Dividend Discount Model Calculator

The Gordon Growth Model (Myron Gordon, 1956) values a dividend-paying stock as V = D1 / (k - g), where D1 is next year's expected dividend, k is the required return, and g is the constant dividend growth rate, assumed to hold forever. The 'forever' assumption is the catch. In practice the model works best for mature, slow-growing dividend payers whose payout has been stable for decades: large utilities, consumer staples like Procter & Gamble or Coca-Cola, regulated telecoms, and big-cap banks during steady periods. It is a poor fit for early-stage growth firms (their growth is far above any plausible discount rate), for firms that do not pay dividends at all (Berkshire, Alphabet, most tech), and for highly cyclical payers (energy majors, miners) where dividend growth swings every year. For a transitioning firm, switch to the two-stage variant in this calculator.

The Gordon formula is the sum of an infinite geometric series of discounted dividends. That series only converges when the growth rate is strictly less than the discount rate. If g equals k, the denominator (k - g) is zero and the intrinsic value is undefined. If g is greater than k, the denominator goes negative and you get a meaningless negative value (the math is saying the dividends grow faster than they are discounted, so the present value is infinite, but the formula reports a sign error). Economically the constraint also makes sense: no company can outgrow the rest of the economy forever, and any g above the long-run nominal GDP rate of about 4 to 5 percent in the US is unsustainable as a perpetual rate. If your inputs violate k > g, the calculator returns zero with a warning and you should switch to the two-stage model, where high growth is only assumed for a finite window.

The two-stage DDM is the right choice whenever a company is currently growing dividends faster than its long-run sustainable rate, but you expect that growth to fade as the business matures. The classic cases are REITs in the middle of an acquisition phase, utilities in regions with strong population growth, healthcare firms with a new blockbuster drug still being rolled out, and consumer brands expanding internationally. You set a high growth rate g1 for the first n years (usually 5 to 10 years, the period over which you can credibly forecast), then drop to a terminal rate g2 at or below nominal GDP growth (commonly 2 to 4 percent). The terminal value at the end of year n captures everything after the fade and is the single biggest driver of the result, typically 60 to 80 percent of the intrinsic value, so be conservative with g2.

The Capital Asset Pricing Model (CAPM) is the standard way to estimate k for a public stock. The formula is k = rf + beta x ERP, where rf is the risk-free rate (the 10-year US Treasury yield, currently around 4 to 5 percent), beta is the stock's sensitivity to the overall market (1.0 means it moves with the S&P 500, above 1 is more volatile, below 1 is less; published on Yahoo Finance and most broker sites), and ERP is the equity risk premium investors demand for holding stocks over Treasuries, historically about 5 to 6 percent. A defensive utility with beta 0.6 and a 4.5 percent risk-free rate gives k = 4.5 + 0.6 x 5.5 = 7.8 percent. A volatile tech stock with beta 1.4 gives k = 4.5 + 1.4 x 5.5 = 12.2 percent. Higher beta means higher required return, which compresses intrinsic value.

The zero-growth DDM is V = D / k, which is the present value of a level perpetuity. It is the right model for preferred stock, because preferred dividends are contractually fixed and never grow. The most common application is valuing traditional preferred shares from banks and insurance companies, which pay a constant quarterly dividend until called. For example a preferred share paying $5 per year with a required return of 7 percent has an intrinsic value of $5 / 0.07 = $71.43. The same formula also fits truly mature common-stock payers whose dividend has not grown in years (some regulated utilities, slow-growing tobacco names), though in practice you should still run the Gordon model with a small positive g to capture inflation-matching increases.