Mark Sadowski nails it again

The GDP data just came in, and Mark Sadowski continues to demolish the experts:

Actual RGDP growth:  minus 1.0%

Mark Sadowski’s initial forecast:  minus 1.4%

Initial Consensus:  PLUS 1.0% to 1.5% (depending on source)

NGDP came in at 0.3%, slightly below Mark’s 0.5%.  I don’t have the consensus, but it must be around 3.0%

In the 4th quarter of 2013 the consensus RGDP growth rate was around 3.8% or 3.9%.  Mark forecast 2.5% and the actual was 2.6%.  Mark forecast 4.2% NGDP growth, and the actual was 4.2% NGDP growth.

I can only find one other Sadowski GDP forecast, for the second quarter of 2013.  The consensus called for 1.0% RGDP growth, Mark forecast 2.0%, and the actual was 2.5%.

If he’s not snapped up by an investment bank in the next 30 days I am going to have to revise my views on the EMH.  Don’t make me do that.  Mark understands government data better than anyone I’ve ever met.

PS.  I notice that the fall in NGDI was entirely due to a collapse of corporate profits, which fell 10% in the first quarter (from the previous quarter.)  That’s an annual rate of decline of 34.4%!!  Does anyone know the explanation for that? Was it measurement error?  Does it comport with the numbers being reported on Wall Street?


Tags:

 
 
 

65 Responses to “Mark Sadowski nails it again”

  1. Gravatar of Ironman Ironman
    29. May 2014 at 07:15

    It’s not a measurement error. We’ve been following the story on Wall Street for some time – this link will take you to our chain of analysis during the past several months, while this link will take you to our most detailed discussion of the things dragging down GDP in 2014-Q1.

    In a nutshell, the slowdown in earnings on Wall Street pre-dates the arrival of cold weather in the U.S. The most negatively impacted industries are mining (China slowdown), educational services (regulatory burdens), health care and social assistance (Obamacare implementation, real estate, rental and leasing (adverse interest rates from QE taper plans/implementation).

    Note that real estate has increasing become a drag since last summer, when the Fed set expectations that it would initiate QE tapering. Since then, fluctuations in mortgage interest rates have increasingly set the pace of sales.

  2. Gravatar of Kevin A Kevin A
    29. May 2014 at 07:16

    It would be nice to have Sadowski weigh in on Piketty’s data foibles and the materiality of those foibles on Piketty’s claims. It seems to be somewhat in his wheelhouse.

  3. Gravatar of Aaron W Aaron W
    29. May 2014 at 07:57

    Speaking of measurement error, I’m curious about these GDP estimates. What are the confidence intervals on their estimates anyway? I’m having a hard time finding any information about this. And this is a common problem with economic data I’ve found. Physical scientists have to report measurement error all the time, why do economists avoid mentioning it?

  4. Gravatar of Mark A. Sadowski Mark A. Sadowski
    29. May 2014 at 08:19

    @Kevin A,
    I’d love to weigh in but I prefer to wait until I’ve had a actually chance to read the book.

    @Aaron W,
    There’s two main ways of measuring the size of the revisions of the components of national income and product accounts: 1) Mean Revision (MR) and 2) Mean Absolute Revision (MAR). For rate targeting MAR is the more appropriate measure, and in fact the MAR of inflation is usually smaller than the MAR of NGDP. However, for level targeting MR is more appropriate.

    Interestingly, at least in the US (Page 27):

    “The MRs for the price indexes for GDP and its major components are generally not smaller than those for real GDP and current-dollar GDP and its major components.”

    http://www.bea.gov/scb/pdf/2011/07%20July/0711_revisions.pdf

    In fact over 1983-2009 the MR for the final revision to quarterly NGDP is 0.14 whereas over 1997-2009 the MR for the final revision to the GDP deflator and the PCEPI is 0.20 and 0.12 respectively. And over time the revisions have trended downward:

    http://www.bea.gov/scb/pdf/2008/02%20February/0208_reliable.pdf

    So I suspect that the MR for NGDP is actually smaller than PCEPI over 1997-2009.

    Which means all the stuff you hear about NGDP revisions being larger than inflation revisions, and hence NGDP Level Targeting (NGDPLT) is less practical than Inflation Targeting (IT), is pure grade A garden fertilizer, in addition to being totally beside the point.

  5. Gravatar of John Hall John Hall
    29. May 2014 at 08:20

    The consensus was not saying 1-1.5%. Maybe a few months ago they were saying that, but not recently. Bloomberg had -0.5% as their consensus estimate. Macroeconomic Advisors’ forecast was -0.8%.

  6. Gravatar of Mark A. Sadowski Mark A. Sadowski
    29. May 2014 at 08:25

    Incidentally, the biggest difference between my nowcast and the preliminary NGDP estimate was the change in private inventories. Well, with this estimate, that difference has largely been revised away.

  7. Gravatar of ssumner ssumner
    29. May 2014 at 08:31

    John Hall, I am talking about the consensus before the first announcement. Obviously once that announcement came in the consensus changed sharply. But what we want is people who can forecast before the first announcement.

