Financial Modeling Data Collection & Integration Graph Rates of Return Efficient Portfolio Frontier Regression CAPM Vitex.xls
Market Model: Regression
Prepare Spreadsheet New worksheet Label it R(BRK-b) Data Returns + Candlestick R(BRK-b)
Prepare Spreadsheet Relay data from Returns tab
Reference the cell by either H3 or RR Names Name cells Cell’s name is H3 Rename the cell RR Reference the cell by either H3 or RR
Names Select cell H4 In the Name Box change H4 to RR Name Box OP H4 fx
Names This replaces the name H3 with RR. From now on, anyplace you would use H3 you can use RR instead. rr RR RR fx
Name Manager To see, edit, and manage names used in the spreadsheet use Name Manager Ctrl+F3
The Market Model Estimates the degree to which returns on the stock depend on returns to the market. observed observed Estimated using regression analysis
Dynamic Regression Alpha = Intercept(D:D, C:C) (Company returns, Index Returns) Beta = Slope(D:D, C:C) (Company returns, Index Returns)
Dynamic Regression Statistics Multiple R = Correl (D:D,C:C) Standard Error = Steyx (D:D,C:C)
Dynamic Regression Statistics R Square = (Multiple R)2 Observations = Count(D:D) Estimated Parameters: 2
Dynamic Regression Statistics Adjusted R Square H4 =RR-((k-1)/(N-k))*(1-RR) is much easier to debug than H4 = $H$4-(($H$8-1)/($H$7-$H$7)*(1-$H$4) … from my Econometrics Text Book
Dynamic Regression Statistics t-Statistic H13 = rho * SQRT((N-k)/(1-rho^2)) … from my Econometrics Text Book
Dynamic Regression Statistics © Oltheten & Waspi 2012
Characteristic Line © Oltheten & Waspi 2012
Characteristic Line Each marker is one monthly observation Line constructed from calculated alpha and beta Y axis measures returns on the Equity of our company X axis measures returns on the Market © Oltheten & Waspi 2012
Observations Generate observed values From index and company returns generate an XY scatter Format to fit the professional style of the rest of the model © Oltheten & Waspi 2012
Observations Format series to fit Dynamic title Cleanup axes
Observations Rule #13: Always verify. Verify that the index is on the x axis and the company is on the y axis. September 2008: -9.078%, 12.635% October 2008: -16.942%, -12.628% © Oltheten & Waspi 2012
y=returns on the company Characteristic Line Use dynamic α and β to generate characteristic line (no trend lines!) y=returns on the company x = returns on the index
y=returns on the company Characteristic Line beta = slope alpha=intercept x = returns on the index y=returns on the company © Oltheten & Waspi 2012
Characteristic Line Define the characteristic Line from the minimum to the maximum index value. = min(C:C) = average(C:C) = max(C:C) © Oltheten & Waspi 2012
Characteristic Line Calculate the predicted company return. at S&P= -16.942% BRK-b = 1.0475% + 0.571652(-16.942%) = - 8.6376843893881900% = - 8.638% Note that excel calculates everything to 16 decimal places unless the number is specifically rounded = alpha + beta * $G16 © Oltheten & Waspi 2012
Characteristic Line Add the Characteristic Line to the graph BRK-b - 5% - 8% 1% 6% 3% 2%
Characteristic Line x values y values = 'R(BRK-b)'!$F$14 = 'R(BRK-b)'!$G$16:$G$18 = 'R(BRK-b)'!$H$16:$H$18 x values y values
Characteristic Line 9.393%, 6.417% -16.942%, -8.638% © Oltheten & Waspi 2012
Format Data Series 'Characteristic Line' as a line with no markers © Oltheten & Waspi 2012
Slope looks like a positive 0.57 Reality Check Maximum S&P: 9.393% Intercept is 1.0475% Slope looks like a positive 0.57 Minimum S&P: -8.638% © Oltheten & Waspi 2012
Analysis Toolpak NOT Analysis Toolpak - VBA Test Rule #13: Always test your model Run the static regression using [Data] [Data Analysis] [Regression] If you don't have Data Analysis then use Developer/ Add Ins to add Analysis Toolpak Analysis Toolpak NOT Analysis Toolpak - VBA
Test Verify that the results match exactly Rule #13: Always test your model Test Verify that the results match exactly
Test Rule #13: Always test your model
Just delete the entire tab Test Rule #13: Always test your model Remove the static regression this is a test procedure, not part of the deliverable. Just delete the entire tab
Regressions
Regressions Copy the entire worksheet R(BRK-b) to R(LLY) Change the links so that it pulls LLY data When the data changes to LLY returns the regression statistics automatically recalculate
Regressions Copy (Ctrl+C) and paste (Ctrl+V) the graph. The graph will still be linked to BRK-b. Select the data series (click on one of the markers) In the formula bar you will see Change the BRK-b to LLY Do the same for the characteristic line =SERIES('R(BRK-b)'!$D$2,'R(BRK-b)'!$C$3:$C$38,'R(BRK-b)'!$D$3:$D$38,1) =SERIES('R(LLY)'!$D$2,'R(LLY)'!$C$3:$C$38,'R(LLY)'!$D$3:$D$38,1)
Regressions When you click on the data series If you don't see the highlighted data then the graph is still linked to another tab Excel will highlight the data used in the graph.
Regressions When you click on the characteristic line If you don't see the highlighted data then the graph is still linked to another tab Excel will highlight the data used in the graph.
Regressions 50/50 Portfolio
Regressions Minimum Variance Portfolio
Regressions Note what happens to the named variables These names apply everywhere else These names apply only to tab R(LLY) These names apply only to tab R(MVP) These names apply only to tab R(Portfolio)
Market Model