I look at a variety of stock market valuation methodologies. Below are several, but I look at many others as well to get a sense of valuation. Of course, valuation means absolutely nothing in the near-term. It is, however, the single biggest determinant of long-term returns in the stock market.
Assuming, of course, we do not go the way of Argentina.
One model I use is a forward price/earnings model. I calculate normalized earnings by multiplying a normalized profit margin by sales per share, assume 6% profit growth for the next 20 years – which is the long-term growth rate for earnings – apply a multiple of 15x – which is the long-term average PE of the market – then discount this 20-year expected S&P 500 target back to the present to calculate the expected capital gain. I then add the current dividend yield to get my total expected return. Currently, this model is forecasting a total return to equities over the next 20 years of 6.7%, well below the long-term average of 10%.
Another model is a dividend model. I forecast normalized earnings one year out and multiply it by an expected payout ratio – which is higher than the actual current payout ratio. I then add the long-term growth rate in dividends. This model is forecasting expected equity returns of 6.3%. (For the record, the current dividend yield of the S&P 500 is 1.8%. The dividend yield for the S&P 500 never went below 2% before 1997, around the time when markets started going bonkers.)
Modeling has its problems, of course. Just like all valuation methodologies, I use models for approximation, not precision. I’m not trying to hit a hole-in-one. I’m just trying to hit the green. (He can’t keep the ball on the fairway. – ed.) And the approximation is that stocks are expensive.
I also use a normalized PE. Normalized earnings is calculated in the same way as above. Sales per share of the S&P 500 is currently $905. Assuming a normal profit margin of 7% - the average of the past 20 years has been 5.5% - normalized earnings per share is currently $63. Thus, the market multiple on normalized earnings is 18.5x, well above the long-term average of 15x.
Trailing earnings are currently $53, forward “operating” earnings are $77. Wall Street consistently over-estimates operating earnings relative to reported earnings by 10% to 20%. If you adjust for Wall Street’s excessive optimism, forward earnings are $62 to $70. Using these two numbers, the market is trading at 16.7x to 18.8x forward earnings. While the trailing historical PE multiple has averaged 15x, the forward PE has averaged 14x. That means the market is 19% to 34% over-valued.
If one looks at analysts’ estimates for individual companies, the median forward PE of the Russell 3000 is 15.9x. Since this is operating earnings, we have to adjust by 10% to 20% to obtain reported earnings. Thus, the median PE for companies in the Russell 3000 is 17.7x to 19.9x. Given that the historical forward PE has been 14x, the market is 26% to 42% over-valued.
In the asset-driven economy, where the Fed has unleashed oceans of liquidity, none of this matters. The market is riding a rising wave. But when the wave starts to break – when the liquidity starts being withdrawn – it will matter greatly.