    Ironman, I think you may be wrong. This link says that corporate profits are actually up in Q1, indeed more than expected:

    “Earnings estimates have rebounded, however, as more companies have reported results. First-quarter profit growth for S&P 500 companies is seen at 3.7 percent, based on actual results and estimates for companies yet to report, compared with a forecast for 2.1 percent growth at the beginning of the month, Thomson Reuters data showed.”

    http://www.reuters.com/article/2014/04/29/us-markets-stocks-idUSBREA360QI20140429

    I still don’t see where the BEA gets it’s estimate that corp. profits fell 10% in Q1.

  8. Gravatar of Mark A. Sadowski Mark A. Sadowski
    29. May 2014 at 08:42

    “John Hall, I am talking about the consensus before the first announcement.”

    I was wondering what on earth John Hall was talking about. For example here’s what Macroeconomic Advisers said immediately *after* the preliminary estimate was released:

    “First-quarter GDP growth of just 0.1% was even lower than we had estimated (1.0%).”

    http://www.macroadvisers.com/2014/05/q1-gdp-growth-just-0-1-amid-weaker-than-expected-figures-on-defense-spending-exports-business-fixed-investment/

  9. Gravatar of Gabe Gabe
    29. May 2014 at 09:24

    You should revise your EMH beliefs. I have more respect for you to see you can admit that.

  10. Gravatar of Ironman Ironman
    29. May 2014 at 09:38

    ssumner, we’re looking at data for the whole stock market – not just the S&P 500. Most of the carnage has been visited upon considerably smaller cap firms.

    http://www.bloomberg.com/news/2014-05-15/u-s-stock-index-futures-little-changed-before-data.html

  11. Gravatar of ssumner ssumner
    29. May 2014 at 09:58

    Ironman, Sorry, but I don’t see any evidence in the article you link to that supports your argument. And the S&P 500 is about 75% of the entire market cap. How could S&P profits be up 3.7% and overall profits down 10%? You would need some hard data to convince me of that, and I don’t see it in the article you linked to. Overall profits in the other 25% of the market would have had to have fallen in half. Is that plausible?

    Gabe, Given that I was joking, I suggest you sharply dial down your respect for me, back to its original level.

  12. Gravatar of Ironman Ironman
    29. May 2014 at 11:01

    SSumner, that’s *exactly* what has happened. Unfortunately, the proof is in the data, as it would seem that media reports have largely missed the relative plunge in earnings for small cap companies that has taken place. Starting with the most recently reported P/E ratio for the Russell 2000:

    http://online.wsj.com/mdc/public/page/2_3021-peyield.html

    As of 23 May 2014, the Russell 2000 had a value of $1126.16 with a P/E ratio of 80.19, indicating earnings of $14.04.

    Stepping back one year earlier, to 23 May 2013, the Russell 2000 had a value of $984.28 with a P/E ratio of 35.85, indicating earnings of $27.46.

    Sometime during the past year, the Russell 2000’s earnings have been slashed by roughly 50%. The following chart provides some clues as to when that situation developed:

    http://themcgowangroup.com/wp-content/uploads/Russell-2000-with-PE.jpg

    Since mid-to-late December, the P/E ratio for the Russell 2000 has hovered around the 80 mark, suggesting no significant recovery.

  13. Gravatar of ssumner ssumner
    29. May 2014 at 11:32

    Ironman, That P/E of 80 is for a trailing 12 month period, so I don’t see how that can explain things. What are the actual earnings of Russell 2000 companies, in dollar terms?

  14. Gravatar of SG SG
    29. May 2014 at 11:38

    Is there even a market for GDP predictions? How would I go about trading on Mark Sadowski’s forecasting prowess?

  15. Gravatar of Vivian Darkbloom Vivian Darkbloom
    29. May 2014 at 11:53

    The Russell site lists the Russell 2000 PE a bit under 22 (ex negative earnings).

    https://www.russell.com/indexes/americas/indexes/fact-sheet.page?ic=US2000

    Yahoo lists the Russell 2000 PE at 19 for IVM, the tracking iShares Russell 2000 ETF as of April 2014.

    http://finance.yahoo.com/q?s=iwm

    These numbers are hard to reconcile with a PE of 80.

  16. Gravatar of benjamin cole benjamin cole
    29. May 2014 at 12:07

    All hail Mark Sadowski…and the MM perspective. Making predictions is really hard, especially about the future. But with the MM outlook and a reading of what a nation’s central bank will do, and thorough grounding in empirical observations…

  17. Gravatar of Vivian Darkbloom Vivian Darkbloom
    29. May 2014 at 12:14

    “Making predictions is really hard, especially about the future.”

    I think Yogi Berra said that first.

  18. Gravatar of John Hall John Hall
    29. May 2014 at 12:19

    @Mark,
    Yes, I was referring to the most recent forecasts. I had pulled up the Bloomberg consensus (on my group’s handy-dandy BB machine) and from emails that Macroeconomic Advisers sends me (really to me as a member of my firm). My point was more that the consensus is typically looking at data as it is coming in. As the data got began to suggest weaker and weaker inventories, the consensus moved lower and lower, to eventually pretty close to where it came out. Nevertheless, Bloomberg’s April survey has the lowest forecast for Q1 GDP at +0.2%, so you would have been a huge outlier (to your credit, I suppose).

