from Follow the Money

The IMF did its job (more or less) last time around …

September 29, 2006

Blog Post
Blog posts represent the views of CFR fellows and staff and not those of CFR, which takes no institutional positions.

More on:

Emerging Markets

The IMF received its share of criticism over the past two weeks. 

The IMF governance structure is dated.  Europe is over-represented on the IMF board (and isn’t inclined to allow much change).  Asia is under-represented.    Countries guard their position on the board jealously.   The difficulty getting agreement on a modest ad hoc quota increase (Brazil and India objected because they were not among the winners) doesn’t necessarily bode well for the next set of more ambitious changes.

The IMF remains strangely (given its original mandate) unwilling to criticize countries with inappropriate exchange rate pegs (its silence on Saudi Arabia’s peg is a case in point; the IMF only delivers criticism in its regional outlook); hopefully the G-7’s call for the IMF to update its guidelines for exchange rate surveillance will spur a bit of change.

The IMF’s advice on how to reduce the surpluses of the world’s big surplus countries and the deficit of the big deficit countries is generally unheeded.    The US hasn’t shown any real commitment to balancing its budget over the economic cycle.  China has let its real exchange rate depreciate this year, even as its trade surplus exploded -- not that you would know about China’s growing surplus if you just read the IMF’s public reports.   

For that matter, the markets -- at least after June -- don’t seem to share the IMF’s concern about imbalances.  Market players are bidding up the currencies of countries with large current account deficits (New Zealand, Iceland, the US – v. at least against the yen), and pushing the currencies of countries with surpluses down (Japan).

The IMF isn’t – despite what some argue – outgunned by the private markets.  At least not in the emerging world.  The $25b the IMF provided to Turkey is far more than the international sovereign bond market ever supplied Turkey (once you net out the bonds held by turkey’s own banks, which are effectively a foreign-currency denominated domestic loan).   But it is outgunned by the huge stockpiles of reserves held by many emerging markets.  $200b and change in loanable funds isn’t what it used to be. 

The IMF’s model for generating the income needed to pay its staff is in a bit of trouble.  The IMF used to pay its staff out of interest in got from lending to the big emerging economies ….  

That in some ways is a shame.   The IMF staff still do more comprehensive analysis than just about anyone- I challenge my friends in the markets to match the IMF’s analysis of petrodollars or its assessment of Lebanon’s balance sheet risks.  

Ironically, though, the IMF’s current absence of income is evidence of some real successes.    The IMF isn’t in the business of providing long-term financing.  It is in the business of providing short-term financing to supplement the reserves of cash-strapped emerging economies.   That is an important role – there is a reason why emerging economies concluded that they need to hold more reserves …

And lo and behold, if you look at the IMF’s balance sheet over the past ten years, it basically has performed its mission.   If you want to look at the supporting evidence (including some fancy charts), read on. 

Over the past ten years, the IMF has lent counter-cyclically, supplying emerging markets with reserves when private market financing dried up. And – as one would expect – those emerging economies have paid the IMF back as private flows resumed and as higher commodity prices provided many emerging economies with a windfall.   

That means IMF lending increased during periods of turbulence – Mexico in 95, Asia, Russia and Brazil in 97-98 and Argentina, Turkey, Uruguay and Brazil again in 2001-02.     Those surges of lending show up clearly in a chart of the IMF’s non-concessional loans outstanding.


Incidentally, if US bilateral lending was added to IMF lending in 1995, the peak would be a lot sharper – the US lent its funds out fast, and got repaid far faster than the IMF.

The following chart presents the same data in a slightly different way.  It shows the one year change in total IMF loans outstanding.  It is in dollars billion – not SDR – so it tends to slightly understate IMF lending when the dollar is strong and slightly overstate IMF lending when the dollar is weak.  But that source of error is small (and few folks think in terms of SDR).

In Bailouts and Bail-ins, Nouriel and I argued that the IMF’s lending subtly changed between 97-98 and 01-02, as the countries that the IMF lent to in 01-02 were more indebted than the countries the IMF lend to in 01 and 02.    This experiment lending to more indebted countries – countries that needed a sustained period of adjustment to bring their debt ratios down – looks to have worked out better than Nouriel and I expected.  

Argentina is obviosly the case that didn't work.  IMF financing was -- mistakenly in my view -- used to put off a necessary depreciation in the peso and a necessary debt restructuring.  But even Argentina ended up in a position where it could repay the Fund.   And Brazil, Turkey and Uruguay all avoided default.

Global conditions took a turn in emerging markets favor.  The dollar’s decline helped those emerging economies with lots of dollar debt (particularly if they exported a lot to Europe).  Turkey is a case in point.   Commodity prices rallied, big time.  Low rates in the “center” fueled a new wave of capital flows to the periphery.   Global growth was exceptionally strong. 

I think though the evidence does still suggest that lending to more indebted countries is more risky.  Even with very favorable conditions, the countries that took out big IMF loans in 01-02 have repaid the IMF a bit more slowly than the countries that received large amounts of IMF money in 97-98 (Russia actually didn’t get that much new money in 98, which no doubt helped … ).   

Consider the following graph, which plots the IMF lending surge in 01-02 against the 97-98 surge for comparison’s purpose.   >imf_lending_in_9798
The big loans stayed outstanding for longer in 01/02.  That may reflect the slower buildup of the crisis.  Brazil took out a loan in 2001 to guard against contagion from Argentina, got into real trouble in 2002 and ended up borrowing a lot of money in 2003 to rebuild it reserves.  Brazil consequently explains most of the surge in IMF lending in the q8-q12 period of the most recent wave of crises.   But I suspect it also reflects the fact that the IMF is lending to more indebted countries. 

Consider a graph showing both IMF and US loans outstanding to Mexico and to Turkey when both are plotted side to side.  Turkey had a lot more debt.  And it has taken a lot longer to repay.>img_and_esf_loans_outstandi
However, even Turkey now looks to be in a position where it will be able to repay the IMF when the time comes.  That is how it should be. 

The IMF shouldn’t have large loans outstanding when times are good.  Its job is to be ready to lend when times aren’t so good. 

Alas, that isn’t the only role the IMF should perform either.   Indeed, looking ahead -- given all the changes in the world economy --  providing crisis financing to cash-strapped emerging economies may be the IMF's least important future role.

More on:

Emerging Markets