"Bond on bonds – disaster ahead"

... in which Barcap's Tim Bond forecasts developed market bond yields to double over the next decade. I tend to agree with that forecast, although Tim Bond and I have completely divergent reasons why.

He thinks an aging population will cause a surge in government debt/GDP ratios, which will increase risk premia. I believe that an aging population lowers potential GDP growth and thus, bond yields as well. I believe that inflation is the one main driver of bond yields in developed markets: cf. Japan, with it's government debt/GDP closing in on 200% (this is much higher than anything forecast for Europe and US in the next twenty years).

However, I am much less confident that the current handling of the financial crisis will not bring a spurt of high inflation down the road. It's not yet in the cards, but avoiding that scenario will require extreme toughness from policy makers - something I have trouble believing in.

By the way, as I have argued in many posts in the past six months, I believe in the near term the risks to bond yields are tilted to the downside: in my opinion, there is still more risk on banks balance sheets than on sovereigns balance sheets, and renewed flights to quality cannot be dismissed. For now.

More on the uncertainty surrounding 2010 GDP forecasts

This is a follow-up to the previous post, in which I discussed the fact that my GDP model produces a very large range of possible 2010 GDP growth, depending on the number of lags (past observations) of the data.

To illustrate this, I backward tested my model to estimate year-ahead GDP growth as of each January since 2003, in what is called "pseudo out of sample" forecasts. It simply means that I put my model in the situation it would have been at the time, and thus, it cannot use data not known as of that moment. For example, for the 2004 GDP growth forecast, only data published through January 2004 can be used.

I had my model produce forecasts using 8 lags (the past 8 months of data) to 38 lags. Then, I calculated the standard deviation of these yearly forecasts:

 

To summarize, what this tells is that the uncertainty surrounding the 2010 figure is about five times larger than usual.

The conclusion: either my model is misspecified (which I don't rule out, but it's been very good on a pseudo out of sample exercise), or, fading monetary and fiscal stimulus, combined with an uncertain outlook for private investment and final demand, makes it very difficult to have a strong opinion on 2010 GDP. And it could of course be both.

2009 GDP Forecast and a thought on 2010

With the first GDP estimate for Q4 released, I can now compare the actual performance of U.S. GDP with my model's forecast for 2009. A picture is worth a thousand words, so here's a chart:

This is what my model would have predicted in January of last year (well not exactly as I am using revised data. I believe if I used vintage data the prediction would have been less accurate, although in the same ball park).

Note: the green bands represent the 50%, 75% and 90% confidence bands around the central forecast.

Despite improvements in recent months, my model is still calling for (much) below-consensus growth in 2010. However, because of the extremely wild swings the underlying series have experienced, which are due to unprecedented (in the estimation sample) economic dislocations, I am afraid the model could be being stretched beyond its capabilities.

(Technical note:) What is happening precisely is that, depending on the used number of lags of the data, the forecasts can be extremely different. Lags are how many months of data I let my model use: for example, I could use the past 12 months observations of money supply. I had determined previously that 30 months of observations was a good number of lags, but in any case, if I used 12 months, 24 months or 36 months, the forecasts were not very different. This is no longer the case, and it's a big problem.

However, I believe it is also a reminder that 2010 risks being quite surprising to both bulls and bears. It was easy to forecast a recovery last year when the largest stimulus in history was kicking in. It was also easy to forecast the recession (pardon my lack of humility), when the housing bubble popped. 2010 is a much less obvious story: the contribution from stimulus is going to fade and subtract from growth pretty soon, and private demand seems to be recovering extremely slowly.

Links 7/1/10

There have just been too many interesting pieces in the last few days for me to quote them all on this blog, so for once my dear readers, you are going to have to read it all (trust me it's worth it):

• Bill Gross: Let's Get Fisical, also: Time's interview
  
• Tim Duy: Fed Watch: Out of the gate with a bang
  
• a sell signal from Mark Hulbert: Bullishness now at dangerous levels
• From The Pragmatic Capitalist: The ultimate guide to 2010 investment predictions and outlooks

My 2010 U.S. GDP forecast: 0.7% growth

That is much below consensus. (Note: my model has had an average error of 0.5%.)

Some quick reminders: I do not make one single guess or assumption in this forecast. It comes directly from a model which uses leading indicators of the economy. The single one assumption that is implicitly made is that those indicators, which have been very useful at predicting the business cycle in the past, continue to provide information on the future path of the economy. The forecasts which would have been obtained by my (recently developed) model through back-testing are significantly better than most of the forecasts made by the economists polled each quarter by the Wall Street Journal, and notably, would have predicted the past recession. More on my model here, here and here. More on leading indicators here.

Consumer confidence vs. GDP growth

The highest correlation that can be found between consumer confidence and GDP growth... is not that high (about 0.5):

This makes consumer confidence one of the less-reliable leading indicators of the economy.

U.S. GDP Forecast update

Here are my latest forecasts. As a reminder, my model uses only cold hard statistical data. There are no assumptions made here, except for the one that past relationships between leading indicators of the economy and GDP will continue to hold.