    My bigger issue with Scott’s post is that I have no idea if Scott’s cherrypicking your record or not. If you’re as confident in your forecasting abilities as Scott is, the best way to get recognized is to submit them to BB (or some other consensus gathering service). BB typically does a piece once a year evaluating who the best forecasters are. I don’t know what your background is, but if you’re an academic like Scott I typically see some university groups submit forecasts (looking on ECFC in BB there were a few that submitted forecasts for US GDP).

    Regardless, markets react to the changes in the consensus forecasts throughout the month and whether the real number is different from the consensus when it is released. I can assure you that the screen in BB that a trader would look at to see the numbers in real-time has the consensus directly to the left of the actual GDP. And that consensus would be the -0.5% that I mentioned, not the one Scott mentioned.

    However, equities are up today and had been all month, even when the consensus was moving lower. If you had bought S&P 500 puts in anticipation of weak growth, you would have lost money. You might have made more money betting on fixed income (TIPs yields fell quite a bit since April). I bring this up solely to note that in order to profit from a GDP forecast requires you to be right about GDP and about everything else about whatever instrument you use to express your bet on GDP.

  19. Gravatar of Ironman Ironman
    29. May 2014 at 12:55

    Vivian Darkbloom, the numbers certainly would be hard to reconcile if negative earnings are excluded and only forecast earnings are considered!…

    SSumner, the P/E that was used for the earlier datapoint is also TTM using actual earnings. The linked chart shows how that evolved over 2013 – think about what had to happen and when it had to happen to produce what has been observed with the TTM P/E observations.

    Also – the market cap of the Russell 2000 was $1.838 billion back on 30 April 2014. With nearly the same index value today, the math suggests that the TTM for net earnings for the index’ components would work out to be $22.9 million, which I would presume is not market cap weighted. What kind of negative earnings it took to reach that net figure is not something we can calculate from the available data.

  20. Gravatar of Vivian Darkbloom Vivian Darkbloom
    29. May 2014 at 13:08

    “Vivian Darkbloom, the numbers certainly would be hard to reconcile if negative earnings are excluded and only forecast earnings are considered!…”

    Ironman,

    Both sources I cited use TTM—not forecast earnings. As mentioned, the Russell 2000 site excludes negative earnings. Are you suggesting that negative TTM earnings at the Russell site explains a nearly four-fold differential in PE over a 12 month period?

  21. Gravatar of flow5 flow5
    29. May 2014 at 13:22

    Forecasts are only usable to the extent that Gov’t agencies have enough time to coordinate their efforts in order to correct the trend of nominal-gDp growth rates. Here it is May 29 2014 & we’re still trying to fine tune the data for the 1st qtr of 2014?

    There’s an upcoming 12 percentage point drop in the rate-of-change in money flows from July to Oct this year. No one is yet talking about this deceleration – yet alone acting to prevent it.

  22. Gravatar of o. nate o. nate
    29. May 2014 at 13:29

    Re: the big drop in corporate profits: It looks like there was a major adjustment in something called the “capital consumption adjustment (CCAdj)” related to changes in the tax code. I don’t claim to understand the specifics. More detail here:

    https://www.bea.gov/newsreleases/national/gdp/2014/pdf/gdp1q14_2nd.pdf

  23. Gravatar of ssumner ssumner
    29. May 2014 at 14:39

    John, You said:

    “And that consensus would be the -0.5% that I mentioned, not the one Scott mentioned.”

    You still are missing the point. What I’m talking about is the consensus forecast a month ago, not this morning. My numbers are accurate. The consensus this morning is meaningless.

    Ironman, Unless I see hard data none of that makes any sense to me at all. I’m extremely skeptical. Is the market cap really less than $2 billion? In that case they hardly matter. And $22 million in profits is nothing, as corporate profits run around $400 billion every three months. Where does the huge drop come from? They fell about $40 billion.

    Thanks O nate.

  24. Gravatar of benjamin cole benjamin cole
    29. May 2014 at 15:26

    Vivian D.–some attribute to Casey Stengel first, but it turns out to be a pithy saying probably as old as language…

  25. Gravatar of John Hall John Hall
    29. May 2014 at 16:23

    No I understood what you said and the ponit of it Scott, that’s why I referenced the April survey from Bloomberg, which had the lowest forecast nowhere near where it is now.

    I was trying to explain that 1) a few cherrypicked forecasts don’t really mean much, 2) he may have had the consensus beat initially, but the consensus updated their forecasts based on available information (and forecasters don’t typically take big bets versus the consensus without some information), and 3) if the connections between his supposedly better information and making money is nebulous, then it only really matters for bragging rights, which I don’t particularly care about (a better forecast about a 1 quarter blip in GDP growth is not a huge deal especially when every forecaster had the second derivative right, even if they had the magnitude of it wrong).

