Saturday, January 31, 2009

DofF Forecast: does it hold any water?

I have a serious question to ask of our Government: Do budgetary projections by the DofF in (e.g those contained in their January 2009 Addendum covered here) hold any water?

In particular, no one has yet taken the DofF forecasters to a task of explaining how on earth, with projected:
  • shrinking GDP (-€7.6bn in 2009 in nominal terms relative to 2008) and GNP,
  • negative inflation (-1%),
  • rising unemployment (+2.9 percentage points on 2008) and falling employment (-4%), and
  • rising, not falling, Net Current Expenditure (+4.3bn in 2009),
does DofF come up with a revenue fall-off of just €3.9bn for 2009 relative to 2008 and total revenue as a percentage of GDP actually rising from 33.6% in 2008 to 33.7% in 2009? (Those of you who are impatient enough, see one potential answer at the end of this post)...

These numbers - the backbone of Irish Government plans for the year - are suspiciously incongruous. Not only because they do not seem to add up. But also because we have no reason to trust DofF forecasts on the basis of their historical accuracy.

Do Government numbers hold up to scrutiny?
This week, it came to media attention that the entire Department of Finance employs only one PhD-level economist. As far as I am aware, we do not know:
  1. Where and when did this person obtain her/his degree?
  2. Was her/his degree in the field of macroeconomic modeling?
  3. Has he/she ever published peer-reviewed research in the areas of taxation and/or macroeconomic forecasting?
In other words, we have no idea how qualified this economist is to carry out macroeconomic forecasting, policy evaluations and risk analysis.

Furthermore, per my knowledge, no one knows who exactly is responsible for supervision and execution of forecasting in DofF and what model is being used. Searching DofF website for Chief Economist reveals no such person. We do not know whether forecasting function is, indeed, an established and managerially resourced function of the Department. Ditto on the Risk Analysis side, which requires both an expert in microeconomic risk modeling and macroeconomic risk specialist.

It is simply not sufficient to say that accountancy or previous budgetary experience, potentially possessed by some DofF employees (how many?) qualifies the Department to deliver any sort of economic analysis or projections. Certainly not the ones which can used by the Government to argue about the need for one reform or another.

In fact, to see the absolute poverty of economic policy research output produced by DofF one should go straight to the source: here. They might as well publish these reports in Gaelic only, for no serious economist would recognize this as proper economics.

One example: in the sole document relating to economic reviews and outlook for 2008, titled Irealnd's Contribution to the Public Consultation Process on the Review of the EU Budget (I am not kidding - they couldn't even spell Ireland correctly). Here, DofF's 'Research' team devotes only 4 pages to the entire analysis of a vital fiscal policy process. The issue of EU-wide tax - something that was a hot topic of debate in Ireland throughout 2007-2008 is given 148 words! Of course, DofF gives five times this much to the discussion of CAP - suggesting, perhaps, our Finance boffins are more comfortable in the cow sheds than in the world of macro-finance and macroeconomics.

Getting basic research wrong - something that is being done by virtually all Irish Government departments on a routine basis - is a serious issue. Brandishing as a major reform a promise to get policies onto an 'evidence-driven' platform, as our Government did last week (see here: 3rd bullet point under Taxation heading), while having no capability to prepare proper economic analysis is hardly a responsible way of governing.

When even the mighty fail by poor research

Few months back, I was sent a research note from PIMCO's cult giant, Bill Gross. Gross is an archetypal salesman, in my view, who has fantastic intuitive understanding of the market (which is way more than our public sector mandarins and politicians have). This is, in most instances, sufficient to earn high rates of return and to contain downside risks.

But, it is not enough to do two things -
(a) provide rigorous analysis of your position in the market at any point in time - past, present, or future; and
(b) explain to others why your intuitive searchlight is capable of picking the right opportunities out of the mass of potential investment strategies.

