This page has been archived and commenting is disabled.

San Francisco Fed: "A Recessionary Relapse Is A Significant Possibility Sometime In The Next Two Years"

Tyler Durden's picture




 

Presented without commentary, suffice to say that idiots who look at the LEI and factor for the curve inversion subindicator deserve all they get. This weekend Albert Edwards confirmed that the LEI, absent the treasury curve data input (which obviously can not invert now that even the 2 Year is trading below 50 bps, although as the chart below shows is getting progressively flatter, after today alone it has hit a new post September 2009 tight) already confirms we are in a recession. We hope FRBSF paper authors Travis Berge and Oscar Jorda read Mr. Edwards piece debunking the entire bullish side of their argument.

First, today's 2s10s, which indicates the LEI's curve indicator will send progressively more recessionary signals, not to mention that profit margins for mortgage originators continue to collapse:

And now, the Fed paper:

Future Recession Risks

By Travis J. Berge and Òscar Jordà

An unstable economic environment has rekindled talk of a
double-dip recession. The Conference Board's Leading Economic Index
provides data for predicting the probability of a recession but is
limited by the weight assigned to its indicators and the varying
efficacy of those indicators over different time horizons. Statistical
experiments with LEI data can mitigate these limitations and suggest
that a recessionary relapse is a significant possibility sometime in the
next two years.

By now, there is little disagreement that the Great Recession, as the
last recession is often called, ended sometime in the summer of 2009
(see Jordà 2010), even though the National Bureau of Economic Research
(NBER) has yet to formally announce the date of the trough in economic
activity that marks the beginning of the current expansion phase.
Intriguingly, just as we seemed to be leaving the recession behind, talk
of a double dip became increasingly loud. This recession talk is not
confined to the United States. It has crossed the Atlantic to Europe,
where the recovery has been even slower, especially among countries on
the periphery of the euro area. A quick look at the number of Google
searches and news items for the term "double-dip recession" reveals no
activity prior to August 29, 2009, but a dramatic increase in search
volume since then, especially in the past two months. Such concern is
likely motivated by a string of poor economic news. The spring of 2010
saw considerable declines in U.S. stock market indexes, the contagion of
the Greek fiscal crisis across much of southern Europe, and a stagnant
U.S. labor market stuck near a 10% unemployment rate. It is
understandable that the NBER has hesitated to call the end of the
recession.

This spate of bad news has prompted a heated policy debate pitting
those eager to mop up the gush of public debt generated by the recession
and the fiscal stimulus package designed to counter it against those
who would prefer to douse the glowing recession embers with another
round of stimulus. Domestic and international commentators have engaged
in a lively debate on this subject in the press and blogosphere. The New York Times, Washington Post, Wall Street Journal, Financial Times, and Economist have all featured one or more stories about a possible recessionary relapse in the past few months alone. In this Economic Letter, we calculate the likelihood that the economy will fall back into recession during the next two years.

The Leading Economic Index

The Leading Economic Index (LEI) prepared by the Conference Board (www.conference-board.org/data/bci.cfm)
every month is an indicator of future economic activity designed to
signal peaks and troughs in the business cycle. It comprises ten
variables that can be loosely grouped into measures of labor market
conditions (initial claims for unemployment insurance and average weekly
hours worked in manufacturing); asset prices (the monetary aggregate
M2, the S&P 500 stock market index, and the interest rate spread
between 10-year Treasury bonds and the federal funds rate); production
(new orders of consumer and capital goods, new housing units, and vendor
performance); and consumer confidence.

Figure 1
The Leading Economic Index (LEI)

The Leading Economic Index (LEI)

Note: Gray bands denote NBER recessions.

Figure 1 displays year-on-year LEI growth rates against recession
periods determined by the NBER, showing the strong correlation between
the two. But, in a recent paper, Berge and Jordà (2010) find that the
LEI is no better than a coin toss at predicting turning points beyond 10
months into the future, with most of its success concentrated in the
current month. Other popular indexes of economic activity, such as the
Chicago Fed National Activity Index and the Philadelphia Fed's
Aruoba-Diebold-Scotti Business Conditions Index, turn out to have even
less predictive power than the LEI.

