Saturday, July 11, 2015

IMF: Most misleading sentence ever?

On June 26th -- the day before Greek Prime Minister Alexis Tsipras announced his intention to call a referendum on the Troika's proposed aid-and-conditions package -- the IMF caused a stir with the release of a preliminary draft of its Debt Sustainability Analysis (DSA) for Greece.  Press coverage focused, correctly, on two points.  First, the IMF found that, under then-prevailing conditions, the burden of Greece's debt was unsustainable.  Second, and arguably more important, was this:
If the program had been implemented as assumed, no further debt relief would have been needed under the agreed November 2012 framework.  (p. 1, boldface type in the original)
The press and others quite naturally interpreted this sentence to mean that the Greek government had failed to take certain specific, agreed-upon actions, and that, had Greece taken those actions, the results would clearly have made Greece's debt load sustainable.  I call this a natural interpretation because...well, because that is what it says.

But it turns out that this isn't what the IMF means at all.

The first clue that there is "creative wording" in play -- where "creative wording" is to "wording" as "creative accounting" is to "accounting" -- is in the first clause of the sentence.  Why, we might ask, does the IMF refer to the program not being "implemented as assumed," rather than "implemented as agreed?"  Is this just an awkwardly indirect way of saying the same thing, since the IMF assumed that Greece would implement the program as agreed?

Actually, no:  the use of the word "assumed" turns out to be crucial to real meaning of the sentence. Specifically, the assumptions in question go well beyond expectations of the Greek government's actions to include a series of jaw-droppingly unrealistic assumptions about what the results of those actions would be.  (Aside 1)

The next clue appears two pages later, where we find a subtly different version of the conclusion:
Taking into account all these background factors, if the key program targets had remained achievable, Greece’s medium-term debt profile would have improved by up to 13 percent of GDP. [...] No further relief would, therefore, have been needed under the November 2012 framework.  (p.3 emphasis added)
Leaving aside for the moment whether the program targets were ever "achievable," we still might wish the IMF would get its story straight.  Was the program not implemented, or did the program targets become unachievable?    (I suppose that one might argue that not achieving the impossible is a failure of implementation, if "achieve the impossible" is itself the extent of the plan. But the fault nevertheless seems to lie less with the execution and more with the plan -- and the plan's authors.)  

But enough of generalities and conclusions.  What can the IMF analysis tell us about the specific failures of reality to conform to plan?  For illustration, let's look at my favorite example -- an extreme case, perhaps, but characteristic of the sort of thinking underlying the IMF's program targets.

Perhaps the single greatest disappointment of the IMF's expectations has been the failure of billions of Euros to materialize from the sale of the Greek government's stake in various enterprises.
Over the course of the SBA and the EFF arrangement, receipts have consistently fallen short of program projections by huge margins. The fourth SBA review in July 2011 projected €50 billion to materialize through end–2015. Actual receipts through the first quarter of 2015 were €3.2 billion, about 94 percent below the target.  (p.8, Box 1)
Now, if the Greek government actually had something it could sell to private investors for €50 billion, its debts would certainly look a lot more sustainable.  But what could it be?  And how exactly did the IMF reach this valuation?  The June 26 analysis is actually rather revealing on this point, for we learn the following:
The projected privatization proceeds under the program were €23 billion over the 2014–22 period. Half of these proceeds were to come from privatizing state holdings of the banking sector.  (p. 4)
Let's spell that out:  the IMF expected the Greek government to realize €11.5 billion by selling its stake in Greek banks.  (Question for the reader:  what would you pay for an equity stake in the Greek banking system?  If your answer is €11.5 billion, please send me your contact information - or, to save time, just send your ATM card.)

In summary, when the IMF writes about the the failure to implement the program as assumed, this is the sort of thing is has in mind: the "failure" of the Greek government to find anyone willing to pay €11.5 billion for a stake in banks that require daily injections of liquidity from the ECB simply to keep the cash machines full.

Which brings us to the question in the title of this post.  When you, objective and informed as you undoubtedly are, read the page-one sentence cited at the beginning of this post (the one set in bold type, both here and in the IMF report itself) is this the sort of "implementation failure" that leaps to mind?  Or is the report's most widely cited single sentence also the most misleading sentence ever?

(Aside 1)  For those who have been following the IMF's work on this subject, this will undoubtedly feel quite familiar.  The obstinate refusal of the IMF's economic dreams to come true has been a recurring theme in every IMF analysis of Greece since the agency's original work of imaginative fiscal fiction in 2010.  Apparently, reality owes the IMF a very big apology.      


  1. This is a terrifically clarifying explanation of what seems to be an EU double-think mindset, that first of all it is prudent to require the impossible, and second to expect impossible benefits. (The just released agreement continuing. Fifty billion euro in asset proceeds, indeed. Three and half percent primary surplus, indeed.)

    Or else this is just a tool. I'm not sure.

  2. is 11billion euros for a stake in the greek banking system really as outlandish as you assume ?
    do you have any data, of any sort, or are you merely stating what to you seems to be commonsense ?

    1. That's a fair question. But the ECB has just told the world that the solvency of entire Greek banking system depends on the ability of the Greek government to raise new loans from the Troika. My assumption is simply that the ECB's claim is not completely insupportable.

  3. 11 billion for Greek banks would probably make sense in a world where the Greek economy wasn't tanked and nothing was done to address the export imbalance. Which is the problem with everything about the Euro-zone "aid" to Greece. The actions would be reasonable if Greece had overborrowed in normal times. In times such as these...