  26. Gravatar of Joseph Calhoun Joseph Calhoun
    29. May 2014 at 17:11

    Scott,

    At least part of the change in corporate profits was about tax changes. From the BEA:

    Profits after tax with IVA and CCAdj decreased $239.5 billion, in contrast
    to an increase of $33.8 billion. The first-quarter changes in taxes on corporate income and in CCAdj mainly reflect the expiration of bonus depreciation provisions.

    The first-quarter changes in taxes on corporate income and in capital consumption adjustment (CCAdj) mainly reflect the expiration of both the 50-percent bonus depreciation provision and increased Section 179 expensing limits claimed under the American Taxpayer Relief Act of 2012.

    But it wasn’t all about tax changes. Corporate profits were down across the board, financial and non financial. And financial earnings will probably continue to get hit with their profits from trading still falling and the mortgage departments still struggling. Add in a flatter yield curve and banks are having a tough go.

  27. Gravatar of Ironman Ironman
    29. May 2014 at 17:37

    SSumner, I’m really starting to hate how Russell presents their data. It turns out that $1.828 billion wasn’t the market cap of the entire index – it’s really the weighted average market cap of its components (the highest is $9.9 billion, the median is $0.684 billion).

    So, take the math I previously did, and multiply by 2000, the number of companies in the index. The total market cap for the Russell 2000 is $3.676 trillion, with a TTM P/E of 80.19 at today’s closing value of 1140 for the index puts the TTM earnings at $45.8 billion, which we already have established is about half what the Russell 2000’s earnings were a year earlier. Missing profits found.

    Vivian Darkbloom, the P/E of 19 that you cited is consistent with what the forward twelve month P/E is for the index – see this source for that figure and also the actual trailing twelve month P/E.

    Your other question would be better directed to Russell, where an explanation of why they would exclude negative earnings from their calculation of the TTM P/E is needed. Clearly, if the contribution of negative earnings is included, the effect upon the P/E ration is indeed as others have calculated.

  28. Gravatar of Eric Dennis Eric Dennis
    29. May 2014 at 17:39

    Mark,

    Have you discussed or formally written up your methodology anywhere? Is it driven by qualitative judgments, or is it more formal/quantitative? I and, I’m sure, others are curious.

  29. Gravatar of Kevin Erdmann Kevin Erdmann
    29. May 2014 at 18:09

    Ironman is on to something, I think. These differences in PEs are weird, and could account for significant losses. Why isn’t there an obvious story on this in the press? And, what o. nate posted is interesting, too. It seems the decrease in profits came out of the capital adjustments. It doesn’t seem like both of these things would necessarily be related, but they do both seem odd.

  30. Gravatar of ssumner ssumner
    29. May 2014 at 18:17

    John, I wasn’t cherry picking, I used all the evidence I could find. Why would I want to cherry pick for a hypothesis that seems to undercut my belief in the EMH?

    Joseph, Thanks for the info. But I am puzzled by this comment:

    “But it wasn’t all about tax changes.”

    You might be right, but the data showing profits are falling for both financial and non-financial firms is exactly what you’d expect if it was all about taxes changes. Or did the tax changes only apply to financial firms?

    Ironman, You may have established that Russell 2000 profits fell in half in 3 months—I haven’t, until I see the hard data–not P/E ratios.

    I simply don’t believe that actual profits could suddenly fall in half for 2000 small firms, when 500 big ones are seeing rising profits. I won’t believe it until I see the evidence.

  31. Gravatar of ssumner ssumner
    29. May 2014 at 18:28

    Kevin, That’s why I’m skeptical. If this was true, the story would be 10 times bigger than Piketty’s book, but I haven’t seem any significant discussion.

    Where is the news story saying Russell 2000 profits have fallen 50% in the past three months?

  32. Gravatar of Matt McOsker Matt McOsker
    29. May 2014 at 18:32

    I will give credit due – who cares what the consensus was Mark’s number came in on target for the final result if you ask me. We can however quibble about why growth was negative – but I am not surprised. IMO we are in a slowdown phase, where GDP will be very weak going forward. One thing I have not looked at is trade data. I am not buying any estimates of a major rebound for the remainder of the year. Would not be surprised if we are in recession in a year.

  33. Gravatar of Kevin Erdmann Kevin Erdmann
    29. May 2014 at 19:26

    Scott,

    The fact that the all-in PE ratio for the Russell 2000 is extremely high is in the news. But, this is so peculiar, I would have expected some strange anomaly, like small banks taking very large write-downs for some reason, or some sort of weird thing with some miners or something, maybe some looming biotech revolution with a bunch of small cap labs piling losses into early trials for something. But, in my brief perusal, I just see a bunch of no-nothing finance guys going on about how it’s 1999 all over again or something.

    There is little doubt that something peculiar is going on regarding losses in a subset of small cap stocks, but I just haven’t found the story that explains why.