At least two reasons explain why the LEI's predictive efficacy is
limited. The first is that the index is a one-size-fits-all weighted
average of indicators. By this we mean that weights are designed to
distill the information contained in 10 variables into a single
variable, rather than by selecting weights that would produce the most
accurate turning-point predictions. Second, we find that no single
combination of indicators is likely to predict well at every time
horizon. The predictive ability of each LEI component varies wildly
depending on the forecast horizon. For example, the spread between
10-year Treasury bond and the federal funds rate works best 18 months
into the future, whereas the initial claims for unemployment insurance
indicator works best two months ahead. Clearly, one should give more
weight to the rate-spread indicator than the initial claims indicator
when forecasting in the long run, but less weight when forecasting in
the short run.

A better forecasting approach

We are interested in predicting a binary outcome: Will the economy be
expanding or contracting at a particular future date, given what we
know today? One way to summarize the likelihood of each of these
outcomes is by taking the ratio of the probability of each. This
"odds-ratio," as it is called in statistics, is equal to one when both
outcomes are equally probable, less than one when a recession is more
likely than an expansion, and more than one when expansion is more
likely than recession. In the context of this either-or condition, the
statistical relationship between the odds-ratio and the LEI variables is
used to characterize the probability of recession. As an illustration,
we use this procedure to predict the probability of recession from 1960
to 2010 using contemporaneous LEI data. These are compared with the NBER
recession periods displayed as shaded gray areas in Figure 2.

Figure 2
Probability of a recession using the LEI in real time

Probability of a recession using the LEI in real time

The similarity between the predicted recession dates in this exercise
and the actual NBER recession dates is quite striking, but perhaps not
entirely surprising. Consider the following rule of thumb: call a
recession whenever the predicted probability of recession is above 0.5;
otherwise call an expansion. Such a rule would achieve a nearly perfect
match with the NBER's delineation of expansions and recessions, with
some slight discrepancy in the mid-1960s. A 0.5 cutoff is equivalent to
saying that the odds of a recession are the same as the odds of an
expansion or that the odds-ratio is 1.

The odds-ratio and LEI indicator combination addresses the first of
the two issues we identified when making turning-point predictions with
the LEI, namely the problem of determining appropriate weights for each
indicator. The solution to the second issue, the differential predictive
power of the indicators at different time horizons, is rather simple.
It consists of finding the best combinations of indicators associated
with the odds-ratio between expansion/recession outcomes at increasingly
distant future dates. Finding the best combination at each month over
the next two years generates 24 different combinations of LEI
components. Figure 3 uses this approach with data up to June 2010 to
display the probability of recession for each month starting in June
2010 and ending in June 2012. The horizontal line at 0.5 coincides with
the value at which the odds of an expansion and the odds of a recession
are even, making it a natural cutoff for the probability of a binary
outcome.

Figure 3
Probability of a recession over the next two years

Probability of a recession over the next two years

Figure 3 displays the predicted probability of recession obtained
using these procedures for three experiments. The first experiment is
the benchmark case and uses all ten components of the LEI. It is
represented by a thin blue line in the figure and shows that the
likelihood of a recession is essentially zero over the next 10 months
but that the odds deteriorate considerably over the following year.
However, even at its worst, the probability of recession is never above
0.3, so that expansion is more than twice as likely as recession. Paul
Samuelson once quipped that, "It is true that the stock market can
predict the business cycle. The stock market has called nine of the last
five recessions." Therefore we investigate whether our results are
driven by the recent declines in the S&P 500 index. However,
repeating the previous forecasting exercise while excluding the S&P
500 variable generates essentially the same picture, displayed by the
dashed line in Figure 3.

The last experiment drops the spread between the Treasury bond and
the federal funds rate from the 10 LEI indicators. Historically, this
spread, which summarizes the slope of the interest rate term structure,
has been a very good predictor of turning points 12 to 18 months into
the future. Specifically, an inverted yield curve has preceded each of
the last seven recessions.
However, the term structure may not presently
be an accurate signal. Monetary policy has been operating near the zero
lower bound to provide maximum monetary stimulus. In addition, the
Greek fiscal crisis has generated a considerable flight to quality that
has pushed down yields on U.S. Treasury securities. Indeed, the thick
red line in Figure 3 shows that omitting the rate-spread indicator
generates far more pessimistic forecasts. For the period 18 to 24 months
in the future, the probability of recession goes above 0.5, putting the
odds of recession slightly above the odds of expansion.

Conclusion

Any forecast 24 months into the future is very uncertain. At two
years out, the odds of recession vary from almost three times more
likely than expansion, to expansion being almost five times more likely
than recession, depending on which LEI components are used.
Nevertheless, LEI forecast trends indicate that the macroeconomic
outlook is likely to deteriorate progressively starting sometime next
summer, even if the data suggest that a renewed recession is unlikely
over the next several months.
Of course, economic policy can strongly
influence the outcome. The policies that are adopted today could play a
decisive role in shaping the pace of growth.

