My Photo
Note: Jeff does not accept guest blog posts on A Dash of Insight.

For inquiries regarding advertising and republication, contact

Follow Jeff on Twitter!

Enter your email address:

Delivered by FeedBurner


  • Seeking Alpha
    Seeking Alpha Certified
  • AllTopSites
    Alltop, all the top stories
  • iStockAnalyst
Talk Markets
Forexpros Contributor
Copyright 2005-2014
All Rights Reserved

« Understanding Economic Progress | Main | Talking Stocks, Blogging and System in an Interview with Tadas Viskanta »

May 17, 2011



Does anyone know where you can get historical forward earnings estimates? I do not have access to a Bloomberg and the S&P website only shows current forward estimates, not historical.

Any help would be greatly appreciated. Thank you in advance.

By the way, this is definitely one of my favorite sites to visit.


DFtF -- The Fed program has barely filled in for the growth in money supply needed for normal GDP growth. The liquidity is not abnormal. It has helped stabilize the economy, and in that sense helped consumers and businesses.

I sense that you are suggesting something more direct. I discussed that topic pretty thoroughly in this piece:

I hope this is helpful.


Don't Fight the Fed

So I suppose that the liquidity that allows consumers to keep spending and for companies to continue to generate the kind of earnings mentioned is completely unrelated to the Fed?


Off topic, but, since it came up... I am by nature a very private person. I chose the name Proteus in honor of a childhood hero, C.P. (Charles Proteus) Steinmetz. Steinmetz, like me, was an electrical engineer, but unlike me, he could numerically solve double integrals in his head.
Rich S.


scm - Perhaps I could have been a little more explicit.

By demonstrating that the rally is justified on the fundamentals, I am trying to suggest that it is not a QE II artifact. Naturally this requires acceptance that QE II has a modest effect on rates.

This story is too big for a single article. There are many bogus correlations out there. I have several more pieces planned, and I'll try to sharpen up the conclusion.

As to the ECRI, I'll discuss that more this weekend. I always cite the public index. Lakshman has been citing a proprietary "long-leading" index that they do not publish. It is signalling a global slowdown, but not a recession.

The accuracy of forward earnings estimates (and the market multiple) clearly depends upon the pace of economic growth. I don't think that the ECRI index differs much from the "soft patch" that many are talking about. We shall see.

Perceptive question:)




Two things:

Am I missing something, or did you not really answer the question you posed at the beginning of this article?

Lakshmi was on CNBC a couple of weeks ago, signaling a macro slowdown. Did you happen to catch this, or do you otherwise pick it up in what you model? I ask because you tout ECRI's authority pretty consistently.

Thanks for what you do.


DE -- Yes, I saw this article. If you read it carefully, focusing just on the one-year forecast that I use, he really only has one point. He says that 2009 estimates (made before Lehman in 2008) were really bad, worse than the errors of the 2001 recession. He also notes that 2012 earnings estimates are now reaching the 2009 level and asks "How did that work out?"

It is really not helpful to take the worst period since the Great Depression and use it to make a general case about the accuracy of forecasting. The throwaway line about 2009 does not add anything to the argument. Obviously, we will eventually pass that level in earnings, so do you just keep chanting "2009,2009?"

So that is really the only substantive point of disagreement with my research. (The part about 'gaming' the actual release is not important for the one-year forecast).

He also seems to use the next calendar year rather than the next 12 months, as I do. This means that it is not really the one-year forward estimate from my piece.

I also note that during the time period when estimates have become less accurate, many companies have stopped providing guidance, especially those with little sell-side analyst coverage. This would also affect accuracy.

Just some preliminary thoughts -- maybe more later.

Consider this -- Bianco did not suggest any better method.

Good question. Thanks.



Jeff any thoughts on Jim Bianco's analysis on short term earnings estimates. I don't believe this reconciles with your research on the accuracy of estimates over shorter time periods such as 1 year.

Chris Tinker

I continue to be amazed by those that think Stock Markets are for those who want to take a ten year view from now on listed global companies. As you may recall from past comments, my approach involves using 1 and 2 year forward expectations data to model the implicit value in a share price and to generate an intrinsic value (i.e. what the stock is likely to be worth in 12 months time) from that. It is a pretty complicated approach but works extremely well and works on the basis that a maximum portfolio holding period (prior to review at least) is 12 months. Risk is about the risk to your capital of owning that asset compared to cash or other assets over that period. That is all that is required to make the investment decision today between owning and not owning the stock. You can then work out the risk option of doing so using a real options approach and all of the data you cite above reinforces the bull rally over the last 9-12 months. CAT on this basis now has a 12 month value of $111 so hardly a screaming buy - but not terrible by any means.

