VoxEU Column Global economy

Finding contagion

The subprime crisis became the global crisis when the 2007 financial shock mutated into a full-blown global economic crisis in September 2008. This column attributes the rapid transmission of financial stress to the surprise of the crisis. Using historical data, it shows that crises with a pronounced surprise element tend to result in more widespread contagion.

The “subprime crisis” became the “global crisis” when Lehman Brothers was allowed to collapse. The 2007 financial shock, which was limited to a handful of G7 nations, mutated into a full-blown global economic crisis in September 2008.

In the third and fourth quarters of 2008, many advanced and emerging markets experienced major stress in their foreign exchange, stock, and sovereign debt markets. Emerging markets, in particular, experienced sharp withdrawal in capital flows; equity and debt funds faced significant withdrawals, issuance of debt and IPOs decreased substantially, and bank lending was scaled back (Balakrishnan, Danninger, Elekdag, and Tytell 2009).

In their recent Vox column, Andrew Rose and Mark Spiegel (2009a and b) argue that the evidence does not support the existence of contagion. Indeed, they find that countries more heavily exposed to the US in terms of trade shares and asset holdings actually fared better than others.

What then could explain the rapid transmission of financial stress in the third and fourth quarters of 2008?

Recent research

In a recent paper (Kannan and Koehler-Geib 2009), we present a mechanism showing how the degree of anticipation of a crisis determines whether or not contagion occurs. Crises with a pronounced surprise element tend to result in more widespread occurrence of contagion than crises that are more anticipated, consistent with the empirical evidence in Didier, Mauro, and Schmukler (2006). The empirical evidence presented in our paper strongly supports the existence of this particular channel of contagion even once we have controlled for domestic factors and other channels of contagion suggested in the literature.

While the data in our paper do not cover the current crisis, a first look at the 2008 crisis suggests that the mechanism proposed in our paper – the uncertainty channel of contagion – may well be at work. We organise our argument into three steps:

  • examining the degree of surprise in the current crisis,
  • explaining the proposed mechanism and the supporting empirical analysis, and
  • providing some evidence on the transmission of financial stress in 2008.
The surprise element in the current crisis

The start of the current downturn is often associated with the collapse of Lehman Brothers in September 2008. To what extent did this event take market participants by surprise – in particular, given the prior unravelling of the subprime crisis beginning in August 2007?

Figure 1 presents the evolution of a weighted average of consensus one-year ahead growth forecasts for the G7 countries plus Brazil, India, Mexico and China from 2001 up to the most recent episode. While a gradual downward correction of growth forecasts seems to have been underway with the unravelling of the subprime crisis in August 2007, growth forecasts only dropped dramatically with the Lehman collapse.

We interpret this evidence that while there might have been some anticipation about further economic turmoil, the size of the shock revealed in the collapse of Lehman clearly took market participants by surprise.

Figure 1. Mean one-year ahead forecast for GDP (weighted average G7 plus Brazil, India, Mexico and China, %)


Source: Authors’ calculation, Consensus Economics.

Translating surprise into contagion

How would the surprise element of a crisis explain whether a crisis spreads or not?

The mechanism we propose in Kannan and Koehler-Geib (2009) is a straightforward one. Investors have a view on particular economies based on information signals that they obtain from time to time. Based on these signals, they trade securities issued by firms in a particular country. The signals also inform them about the likelihood of a crisis materialising. When an unexpected crisis occurs, investors begin to doubt the accuracy of their signals and question their view on other economies. In equilibrium, this change in beliefs will increase the aggregate uncertainty regarding fundamentals in other countries in which the investor has invested. As this uncertainty increases, so does the cost of financing faced by firms and this, in turn, increases the probability of a crisis in the other country. The mechanism operates symmetrically – an anticipated crisis event actually reduces the aggregate uncertainty in other countries, thus reducing the probability of a crisis.

We verify the mechanism empirically using a broad data set comprising 38 countries with monthly data spanning from December 1993 to September 2005. The sample of countries, and the associated time frame, enables the inclusion of the following six significant crisis periods into the analysis: Mexico (1994-95), Thailand (1997), Russia (1998), Brazil (1999), Turkey (2001), and Argentina (2001-02). These crises cover both currency crises and episodes of sovereign debt default. In both cases, however, these episodes featured significant drops in stock market returns in the respective economies. Therefore, we use drops in stock market returns to date the crises events.

