Thursday, August 04, 2011

NY Fed Model: 1-in-125 Chance of 2012 Double-Dip

The New York Federal Reserve updated its "Probability of U.S. Recession Predicted by Treasury Spread" this week with treasury yield data through July 2011, and the Fed's recession probability forecast through July 2012 (see chart above). The NY Fed's Treasury model uses the spread between the yields on 10-year Treasury notes (3.00% in July) and 3-month Treasury bills (0.04%) to calculate the probability of a U.S. recession up to twelve months ahead (see details here) using the spread between those two yields (2.96% in July).

The Fed's model (data here) shows that the recession probability peaked during the October 2007 to April 2008 period at around 35-40% (see chart above), and has been declining since then in almost every month. For July of this year, the Fed's recession probability was less than 1% (0.97%) and for July of next year the recession probability is even lower, at only 0.80% (8/10 of 1%).  According to the NY Fed Treasury Spread model, the chances of a double-dip recession through the summer of next year are essentially zero (a 1-in-125 chance for July 2011).


At 8/04/2011 7:06 PM, Blogger Wayne Adams said...

I dunno how sound those models are anymore.

Looking at how much Treasury issues the banks have taken up makes me think that this time things really are different (

Just how different, I don't know.

My guess is that bank traders, now that they have assumed so much US sovereign risk, are being told to keep the spread at profitable levels.

If my guess is right, then I wonder when that gentlemen's agreement is going to get swept to the side.

At 8/04/2011 7:19 PM, Blogger bix1951 said...

Ignorant me watching the TV today pundits are getting on the air saying we are already in a new recession that started in May.
Is this just propaganda for the masses
or what?
I was not selling today.
Who was?

At 8/04/2011 7:34 PM, Blogger Dan Ferris said...

Yeah, Wayne, I agree about models. Financial models seem particularly bad, as models go.

Models are for people who are good at math, not people who want to make money in the market, or be right about economic outcomes.

At 8/04/2011 8:17 PM, Blogger Mike Skiles said...

This was posted here in 2/2008 right at the start of the recession:

Seems like when you make a post about low probability of a recession, then the fluff is about to hit the fan.

At 8/04/2011 8:22 PM, Blogger Mark J. Perry said...

That was a different recession model, the NY Fed model was strongly predicting a recession in early 2008.

At 8/04/2011 8:49 PM, Blogger Nick said...

Why would I believe anything these people say? They haven't gotten anything right in the history of their existence.

At 8/04/2011 10:12 PM, Blogger arbitrage789 said...

My betting is that GDP for Q4 of 2011 and for Q1 of 2012 will not both be negative. (They would both have to be negative for a recession to have begun in Q4 of this year).

One question, however, is why might it be true that the Fed model is “wrong” this time. As it happens, the yield on the 13-week Treasury note is lower than at any time during the period over which the Fed model is based. And so, right there, “this time is different” as far as one of the components is concerned. In addition, there may be an inflation concern built into the yield on the 10-year note. Core CPI right now is pretty flat, but with all the QE and ZIRP we’ve had, that could change.

So, while I don’t think that a recession is going to begin within the next several months, there is also reason to wonder whether the conditions that made the Fed model work so well in the past might not be applicable in this case.


If I had to guess, I'd say we do get a recession in 2013.

At 8/04/2011 10:52 PM, Blogger Cabodog said...

Just today, our local newspaper ran a story about the rebound in the sales of luxury goods.

As in, what recession?

Lots of indicators are pointing to an economy in recovery -- restaurant spending index, luxury goods sales, 5% same-store sales increases...

This has to be more to do with fears of Euro-zone defaults than a slowdown in the economy.

Also, August is a great time for the bears to strike due to the reduced number of players in the market.

At 8/04/2011 11:27 PM, Blogger arbitrage789 said...

I would also say this: if the yield on the 10-year note got down to 2%, I would say the the chances of a recession were greater than 7 out of 10, even if the yield curve were upward-sloping. And if the yield on the 10-year note got to 1%...

...any chance there WOULDN'T be a recession then?

At 8/05/2011 12:13 AM, Blogger Benjamin Cole said...

Well, economic models work until they don't.

We will have a recession unless the Fed undertakes an aggressive monetary policy.

The right-wing obsession with minute rates of inflation is putting a monetary noose around an over-leveraged economy. See Japan for how tight money following a real estate/equity bust works out.

The Fed can undertake an aggressive program of targeting a nominal GDP growth rate of 7 percent. That will require serious QE and perhaps cutting interest on reserves. Actually, this is a school of thought pioneered by Milton Friedman. Conservative economists such as Scott Sumner now champion this approach--not fiscal stimulus.

And, yes, don;t crap in your pants, but perhaps 4-5 percent inflation for a few years. The same sort of inflation we had through the 1980s and 1990s, great boom decades.

Or, we can do a Japan, with the flag of the setting sun, depopulation, and permanently declining equity and real estate values.

At 8/05/2011 1:06 AM, Blogger Wayne Adams said...


That's why I am betting on QE3 in the next year.

With banks funding off of short term treasuries instead of short term mortgages derivatives, the short term treasury rates are likely to remain lower than ever. That's relatively easy for the Fed to continue doing.

A jobless recovery means negative growth expectations are going to pull the 10-year yields down. The only thing that could keep those rates up is inflation expectations. Enter the helicopter.

At 8/05/2011 6:26 AM, Blogger Jason said...

What if the traditional definition of a recession (2 qtrs of negative GDP growth) is invalid in a long term macro-economic cycle, the likes of what we are presently in? The definition has worked for, what, 80 years? What if this time things are different?