All of this fits neatly in with your view and rather than deal with theoretical analysis of medium term DCF models, mean average historic p/e or other assumptions based on views beyond year two that are simply extrapolation and guesswork, investors need to acknowledge that money is made or lost on equity investments over a much shorter time frame.


Mike C -- I strongly recommend that you start your own blog. You might understand better. I do not seem to be able to explain why your brand of off-topic commenting, changing the subject, quoting at length from others in a way that creates an expectation of a response -- all of this -- is just wrong.

I encourage you. Go for it. Start your own blog and invite comments. It is the best way to demonstrate your knowledge and your stock picking success, which you obviously think is the equal of any of us. I promise to read you and to add you to the blog roll.

Meanwhile, you are welcome here. You have made helpful comments in the past. It is your own choice whether to lurk or to comment. I ask only that you use a single name, and that you try to stay on topic. At least be aware that I have dealt with most of the points you raise in the past -- often on numerous occasions.

Before challenging me with the latest revisionist thinking from Hussman (and I feel his pain) you might actually see whether I have discussed those issues -- going back to 1950 and mean reversion in profit margins.

You would have done better to comment on those articles, when it was the actual subject of the post.

I am making a resolution that I will not spend any more time responding to whatever you are going to say next.

Enjoy the last word!



I was amazed at the rampant pessimism last summer. Not only was the double dip crowd in full force, but all of the amateur technicians had jumped into short positions due to the Death Cross and Hindenburg Omen. To say that the market was a coiled spring by the end of August was one heck of an understatement.

After the fact, the bears have a few options to assess the situation in hindsight. Acknowledge that they were wrong on both a fundamental and technical basis, or try to escape responsibility by saying it was all Fed manipulation. Since substantial behavioral research shows that people tend to ascribe success to their own actions and failure to external forces, it's no surprise that they came up with the idea that QE2 caused a rally when, in fact, it was their excessive bearishness that pushed the market to such a low point that a big rally was actually necessary to make up for their mistakes.


inkerton -- Thanks for helping us to sort this out. Everyone who writes about market valuation uses some type of projection of earnings, dividends, or cash flows. My particular disagreement is with those who use backward-looking data, and do not consider interest rates. I named the leading exponents.

Implicit (somewhere) within their approach is some kind of forecast about the future. I have had the challenge to their followers out there for nearly a year, so I am surprised that this particular article got a response. Perhaps I should word it differently to ask what useful thing is being predicted.

As I have noted, I have written many, many articles on these methods, and that is not the main theme here. Maybe I need a FAQ on the topic -- something easy to point to.

Thanks again for your thoughtful and constructive comments.


Mike C


I agree with you on the usefulness of 1-year estimates. That said, there is no correlation between 1-year earnings growth and stock market performance. At the end of the day, there is assertion and there is data. Now a more nuanced view is that forward expectations drive stock performance. Cullen Roche over at has a model built on this although I can't seem to find the specifics on it. I guess he wants to keep the exact recipe secret. :)

There are ALOT of people working on ALOT of different models for stock market exposure. I consider myself kind of like Bruce Lee in that I try to take what makes sense from various models and discard the rest. I think looking at risk and ECRI makes a lot of sense and the work I've done confirms the usefulness of it. I got the ECRI idea from this blog and a link provided here so I thank Jeff for that.

2000: one of the worst decades in the US stock market is beginning.

None of these outcomes were obvious 10 year prior;

This is actually false. In reality, 10-year returns are probably easier to predict then 1-year returns. Once you get out to 10 years, you know earnings CAGR is going to be around 6-7%. Then it is just a question of starting and finishing multiple and simple algebra problem. Go read Grantham's quarterly letter from 2000. He predicted the stock market returns well in advance. Other models by other people also nailed the 2000-2010 decade of returns. That said, knowing the market is going to return 3-5% from 2010-2020 doesn't do me much good from a pragmatic perspective. I'd rather try to catch the big upswings and avoid the big downswings and do much, much better.

Paul in KC

the point is the market cares a great deal about forward earnings; therefore it is important to focus on this aspect of equity valuation; Jeff's article with respect to one year forward PEs very helpful and USEFUL. we can all rationalize about different viewpoints regarding earnings; but predictions 10 years away regarding anything way too tough. Some easy examples; how likely would anyone have believed these statements at the time?

1940: Germany as a nation will cease to exist and THe Soviet Union will have the largest military in the world. Does this seemed likely in June 1940??

1980: The Iron Curtain will collapse in 10 years. Interest rates will be cut in half;

1920: The "greatest depression" in the modern age will occur in the next 10 years.

1910: Modern Europe (meaning the monarchies) will cease to exist.

1990: the greatest decade in the US stock market is beginning.

2000: one of the worst decades in the US stock market is beginning.

