Some More Ideas About Long-Term Valuation

I have spent some time this weekend following up on some valuation ideas from a long-term perspective because I know that this is likely to be (or at least should be) an important issue for some when looking at long term investments in IRA's, 401K's, etc.

The links I put in a recent post all approached the question from a perspective of P/E ratios relative to historical levels maybe taking into account another condition as well. Without going into critiques, I think that looking at P/E ratios alone is not adequate, and looking at only a few decades is also not adequate if trying to understand long-term de-leveraging cycles and periods of truly historic market over and undervaluation.

So I spent some time thinking, referencing books, and looking at charts and websites to try to give an easily understandable perspective on some other measures of valuation. A good place to start in getting background on this info is in Conquer the Crash by Robert Prechter which I have mentioned before. It presents a case for deflation and what to do with investments if you agree with the thesis and think that planning for that would be more prudent than (not) planning for the status quo. So whatever your leanings on the issue, the numbers can be evaluated in their own right for decision making purposes.

To start one measure that captures both a valuation and an inferred sentiment in one is looking at historical dividend yields vs market history. The long term average dividend yield on the Dow 30 has been about 3% over the last 100 years which encompasses both a Great Depression, multiple bear markets, and obviously all the bull in between. Just for a quick background on dividend yield.....Stocks have both a value per share and a yield to the extent that they pay dividends. Many stocks don't pay dividends at all and have no yield. But if they don't, then the investor must have rising share prices to make it worthwhile to own stock. If you get dividends, at least you share some profit along the way even if shares don't rise. A high dividend yield shows that the company is paying a large dividend compared to its share value.

Now looking at history dividend yields rise at market bottoms with most bear market bottoms hitting levels of 6-8% (and 15% at the 1930's/depression bottom). Several factors go into this. Stocks decline obviously. But earnings do also. So companies often cut dividends due to decreased earnings. So it's not really that dividends are rising, but that stocks are falling faster than dividends are. Also, the sentiment side of this is that as stocks perform badly for extended periods of time, investors will need a reason to buy stocks and demand higher yields. So basically high dividend yields will be in part a result of pessimism towards stocks. Because of this, the data can be used as a timing indicator of sorts.

So where are we today? This site has nice charts going back quite a while and to recent data to show stock prices relative to dividend yields on the Dow 30. At the lows in March the yield peaked at about 4.5%. Currently, the rise in the market has valuations at about 2.8% which is lower than the long term based off of the info here. So, even at the worst in March yields did not rise to historically bear market bottom levels. Given that the economic conditions are basically worse than all those other bear markets (maybe not the Great Depression but that remains to be seen), I wouldn't be too quick to assume that the March low was "the" low. Certainly making new purchases now doesn't seem to make a lot of sense by this metric.

While some good historical perspective on P/E's can be gleaned from the links I recently posted I will add a bit to that here, building on some data in Conquer the Crash. The long term average P/E ratio is in the mid teens. Historic bear market lows have seen P/E's reach 7 or 8. Prior to the 2000's and beyond, the bull market peaks saw P/E's in the mid 20's. The 2000's saw P/E's rise to the 30's at the peak prices and then the 40's on the way down. One thing Prechter notes are divergences where P/E's fall below prior lows even as prices make higher highs. That suggests sustained pessimism and undervaluation towards stocks despite price moving higher. That has portended the great bull markets. The other end of the spectrum is when the P/E rises to new highs after markets are well below peak values. This shows sustained optimism that stocks are undervalued and investor willingness to overpay for the real earnings. That had never really happened until the 2000 top. But that is what has happened since and in dramatic fashion.

The P/E continued to rise as the market rebounded to lower highs in the early 2000's. Currently despite stocks being 30% off the recent bull market highs, the actual P/E is near 140! So take what you want from that, maybe the earnings will rebound strongly and quickly put that back in line, but again it may be a sign of lingering optimism and sustained overvaluation towards stocks. So by this measure again, we have not seen historically low P/E's reached in this bear market, and the inference may be that we are certainly not witnessing a longer term disdain for stocks and corresponding undervaluation.

There is even more to that discussion in that P/E reporting standards were changed (I believe in 2001) to exclude interest payments on debt from the earnings number. That is called "operating earnings" and gives a lower P/E (due to a higher earnings amount), making it seem more reasonable. So, that may make the valuations of the last decade even more excessive compared to historical standards.

Another way to approach the valuation issue is to look at the relative valuation of the two main investment classes - mortgage backed securities and credit default swaps....I mean Stocks and Bonds. I suppose there are several ways to do this. One may be to compare the dividend yield on stocks versus high quality bonds or government note like the yield of the 10 year note. I haven't been able to find data on that. Another way is to simply divide the S&P 500 by the yield on the 10 year note. That will help to give a relative indication of whether stocks are overvalued compared to bonds. Looking at the current situation the S&P/10 yr yield ratio is nearly 2 standard deviations away from the average for the last year or so. This is not a real long term measure, but argues for intermediate term overvaluation of stocks possibly.

Now one ratio that did get pretty low was the S&P 500 value/book value. It hit around 1.3 in March. I had trouble finding data any longer than back to the early 70's on this. From that data, 1.3 was very low and should expect a major bottom. Currently it is back to around average book value during that period. So this would argue for a longer term bottom possibly. However, I think it is very important to keep a long term time perspective. I don't know that comparing the current environment only back to the 70's is good enough given what I think are greater comparisons to the 1930's environment. I would like to see more on the truly long term numbers particularly those after deflationary periods.

So hopefully this post will give either actionable information or at least start asking the right questions. In sum, these measures are hinting at overvaluation for the current period after a significant but not historically rock bottom valuation level in March 2009.