a passing grade the student must: related fields. model choice, includes model specification, inference, and and from other courses) where the solution economic and statistical theory (from this course report, as well as orally, should be applied, financial contracts and for transforming models, financial data, ▶ be able to find suitable stochastic models for ▶ work with stochastic calculus for pricing of ▶ understand when and how filtering methods ▶ validate a chosen model, ▶ solve all parts of a modelling problem using ▶ utilise scientific articles within the field and ▶ present the solution in a written technical
Literature Nielsen, J. N. (2015) Statistics for Finance , Chapman & Hall, CRC press. the course home page) Course program. ▶ Lindström, E., Madsen, H., ▶ Handouts (typically articles on
long range dependence? Properties of financial data Evaluate claims on S&P 500 data. ▶ No Autocorrelation in returns ▶ Unconditional heavy tails ▶ Gain/Loss asym. ▶ Aggregational Gaussianity ▶ Volatility clustering ▶ Conditional heavy tails ▶ Significant autocorrelation for abs. returns - ▶ Leverage effects ▶ Volume/Volatility correlation ▶ Asym. in time scales
Autocorrelation in returns No or little autocorrelation. 10 -5 Covariance log returns 10 8 6 4 2 0 -2 0 10 20 30 40 50 60 70 80 90 100 Lag
Unconditional distribution Normplot of the unconditional returns. Normplot unconditional log returns 0.999 0.997 0.99 0.98 0.95 0.90 0.75 Probability 0.50 0.25 0.10 0.05 0.02 0.01 0.003 0.001 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 Data
Gain/Loss asym. Losses are larger than gains (data is Index). This may contradict the EMH, see Nystrup et al. (2016). S & P 500 3000 2500 2000 1500 1000 500 0 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Aggr. Gaussianity Returns are increasingly Gaussian. Interpretation? Normplot log returns Daily Normplot log returns Monthly 0.999 0.997 0.999 0.99 0.997 0.98 0.99 0.98 0.95 0.95 0.90 0.90 0.75 0.75 Probability Probability 0.50 0.50 0.25 0.25 0.10 0.10 0.05 0.02 0.05 0.01 0.02 0.003 0.01 0.001 0.003 0.001 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 -0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 Data Data Normplot log returns Quarterly Normplot log returns Yearly 0.999 0.997 0.99 0.98 0.99 0.98 0.95 0.95 0.90 0.90 0.75 0.75 Probability Probability 0.50 0.50 0.25 0.25 0.10 0.10 0.05 0.05 0.02 0.01 0.02 0.01 0.003 0.001 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 Data Data
Vol. Clustering Volatility clusters. Average cluster size? S & P 500 log returns 0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 -0.25 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Dependence in absolute returns Significant autocorrelation. Long range dependence or other reason? Hint: Nystrup et al., (2015, 2016) 10 -5 Covariance absolute log returns 6 5 4 3 2 1 0 -1 0 10 20 30 40 50 60 70 80 90 100 Lag
Conditional distribution Normplot of the conditional returns (GARCH(1,1) filter). Conditional log returns Normplot conditional log returns 8 6 0.999 0.997 4 0.99 0.98 2 0.95 0.90 0 0.75 Probability -2 0.50 0.25 -4 0.10 0.05 -6 0.02 0.01 -8 0.003 0.001 -10 -12 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 -10 -8 -6 -4 -2 0 2 4 6 Data
No correlation in conditional absolute returns Covariance cond absolute log returns 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 -0.05 0 10 20 30 40 50 60 70 80 90 100 Lag
Leverage effects measure of volatility. ▶ Most assets are negatively correlated with any ▶ One popular explanation is corporate debt. ▶ Makes sense if you are risk averse.
Recommend
More recommend
