The resilience of the international status of the US dollar remains surprising (Frankel 2013). At the peak of the global financial crisis which started in the US, in particular in the last quarter of 2008, US treasury yields fell and the US dollar appreciated. This has created the impression of a stronger demand for US securities in general. The evidence suggests, however, that non-US residents were instead relatively ‘picky’, fleeing into short-term US Treasury bills but reducing purchases of longer-dated Treasury bonds and shedding other US bonds. The repatriation of foreign investment by US residents (Bertaut and Pounder 2009) may have played a more important role for the correction of the US dollar and of Treasury prices.
The analysis of the behaviour of foreign investment in the US during the global financial crisis raises more general questions. It is known that there is a general tendency for cross-border portfolio flows to retrench in times of financial stress (Broner et al. 2013).
- Are there exceptions to this general rule?
- Are foreign investors systematically more likely to sell (or buy) certain assets in times of heightened stress?
- Do they show any systematic tendency in their investment behaviour, such as purchasing foreign securities as safe-haven assets, depending on the country of issuance and the type of securities?
In a systematic analysis of portfolio flows by foreign investors in major economies issuing reserve currencies, such as the US, the Eurozone, and Japan, we show that there are no securities which are consistently safe-haven assets. Safe-haven assets are those experiencing portfolio inflows when risk is on the rise or perceived to be high (Habib and Stracca 2013). In particular, we study the impact of shocks to global risk and global risk aversion (such as after Lehman), as well as shocks with a more idiosyncratic nature (such as the Eurozone debt crisis) on cross-border portfolio flows, taking the perspective of foreign investors. We consider several measures of risk, and not only crisis episodes, disentangling the impact of global uncertainty and risk aversion, and investigating the reaction of foreign demand to a number of idiosyncratic factors. For securities issued in the Eurozone, we also control for the net issuance of securities using high-quality data from the ECB. The analysis leads to three main results:
- First, both crisis episodes and rises in global risk and global risk aversion are normally accompanied by net portfolio outflows, indicating that foreign investors normally ‘run for the exit’ when risk shocks occur, confirming previous findings (Broner et al. 2013).
- Second, there is no asset or country that is systematically more in demand by foreign investors conditional on all crisis episodes and risk shocks.
Nevertheless, money-market instruments issued by the US, Eurozone low-yield countries, and Japan have behaved as safe-haven assets in at least some of the crisis episodes that we study, or following changes in certain risk measures. Foreign purchases of short-dated money-market instruments are more resilient during crises – in a few instances even larger than in normal times (see Figure 1, Chart A). Generally, foreigners continue to be net buyers of longer-dated government bonds during crises, even though they reduce the amount purchased compared to normal times (e.g. reduced purchases of US Treasury bonds in the Lehman crisis), and sometimes become net sellers (as in the case of Eurozone government bonds in 2011). (See Figure 1, Chart B.) It is interesting to note that the peak of the Eurozone debt crisis in 2011 does not seem to have spared the safer Eurozone low-yield countries, which also recorded small capital retrenchment by foreigners compared with normal times.
Figure 1. External liabilities of major economies: Flows by asset class, 1990Q1–2012Q4 (as % of the outstanding stock of total portfolio liabilities in the previous year)
Sources: IMF Balance of Payments, International Financial Statistics, and authors’ calculations.
Notes: EA debt crisis refers to the third quarter of 2011. Lehman crisis refers to the average flow in the last two quarters of 2008. Other crises is the average for the following quarters: 1990Q3, 1990Q4, 1995Q1, 1997Q4, 1998Q3, 2000Q4, 2001Q3, 2002Q3, 2008Q1; see main text for further details. Euro Area (EA) high-yield includes Ireland, Italy, Portugal, and Spain. Euro Area (EA) low-yield includes France, Germany, Netherlands, Finland, Austria, and Belgium (since 2001). Aggregates for EA high-yield and EA low-yield do not net out intra-Eurozone transactions. Data for the euro Area (EA) consolidated external liabilities, netting out intra-Eurozone transactions, are available since 2000 (since 2006 for government bonds and notes).
- Third, we find that following Eurozone-specific shocks, foreign investors appear to leave those economies (in particular Eurozone high-yield countries) where the shock has its epicentre, but the evidence for that behaviour is weaker in the case of US-based crises.
This may partly reflect the global nature of US events, and partly the fact that especially short-term debt instruments issued by US residents (most prominently the US Treasury) are safe-haven assets and in higher demand by foreign investors even after US events such as Lehman or other US-based crises. Therefore, the status of the US, as the country at the centre of the international monetary system, remains somewhat special.
Disclaimer: The views expressed in this column are those of the authors and do not necessarily reflect those of the European Central Bank.
Bertaut, C and L Pounder (2009), “The Financial Crisis and US Cross-Border Financial Flows”, Federal Reserve Bulletin, 95, November.
Broner, F, T Didier, A Erce, and S L Schmukler (2013), “Gross capital flows: Dynamics and crises”, Journal of Monetary Economics, 60: 113–133.
Frankel, J (2013), “The latest on the dollar’s international currency status ”, VoxEU.org, 6 December.
Habib, M M and L Stracca (2013), “Foreign investors and risk shocks: seeking a safe haven or running for the exit? ”, ECB Working Paper 1609, November.