Published in June 2008 (see here: those of you who attended my class last Fall in TCD's MSc in Finance would recognize it) the note contained a rant about US inflation indices. Specifically, Gross expended some 4 pages of small print arguing that
  1. US inflation has been historically higher than measured by the CPI;
  2. True US inflation should be much closer to the 'global' average (including such economically stable and developed countries like Venezuela, Indonesia, Brazil, Philippines, Thailand, Columbia, Turkey, Ecuador and Vietnam - out of a sample of 24 countries chosen, seemingly, to deliver Gross' point).
All of this led to the following conclusion:
"What are the investment ramifications [of the 'fact' that U.S. inflation is closer to worldwide levels than previously thought]? With global headline inflation now at 7% there is a need for new global investment solutions, a role that PIMCO is more than willing (and able) to provide. In this role we would suggest: 1) Treasury bonds are obviously not to be favored because of their negative (unreal) real yields. 2) U.S. TIPS, while affording headline CPI protection, risk the delusion of an artificially low inflation number as well. 3) On the other hand, commodity-based assets as well as foreign equities whose P/Es are better grounded with local CPI and nominal bond yield comparisons should be excellent candidates. 4) These assets should in turn be denominated in currencies that demonstrate authentic real growth and inflation rates, that while high, at least are credible. 5) Developing, BRIC-like economies are obvious choices for investment dollars."

Lacking:
  • serious analysis - Gross tweaked the evidence to support his own premise;
  • proper investigation of academic and practitioner research - Gross ignored the fact that several Congressional and academic investigations since the early 1990s have concluded that CPI actually overestimates the true extent of inflation in the US by between 0.5% and 1% pa,
he produced a call to arms for investors that cost PIMCO and those who follow its strategy an arm and a leg. How? Gross' advice - issued in June 2008 -
  • has missed a significant H2 2008 rally in Treasuries, Munis and TIPS;
  • calling for heavier weighting for commodities-linked economies came at the time of extreme valuations of these economies (e.g Russia and Brazil both have peaked in June-August 2008), before they fell off the cliff in H2 2008;
  • led to an unprecedented cancellation of dividends by several PIMCO munis funds - the first time in known history any fund suspended payouts for what is, in effect, a monthly yield-generating securities class.
I do enjoy the fact that, being criticised at the time for arguing against Gross' June note, I did turn out to be right about both his call on inflation (he was concerned with hyper-inflation as the world was teetering on the verge of deflation) and on emerging markets.

Back to DofF numbers
But I am not telling this story with some malice towards Gross or PIMCO in mind. At the very least, the man can spell Ireland better than our DofF boffins can. Instead, I am using it as an illustration as to the importance of proper research in backing any strategy - investment and/or policy-related. PIMCO's operations are much more superior to what is going on in our DofF and the rest of civil service when it comes to the quality of research and analysis. This implies that if people like Gross can get things spectacularly wrong, people that occupy our DofF - quipped with one token PhD level economist - simply have no chance at getting anything right.

Remember their latest numbers:
  • shrinking GDP (-€7.6bn relative to 2008),
  • negative inflation (-1%),
  • rising unemployment (+2.9 percentage points on 2008) and falling employment (-4%),
  • a revenue fall-off for the Exchequer of just €3.9bn for 2009 relative to 2008, and
  • a total revenue as a percentage of GDP actually rising from 33.6% in 2008 to 33.7% in 2009
Well, of course to get these things to add up, one has to assume that tax increases passed in the Budget 2009 will not reduce tax revenue. In other words, that the Laffer Curve does not work. We shall see, of course, but empirical studies provide little comfort that such an assumption is a reasonable one. Ditto the numbers on retail spending in the NI and South of the border, SuperQuinn's plan to shut down supermarket located near Newry and loads of anecdotal evidence showing that Irish shoppers are fleeing the Republic for that VAT heaven of NI.

This spells serious trouble for the Government. Suppose that due to increases in the income tax, VAT and other taxes, the revenue were to decline by, say, 2.1% of GDP - as it did in less recessionary 2008. This would imply that tax increases will still be contributing positive revenue growth for the Exchequer, although on a much smaller scale. In such a scenario,
  • the net Exchequer borrowing will jump from 6.3% of GDP to 8.4% of GDP,
  • the General Government Deficit will rise by €3.8bn in 2009 - from 9.1% of GDP projected by DofF assuming €2bn in savings goes through, to over 12%.
Now, suppose tax increases wipe out any revenue gains by 2010 - the deficit will then rise to above 13% of GDP in 2009 and 15% in 2010.

Add to this the fact that while DofF was basing its numbers on -4% growth rate in GDP for 2009, the economy quite probably will contract by at least 5% - balooning potential deficit to 15-16% this year.

A scary thought, indeed, because even the IMF will not lend Mr Cowen a penny with such financial performance on the plans. So much for Brian, Brian&Mary's 'evidence-based' policies...

No comments: