Use access key #2 to skip to page content.

More On The ECRI Leading Indicator

Recs

14

September 17, 2010 – Comments (8)

Like I have talked about in a number of posts recently, when you look at leading macro indicators, the sustainability of this recovery is called into question. Most headline indicators that the crowd is focused on are either coincident or lagging.

-----------------------------------

MORE ON THE ECRI LEADING INDICATOR
17 September 2010 by Comstock Partners

http://pragcap.com/more-on-the-ecri-leading-indicator

Last week, toward the end of our comment on consumer deleveraging, we mentioned that the year-over-year change in the ECRI Weekly Leading Indicator had strongly suggested the distinct possibility of recession.  In answer to some questions about it, we would like to provide a bit more detail this week.

When we were writing last week the latest release showed the indicator declining 4.11% from a year earlier.  We therefore searched the historical data to determine what happened to the economy at other times when the indicator had fallen by that amount or more.  We found that over the last 42 years this has occurred seven times, and in all seven instances a recession started shortly before or after the signal.  This week we re-examined the data, except that this time we looked for a decline of 3.50% or more, and found that the lead times were even better in four of the seven occurrences.

We can make a number of observations from the data.  In all seven instances where the index fell 3.5% or more from a year earlier a recession occurred shortly before or after the signal.  There were no occasions where the index declined 3.5% without a recession.  In two cases the signal led the recession, in three cases it followed, and two times it occurred in the same month.  It ranged between a five-month lead and four-month lag. The average and median lead times were zero.  In all instances the market had peaked before the signal, anywhere from one to ten months, with an average of five and a median of two.  We note again that ECRI Managing Director Lakshman Achuthan has not officially called a recession, although he has stated that, based on his indicator, there was more than a 50% chance of one.

Although we would not rely on any single indicator to form an opinion, the ECRI Leading Index strongly supports our view as discussed extensively in prior comments that the economy, at best is headed for a severe slowdown, and, at worst,  another recession.  It also makes it much more likely that the April peak in the S&P 500 will turn out to be the 2010 high, and that the performance of the economy in the period ahead will be highly disappointing to investors looking for a normal economic recovery.

8 Comments – Post Your Own

#1) On September 17, 2010 at 1:38 PM, rofgile (99.31) wrote:

Really, its a leading economic indicator, Binve?

Compare the ECRI vs the S&P index - it looks coincident to me.

From this page: http://www.favstocks.com/ecri-watch-improving/1724858/

 

From your article:

 

"We can make a number of observations from the data.  In all seven instances where the index fell 3.5% or more from a year earlier a recession occurred shortly before or after the signal.  There were no occasions where the index declined 3.5% without a recession.  In two cases the signal led the recession, in three cases it followed, and two times it occurred in the same month."

 - Only in 2 out of 7 calls was the index leading with a signal before a recession.  The other 5 calls were at or after the start of the recession.  That also sounds like a coincident, not leading indicator.

- We are still showing economic growth, and corporations are still increasing earnings.  This doesn't support the ECRI saying we are headed for recession or double dip.

- Based on China's strong growth, moderate inflation, lowering of initial jobless claims, increasing world trade ->  I would say that things are still between pretty decent - to a little slower growth than optimal.

 I guess we'll find out with the 3rd quarter GDP?   When is that released, end of October?  

 -Rof 

Report this comment
#2) On September 17, 2010 at 1:45 PM, binve (< 20) wrote:

rofgile ,

Yes it is. If you bothered to read the article, it is the ECRI Leading Index that is being discussed.

Like I said at the very top of the post: Most headline indicators that the crowd is focused on are either coincident or lagging.. You are looking at the smoothed ECRI, which is a *coincident* indicator.

If you want another souce (that I have provided in previous posts), here it is: http://dshort.com/articles/ECRI-Weekly-Leading-Index.html.

From Doug Short's site:

The question, of course, is whether the latest WLI decline is a leading indicator of a recession or a false negative. The published index has never dropped to the current level without the onset of a recession. The deepest decline without a near-term recession was in the Crash of 1987, when the index slipped to -6.8.

We are currently at -9.2..

