Yves here. This is a tidy and useful addition to the literature on how high levels of international capital flows generate financial instability.
By Eduardo Olaberría, Economist, OECD Economics Department. Originally posted at VoxEU
Policymakers have long been concerned that large capital inflows are associated with asset-price booms. This column presents recent research showing that the composition of capital inflows also matters. The association between capital inflows and asset-price booms is about twice as strong for debt-related than for equity-related investment. Policymakers should therefore pay attention to the composition of capital inflows, since debt-related inflows may still undermine financial stability even if they do not result in an overall current-account deficit.
For decades, policymakers’ perception has been that large capital inflows can fuel booms in asset prices. If this were true, bonanzas in capital inflows would imply an important risk to financial stability, since booms in asset prices are leading indicators of financial crises. However, as noted by Reinhart and Reinhart (2008: 50), despite being widespread among policymakers, until recently this perception was based mainly on anecdotal evidence.
Recently, a number of empirical studies (e.g. Aizenman and Jinjarak 2009 and Ferrero 2011, 2012) have studied the association between large capital inflows and asset prices – in particular housing prices. Focusing on the current account as a proxy for capital inflows, they found that the association is, indeed, positive and significant. These findings are clearly a step forward. However, focusing only on the current account can be misleading. In recent research (Olaberría 2012 and Jara and Olaberría 2013), we present strong evidence emphasising that to understand the association between large capital inflows and booms in asset prices, we need to go beyond the current account and look at the disaggregated flows. We argue that these facts are consistent with theory, and highlight why they matter for economic policy.
Asset Price Booms and the Composition of Capital Inflows
Using a panel of quarterly data for 35 countries covering the period 1990–2010, in Jara and Olaberría (2013) we estimate the unconditional probability of observing a boom in housing prices, and the probability of observing a boom in housing prices conditional on having:
1. large current-account deficits;
2. a bonanza in net FDI inflows;
3. a bonanza in net portfolio equity inflows;
4. a bonanza in net portfolio debt inflows;
5. a bonanza in net banks and other inflows;
6. a bonanza in 4 and 5.
Bonanzas are defined as episodes of extreme net capital inflows – when domestic or foreign investors substantially increase capital inflows into a country relative to their historic levels (see Forbes and Warnock 2012). Asset-price booms are defined as periods of unusually large price expansions (a method used in Mendoza and Terrones 2008 to identify credit booms).
The main results, reported in Figure 1, show that having a large current-account deficit does indeed increase the probability of observing a boom in housing prices, but that the increase is not very significant (17.7% vs. 15.4%). In contrast, the figure shows that the probability of observing a boom in housing prices more than doubles (increases to almost 40%) during periods of bonanzas in debt-related inflows. (Olaberría 2012 found similar results for booms in stock prices.)
More generally, our research provides a systematic empirical analysis of the association between capital inflows and booms in asset prices (for both housing and stock prices). Controlling for other macroeconomic factors, and using different estimation methodologies and instrumental variables, our research shows that the association varies across capital inflow categories – being about two times higher for debt-related investment than for equity-related investment. In addition, we find that the association is weaker in countries with more flexible exchange-rate regimes and better quality of institutions. Finally, we find some evidence that capital controls may reduce the association between large capital inflows and booms in housing prices, but the evidence is weak and not robust to different methodologies.
Figure 1. Conditional and unconditional probability of observing booms in housing prices
Source: Authors’ own calculations based on data from IMF and Bank for International Settlements.
These findings are consistent with theory. Theoretical models linking booms in asset prices with large capital inflows start with the idea that, because of financial market imperfections – such as adverse selection and moral hazard – an economy’s borrowing capability is limited by the value of its assets. When large capital inflows enter an economy, the demand for assets that are in rather fixed supply increases, and asset prices rise, increasing the economy’s credit limit. Increases in the credit limit promote new rounds of capital inflows, potentially evolving into an asset price boom through a circular process in which higher asset prices make the financial conditions of the economy appear sounder than they actually are, promoting more borrowing and pushing asset prices even higher.
According to Aoki et al. (2009), this theory applies mainly to debt-related flows – which are more likely to suffer from problems of adverse selection and moral hazard, and can exacerbate cycles in asset prices by encouraging excessive risky lending during booms – and not necessarily to equity-related inflows. In fact, according to Krugman (2000), equity-related inflows – i.e. FDI – could help flatten cycles in asset prices.
In sum, the composition of capital inflows matters. While large current account deficits driven by FDI inflows are less likely to be linked with booms, large inflows of debt-related investment are likely to put pressure on asset prices – even when the current account is in surplus. (Although countries with current-account surpluses are better equipped to respond to the risk than countries with large current-account deficits.)
Implications for Policymakers
In general, to monitor the threat of capital inflows, policymakers look at the current-account balance. If it is in surplus (as is the case in Germany today) or the deficit is low, they assume that capital inflows do not pose a risk to financial stability If, on the other hand, the deficit is large, they worry about the potential risk and, only then, they may look at the composition of capital inflows.
Our research shows that this view is misleading. To better measure the risk of capital inflows, monitoring their composition matters even when the current account is in surplus. To emphasise the relevance of this, consider the example of Denmark during the years 2006–2008 (see Figure 2). Denmark had a positive current account that was driven by large outflows of equity investment (net FDI and portfolio equity). However, the current-account balance was hiding a bonanza in debt-related capital inflows (net inflows of Portfolio debt and bank loans). Not surprisingly, during the same period Denmark had a boom in housing prices. Were these two things related? Our results suggest that they probably were; however, to answer this question more research is needed.
Similarly, Germany today has a large current-account surplus, but policymakers shouldn’t assume that because of this fact capital inflows are not a concern. The positive balance is explained mainly by FDI and portfolio-equity investment abroad, but Germany is also receiving large inflows of debt-related investment, and some observers suggest that house prices have been rising fast – in particular in some urban areas. Once again, our research cannot say if these facts are related, but it suggests that policymakers should monitor them closely.
Figure 2. Current account and net capital inflows by type in Denmark, 2000–2009
Source: Authors’ own calculations based on data from IMF.
See original article for references