Obstfeld on FX Reserves and Panic of 2008

I have been reading an NBER Working Paper by Obstfeld, Shambaugh and Taylor Financial instability, reserves, and central bank swap lines in the panic of 2008. These are well respected authors, so I was quite disappointed to find that they have made several errors and missed several key features of what they call the “panic of 2008.”

Most important of these is the fact that currency depreciations during 2008 were driven to a very great extent by the foreign currency liabilities of the banks and of the corporate sector. This reality was staring them in the face in the form of Iceland, but they amazingly treated Iceland as an outlier and dropped it from all their analysis. They seemed to have forgotten that in risk management, the outlier is the data and everything else is a distraction. Iceland was an extreme case, but the short dollar position of banks and companies were a critical factor behind currency depreciations in the three large emerging economies that Obstfeld et al plot in Figure 1 (Russia, India and Korea).

Failure to consider the exposure of the banking system leads them to under estimate the reserve needs of emerging economies. They make the statement that countries like India “do have foreign reserves sufficient to allow them to act as crisis lenders to foreign governments.” This is simply not true.

Obstfeld et al make another mistake in asserting that for countries like Korea, the swap lines from the US Fed served only a signaling purpose because these countries had plenty of reserves and the magnitude of the swap line was not meaningful in relation to the reserves. This again is simply false. Earlier this month, the Wall Street Journal quoted a finance ministry official as saying that the Bank of Korea had drawn down more than half of the swap line and that it might need a second or third line. Korea is really short of reserves and it has also been reported that not all of its reserves are sufficiently liquid.

It is distressing to find such serious errors in a paper by economists of such high standing who have done so much of widely cited work in this field. I know a working paper is supposed to be for dissemination in preliminary form and is not necessarily subject to peer review, but still ...

Posted at 6:25 pm IST on Fri, 27 Mar 2009         permanent link

Comments