Those of you who followed John Kish's "Intrinsivaluator" will recall that John had a series of IVs, which he labeled "conservative," "more conservative," "liquidation," etc., depending on the discount rate used. They all tracked BV extremely well (log-log plots, r^2 = 0.99), but they were offset from one another. Which was right? I could think of two approaches to answering that question. One was to use a discount rate equal to the total return of the stock. Only a discount rate equal to the total return can bring a future stream of cash flows back to the present value. It's like buying a 6%, 30-year bond for $100. Only a discount rate of 6% will bring the future stream of interest payments back to $100.

The other approach was to use the stock price as a crude estimate of IV, and to plot price versus BV over a very long time period. In the long term the market is a weighing machine. It turned out that the stock price starting in 1965 did give a very good fit when plotted against BV, with short term ups and down along the way. In the short term the market is a voting machine. Mungofitch strongly objects to this method, but whether the method is reasonable method or unreasonable, the price data do at least track BV extremely well (r^2 = 0.980).

Estimated IV using DCF or two-column track BV extremely well (and FWIW stock price also tracks BV extremely well). I think that BV based valuations deserve strong consideration alongside valuations based on cash flows, operating earnings or look-through earnings. In fact, estimating IV as a simple multiple of BV works quite well, as long as one has a logical and proven method for choosing the correct IV/BV multiple.