  34. Gravatar of A A
    29. May 2014 at 21:25

    You don’t want to confuse an equity investor’s perspective with an economist’s view. If you look at consolidated Russell 2000 EPS to infer about aggregate demand, then you are making assumptions about the dispersion of losses, revenue declines, and index composition. There isn’t a direct route from falling reported earnings to the state of the economy. It also begs the question of why the revenues and operating profits of the S&P 600 have not similarly plunged.

  35. Gravatar of Vivian Darkbloom Vivian Darkbloom
    29. May 2014 at 22:47

    The Russell 2000 represents between 8 and 10 percent of total market capitalization. Ironman, I don’t doubt that you are “looking at the entire market” as expressed in the first comment above, but your focus here is on a small percentage of the overall market.

    Stocks in the Russell 2000 Index (and other indices) are adjusted annually (“reconstituted”) to throw out stocks that have grown sufficiently to disqualify them from the index and to add stocks that have either grown sufficiently to include them or have fallen from mid-cap status.

    The reconstitution is announced June 14 and takes effect June 28 of each year.

    For 2013, there was a change of about 1750 out of 6000 stocks listed by Russell as “small cap”. Only 2000 stocks are in the Russell 2000 index, so it is not clear how many of the stocks in the actual 2000 index changed. Perhaps there is a source somewhere listing that without having to do it manually, but I have not found it. It appears that there is a much higher turnover in the Russell 2000 index than in the large cap indices.

    So, when one concludes that profits among Russell 2000 stocks have fallen dramatically in 2013, we are not necessarily talking about the same stocks for year-to-year comparisons. Much of the change could be due simply to the reconstitution of the index. And, that change might in large part be due to former Russell 2000 stocks having become to profitable and therefore valuable to continue in the index.

    The two sources I cited (Russell 2000 itself and Yahoo) do use TTM, albeit Russell 2000 discloses it ignores loss companies. Yahoo does not mention losses.

    I am at a loss to explain why, given our supposedly sophisticated financial markets and information systems, the data is so opaque and seemingly inconsistent on such a fundamental matter. My hunch, however, is that as much of the discrepancy has to do with the reconstitution of the index than it does with actual earnings.

  36. Gravatar of Integral Integral
    30. May 2014 at 00:23

    Sentences to make your blood run cold: “Fed’s George wants rate hikes soon, and not too gradual.”

    http://www.reuters.com/article/2014/05/30/us-usa-fed-george-idUSKBN0EA05E20140530

    Worth a mention, Scott.

  37. Gravatar of ssumner ssumner
    30. May 2014 at 02:35

    Matt, I’d be very surprised if we had a recession this year.

    Everyone, It seems to me there are two possibilities:

    1. The S&P is wildly unrepresentative, despite being 75% of the market.

    2. The BEA defines profits differently from the corporate sector.

    If the answer is #1, then I suspect that something weird is going on, like massive write-offs for tax purposes in early 2014, which have nothing to do with profits on current operations.

  38. Gravatar of ssumner ssumner
    30. May 2014 at 02:38

    Thanks Integral.

  39. Gravatar of Emad Mostaque Emad Mostaque
    30. May 2014 at 02:43

    EPS for the Russell 2000 has been the following historically on a quarterly basis (source: Bloomberg, Russell):

    Q1 14: 6.08
    Q4 13: 6.69
    Q3 13: 7.75
    Q2 13: 8.04
    Q1 13: 8.10
    Q4 12: 7.75
    Q3 12: 7.50
    Q2 12: 7.48
    Q1 12: 7.23
    Q4 11: 8.24
    Q3 11: 6.96
    Q2 11: 7.23

    EPS expectations for this quarter on Bloomberg consensus are for 10.44 this quarter, 12.27 Q3, 13.36 Q4 (!)

    Top line growth is expected to be 3% per quarter in this period, so this is mostly margin expansion. To put this in context, Q2 12 Sales per share were 226, Q1 14 232

    While there is turnover in the Russell 2000, the bottom up picture looks very similar.

  40. Gravatar of Mark A. Sadowski Mark A. Sadowski
    30. May 2014 at 03:37

    Eric Dennis,
    “Have you discussed or formally written up your methodology anywhere? Is it driven by qualitative judgments, or is it more formal/quantitative?”

    I briefly discussed my methodology in the comments section here back in July when I released the results of my first nowcast. It is purely quantitative and it is relatively simple.

    I use the available monthly data to construct bridge equations relating the major components of NGDP to the monthly data. I then estimate current quarter GDP using the monthly data. Where actual data is available (e.g. PCE), I simply use that, and where there is no data at all (e.g. month three of a quarter for a given time series at monthly frequency), I simply extrapolate based on the existing trend.

    As I said the model is relatively simple. Whereas the most sophisticated nowcasting current quarter GDP models divide GDP into as many as 200 components, I only divide GDP into nine. Whereas the major modelers are using ARIMA, I am only using OLS.

    Conceptually the major difference between what I do and what the major modelers appear to be doing is that I nowcast NGDP first and derive my RGDP nowcast from that, whereas the major modelers appear to be nowcasting RGDP first and derive their NGDP nowcast from their RGDP nowcast.