 

- advertisements -

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Mon, 08/09/2010 - 15:03 | 511186 Cursive
Cursive's picture

Of course, economic policy can strongly influence the outcome. The policies that are adopted today could play a decisive role in shaping the pace of growth."

This is not news other than to show that these idiots can't forecast or control anything, but think they can still control the Ponzi their fore bearers unleashed.

Mon, 08/09/2010 - 15:47 | 511308 Popo
Popo's picture

Agreed. Anyone can see they have failed at 100% of their forecasts, and their only tool has failed them.  History will be cruel to these imbeciles I suspect. Their ultimate failure is their belief that they have control.  What incredible delusions of grandeur these college professors have.

 

 

 

 

Tue, 08/10/2010 - 04:22 | 512118 ATG
ATG's picture

"the likelihood of a recession is essentially zero over the next 10 months"

Oh yeah?

All the thousand plus Fed economists laid end to end may still not reach a correct market conclusion or decision...

Mon, 08/09/2010 - 15:03 | 511189 LoneStarHog
LoneStarHog's picture

God these clowns are such a waste of good Carbon-based material.

Mon, 08/09/2010 - 15:04 | 511191 Bankster T Cubed
Bankster T Cubed's picture

relapse, eh?   what a bunch of assholes

Mon, 08/09/2010 - 15:05 | 511196 truont
truont's picture

A "recessionary relapse" within the next two years?

It's happening right now, you econotards!

Mon, 08/09/2010 - 15:18 | 511233 No More Bubbles
No More Bubbles's picture

Exactly what I was going to say, but you beat me to it.

Of course, it really is going to be RECOGNITION of the DEPRESSION we've already been in for 2 years.

They were not referring to the "Great Depression" as such in the 1930's, that description came later, much later.......

Tue, 08/10/2010 - 05:49 | 512165 mephisto
mephisto's picture
San Francisco Fed: "A Recessionary Relapse is a Significant Possibility Sometime in the Past 3 Months".

There, fixed it for you.

 

Mon, 08/09/2010 - 15:09 | 511211 bruce wayne
bruce wayne's picture

GDP goes from 5.7% to 3.7% to (assuming JPM is right) 1.7% within six months but there is no chance it will be negative within the year?  Ignore the straight line down (kinda like the ECRI WLI...)

Mon, 08/09/2010 - 15:10 | 511212 Mongo
Mongo's picture

Does a bear crap in the woods?

Mon, 08/09/2010 - 15:16 | 511231 israhole
israhole's picture

So, it'll happen sometime after May next year.  

They've gotten everything else right, why not this?

Mon, 08/09/2010 - 15:21 | 511240 wowser22
wowser22's picture
Ben Bernanke, Please Send Me Some Green! :

http://www.youtube.com/watch?v=vqSfnH51KX8


Mon, 08/09/2010 - 16:52 | 511461 Pedro
Pedro's picture

Send me some green too.  I need to reload on my puts for next spring, since this falls puts may end up worthless (again).

Mon, 08/09/2010 - 15:24 | 511250 Cult_of_Reason
Cult_of_Reason's picture

same perma-bulls and perma-cows who say all is well, now pop up on cnbc to demand the immediate nuclear qe2 option from the fed tomorrow. the say, "all is well and dow is going to 36,000; but we need the fed to start buying treasuries immediately to avoid the imminent double dip recession and unemployment above 10%".

P.S. Rick Santelli for President!

Mon, 08/09/2010 - 15:25 | 511257 Mad Mad Woman
Mad Mad Woman's picture

More idiotic drivel from the idiots at the Fed.

Mon, 08/09/2010 - 15:26 | 511258 cougar_w
cougar_w's picture

next two years

Translation: Don't worry about it, go shopping.

Mon, 08/09/2010 - 17:04 | 511471 MarketTruth
MarketTruth's picture

Many millions are doing just that... with their food stamps, which is run/operated by the Federal Reserve's retail arm JP Morgan no less! Seriously.

Mon, 08/09/2010 - 15:30 | 511267 Screwball
Screwball's picture

People actually get paid to write this shit?  Unbelievable.

Mon, 08/09/2010 - 15:33 | 511276 Turd Ferguson
Turd Ferguson's picture

Someone posted this earlier. I re-post it below. An interview with John Williams.

http://www.theenergyreport.com/pub/na/7005

Rather than waste time with these sell-side douchebags, print this interview and keep it for posterity.

