Dividend Regressions: January 2012


Variables used in the regression

  1. Dividend Yield = Dividends per share in most recent year/ Current Stock Price
  2. Dividend Payout Ratio = Dividends / Net Income
  3. Insider Holdings = Shares held by insiders/ Primary number of shares outstanding
  4. Institutional Holdings = Shares held by institutions/ Primary number of shares outstanding
  5. Standard deviation = 3 year standard deviation in stock returns (or ln(stock prices). You can get this from the complete dataset for US companies under updated data.
  6. ROE = Return on equity in most recent time period (Net Income/ Book value of Equity)
  7. Expected Growth in EPS over next 5 years = Consensus analyst estimate (or your own) of expected growth in EPS . If you don't have an analyst estimate, use your own estimate of expected growth.
  8. Market Debt to Capital = Debt/ (Debt + Market Value of Equity): If you have market value for debt, use it. If not, use book value of debt and market value of equity.

Dividend Yield Regression

div reg 1

div reg 2

div reg 3

 

Assume that you want to estimate the dividend yield for a firm with the following characteristics:

Institutional holdings = 75% of outstanding stock

Regression beta = 1.20

Expected Growth in EPS over next 5 years = 12%

ROE = 10%

Expected Dividend yield = .047 - 0.055 (.12) -0.001 (1.20) -0.023 (.75) - 0.007 (.10) = .02125

If your predicted value is less than zero, your predicted dividend yield is zero.

Dividend Payout Regression

 div reg 4

div reg 5

div reg 6

Assume that you want to estimate the dividend payout ratio for a firm with the following characteristics:

Institutional holdings = 75% of outstanding stock

Regression beta = 1.20

Expected Growth in EPS over next 5 years = 12%

ROE = 10%

Expected Dividend payout ratio= 0.45 - 0.441*.12 - 0.073 *1.20 - 0.223*.75 - 0.182*.10 = 0.12403 or 12.40%

If your predicted value is less than zero, your predicted dividend payout ratio is zero.