Tracking the recession

Here are links to perspectives from others on where the economy stands at the moment.

Rising delinquencies. Here’s the assessment from Freddie Mac’s 10-Q
(hat tip to Calculated Risk):

We also observed a continued increase in market-reported delinquency rates for mortgages serviced by financial institutions, not only for subprime and Alt-A loans but also for prime loans, and we experienced an increase in delinquency rates for all product types during the first quarter of 2009. This delinquency data suggests that continuing home price declines and growing unemployment are significantly affecting behavior by a broader segment of mortgage borrowers…. Our loan loss severities, or the average amount of recognized losses per loan, also continued to increase in the first quarter of 2009, especially in the states of California, Florida, Nevada and Arizona, where home price declines have been more severe and where we have significant concentrations of mortgage loans with higher average loan balances than in other states.

ADS index. The Philadelphia Federal Reserve has added some features to its website for the Aruoba-Diebold-Scotti Business Conditions Index (the background for which I described here) which make it a bit more accessible. The index currently stands at -2.02. I’d want to see it rise above -1.0 before describing the upward trajectory since January as signaling a clear improvement. For readers’ convenience I’ve also included a thumbnail of the latest graph of the ADS index on our main page. The figure below should be a live link to the latest values any time you reload this page:




Aruoba-Diebold-Scotti Business Conditions Index.



Tim Duy’s not optimistic. Here’s what the always insightful Tim Duy thinks about prospects for recovery:

I suspect we have a long path ahead of us on the structural challenge poised by overleveraged households– suggesting that the green shoots we hear so much about will yield little more than stunted growth.

Music to central bank by. And via Greg Mankiw, check out Merle Hazard and Bretton Wood singing “Inflation or Deflation”:



Stanford Professor John Taylor plays straight man for the act here.

StumbleUponLinkedInReddit

11 thoughts on “Tracking the recession

  1. Robert Bell

    JDH: I’m glad you mentioned Tim Duy. I felt his explanation was a natural complement to your earlier posts about the 2008 oil price shock – i.e. like you he is decomposing the problems into a short term cyclical effect and a long term structural effect.

  2. moneymouth77

    There is a lot to be said for that excellent musical interlude ;) However, the point is made, and amplified, by what the Bank of England has said today about banks not wanting to take any more risks in the face of ongoing recession. We need more definite prognoses!

  3. aaron

    The sharpest part of the oil spike is 08, but the oil shock really started at the beginning of 05 / end of 04.

  4. DickF

    A different perspective.
    I think we are in a very special time…the only really analogous time was the Great Depression and there are many differences. The GVM [index] was based on relationships between the gold price the 10yr note and how they translate into inflation driven price change that would show up in the index measures of price change [CPE].
    I think that all prior historic relationships are breaking down. You will see talk in these threads about the relationship of the oil price to gold, or other commodity, say wheat prices to oil or to gold or steel or whatever. You would always expect that interest rates would bear some relationship with the price of the dollar. All these relationships are deviating from their prior patterns. I believe that the years of currency fluctuation, of inflation followed by deflation, that goes back for decades now, has so compromised the worlds productive infrastructure, its invested capital base, through a chronic condition of under investment in long cycle infrastructure that we no longer have enough capacity to manage supply and demand changes in our production and storage systems. I think this makes the price of commodities volatile for reasons that are not driven by the change in currency immediately or in present time…even though the condition of lack of capacity is a long term cause of inflation/deflation cycles. This is a long winded way of saying the many of the prices in the PCE that make up the index are no longer driven so much by inflation but are now driven by the problems of a decaying production supply chain. I think this is part of the reason that the variance between the PCE and GVM are wide.
    Another reason is the unprecedented Government involvement in the credit markets, with the 10 year and the equally unprecedented market interventions that are driven by National Actors in the markets…notably China in both food and industrial commodities but also a host of other Nations especially with regard to food commodities. There is less capacity in our system and National actors are playing for supply commitments on both an as needed and a strategic basis…this is creating price volatility and unusual moves.
    Because of all this, I look at the trend of the GVM and not the absolute numbers. I do not put special meaning in the drop between April and May. I pay more attention to the continuing decline that goes on and on and the apparent mirroring of that trend in the actual PCE.
    We could experience a continuing price deflation even as energy prices spike upwards…certainly these contrary relationships cannot last forever but they seem likely to occur as our global economic system continues to shudder under the tortures of schizophrenic monetary and fiscal economic policies.

  5. GK

    Initial claims were 637K. It looks like the recovery will not happen before Q309 (which has been my consistent prediction).

  6. Ironman

    GK: Very possibly.
    Earlier today, I looked at what investors are expecting in terms of an economic recovery.
    Although in that post, I looked purely at trailing twelve months (TTM) dividend futures data, I took a step back and looked at the quarter dividend futures data. Here’s what I found – all data is dividends per share (DPS) for the S&P 500, through 14 May 2009:

    Year-Qtr   Qtr DPS    TTM DPS
    2009Q2     5.492      25.652
    2009Q3     5.324      23.926
    2009Q4     5.541      22.317
    2010Q1     5.613      21.970
    2010Q2     5.368      21.846
    2010Q3     5.677      22.199
    2010Q4     5.699      22.357
    

    While the TTM data suggests a bottom in 2010Q2, the quarterly data indicates a recovery in corporate earnings beginning between 2009Q3 and 2009Q4, with a one-quarter dip in 2010Q2. If that holds, I could see the NBER finding that a recovery would have begun in 2009Q3.

  7. aaron

    I would have totally agreed 6 months ago, but after the bank bailout and Obama’s “stimulus”, if there’s a recovery in Q3 it won’t be significant.

  8. Steve Kopits

    We have now begun to debate the shape of the recovery, and that’s a good sign. Whether it begins in June (Q2) or July (Q3) seems a bit academic to me. What matters is the shape.
    Could someone please give us a decent piece on the path of de-leveraging? It is the central issue in the ‘Deep Repression’ argument and I have not seen one decent piece of analysis on it yet.
    More generally, I have to note my deep disappointment with the economics community in this whole crisis. The analysis has been wrong (Greenspan Oct 07: ’1-in-3 odds of recession’); non-existant (who resigned in protest from the Fed or Treasury over growth in US indebtedness?); irrelevant (see Menzie’s post in the last couple of weeks about papers presented at an econ conference); and just plain weak (so, OK, what does the de-leveraging path look like–not in Model 1-to-7 weasel words, but in a ‘here’s-my-view’ form that I can explain to my mother, or heaven forbid, my clients).
    Forgive me, but this crisis has appeared to have exposed either a lack of determination or a lack of technical skill of economists at our government and academic institutions, and that is a luxury none of us can afford.

  9. Stacey Dawkins

    Ironman: Where did you find that table? I had came across TTM dividend forecast data but couldn’t find a quarterly fcast data.

  10. Ironman

    Stacey,
    I calculated the data using IndexArb’s dividend futures data. To find a given quarter’s expected dividends per share payout, you need to subtract the previous quarter’s figure from the figure for the quarter of interest.

Comments are closed.