We find that the incidence of the Mexican, Thai, and Russian crises – identified by the literature to be surprise crises – increase the level of uncertainty regarding fundamentals in other countries (measured as the dispersion of one-year ahead GDP forecasts by analysts). The effect is found to be stronger for countries within the same region. In contrast, we find that anticipated crises (the Brazilian, Turkish, and Argentine crises) actually decrease the level of uncertainty in other countries, again with a stronger effect in the same region.1 The importance of changes in the level of uncertainty is captured by our second main finding, which is that higher uncertainty is associated with a higher probability of the occurrence of a crisis. These findings are robust against the inclusion of a large variety of control variables.

Evidence of transmission of financial stress

What can we learn from the development of the suggested measure of uncertainty and stock market returns in terms of the transmission of financial stress during the current downturn? Figure 2 shows the evolution of a weighted average of our uncertainty measure for the G7 countries plus Brazil, India, Mexico and China from 2001 up to the most recent episode.

Figure 2. Standard deviation of one-year ahead forecast for GDP (weighted average G7 plus Brazil, India, Mexico and China

Source: Authors' calculation, Consensus Economics.

Aggregate uncertainty in the world economy nudged upwards in August 2007 when the subprime crisis first came to light. The collapse of Lehman, however, marked a distinct increase in the aggregate measure. While the disagreement about future economic growth has diminished, it remains at an elevated level. This marks a difference to the uncertainty measure of Bloom and Floetotto (2005), who analyse implied stock market volatility and show that this measure has already fallen substantially.

In terms of stock market returns, an interesting picture emerges. Beginning in August 2007, stock market returns increased worldwide, and only fell significantly in September 2008 and the following months. As illustrated in Figure 3, this pattern was even more pronounced in emerging than in industrialised economies, reflecting the initial discussion of decoupling of emerging markets from the financial centre. The strong drop in stock returns after September 2008 is a piece of evidence in favour of the type of contagion that we analyse in our paper.

Figure 3. Stock market returns (MSCI GDP-weighted aggregate indices)

Source: Authors' calculation, MSCI.

Taken together, the surprise element of the crisis, the development of uncertainty, and development of stock market returns seem to suggest that the uncertainty channel of contagion may well have been at work in spreading the current crisis around the globe.

What comes next?

One implication of our paper is that surveillance activities have to be an important element in any effort to promote international financial stability. Apart from helping prevent crises in the first place, closer monitoring of an economy’s fundamentals can minimise the risk that a surprise crisis in one country spills over to others. The still-elevated level of uncertainty or disagreement about the state of the various economies leaves us with a note of caution. With fiscal deficits rising, in some cases coupled with concerns about long-term affordability, this uncertainty might reflect concerns about inflation, looming interest rate increases, or sovereign defaults, which all affect the cost of financing.


1 In the paper, we document the various ways researchers have determined whether the crisis was an anticipated one or not.


Balakrishnan, Ravi, Stephan Danninger, Selim Elekdag and Irina Tytell (2009), “How Financial Stress Spreads,” VoxEU.org, 27 April.

Bloom, Nicholas and Max Floetotto (2009), “The Recession Will Be Over Sooner Than You Think”, VoxEU.org, 12 January.

Didier, T., P. Mauro, and S. L. Schmukler (2006), “Vanishing Contagion?”, IMF Policy Discussion Papers 06/01, International Monetary Fund.

Kannan, Prakash and Fritzi Koehler-Geib (2009), “The Uncertainty Channel of Contagion,” IMF Working Paper 09/219.

Rose, Andrew and Mark Spiegel (2009a), “Cross-Country Causes and Consequences of the 2008 Crisis: International Linkages and American Exposure,” CEPR Discussion Paper 7476.

Rose, Andrew and Mark Spiegel (2009b), “Searching for International Contagion in the 2008 Financial Crisis”, VoxEU.org, 3 October.

9,659 Reads