Introduction Fin250: Lecture 2 Fall 2010 Reading: Brooks, chapter 1
Outline Software tools Forecasting basics (key things) In sample bias Forecast objectives Changing relations Financial forecasting basics Definitions Cautions
Software Statistical packages Stata, Eviews, Rats SAS, TSP, SPSS Computer languages C, C++, Java, V-basic, FORTRAN, GO Matlab, Gauss, S-plus, Octave, R Other Excel Technical trading tools
Excel Advantages and Disadvantages Advantages Every knows/has it Easy to use Disadvantages Not powerful Hard to do sophisticated problems Few advanced tools
Matlab: Advantages and disadvantages Advantages Powerful Relatively easy to use Great graphics Nice tools Disadvantages Programming Expensive Not everyone uses
Key Things About Forecasting In sample bias Forecast objectives Changing relations
In Sample Bias Use data for two purposes Estimation Testing Using the same series for both often makes testing look better “Forecasting what you already know” One answer “out of sample experiments”
Forecast Objectives Many forecasting objectives Statistical Mean squared error R-squared Economic Trading rule profitability Risk measures
Changing Relations Data features change over time Model updating Useful lifespan
Forecasting Definitions Time series Predict prices and volatility over time Cross section Comovements Asset prices tend to move together CAPM/beta/correlations Pairs trading Sometimes mixed
Forecasting Cautions Forecasting (financial and otherwise) is messy Predicting the stock market is difficult