One-time unexpected short-term fiscal effects, such as the Cash for Clunkers program, may thus not be captured, and I won't try to make up for it by artificially boosting my model's forecasts. The public demand component is accounted for through other variables, and I do expect my model to capture most of it over the medium-run.

GDP Forecasts:

Next 12 months: 1.5% growth
Q4 2009: 1.1%
Q1 2010: 0.8%
Q2 2010: 1.0%
Q3 2010: -1.3%
Q4 2010: -2.1% (as always, this last one is to be taken with a grain of salt as my model is not made to make forecasts more than 12 months out).

Below is a graph of my monthly GDP forecasts (annualized), along with the 50%, 75% and 90% confidence bands (click to enlarge):

Leading Economic Indicators Index increases again

I am a big fan of the Leading Economic Indicators, however I like much less the Index of Leading Indicators. The reason is because of the way it is constructed: the weights of each component are designed to smooth out the index (precisely, each one of the weights is calculated as the inverse of the component's volatility). As such, the weights don't reflect the predictive power of the indicators, nor their lead length relationship with the economy.

Secondly, the weights are recalculated each year and the components themselves have changed about once a decade.This invalidates historical comparisons.

Still, I prefer to see a rising Index rather than a falling one (click to enlarge):

Unemployment might be peaking sooner than you think

Although I am more bearish in my GDP forecast than many others (see these posts), I don't believe that should necessarily translate into unemployment peaking in the mid-teens. See this graph (click to enlarge) :

Using this transformation of the data, since the growth in initial claims for unemployment insurance has peaked, the unemployment rate could follow soon. How slowly it will go down is another question.

Building permits and U.S. GDP growth

The highest correlation that can be found between building permits and subsequent year-over-year GDP growth is 0.64:

(Note: This is a 15-months rate of change, and Building Permits are advanced 6 months.)

This illustrates one of the missing private demand leg in this recovery.

GDP forecast update

Here is my updated forecast for U.S. real GDP growth (chart below). My model doesn't use assumptions about any economic variables, the behavior of households, fiscal policy, etc.: it just uses cold hard data. This data in turn is supposed to lead the state of the economy by a few quarters.

A few comments:

• For the next twelve months, real GDP will grow by (1%) 1.1% according to my model. This is much below the consensus.
• Although most comentators have already declared the end of the recession, my model says we could see negative growth in some quarters.
• Don't be too alarmed by the huge drop forecasted for the last quarter of 2010. The model isn't supposed to be able to do well that far in the future (it does best at a horizon of 6 months or so).
  

So I hope my readers will not be surprised by disappointing GDP figures, notably starting mid-2010.

Note: the green bands represent the 90%, 75% and 50% confidence around the central forecast.

See below for back-testing against other forecasters for the 2005-2008 period. The model fares much better when 2008 is included as it predicted minus 2% growth for 2008, while no other economist polled by the WSJ called for negative growth. I called my model VAR(30).

ranking among WSJ-polled forecasters

Firm	Average absolute error (%)		Rank		Excluding 2008 (%)		Rank
VAR(30)		0.52		1		0.57		16
UCLA Anderson Forecast		1.13		2		0.22		1
Merrill Lynch		1.14		3		0.43		7
UBS		1.17		4		0.34		2
The Northern Trust		1.24		5		0.50		10
Lehman Brothers		1.24		6		0.50		11
Standard and Poor's		1.25		7		0.42		4
Vanderbilt University		1.26		8		0.43		6
Perna Associates		1.27		9		0.50		12
Decision Economics Inc.		1.28		10		0.48		8
Goldman Sachs & Co.		1.28		11		0.58		17
Global Insight		1.28		12		0.42		5
Swiss Re		1.32		13		0.51		13
Maria Fiorini Ramirez Inc.		1.35		14		0.54		14
Econoclast		1.36		15		0.36		3
Morgan Stanley		1.38		16		0.95		38
Credit Suisse		1.39		17		0.73		26
Wells Fargo & Co.		1.40		18		0.61		18
Mortgage Bankers Association		1.45		19		0.64		21
Comerica Bank		1.48		20		0.65		22
RSQE, U. of Michigan		1.48		21		0.55		15
Median Survey forecast		1.50		22		0.64		20
FedEx Corp.		1.55		23		0.61		19
The Conference Board		1.55		24		0.71		25
Wachovia Corp.		1.55		25		0.48		9
Barclays Capital		1.58		26		0.69		24
Economic Analysis		1.60		27		0.75		28
Bank of America		1.61		28		0.83		29
High Frequency Economics		1.62		29		0.88		35
Keystone Business Intelligence India		1.63		30		0.85		32
AllianceBernstein		1.64		31		0.74		27
Hanmi Bank		1.67		32		0.87		34
Wayne Hummer Investments LLC		1.67		33		0.67		23
Nomura Securities International Inc.		1.69		34		0.86		33
National City Corporation		1.76		35		0.85		31
Moody's Investors Service		1.77		36		0.94		36
Eaton Corp.		1.77		37		0.83		30
Bear Stearns & Co. Inc.		1.83		38		1.02		39
Deutsche Bank Securities Inc.		1.84		39		1.13		41
Macroeconomic Advisers		1.84		40		0.94		37
National Association of Realtors		2.09		41		1.13		40