So you disagree with Buffett that in the long-term the stock market is a weighing machine?
slacker, it is beneath you to put words in my mouth. leave that to others who have no scruples.
actually, i believe it was Graham who originated the voting machine/weighing machine wag, with which Buffett (and i) agree.
and we are probably safe in saying that what is weighed in the end is intrinsic value. the problem is, what is intrinsic value?
there are numerous ways to try to assess this. discounting cash flow is one of the methods. you know i have some familiarity with this approach based on my assessment of QCOM's present value at various discount rates. and you also know that, based on such an approach, my assessment of QCOM's value is rather below its current market price (although QCOM is now much closer to the assessed value than when i provided those figures).
the problem with discounting cash flow (as i see it) is that one must make various assumptions (e.g., what discount rate and what earnings growth rate to use). the reality is that we have no way of knowing ex ante how accurate these estimates will be, even though the estimates have an enormous impact on the assessed present value.
the difference between the starting assumptions used is the reason why somebody like myself and Piecyk could employ similar methodologies and yet come up with radically different assessments of NPV.
(of course, i have the benefit of understanding that the discount rate is synonymous with expected return [something many tech investors seem not to understand, thus leading them to employ extremely below-market discount rates while expecting above-market returns, which they unfortunately do not realize is a contradiction in terms], but that accounts for only part of the difference.)
the thing is, if we want to ignore options as a present expense, then we must make an assumption about their future dilution to cash flow. this is something almost no analysts do.
if we look at the reality of what forward real GDP growth is likely to be (or better yet, real per-capita GDP growth [which is lower] as this is more closely related to SPX earnings growth), and extrapolate SPX earnings growth accordingly, we will come up with a real growth figure of 1-2% if we ignore options. if we look at options dilution, we may find that options are diluting current shareholders at 2-3%. when you add the headwind (expected real growth ex-options) to the tailwind (expected options dilution) what do you get?
mind you, this approach ignores any noncash options expensing and simply treats options as a dilutive parameter. but the effects are significant across the broad market.
of course, it may be the case that some small minority of the SPX will grow real earnings at some vast multiple of its options dilution rate over the next two decades. but who has a crystal ball to know in advance which cos these are. stated differently: how lucky do you feel?
Do you also disagree that the long-term value of a company is determined by it's discounted cash flow?
i think the above has presented my position, but for review purposes: if you could accurately predict the cash flow of a co (or the market) over several decades, i think you could accurately predict the value (if not the market price) over that time and discount it back to the present. but this is something nobody can accurately do.
what we are left with is the Gordon Equation, which starts with something real (the current dividend yield) and increases that return at the rate of real historical earnings growth (which is of course much lower than most bulls realize).
it is my guess that this approach will be more accurate over time than trying to DCF the market or individual companies.
alternatively, one could simply look at the earnings yield and call that the real return. this is where stock options expensing becomes an issue, since market earnings may be overstated due to the options expense (in other words, the real growth rate will be lower than the historical one).
separately, many "strategists" try to claim the market is cheap based on a comparison of this (perhaps overstated due to options-non-expensing) earnings yield to the yield on the 10yr bond. but these people do not understand that there is no hard and fast relation betweeen the equity earnings "yield" and the bond yield.
no method is perfect. |
