Monday, 13 January 2014

The Cross-Currency Liquidity Exposures of Banks

I recently attended a very interesting presentation by William Allen, formerly of the Bank of England, on international liquidity.

Much of this concerned the use of central bank swap lines in response to the financial crisis.  One of the immediate consequences of the collapse of Lehmans, was a substantial repatriation of short term cross-currency investment.  For example, European banks had large short term dollar borrowings funding longer term dollar assets.  The pressure on US money market funds led to a lot of this funding being pulled, leaving the banks trying to fund the gap.  And although the Fed was providing liquidity in the US, the state of the interbank market meant that not much of that was finding its way to the foreign banks.

The interesting thing about this is that the liquidity crisis affecting banks in many countries was substantially taking place in a currency that was not their domestic one.  And whilst a central bank's ability to provide liquidity support in its own currency is theoretically unlimited, that is not the case in a foreign currency.  Central banks with large foreign reserve holdings may be able to cope.  Otherwise, the central bank swap lines provide a way for them to get hold of the currency.  Thus, to deal with the pressure on European banks post-Lehmans, the ECB exchanged euro for dollars with Fed, using the dollars to provide liquidity to the banks.

Nevertheless, it highlights an important issue when it comes to thinking about the role of banks in the economy.  Many economists would consider that the health of the banking sector is an important factor in the state of the economy due to the consequences for the flow of credit to the non-financial sector.  But even those models that attempt to incorporate some representation of banking tend to assume an essentially domestic operation.  In reality, banking is very international.  Over 50% of the assets of UK resident monetary financial institutions are denominated in currencies other than sterling (over 30% for those MFIs that are also UK owned)*.  The implications for credit exposures are reasonably well understood as a result of global fallout from problems with US sub-prime.  The implications for liquidity are perhaps less well known, but it raises important issues about the abilities of central banks to deal with bank runs.  The monetary authority may exercise a high degree of control over its national currency, but not necessarily over its national jurisdiction.

Overall, there was good central bank co-operation in the crisis, which went a long way to preventing things getting even worse.  However, cross-currency exposures of banks will remain an important risk issue going forward.  

* Figures for November 2013.  Source: Bank of England.


  1. Your point is good, but I would emphasise that the size of international exposures within a banking system is the result of conscious decisions, and it could be controlled if policymakers really wanted to do so.

    For example, the U.K. is an outlier in terms of international exposure of its banks (as is Switzerland and a few other smaller European countries). On the other hand, heading into the crisis, Japanese and Canadian banks largely sidestepped the USD entanglements that plagued the Europeans. In the case of Japan, the banks were recovering from previous misadventures and stayed out of trouble. The Canadian banks faced relatively tighter regulation and were also cleaning up some previous messes, and so kept their foreign exposures manageable. (However, the Canadian money markets seized up, so some emergency liquidity operations did have to be undertaken to deal with Canadian domestic problems.)

    The only way that central banks can deal with this is to be willing to throw the international arms of their banks under the bus (like Iceland), or pre-emptively regulate international exposures. I sense an unwillingness of policy makers to pursue either option, so we will probably have to live with this source of potential problems for some time.

    1. Yes, I was perhaps being a bit disingenuous giving the UK figures, knowing full well that they're not typical. My excuse is my rather UK centric focus. Japan did, I believe, experience significant cross-border flows during the crisis, but this was mainly repatriation of carry trade positions, so it did not give rise to the sort of problem described here.