None of these outcomes were obvious 10 year prior; that sort of thinking is fun and may even be useful in determining responses to certain outcomes;but our predictive abilities in these time increments are poor at best.

Mike C

Small correction for accuracy sake...I did comment on one other post under a different name. Other then that, I have not commented for a great number of months, and will return to lurker mode.

Mike C

I hope that you will understand and respect how I am trying to operate my blog. Other bloggers are far more aggressive or do not take comments at all. I embrace and encourage discussion on the topic at hand.

Yes, I can understand and respect that. I would highlight that I have NOT commented on your blog for several months. Truthfully, I do not think it is possible for you and I to have a constructive dialogue for either of us so I have refrained for that reason. My approach to your blog has been to read it, get what I can of value (the new risk metric is novel and interesting), and keep my mouth shut. That said, this sentence did successfuly bait me into a response:

"I continue to invite any loyalist for another method -- particularly those espoused by Hussman or Shiller -- to show that they can predict earnings one year in advance with greater precision."

Personally, I think accuracy is an important thing, and I just felt compelled to respond because absolutely nothing in the Hussman or Shiller approach has anything to do with successfully predicting one-year earnings. As Inkerton demonstrated an ability to clearly understand their methods are about long-term stock valuations and long-term return forecasts, not 1-year forecasts of any sort.

What I do find perplexing is you mention you don't want to engage in refutations of someone else, or that you have covered Shiller before, etc. So why keep bringing them and their "methods" up in the discussion. You could have easily made your point about 1-year earnings forecasts and QE2 without mentioning them.


Mike C -- Last year I spent many hours investigating a topic that everyone was getting wrong -- a piece of Wall Street truthiness. (and by "everyone" I mean this in a general sense, not as a literal assertion that no single person in the world disagreed) Identifying the accuracy of one-year forecasts is an important research finding.

I shared the results freely, without any compensation from anyone.

When I say that "everyone" uses this time frame, I mean that it is widespread, a generally accepted standard. You hear it from almost every analyst interviewed and see it in any analysis of a stock. Most of us do not try to predict much farther in advance. My article pointed out that longer-term predictions were not as accurate.

To call this a red herring is unfair and inaccurate, especially when you admit you have no data of your own. To the casual or first-time reader it diminishes the research work I have done.

My invitation pertains to identifying better methods of forecasting earnings, particularly in a relevant time frame. It does not mean that I am going to engage in a refutation of someone else's recent commentary.

I spend a few hours most nights developing a topic and writing an article. This one is about QE II -- a topic that many readers asked me to discuss.

You want to talk about many other things, and you do so in a very literal fashion. The off-topic arguments are all themes that I have discussed in prior articles, but I cannot spend hours debating you in the comments.

I hope that you will understand and respect how I am trying to operate my blog. Other bloggers are far more aggressive or do not take comments at all. I embrace and encourage discussion on the topic at hand.



So given your quote, "I am not making a ten-year forecast about any stock or about the market as a whole," I think it is both unfair for others to set you up in opposition to Hussman, and it is somewhat unfair for you to call for a follower of Hussman, as you do, to "show that they can predict earnings one year in advance with greater precision." That is because, as far as I know (I'm not an afficionador or a shareholder) Hussman is not attempting to predict earnings one year in advance at all. Calling a follower of Hussman to predict earnings one year in advance is, as far as I understand it, simply a non-sequitur. The two of you are doing completely different things.

Thus, it is perfectly within the realm of possibility both that you are 100% correct about the moves the market, or your picks, will make within the next year or two, and also that Hussman or Shiller is simultaneously 100% correct about the expected returns of holding the S&P for ten years, if bought today. I'm not saying he or they are correct, I'm just saying that you are doing a completely different analysis, based on completely different expected holding period.

That isn't to say there aren't other problems with Hussman. An additional one, mentioned neither by you nor by me so far, is the fact that his actual holding period for particular stocks (as far as I know) does not appear to match the long term period on which his analysis focuses, it is much, much shorter, which to me is also highly problematic, as I think the holding period of stocks or even of an index should roughly match one's prediction period. Yours does: short-term predictions, short-term holding periods.


PWilson -- You have a very good point. The effect of the Fed policy is psychological as well as technical.

This is very difficult to measure. Many of those who thought QE II would have no effect are the same asserting that there will be a disaster when it ends. Bernanke (probably unwisely) claimed great success as measured by the stock market gains.

Much of economic activity relates to confidence -- both on the part of individuals and also business leaders.

Your raise a good and difficult question.




It seems to me that the pundit consensus is overwhelming bearish. In fact, during the three decades I've been managing money, I cannot remember such a barrage of negative commentary (before a move in the market). I cannot help but think this is a prelude to a major move higher.

The comments to this entry are closed.