Mon, 08/09/2010 - 15:42 | 511299 win
win's picture

Remember that recession that started in Dec 2007? And was over in July of 2009?

Yea, well these are the same guys that:

a) didn't see it coming

b) did not discover it until Dec 2008 - a year after the fact

I expect to hear very shorty that the recovery of July 2009 never really happened

There guys are
very bad at predicting the future
 worse at predicting the past.

Mon, 08/09/2010 - 15:45 | 511304 stoverny
stoverny's picture

Whatever, Dow's up another fifty and Zero Hedge is running Jim Cramer banner ads.  ITEOTWAWKI, and I feel fine.

Mon, 08/09/2010 - 16:00 | 511338 truont
truont's picture

The more we type Cramer in the blog & comments, the more we rant about Cramer and repeat Cramer's name over and over, the more likely the Google algo-ads will push Cramer advertising to ZH. 

The algo-ads cannot distinguish between positive Cramer comments and negative Cramer contents.  The algo-ads just count the "hits"--they just count how many times Cramer's name is mentioned.

(Cramer count=8)

Mon, 08/09/2010 - 18:00 | 511542 cougar_w
cougar_w's picture

Ads are served locally, the javascript reads the incoming HTML and sends the results back to Google, which puts ads in the slots. Someone needs to write a FF extension that finds all instances of "Cramer" and changes the speeling, to say "Krammer" or "Cramersky" something.

OK all you twitchers out there, get on it. The revolution needs you!

Mon, 08/09/2010 - 20:06 | 511694 Citizen of an I...
Citizen of an IKEA World's picture

Isn't the idea to pull Cramer ads into this website, where Cramer is understood as a laughing stock whoretool, and make Cramer pay for the exposure to people who despise him*?

 

* Despise Cramer, that is.

 

(Including this note, Cramer count=5.)

Mon, 08/09/2010 - 20:55 | 511746 New_Meat
New_Meat's picture

Does this mean that Cramer is paying Google (of course) which is in turn paying ZH?

We need to talk that book!  And Marla benefits!

Cramer++;

- Ned

Mon, 08/09/2010 - 15:51 | 511325 banksterhater
banksterhater's picture

All Fed clowns, current & retired should be made to shut the F up! They have meetings for a reason, they all want 15 min of fame, that Angel has to be the #2 moron behind G-Spasm.

Mon, 08/09/2010 - 15:55 | 511333 Dr. Sandi
Dr. Sandi's picture
San Francisco Fed: "A Recessionary Relapse Is A Significant Possibility Sometime In The Next Two Years"

...which then goes on to predict that the sun could rise in the east at least once during the same time frame.

Okay, I made the second part up. They probably don't have the resources for a prediction of that complexity.

Mon, 08/09/2010 - 15:55 | 511337 plocequ1
plocequ1's picture

Dow is up. 6 minutes to Chartfest

Mon, 08/09/2010 - 16:02 | 511355 redarrow
redarrow's picture

Which means that there will be no QE tomorrow. They will not want to waste the last bullet before more evidence of slowing hits. At the moment all that may happen is that there will be more pledges of holding down interest rates for an extended period. rollover of the maturing bond payments and thats it. 

Mon, 08/09/2010 - 16:42 | 511448 banksterhater
banksterhater's picture

And the rumor about not paying interest on reserves is bullcrap because the banks said they'll just put the money into Treasuries, there's little demand from qualified borrowers for loans.

Mon, 08/09/2010 - 16:06 | 511371 digalert
digalert's picture

Someones wandering from the reservation, must act now. Perhaps a ride in Air Force One will snap them back in line.

Mon, 08/09/2010 - 16:39 | 511440 reading
reading's picture

That's funny, Timmy G just said the recovery was going great!  Better than expected in many areas... hmmm who should I believe?  I mean they're all so damn smart...

Mon, 08/09/2010 - 19:18 | 511637 Quinvarius
Quinvarius's picture

How many times does the Fed have to repeat we are near collapse before the market starts to trade off of that observation?

Mon, 08/09/2010 - 20:57 | 511749 New_Meat
New_Meat's picture

I'm looking at the figures, particularly the gray "recession" stripes.  I don't think that the NBER has pronounced the end of the '07 recession. 

In fact, I'm sure that they have not predicted the winner of the '94 World Series.

- Ned

Tue, 08/10/2010 - 08:15 | 512245 gridlocked
gridlocked's picture

What will this be recession squared?

Do NOT follow this link or you will be banned from the site!