We are still suffering from the lingering effects of debt overload, a massive generational shift, the effects of a changing global economy while our government is mired in the failed notion of post world war II economic planning. I think we can see 20-30 more years of middling growth while we come to terms with the reality of entitlements, global competition and reduced political and economic influence. Not to mention the policy insanity of the Euro-zone...

At 8/05/2011 8:01 AM, Blogger morganovich said...

for an index that totally missed the recessions in 2000, 1982, and 1976 to claim a 1 in 125 chance of recession is laughable.

i'd be willing to be that if you backtest their predicted recession chances against the actual recessions, you'd get error bars of +/- 50 in 100.

this is just a rate spread model.

the sheer size of currency and bond market manipulation going on at the moment makes it a pretty poor tool to use.


"My betting is that GDP for Q4 of 2011 and for Q1 of 2012 will not both be negative. (They would both have to be negative for a recession to have begun in Q4 of this year)."

this is largely an artifact of the post 1992 changed in CPI and GDP deflator.

if it were 1980 and were were having these conditions, we would still be be (according to the BLS etc) deep in the first recession.

the only reason this is being called "recovery" is that GDP deflators are so low.

it's not surprising that the 2 slowest employment recoveries since ww2 have been the two after these changes.

for any given set of objective facts, reported recessions are much shorter and shallower than they used to be.


the bond dealers are being told that they need to buy big slugs of bonds at low rates if they want to keep their status as primaries.

to entice them, they are being given cheap money from the fed and allowed to gear up (10:1) on purchases to make it work on their sheets.

this is what has driven the much vaunted "profits recovery" of the S+P.

take out financials, and there has been very little recovery all along.

At 8/05/2011 8:12 AM, Blogger morganovich said...


a lot of folks were selling in july.

i get letters from a couple dozen hedge funds, and nearly everyone was cutting exposure and going to a negative risk posture, especially for europe.

yesterday there was a massive margin call cascade. desks at single brokers had 100's of millions in margin call sales.

that's why you saw such a wipeout in the last hour.

i think those who did not sell in july were all waiting for the debt ceiling deal to create a market bounce and hoping to sell into that.

when everyone is waiting for good news to crate a bounce so they can sell, you get no bounce.

we are now in very difficult territory for the tape. momentum is broken, but stocks are not value investor cheap.

the economic data is bad and central banks have lost credibility.

the market has laughed at the ECB and BOJ. our fed is not better.

CB's have been playing a game of doubling down.

problem? add liquidity. bigger problem? add more.

it's like a blackjack player doubling his bet after ever losing hand and saying "i just need one win to make it all back".

eventually, the doubling crashes you into table stakes and you lose your short.

with CB's it's worse, as their actions are actually not independent of the outcomes and they are causing and increasing the magnitude of the next problem trying to fix this one.

they run into limits on debt and stimulus etc and they wind up looking at a worse crash and a pile of debt they have run up.

it's not a pretty situation.

this could get messy.

At 8/05/2011 8:47 AM, Blogger Eric H said...

1 in 125 huh? From the same guys that only had a 40% chance of the current recession? The one that allegedly ended in 2009 according to the same prognosticators?

Can't double dip if you don't get out of the first one.

Oh, and Benjamin, since inflation is flat, I'll gladly pay you 2010 king dollar prices for all your silver bullion, or gasoline, or wheat berries. Or, since you are fond of cell phone service, $39.99/month for 1000 nationwide cell phone minutes with unlimited mobile-to-mobile and nights and weekends starting at 7pm. That's a 2007 price, but hey, inflation is minute...

At 8/05/2011 8:55 AM, Blogger rjs said...

has that model ever been used at the zero bound?

during a liquidity crisis?

At 8/05/2011 11:37 AM, Blogger bart said...

The Fed model has had 3 misses since 1960. The NTRS (Paul Kasriel) model has had none (as in zero misses), and mine based on it (plus inflation "adjustments") shows we're in a short one right now.

At 8/05/2011 11:47 AM, Blogger Benjamin Cole said...

Eric H-

I get unlimited nationwide for $50 a month from Boost Mobile. I can literally keep an open phone line, 24/7 from Los Angeles to NY for $600 a year--and carry it around with me.

And take pictures and send them. And record my GPS location.

And record verbal notes--there is other stuff on the phone but I don't know how it works yet. I can go online but it is too slow for most uses.

You want to know my GPS location? I am dying to tell someone.

At 8/05/2011 1:06 PM, Blogger Junkyard_hawg1985 said...

While the yield curve in the U.S. is upwardly sloping, the yield curve in India and Brazil are in the process of inverting:

Even though our yield curve is not inverted, our yield curve is almost identical to 1936 when we had a severe downturn in 1937.

At 8/05/2011 11:08 PM, Blogger Eric H said...


No commodities then?

Here's another prediction by the Fed gone wrong.

At 8/05/2011 11:17 PM, Blogger Eric H said...

With your <a href=">coverage map, I don't think you could keep a line open too long. Not worth the 25% inflation to me.

At 8/07/2011 7:05 AM, Blogger VangelV said...

What a joke it is to continue to reference models that have proven to be no better than a coin flip or a trip to your local astrologer. The way I see it, there is a better than 1-in-125 chance of an outright collapse in 2012 and a better than 1-in-125 chance of a major currency disruption or a significant event in the Middle East that takes the economy down. Sorry Mark but there seems to be a disconnect between your academic training and your ability to see reality as it is.


Post a Comment

<< Home