Econometrics : Capital Asset Pricing Model regression - Dissertation Example

Explanation (in detail) Excel file (CAPM 73-03). It is so d because the Capital Assets Pricing Model (CAPM) is created on 1973:01 – 2002:12 data and forecasted for 2003:01 – 2005:12. Chronologically the first worksheet was *1 Initial Data*. *1 Initial Data* worksheet In B-P columns there are companies’ share prices. Q column is for P&S 500 values. Treasury bill is in the R column. As Treasury bill is given in *%* and each value is given for a certain day (for instance, 03.01.1973), then we have to make a monthly average value and take it in fraction (just to divide into 100%). The formula is , where is at the beginning data, - at the end. The formula for data in T column is , where is at the beginning data, at the end. (It’s necessary to divide the difference into to get the fluctuation level). In U column we have . *2 Share return * worksheet Here in B-P columns you can find the data from the 1st worksheet. Q column contains the risk-free return. Then we find the difference between each company share prices and this Q data. It’ll be the left side of our main equation. *For Regression* worksheet Here we just input the data from previous sheets. *CAPM* workfile Our workfile reads data from this CAPM 73-03.xls, from the * For Regression* worksheet. - is our main formula, or The is the column B (*RiskPremium*) is for each company In Series sheets, which are made for each company, you can find the share fluctuation for each company (for instance in *americanexpress*). In equation sheets you see the equation coefficients and main statistics for each company. Graph files help us to calculate forecasts. In our workfile you can see, for instance, eq01amex – equation for American Express. Press it. There you can see 2 lines : Variable Coefficient Std. Error t-Statistic Prob. C 0.024103 0.004902 4.917279 0.0000 RISKPREMIUM 1.360567 0.061436 22.14607 0.0000 Our Beta is first value of risk premium (you may change the name, if you want.) So for American Express Standard error is small (that proves the accuracy of CAPM model). So on for each company. Forecast for 01/2003-12/2005 you can find in graph01amex worksheet. There are some statistics there – For instant *Root Mean Squared Error* is 0.0299830.03 is small enough. It means that we’ve got an adequate and accurate model. It confirms the veracity of CAPM-model with high probability. Read More