    I started doing this mostly because I was disatisfied with the accuracy of the major modeler nowcasts. I am genuinely surprised at how well my simple model has performed in the past year.

  41. Gravatar of Ironman Ironman
    30. May 2014 at 04:31

    Vivian Darkbloom, reconstitution of the index was my initial thought too, but that was something that happened in June 2013 – not December when the major shift in P/E happens.

    Even so, what we see of the index’ TTM P/E over 2013 shows it rising from a level of 25 fairly steadily to about twice that level, followed by a major shift upward to 80 in mid-to-late December 2013. Since, it has floated around that elevated level. Since a TTM represents a rolling measure, what that tells us is that higher earnings were progressively replaced by much lower earnings in the P/E measure – dramatically so in December 2013.

    I can make the argument that the previous year’s earnings were highly elevated as part of the “Great Dividend Raid of 2012” – where companies pulled dividends from 2013 into 2012 to avoid the fiscal cliff tax hikes set to occur in January 2013. Those actions would cause the level of reported earnings in late 2012 to rise substantially, which would then drop out and lower earnings once the anniversary of the raid passed. In the meantime, those elevated earnings artificially lowered the Russell’s P/E until they dropped out.

    Now, here’s the thing. The fiscal cliff tax deal from 3 January 2013 created a very strong incentive for the owners of these companies to shift their compensation so that they received a larger portion of it in the form of dividends rather than as salaries. Recorded earnings would rise in response, since the transfer of funds from paying salaries (cost) to dividends would drive that effect – even if revenues didn’t rise. That’s indeed what we observe with large cap companies – there’s no reason to think the small cap companies behaved any differently – if anything, since their owners would have more influence over their dividend policies, they would be more likely to take that action.

    So we shouldn’t see such a large collapse in earnings in the Russell 2000 in December 2013 unless their earnings likewise collapsed at that time. The data for the whole market shows that the number of companies acting to cut their dividends began to spike at this time (that’s the whole market data we referred to earlier), where we identified Real Estate Investment Trusts as being among the hardest hit at the time. Simply put, they acted to cut their dividends because QE-taper related interest rate hikes had so negatively impacted their earnings they could no longer sustain their dividend payments.

    We think it’s the confluence of these things that caused the Russell 2000’s P/E ratio to spike up to 80. That its still there points to an elevated level of economic distress for these smaller cap companies.

  42. Gravatar of Vivian Darkbloom Vivian Darkbloom
    30. May 2014 at 05:46

    Ironman,

    Thanks for the thoughtful response.

    As regards the June reconstitution and ’round about December drop in earnings, it would be my assumption that earnings trail (“rolling earnings” as you put it). In other words, earnings that are reflected and reported in the Russell 2000 would not be altered until the next quarterly period—say, September 2013, at earliest. This is complicated by the fact that not all companies have the same accounting year. Or, is it your understanding that when the index is reconstituted, earnings are *immediately* updated to reflect the trailing earnings of the newly constituted index (i.e., of earnings of companies added to the index even though those earnings were reported before they joined the index). You will see, for example, that per the earnings reported in the prior comment by Emad Mostaque, there was already a significant drop from the 2nd to 3rd quarter of 2013.

    I think it is a mistake to try to identify one cause for most of these type of phenomena, because there are often many, but I continue to think reconstitution of the index likely has something to do with the (unusual) drop in earnings of that index. I agree that tax considerations could be another and that it is likely a confluence as you appropriately stated.

  43. Gravatar of Ironman Ironman
    30. May 2014 at 05:47

    Emad Mostaque, thanks for tracking the EPS data down! Now, we just need to know the approximate number of shares that floating about in each quarter!…

  44. Gravatar of James in London James in London
    30. May 2014 at 05:47

    Confusion certainly reins out there with 10 year bond yields back on the floor, yet S&P500 shooting ever higher. Maybe it is just nirvana, strong growth with permanently low inflation?

  45. Gravatar of James in London James in London
    30. May 2014 at 05:53

    OK, Scott, not “on the floor”, at a 12 month low, and still well above the 1.5%-2.0% range seen for two years between Summer 2011 and Summer 2013. However, it is still strange to see the 10 year yield falling and the S&P hitting record highs. Unless it is nirvana.

  46. Gravatar of ssumner ssumner
    30. May 2014 at 09:06

    Thanks Emad, That’s more like what I would have expected.

    James, See my new post.

  47. Gravatar of JP Koning JP Koning
    31. May 2014 at 03:30

    If you look at Russel 2000 EBITDA per share over the last few years, you don’t get the same weakening trend that the EPS numbers show. Which means that increased depreciation has probably been responsible for the shrinking Russel 2000 earnings numbers (and its rising PE). This is consistent with the introduction of special programs during the post-credit crisis period designed to allow small firms to save on taxes by accelerating depreciation. With these programs expiring, small firms will no longer be able to protect their earnings with depreciation shields… so going forward we’d expect the Russel 2000 EPS number to rise, which is also what the consensus seems to be forecasting.

    (A good reminder to be wary of earnings and focus on cashflow.)

