More on the valuation of the US stock market

More on the valuation of the US stock market
I have gotten very nice feedback on my post on why there is no bubble in the US stock market and a couple of great questions. Here is one of the questions from a reader who will remain anonymous:
...I am a fixed income guy without a bone in this fight, but i do have the following question:  How would you reconcile the model to some valuation metrics such as CAPE and even the simple trailing PE that suggest that US stocks, if not in a bubble or even expensive, are certainly not cheap.  How do you reconcile the purported cheapness of stocks with a declining rates of EPS growth and with PE expansion being the main driver of stock prices today?  I know about the low inflation environment driving PE expansion, but at some point, earnings growth has to be a contributor to growing stock prices.  And that is just not happening.  One final note, I do agree with you (and Scott Sumner) that monetary policy is too tight.  But this can't be good for stocks either when stock price are dependent on PE expansion and EPS growth is not there.  if you could comment on that I would appreciate it very much.  thank you for your time.
The simple answer is that there are numerous way of valuing the stock market. Different metrics give different results. If that was not the case then we would not have this discussion. I do acknowledge that some of these methods indicate the US stocks are slightly overvalued, but none of these methods to my knowledge indicate any large overvaluation and certainly nothing that you would considered a bubble. Just take price/earnings - the P/E ratio for S&P500 is now slightly above the long-term average, but likely within one standard deviation of that average (I didn't spend much time looking at the P/E numbers...) In that sense my model for the S&P500 is pretty much in line with other more conventional valuation methods. Fundamentally, however, I think the real uncertainty regarding stock valuation is how to forecast future earnings. In my model I use private consumption expenditures as a proxy for both nominal spending growth and earning growths (over long periods these move hand-in-hand). However, this is basically a backward-looking measure and as such it tells us nothing about future earnings, which really is the important thing for investors. I do use ISM New Orders as a proxy for near-term earning expectations, but I do not in any way capture long-term earnings expectations. However, I fundamentally think that the expectations for long-term earnings growth reflect long-term expectations for nominal GDP growth as the capital-labour ratio tend to be fairly constant over time. Here it is important to remember that both stock prices and earnings are nominal rather than real. Hence, effectively it is Fed policy, which determines earnings growth over long periods. This of course also means that the Federal Reserve if it wanted to could increase earnings expectations by increasing NGDP growth expectations, which would increase stock prices. However, this is not a monetary induced bubble. A bubble is an increase in stock prices, which does not reflect higher earnings expectations. This also means that even if S&P500 is fairly priced or even cheap it could still drop if there was a change in expectations about future earnings growth - for example if the Fed started scaling back quantitative easing prematurely and too aggressively. In that sense it should be remembered that my model indicates that the S&P500 was fairly valued in 2008, but nonetheless S&P500 collapsed as earnings (NGDP) expectations collapsed. That of course could happen again. On the other hand what we have seen over the past year in the US stock market in my view to a very large extent reflect a change in expectations about future NGDP growth as the Fed finally move from a discretionary and quasi-deflationary policy to a  



Sign up now to receive the latest blog posts and news about our research.