Report this comment
#3) On September 17, 2010 at 2:19 PM, rofgile (99.31) wrote:

I'm not questioning that it is called the "ECRI leading index".  I was questioning whether it really lead economic data.  Does it dip before you see gdp dip, same with does it rise before gdp rising?  I think your Doug Short article

In your Doug Short article there is a graph of it vs GDP.

Overall, I would have to agree that it appears to have some predictive value - it appears to lead between 1980-1985 and 2007-2009.  

However, it is interesting to look at how large its swings are between the 2007-present.  I think it is overswinging GDP changes.  In the period from 1980-2005 it matches GDP much better than it has in the last 2007-present period.  It may likely be overswinging on the downside and giving a false signal, Binve.  The downside of the ECRI is more extreme in 2008-2009 by almost 3-fold, same with the upswing from 2009-2010.  This is far different than how the chart looks in the 1980-2005 period.  

I did read your article, I just am challenging your conclusion that the current ECRI downswing is predictive of a downturn. 

 -Rof 

Report this comment
#4) On September 17, 2010 at 2:33 PM, binve (< 20) wrote:

>>, I just am challenging your conclusion that the current ECRI downswing is predictive of a downturn.

Fair enough.

This is not the only leading indicator forecasting weakness

- New orders minus Headline ISM: http://caps.fool.com/Blogs/market-still-deluding-itself/446288

- CMI Growth: http://caps.fool.com/Blogs/what-is-the-consumer-metrics/427662

Report this comment
#5) On September 17, 2010 at 2:47 PM, davejh23 (< 20) wrote:

"In the period from 1980-2005 it matches GDP much better than it has in the last 2007-present period...The downside of the ECRI is more extreme in 2008-2009 by almost 3-fold, same with the upswing from 2009-2010.  This is far different than how the chart looks in the 1980-2005 period."

Absent gov't deficit spending, perhaps it would match much better.  Deficit spending plugged a 10% hole in GDP last year, and as modest private sector growth resumed, continued deficit spending was still running at ~10% of GDP.  It looks like this would be spot on on the downside if you pulled out deficit spending...don't know about the huge upswing...

Report this comment
#6) On September 17, 2010 at 2:49 PM, rfaramir (29.31) wrote:

What rofgile said was: "Compare the ECRI vs the S&P index - it looks coincident to me."

The S&P index itself, as a rough measure of the 'market' is also a leading indicator of the 'economy'.  So if they are coincident, it is no coincidence.  :-)

What measures the 'economy' is GDP, as imperfect a measure as that is.

Report this comment
#7) On September 17, 2010 at 4:36 PM, rofgile (99.31) wrote:

#6 rfaramir:

 That's a good point, the graph in #1 vs #3 is almost an apples-to-oranges comparison.

 #1 is comparing "some annualized smoothed ECRI" which lacks the huge recent swings seen in the actual ECRI of #3.  #1 compares against S&P index, while #3 compares against actual GDP.

 I would totally retract #1 comment of mine from this discussion, since I basically off-base (since I am comparing against S&P - which I agree could be forward looking) and this graph in #1, I don't understand how "annualized smoothed ECRI" is generated from the webpage.  Sorry Binve - I agree that it is a leading indicator and don't debate that.

  #5 davejh23 - I think that's a likely good explanation.  Without govt spending, GDP in the US likely would match well with the ECRI indicator on both the downside *and* upside.  

  For a good example of this, look at countries like Germany and Japan that had far worse GDP hits than the US in the 2007-2010 downturn.  The NYTimes had a nice piece today comparing our downturn versus those countries.  Countries like Germany that had a huge negative GDP, now have huge positive GDPs that make it look like their economies are on fire (but are may be just catching up to baseline).  

  ----

  So now that govt spending is ending... maybe we'll begin to track with ECRI better?  (or is govt spending ending? seems like I hear rumblings of more action).  I am kinda glad we didn't track well with ECRI - we would have had a less moderate downswing and upswing - and the downswing and upswing we did have felt quite extreme to me!

 -Rof 

Report this comment
#8) On September 17, 2010 at 4:38 PM, rofgile (99.31) wrote:

"Countries like Germany that had a huge negative GDP, now have huge positive GDPs that make it look like their economies are on fire (but are may be just catching up to baseline)"

 should read,

 "had a huge negative change in GDP, now have huge positive changes in GDP"..

 -Rof 

Report this comment

Featured Broker Partners


Advertisement