  48. Gravatar of ssumner ssumner
    31. May 2014 at 05:05

    Thanks JP.

  49. Gravatar of Andrew Andrew
    31. May 2014 at 08:08

    Is there anywhere to follow Mark besides this comment section?

  50. Gravatar of Mike Sax Mike Sax
    1. June 2014 at 10:44

    Kevin A, I find the foibles of those pointing out Pketty’s alleged foibles far more impressive. Like in that FT piece, Chris Giles makes the quite serious charge of ‘cherry picking’ while doing some much more troubling data picking of his own in the British data that allegedly shows that there’s been no increase in British inequality since 1980 and might even have been a contraction in it.

    http://diaryofarepublicanhater.blogspot.com/2014/05/thomas-piketty-and-his-critics.html

  51. Gravatar of Major-Freedom Major-Freedom
    1. June 2014 at 11:04

    Mike, Piketty’s book is so full of errors that even progressive and liberal economists are publicly correcting it.

    I chuckled after reading you say “alleged” foibles, as if the book is pure perfection and without any errors. The cognitive dissonance is palpable. It is obvious you are only defending it because its conclusions and hatred of economic freedom fit into your ideological narrative.

    Ha, “alleged.”

  52. Gravatar of Vivian Darkbloom Vivian Darkbloom
    1. June 2014 at 12:30

    JP Koning,

    I think you may be overstating the case that tax depreciation programs are depressing earnings reported for financial statement purposes.

    While it is true that Congress enacted several accelerated depreciation programs and expensing enhancements under section 179 after the crisis, and these programs would disproportionately affect small businesses (due to the caps involved), we should keep in mind that these accelerated provisions, while they directly affect tax accounting, do not directly affect financial accounting (the latter is the basis on which companies file their financial statements). It may be true that these businesses did accelerate investment in 2013 in order to save money on taxes. Thus, I wonder if you could walk us through the accounting consequences you believe flowed from these tax code provisions and how that would have affected reporting for financial statement purposes?

    Also, the following was reported by Bloomberg recently:

    “The Russell 2000 last month reached a valuation of 10.8 times its members’ annual earnings before interest, taxes, depreciation and amortization, according to data compiled by Bloomberg. That was the highest since at least 1995 and compares with an average weekly ratio of 7.7 times Ebitda, the data show. The valuation was at 10.2 at the end of last week.”

    http://www.bloomberg.com/news/2014-04-27/hedge-funds-short-small-caps-most-since-04-russell-falls.html

    Nota bene that the reference above is to EBITDA. Thus, while I think the tax story may explain some investment brought forward due to tax law changes (and scheduled expiration), the above indicates historically high ratios to EBITDA which can’t be explained by tax law provisions alone,

  53. Gravatar of Major-Freedom Major-Freedom
    2. June 2014 at 07:31

    I think the most easily understandable set of ideas that can enable an understanding that EMH is flawed, can begin with a recognition that market prices encompass information from “active” investors, meaning investors who believe that value can diverge from observable price, meaning investors who reject EMH.

    Therefore, given that market prices are in part (I would say the most part, but it is actually not important just how much; as long as it is positive and non-zero) a function of anti-EMH investor expectations, it follows that pro-EMH intellectuals must concede that if market prices provide the most accurate and up to date information as per EMH, then taking the fact that market prices contain anti-EMH expectations, implies that the most accurate and up to date information includes anti-EMH information.

    This logic of course does not work the other way (which is yet another positive sign for anti-EMH), because to grant that market prices are in part a function of pro-EMH investor expectations, is still to concede anti-EMH, since anti-EMH doesn’t claim to be a universal explanation of market prices, only at most a “sometimes” explanation. After all, anti-EMH just means “I reject EMH as stated.” Anti-EMH allows for occasions where prices and value perfectly overlap. Pro-EMH on the other hand is absolute (universal). That is what anti-EMH rejects. It rejects the absolutism of EMH.

    In short, the conclusions of EMH contradict the premises of EMH.

    Active investors can make above average returns by trading with other active investors who price securities using inferior models of valuation. Does this imply that one and the same investor has to forever be immune from below average or average returns? Certainly not! For that active investor has no guarantee that other active investors will discover and utilize superior models of valuation. All anti-EMH promises is that active investing CAN reap above average returns, if one’s valuation models are superior to other active investor’s valuation models. It would be absurd to believe that all active investors have equally good valuation models.

  54. Gravatar of Major-Freedom Major-Freedom
    2. June 2014 at 07:34

    “…no guarantee that othwr active investors will FAIL to discover and utilize superior models of valuation.”

  55. Gravatar of Daniel Daniel
    2. June 2014 at 12:29

    Major_Moron is too stupid to understand the EMH.

    In other news, the sun rises in the east and water is wet.

  56. Gravatar of Major-Freedom Major-Freedom
    2. June 2014 at 12:53

    Daniel:

    You’re still immature, and still without any subatantive rebuttal.

    In other news, the day ends with a “y.”

