Friday, February 13, 2009

Updated S&P EPS - Throwing in the Refrigerator...

... and the oven, and the dishwasher, etc.

Apparently, the media is finally catching up with the S&P EPS estimates that have been out for a few days now and which I wrote about in the Kitchen Sinks & Miracles post. It would seem that this quarter is going to see everything thrown in as - amazingly - S&P has actually revised EPS for Q4/08 downward again. This is at least the 2nd update of EPS this week and this latest one holds some significant changes. And perhaps not coincidentally, there have been new discussions of valuation of SPX so I'm going to do a little updating on that topic as well.

First, let's have a new look at the SPX EPS estimates for both operating (opEPS) and as reported EPS (arEPS). Just this week opEPS for Q4/08 was revised from $6.33 on 2/10 to $5.77 on 2/13. The arEPS values were arguably even worse as it went from -$8.79 to -$10.44. Evidently, the hits just kept on coming. Perhaps more ominously is that this last revision took down Q1/09 arEPS estimate -15% from $9.87 to $8.36. And looking at the chart, this is the first revision of arEPS in some time that shows a significantly lower outlook for all of 2009 and into 2010. On average, the reduction was roughly -22.5%. However, the opEPS reductions was not nearly as severe, which would imply that S&P is expecting more write-downs, charge-offs, reductions in goodwill or whatever continuing for the next several quarters.

As in the prior post, here's a look at the EPS estimates over time. As already noted, the arEPS negative value is unprecendented. What is very interesting to me, from a trading perspective is the comparison between the tech implosion and current forecasts. The projections for Q4/09 opEPS show a recovery only to the levels of Q3/05. The case of arEPS Q4/09 has a projected value that puts it somewhere in between Q2 and Q3/02. Interestingly, the current level of SPX roughly corresponds to the SPX level between Q2 & Q3/02. I'm not quite sure what to make of this. (I won't even touch the effects of inflation/deflation here - that's far too expansive a topic for my little blog.) At today's SPX close, the PE based on opEPS - assuming no further revisions to the current quarter's EPS report - is 14.93 Using arEPS - which again, is what S&P itself will actually use for the PE of the index - the PE is 29.86!

When I wrote the last post on SPX earnings, I only briefly touched on the historical PE and just made some references. Then I looked back at my post archive and realized that some of the data I was pointing to was actually in some work I had done prior to starting this blog. I'm going to correct that here and update it a little bit.

At the right is a chart showing the trailing-twelve-month PE value using both arEPS and opEPS for calculation since 1988. Also included are the reported EPS values. (I thought it better to clutter the chart than the entry itself.) First, the average PE(arEPS) = 19.25 and the PE(opEPS) = 22.81. S&P only has data for opEPS going back to 1988, hence the limitation. In an earlier post on SPX and P/E, I looked at Shiller's historic data of arEPS and noted the average P/E from 1950 to Dec. 2008 was 16.59 with a standard deviation of 6.76. Considering the above mentioned PE values, neither metric is in "value" territory. However, look closely at the chart - specifically in the 2002-2003 period - when the market finally bottomed the PE(arEPS) was a still-elevated 27.14 and the PE(opEPS) was 18.51. Both of these are above the average. So what changed to put in the bottom? My guess is a recovery in arEPS. Yes, there was still another quarter before the true bottom in arEPS but by then, opEPS had shown some steady (if small) improvement.

My point in discussing this is that when various bears talk about how the market is overvalued and then they project SPX will be at such-and-such a level based on some multiple, they are suffering from similar biases as bulls. There is no "correct" multiple. Sure, you can hazard a guess based on history but look again at the chart right in the middle of the tech bust - the PE(opEPS) reached the mid-40s! That to me, signals that market players refused to come to terms with the reality of a burst bubble and corporate earnings although eventually they threw in the towel - sort of - and the market reached a bottom. So what you find right now is a debate of sorts. On one side you have bulls performing a sleight of hand claiming the market is cheap because PE(opEPS) is below the average PE(arEPS). And on the other you have bears claiming that the market will have to go down because it should have a PE of 15 and earnings going forward are awful. Granted, I have more sympathy for the bears, but to slap a 15 PE on SPX because that's what you consider a "fair value" is absurd in the face of popular delusion. The evidence that the bottom can form even when stocks are overvalued relative to historical averages is right there. Could multiple contraction occur? Certainly, and if popular trust is crushed by more revelations of criminal and rigged behaviour on the part of market makers, it could happen. (This was a major factor in Japan in their post-bubble era.) This would require a change in American popular sentiment that I suspect is no longer possible.

My (currently under-developed) opinion: In the next few years, the market will return to a situation in which investing is done based much more on dividends than on capital gains.

Always remember: "The market can stay irrational longer than you can remain solvent."