I have been tracking valuation over time since seeing a CNBC interview with David Tepper 4 years ago. CNBC was grilling him because SPX hadn't reached his price target. He was emphatic - "I didn't say such and such a number, I said 18x forward earnings - and it reached that level. At the time 18x forward earnings was the price target you cited." Or words to that effect.
So this endeavor started as a project for the small fund where I was the primary strategy consultant, just to keep up with Wall Street thinking. This is also how they do their year end targets - calculate earnings estimates, assign multiple, voila, target 2400 or whatever. It doesn't mean they will be right - in fact consensus probably more likely to be wrong - but I liked to know what they were thinking as a group.
As it turned out 18x forward earnings played a major role in the SPX 2015 top and distribution and subsequent drop. Oh yeah, pivots were awesome too. Check the home page for that (NDX top on level).
When the fund closed down and I started this site to show some chops, I had to reverse engineer this project since the Bloomberg was no longer available. I began with Wall St Journal data, but then saw Ed Yardeni cite Thomson Reuters so made the change. Then I got frustrated with week to week chop in the cited P/E, and that issue was solved with a 10 period moving average.
Over the past year I have consistently said that I expect to see at least 18x forward earnings for a major top. For a while I was saying 18x-20x would be appropriate for a euphoria high. At some point after the Trump election this was reduced to 18x-19x. This target area was tagged on 3/1.
The Thomson site listed 17.99 as of 3/10 week data collection, and this week dropped down to 17.76. That is the raw data, not the 10MA. At the price high of 2400.98, the 10MA of 18x level was 2388. So the level I have been looking all this time was tagged. The 10MA value has continued to climb a bit since then, with a current value of 2399. Price highs this week were 2390 and 2388 - so not quite. Also we haven't had a weekly close above 18x 10MA value. But the market has clearly slowed down at this area. Oh yeah, pivots in play too.
Here is a more graphic version put out by Ed Yardeni of what am I doing with this project. According to his data, this is the first time we are seeing 18x forward earnings or near it in quite some time. But I know what was on the Bloomberg when using. I don't know the reason for the discrepancy but maybe some CFA can explain.
I'm also interested in the rate of change and not just the level. Traders just call this slope. Here is my non flashy chart of what I am getting at. If I put the price on this, the slope is harder to see so this is the pure price of the SPX implied by the 10MA of 18x forward earnings.
As long as this continues to climb, then price can climb too while maintaining this rather full valuation. But when that starts to look more sideways, caution warranted, and if sloping down, watch out.
Here's a chart with SPX added. The reason it doesn't look like the level has tagged is because the market has not had a weekly close on the level. 3/1 high tagged the area and retreated, and recent highs were a shade under as well.
Sum - I consider the market fully valued here. We may see a professional selling reaction from 18x forward earnings and so far it has paused the rally when seemingly nothing else could. But as long as the slope is increasing - ie earnings are going up - then I cannot call it a negative.
Citigroup Economic Surprise Index was another item available on the Bloomberg - very handy with several different versions (Japan, EU, China, EEM, etc). Yardeni has somehow been able to make the main USA version available and I have been linking to it each week.
This is still going up and a definite plus.
So if we had some simple scale on 0 to +/-5 were 0 was neutral, 5 the best, -5 the worst, let's say valuation is about a 1. This will be downgraded to zero if the slope is sideways. Fundamentals are better and reaching a fairly high zone in the Surprise chart, let's say 4.
Combined value 2.5, moderate positive. These numbers are arbitrary. Just giving the math types some sense of what I am thinking about these two combined.