The global financial crisis that started with the meltdown of the US subprime mortgage market in 2007 was preceded by a protracted period of growth in debt and leverage ratios in an environment of increasing world financial integration, low real interest rates, and growing US external deficits. Figures 1–3 show that, since the early 1980s, in tandem with the early stages of financial globalisation, credit in the US economy grew substantially and an increasing fraction of this credit boom was fuelled by foreign lending.
Figure 1 plots the gross credit market debt as a percentage of GDP for the total US domestic nonfinancial sectors and its three components.1 Total debt started to increase at the beginning of the 1980s, and since then it has increased substantially. Total debt reached almost 230% of GDP in 2007, nearly double what it was throughout the 1960s and 70s. Gross public debt as a share of GDP did not change much, but it is well known that foreign holdings of public debt grew much larger than domestic holdings, particularly with the surge of foreign reserves in emerging economies.
The removal of controls on international capital flows and innovations in financial markets resulted in unprecedented growth in stocks and flows of foreign assets. As shown in Figure 2, the US experienced a dramatic increase in gross foreign borrowing starting in the 1980s, with a significant acceleration in the last decade. The US economy also increased its gross lending to the rest of the world, but the expansion of US lending to foreign countries has been small compared to the growth in foreign borrowing. Thus, a substantial increase in net and gross foreign borrowing has contributed significantly to the growing debt of the US nonfinancial sectors.
This point is more clearly illustrated in Figure 3. The debt that is “domestically generated” is the total debt of the domestic nonfinancial sectors minus the debt owed to foreigners (i.e. “Foreign lending” in Figure 2). As Figure 3 shows, the rapid increase in domestic borrowing since the mid-1980s was possible, to a large extent, because of foreign borrowing. The fraction of total borrowing financed by domestic lending did not change much during the same period. Moreover, prior to the mid-1980s and going back as far as the Flow of Funds data allows, there is virtually no gap between the two lines in Figure 3, indicating that foreign inflows were of little relevance for the financing of US credit. That is, prior to the mid-1980s, the US was roughly the textbook definition of financial autarky. Using gross or net credit-market positions in foreign borrowing makes little difference.
The fact that at least half of the credit boom observed in the US economy was financed by foreign borrowing indicates that a thorough understanding of the process that resulted in high US leverage and of the forces at play in the financial crisis requires the understanding of the role played by global capital markets. In particular, did the globalisation of financial markets contribute to the current crisis? Or to be more precise, taking as given the numerous distortions and imperfections of financial markets that have now become evident, were their effects on the credit expansion and collapse inside and outside the US amplified because of international financial integration?
Addressing the question
In “Financial Globalisation, Financial Crises and Contagion,” we address this question by using a model that can rationalise both the expansion in domestic credit within the US and the growth of its liabilities vis-à-vis the rest of the world (Mendoza and Quadrini 2009). To this end, we use an extension of the framework developed in Mendoza, Quadrini & Rios-Rull (2009), which has proven useful for explaining these two features of the data. In that framework, a surge in domestic credit and a sustained decline in net foreign assets in the US result from financial integration that connected the US with countries lagging in their domestic financial development. The main feature of our new framework is that it gives an explicit and central role to the banking sector in allocating credit between “savers” and “borrowers”. Banks face capital requirements based on the mark-to-market rule, but can also originate loans that avoid this requirement at a higher cost. We study the impact of a negative shock to the equity of banks in the US on the allocation of credit and asset prices globally.
Looking first at the long-term properties of the model, we find that financial globalisation increased US foreign borrowing by 64% of GDP. Next we conduct a simulation exercise where we study the impulse response of asset prices, interest rates, and consumption to a “small” unanticipated shock that reduces the value of banks’ equity in the US (by about 0.5% of the value of worldwide loans). We find that this shock induces a decline in asset prices of about 12%. Given the globalisation of financial markets, the asset price decline is a global phenomenon affecting the rest of the world even if the initial shock originated in the US.
We then ask whether the asset price impact of the shock would have been different if the US economy was not financially integrated. There are two channels by which the international mobility of capital affects the severity of the shock. On one hand, capital mobility allows the US to access a larger financial market, so that when the shock hits there is a larger market in which assets can be sold to keep up with credit constraints. Ceteris paribus, this reduces the impact of the shock on the US although it increases the impact in other countries (due to contagion). On the other hand, when the financially globalised countries have heterogeneous domestic financial characteristics, capital markets liberalisation results in higher leverages for countries with more developed financial systems like the US, which amplifies the impact of the shock.
In our baseline calibration, we find that the first effect dominates. Thus, the asset price impact of the US financial shock would have been larger had the US economy been in financial autarky. This, however, is a “hypothetical” experiment that has three important caveats: (a) if the US was financially integrated with countries at the same level of financial development, the rise in leverage would not emerge and the (worldwide) asset price effect of the shock would be significantly smaller; (b) if the US had remained in financial autarky, the credit boom would have been smaller and, with lower leverages, the crisis would have been milder; (c) financial globalisation allows the US shock to have negative real and financial effects on other countries. In this sense, we conclude that higher leverages and financial contagion made possible by capital markets liberalisation did make the worldwide impact of the financial shock worse.
Can we blame globalisation for the severity of the current crisis? If we accept the view that the shock originated in the US, then other countries would have been better off not being globalised since these countries were adversely affected by the US shock. However, if the focus is on the consequence of the shock for the US economy alone, then globalisation has not made the crisis worse. The impact on the US economy has been smaller because the crisis has been shared with other countries. Of course, one can also claim that globalisation increased the likelihood of a financial shock, which is an issue that we did not address in the paper.
What type of policy lessons can we learn from the exercise? Because in our setup a financial shock propagates through banks when their capital requirements become binding, a regulatory environment that makes the requirement pro-cyclical would act as a powerful stabilisation mechanism. A capital requirement for banks based on the mark-to-market principle, although justified by many considerations that are not addressed in the paper (e.g. moral hazard problems), makes the capital requirement counter-cyclical, amplifying the impact of financial shocks. In our paper, we considered an illustrative example in which we changed the mark-to-market accounting principle for a criterion that marks bank assets to their long-run value. This reduces the decline in asset prices in response to the shock to bank capital from 12% to 6%.
1 Data are from the Flow of Funds of the Federal Reserve Board. Debt includes only liabilities in the form of credit-markets instruments.
Mendoza, Enrique and Vincenzo Quadrini (2009). “Financial Globalization, Financial Crises and Contagion”. NBER Working Paper No. 15432 (forthcoming in the Journal of Monetary Economics).
Mendoza, Enrique, Vincenzo Quadrini and Victor Rios-Rull (2009). “Financial Integration, Financial Development, and Global Imbalances”. Journal of Political Economy, 117(3).