Attempting to determine the intrinsic value of the S&P 500 via discounted cash flow is one of my favorite and least favorite exercises at the same time. I like to keep tabs on the overall macroeconomic environment, and determining the intrinsic value of the primary index on which we base a “market portfolio” is one such method. But I dislike it in the sense that it’s very difficult to value precisely and the final reading can fluctuate significantly based on very minor adjustments to the inputs. It leads to an internal debate – what data should you consider for the proper rate for earnings growth (Past five years? Ten years? Since 1970?); what is the expected cash payout for companies; what about a long-term growth rate? So it’s like anything else, make realistic assumptions, practice conservatism when in doubt, and play with the inputs to determine what the market might be implying about certain variables. The exercise is generally quite simple. I only need five main inputs. Expected earningsExpected payout from these earningsExpected growth rate in earningsExpected long-term growth rate in earningsEquity Risk Premium For expected earnings, I assume yield stands at around 5.20% based on recent data and use 2100 as the starting point for the index. In other words, first year expected earnings are 109.20 (2100 * 5.20%). Expected payout from these earnings has varied wildly historically and can make a big difference on the eventual result depending on what value is chosen. In these past couple years, the payout ratio has actually been above 100% -- i.e., companies have paid out more than 100% of their earnings back to investors through dividends and buybacks. Average payout over the past five years or so has been around 77%. Historically it’s been close to 74%. I assume 75%, favoring reversion to long-term historical trends over something that isn’t sustainable. Over the last decade, the growth rate in earnings has been approximately 4.3%. This is a bit low relative to long-term historical odds due to the effects of the financial crisis. If we go by current return on equity (ROE), we would be in the upper-4% range. I prefer conservative assumptions so I will go with 4.3%. The long-term growth rate in earnings is the hardest to predict. Assumption deviations of 50 bps can throw the valuation off by 15% or more. For this reason, I run this variable as a sensitivity consideration in intervals of 50 bps, running from 1.50% (too low) through 4.50% (probably too high), while holding everything else constant. For the equity risk premium, I am going with 5.3% to mirror what’s been implied over the past ten years of market data. My base-case assumption is that the long-term earnings growth rate of the U.S. economy (real growth + inflation) is 3.5%. Assuming the expected return of the market is 7.8% (30-year treasury yield + equity risk premium), this becomes our discount rate for free cash flow (to equity) over the projection period. The discount rate for our terminal value becomes that 7.8% figure minus the 3.5% growth rate. Adding these two together, I come up with an intrinsic estimate of 2,007.09 for the S&P 500. By fiddling with the growth rate in 50-basis point increments, we get the following type of curve shown below. It is naturally a bit parabolic due to the compounded effects of having a higher long-term earnings growth rate.