Double-Dippers Are All Wet: Towel Off Already
Recession watch: Probability of recession for June 2011, based on the yield curve, is 0.2% (see chart above, data here). Probability has been higher than 30% before every recession since 1967.
Professor Estrella is featured in this Bloomberg article "Double-Dippers Are All Wet Ignoring Yield Curve":
"The yield curve has inverted prior to all of the last seven recessions, with no false signals since 1967, according to Estrella, whose website provides all kinds of research and data for the uninitiated.
Estrella uses the monthly average spread between the 3- month Treasury bill and the 10-year Treasury note to filter out the noise. The lead time between the appearance of a negative monthly spread and recession can be anywhere from three to 18 months. In the most recent instance, the spread turned negative in July 2006, and the U.S. economy slipped into recession in December 2007, according to the National Bureau of Economic Research, the official arbiter of the business cycle.
Slowdown, yes; recession, no. That’s the message of the yield curve. Its track record is impeccable. It beats forecasters, econometric models, even the Fed, which seems to resist the inherent message in the spread. For all those double-dippers still splashing around in the pool, it’s time to get out, towel off and learn to love a slow recovery."
HT: Larry Kudlow
12 Comments:
first off, this has been an historically erratic indicator, completely missing the 2001 recession until it was already underway. "it's track record is impeccable?" that's a joke. just the chart attached shows that to be untrue except as a trailing indicator in 2001.
secondly, the bond yield spreads it uses as inputs are currently heavily manipulated. never before has the fed engaged in the sorts of bond purchases at auction that we have seen in the last 2 years. the fed bought 40% of many auctions.
so, given that the inputs to the model have been altered by interventions, why should we expect the model to behave as it usually does? (which is still not terribly impressive) garbage in, garbage out.
even if you do trust the model, why would you trust the inputs right now? if you alter the data going into a physics equation, you get the wrong answers too.
there is compelling reason to believe that the yield curve does not mean what it has historically meant and that it's price relationships have been altered by a unprecedented exogenous influence.
if that is so, then then this model is of little value and quite possibly worse than nothing. a bad map is worse than none at all.
meanwhile, rail cars dropped for the second straight month. retail sales did likewise. mortgage applications dropped to 15 year lows.
if you believe that the possibility of double dip is 0.2%, you ought to be willing to give some pretty long odds on a one year double dip bet. even 100:1 would be dramatically in your favor.
i'll happily take the other side.
why do i doubt professor estrella would be willing to make such a wager?
I notice downbeat reports about the decline of retail sales today.
But no real discussion of what consumer spending(supposedly 70% of the economy) is.
Three components...
durable goods
non durable goods
services
Largest component is services
retail sales can drop at the same time consumer spending increases
Also, the 70% number is not sacred
it might be better for that number to drop and for business spending to increase
Retail sales?
If consumer spending is 70% then what is the other 30%?
20% government
10% investment
government and consumer spending could be less
investment could be more
bix-
consumer spending (change from previous month) was
0.0 in april
0.2 in may
http://www.bea.gov/newsreleases/national/pi/pinewsrelease.htm
significant slowdowns from earlier in the year.
personal income growth has exceeded expenditure, indicating either increased a savings or the pay down of debt. (which is going to be a long and necessary trend)
this points as well to a significant slowdown, though not yet actual recession. by q4 2010 and q1 2011 the year on year GDp comparisons will begin to get difficult again and we shall see what sort of recovery we have.
Mr. Morganovich,
Just wanted to say your comments are always intelligent and knowledgeable and interesting.
From the Rockefeller Institute;
Confirms Rising Retail Sales a Mirage
http://www.rockinst.org/pdf/government_finance/state_revenue_report/2010-07-13-SRR_80.pdf
The report confirms that rises in state sales revenue collections have been due to increases in taxes and not increases in sales.
This is why same store sales is not a good indicator on the health of the economy. Look at actual DOLLAR sales or merchandise and they are still nowhere near pre-recession levels.
Just trying to show both sides. I want to be convinced that jobs will be plentiful and dollars will flow freely. I am just not there yet.
bix-
thanks.
of further interest, 60-70% of GDP growth in Q1 and Q2 came from inventory increases.
absent inventory, Q1 GDP growth annualized to only 0.82%.
inventory levels now appear to have leveled off or even gone into slight decline.
if such such low growth in consumer spending persists (and june retail sales were not terribly encouraging), it would not take a great deal of inventory contraction or a particularly large credit hiccup to push the recovery into recession in the coming quarters.
0.2% chance of a double dip seems a very aggressive statement in such a situation.
This is one of those times when I wish I could draw cartoons. A couple of people would be looking into the stars of the night sky...lo and behold, there would not be one big dipper, there would be two!
how the heck does yield curve analysis from prior instances apply to this instance? fed funds are at 0!!!! how can you invert a yield curve at 0? look at the REAL rates here and you will indeed see inversion.
Morganovich,
Excellent write-up. In support of your statement about the yield curve not being a great predictor was the depression within the depression of 1937. The yield curve was positive throughout:
http://static.seekingalpha.com/uploads/2010/6/23/saupload_20100621curve20s.png
At this point I am wondering if the second dip started in May. The NBER always waits a long time to make the call because it is a difficult to see the beginning when it occurs. The signs for a drop beginning in May include:
1) Total employment based on the household survey started falling in May (April peak).
2) Retail sales started falling in May (April peak).
3) Home sales have fallen sharply starting in May (April local maximum).
4) Auto sales fell in June from March localized peak).
5) First time unemployment claims (4 wk avg) hit a localized bottom the week of May 8th.
mj-
you make an excellent point about the impossibility of inversion if short term rates are zero.
when you discuss the real rates, at what are you looking? i'd be interested to see how you came to that conclusion about real rates (which are disturbingly negative all over the world)
I feel like I am just piling on at this point, but it is just too hard to ignore.
Using the "impeccable" precedence of an INDICATOR (which in itself is nothing more than a semi-loose relationship over far too short of a time horizon) when the inputs are so badly distorted is beyond foolish.
So by this theory, shouldn't the government use all forces necessary to never allow an inverted yield curve ever again? Will that stop all future recessions?
Post a Comment
<< Home