  57. Gravatar of Daniel Daniel
    2. June 2014 at 14:06

    given that market prices are in part […] a function of anti-EMH investor expectations

    So if people don’t believe in something, that something is false ?

    Seeing as how Austrian economics is a marginal cult inhabited by loons like yourself, your logic says is must be false (stopped clock, twice a day).

  58. Gravatar of Daniel Daniel
    2. June 2014 at 14:11

    By the same logic

    Some people believe the world is flat. Their very existence disproves the roundness of the earth, since if the world was indeed round everybody would agree it is.

  59. Gravatar of Major-Freedom Major-Freedom
    2. June 2014 at 15:20

    Daniel:

    “So if people don’t believe in something, that something is false ?”

    I don’t see how that would be implied. I’m just talking about the information contained in prices. All beliefs that people have and utilize to affect bids and asks, no matter how factually accurate they are, become objectified in prices. If people don’t believe X, then prices will not contain information to the contrary.

    “Seeing as how Austrian economics is a marginal cult inhabited by loons like yourself, your logic says is must be false (stopped clock, twice a day).”

    Not an argument. This is just what you need to say to give you your addict fix. But unfortunately the high doesn’t last long, which is why you need to do it often. Your loser life, not mine.

    “By the same logic”

    You haven’t shown any understanding of the logic.

    “Some people believe the world is flat. Their very existence disproves the roundness of the earth, since if the world was indeed round everybody would agree it is.”

    That isn’t the same logic at all.

  60. Gravatar of Daniel Daniel
    3. June 2014 at 01:45

    How is the logic different ?

    You’re saying that people disbelieving in EMH somehow falsifies it, because they act as if it is in fact false.

    So if I behave as if the Earth was flat, then that means the Earth is indeed flat.

    If I behave as if you are moron, then you indeed are (well, that one is true).

    I think I’m starting to like this logic.

  61. Gravatar of Major-Freedom Major-Freedom
    3. June 2014 at 03:19

    Daniel:

    “How is the logic different ?”

    “You’re saying that people disbelieving in EMH somehow falsifies it, because they act as if it is in fact false.”

    EMH is false because it self-contradicts. EMH makes a claim about prices that is inconsistent with EMH.

    EMH is not false merely because people believe it is false.

    Think of it this way. If I said that “People believe the world is round”, then that is falsified by the existence of people who believe otherwise. This is not an argument that the world is flat merely because people believe it so. The pointing to the belief is just a reference to a fact that falsifies the claim.

  62. Gravatar of Daniel Daniel
    3. June 2014 at 04:31

    EMH is false because it self-contradicts.

    Then Misesian bullsh*t is false, because it self-contradicts. And the people who buy into it must be irrational (that’s an euphemism for stupid).

    EMH makes a claim about prices that is inconsistent with EMH.

    Please show me the people who got rich by regularly beating the market.

    Some people were dropped on their heads. You were thrown against a wall.

  63. Gravatar of Daniel Daniel
    3. June 2014 at 04:41

    I’m still struggling to grasp this one

    “some people act as if the EMH is false, therefore it must be false”

    What does that even mean ? Is this the best kind of criticism against the EMH people have ?

    Is that how theories are disproved ? By listening to your gut feelings about them ?

    I feel like I’m losing brain cells just by interacting with morons like you.

  64. Gravatar of Major-Freedom Major-Freedom
    3. June 2014 at 08:24

    Daniel:

    “EMH is false because it self-contradicts.”

    “Then Misesian bullsh*t is false, because it self-contradicts. ”

    How? You haven’t shown it self-contradicts. Do you even know what that means?

    “And the people who buy into it must be irrational (that’s an euphemism for stupid).”

    This is a contingent statement predicated in a claim that you have not substantiated.

    “EMH makes a claim about prices that is inconsistent with EMH.”

    “Please show me the people who got rich by regularly beating the market.”

    I do not claim that one and the same person must get rich by regularly beating the market.

    “Some people were dropped on their heads. You were thrown against a wall.”

    Some children were emotionally and/or physically abused during their upbringing. Which appky to you? I’m guessing the psychological torment you went through was relatively pronounced. You sound exactly like a victim of bullying “taking ownership” of bullying, because it’s the only method you have been taught to use when there are disagreements.

    This is likely why you hate Austro-libertarianism so much. It calls for removing the figure of mommy and daddy that you never had growing up, the government.

    “I’m still struggling with this one”

    “some people act as if the EMH is false, therefore it must be false”

    I never made that claim.

    “What does that even mean ? Is this the best kind of criticism against the EMH people have ?”

    You mean your imaginary straw men?

    “Is that how theories are disproved ? By listening to your gut feelings about them ?”

    Project much?

    “I feel like I’m losing brain cells just by interacting with morons like you.”

    Not an argument.

  65. Gravatar of Bitcoin Faucet Rotator Blog It was the best of times, it was the worst of times Bitcoin Faucet Rotator Blog It was the best of times, it was the worst of times
    10. April 2015 at 05:43

    […] Scott Sumner has pointed out twice now, the huge decline in GDI is almost entirely due to a fall in corporate profits. Whereas […]

Leave a Reply