S&P 500 2009 Estimates Are All Over the Map

Quite an interesting piece from Bob Pisani's [CNBC] "blog" - the vast difference in 2009 estimates for the S&P 500 companies in the "top down" community versus "bottoms up". In English, top down are the macro economists or strategists who take a world view, whereas the bottoms up guys are generally analysts who follow a sector/stock group, and therefore usually have a very narrow world view. I try to do both as this website shows every day...

Here are the earnings power of 2009 based on both parties:

Bottom-up: $91.85

Top-down: $62.98

Now keep in mind the "bottom up" crowd is the one we like to poke fun at often (well we've been poking fun at both groups a lot since they've both mostly missed the boat)...but they bring it upon themselves. Last spring I was pointing out how analysts (bottom up) had huge Q3 and Q4 2008 estimates for earnings - remember back then "2nd half 2008 recovery!" was all the rage among the pundits, and analysts were among the biggest drinkers of Kool Aid of all. Fourth quarter 2008 was so unseemly, they had predicted 60% year over year growth over fourth quarter 2007. That is now looking so wrong one should fear for their jobs - but in this world of analysts there is little to fear if you ALL got it wrong.

Hence the pack mentality that makes their numerical information many times useless. The mentality of the pack rats is if you are wrong, and we are all wrong together - then I can go to my boss and say "hey no one saw this coming so therefore my foolish guesses were still in line with the other lemmings" - so why should you ever take a stand and be a major contrarian? All it can do is make you look foolish if you don't nail it and you are the only one to be either very positive versus the group or very negative.

Sure you might make a few outstanding calls by being an outlier, but why take the risk and risk the cushy job by coming out on the wrong side? What is making things worse, is since the scandals in the analyst community in the early part of the decade, this has gone from a "revenue generating" part of a bank to a "cost center" and hence in the high paying world of finance, the worse stay in the banks and the better ones go off to institutions where their sector expertise gets its just reward (if they are good of course). I wrote a long piece on this a year ago but that's not the point of the post.

The point is what should the multiple be on these earning estimates? And which one is more accurate? One could argue 10 PE ratio to make it easy - if you use the top down analysts that means S&P level of 630 lies ahead (remember we've been fighting S&P 840/850 lately so that's a long way to go down). If you believe the bottom up guys - hey we're cheap! 920 is the trough even on 2009 estimates... buy buy buy.

Obviously from there you have to ask questions such as ... will 2009 even be the trough earnings which many now assume. Or will 2010 be worse? If 2010 is not worse - how much better will it be - if not much, then you probably keep a very low multiple out there. If however you expend a solid rebound in 2010 than 10 x $63 EPS = 630 does not make sense as a level to sit at or even target for very long as the rebound will be rather swift.

An even bigger question, is what PE multiple do we get in the year(s) ahead if we enter a period of substandard growth? In the early 80s to late 90s we had major multiple expansion as the US enjoyed a very good period for most of that time frame. But as we are a maturing economy should our stock market begin to return to a lower permanent PE ratio? Not 10, but perhaps 14-15, instead of 18+?

So there are many questions out there, but when we see the day to day random nature of this market and ask each day what exactly is the market discounting into the future - you can see the huge variance in the data set the market is reviewing. And then all the variables over and above that we need to figure out in the year(s) ahead.

Comments

[Steven Hansen:]

as there was not one group of economists that have correctly predicted the economy even three months ahead in 2008, how could analysts even begin to predict how it effects each company.

the herd mentality happens to investors and traders as well. so everyone can be in one big herd bellowing and snorting. in the end the results are the same whether everyone hears the thunder months away, or everyone just waits to have reality slap them in the face.

2008 Nov 20 05:00 AM

[kotika98:]

you should call the $90 estimate a "normalized earnings" estimate and a 10-12 multiple sounds like a fair multiple on that. The $60 estimate is "trough earnings" and one should not put multples on it, as Cramer said it in the book - buy cyclicals at the bottom when they have the highest multiple, and sell when it at the top when it has the lowest. Makes perfect sense, come to think of it.

So, if you believe that the earnings can grow 4-5% long term and you think they will be in 10 years at 60*(1+4%*10) = 84 then 600 is your S&P target. But if you think earnings will recover to 90 in 2010 and by 1019 will be 90*(1+4%*9) = 120 then your target on the S&P is 1000-1100 or so.

2008 Nov 20 09:54 AM

[Smarty_Pants:]

Of course these numbers could also be Panglossian to the extreme. What if things get worse (as they seem to be on a daily basis) and earnings are only $45 and multiples shrink to 6 as they did in the Great Depression?

S&P 500 bottoming around 270?

Might happen.

2008 Nov 20 03:26 PM