A Discount Rate Primer

Bloomberg has an essay about an alternative means for discounting future events.  The author provides a link to his blog, which includes a link to the paper on which the Bloomberg essay is based.  The gist of the paper, and thus the essay, is that discount rate may not be consistent throughout the discounting period, and the discount rate may change over time.  The paper then develops a more complex discounting mechanism which specifically accounts for the uncertainty of the discount rate.

The discounting function is the fundamental equation of finance.  It provides the comparison yardstick by which we can evaluate the wisdom of current expenditures for future benefits.  The concept of a discount rate underlies such ideas as a required rate of return, cost of capital, and bond duration.  The one number represents all of the factors which make a future event less valuable than an immediate one.

We can agree that the market prices different discount rates for different discreet periods in the future.  The sloping yield curve of the bond market is evidence of that.  The market also gives us very precise measures of the discount rate related specifically to the time value of money for up to thirty years: the US Treasury STRIPS market.  The market can also give us some indications and estimates of a discount rate for longer periods (corporate bonds have been issued with up to 100-year maturities) but these throw in some default risk and reinvestment risk premiums which are difficult at best to identify individually.

In the context of evaluating appropriate discount rates for very long term analyses, which seems the purpose of the exercise of considering alternative means for arriving at a discount rate, we don't need an estimate for the entire period.  The STRIPS market already tells us that the value of a good thirty year on the future is 25.255% of a good today  Now the estimate need only be made to that point thirty years from now.  And no matter what discount rate is applied from that thirty-first year forward, the current price, the present value, of the good will be less than 25.255% of the ultimate value of the good.  For example, if the second thirty years was to be discounted at 1%, one would be willing to forgo only 18.74% of the value of good today in return for the promise of the full value of the good in 2071.

So let's not get confused: Receipt of a good in the future is always less valuable than receipt of the good currently.  No matter what the discount rate, a good delivered in the distant future will always be significantly less valuable than an immediate good.  There are empirical means for determining a precise discount rate for up to thirty years, and very good estimates for many years more.  And even if we agree on the weaknesses in our estimation methodologies, more sophisticated calculations are not going to change these fundamental relationships.