thumb|350px|Subprime Crisis Diagram
Financial contagion refers to "the spread of market disturbances—mostly on the downside—from one country to the other, a process observed through co-movements in exchange rates, stock prices, sovereign spreads, and capital flows". Financial contagion can be a potential risk for countries who are trying to integrate their financial system with international financial markets and institutions. It helps explain an economic crisis extending across neighboring countries, or even regions.
Financial contagion happens at both the international level and the domestic level. At the domestic level, usually the failure of a domestic bank or financial intermediary triggers transmission when it defaults on interbank liabilities and sells assets in a fire sale, thereby undermining confidence in similar banks.
An example of this phenomenon is the subsequent turmoil in the United States financial markets. International financial contagion, which happens in both advanced economies and developing economies, is the transmission of financial crisis across financial markets for direct or indirect economies. However, under today's financial system, with the large volume of cash flow, such as hedge fund and cross-regional operation of large banks, financial contagion usually happens simultaneously both among domestic institutions and across countries. The cause of financial contagion usually is beyond the explanation of real economy, such as the bilateral trade volume.
The term financial contagion has created controversy throughout the past years. Some argue that strong linkages between countries are not necessarily financial contagion, and that financial contagion should be defined as an increase in cross-market linkages after a shock to one country, which is very hard to figure out by both theoretical model and empirical work. Also, some scholars argue that there is actually no contagion at all, just a high level of market co-movement in all periods, which is market "interdependence".
Causes and consequences
Financial contagion can create financial volatility and can seriously damage the economy and financial systems of countries. There are several branches of classifications that explain the mechanism of financial contagion, which are spillover effects and financial crisis that are caused by the influence of the four agents' behavior. The four agents that influence financial globalization are governments, financial institutions, investors, and borrowers.
The first branch, spill-over effects, can be seen as the negative externalities. Spillover effects are also known as fundamental-based contagion. and Lagunoff and Schreft (2001) analyze financial contagion as a result of linkages among financial intermediaries. The former provide a general equilibrium model to explain a small liquidity preference shock in one region can spread by contagion throughout the economy and the possibility of contagion depends strongly on the completeness of the structure of interregional claims. The latter proposed a dynamic stochastic game-theoretic model of financial fragility, through which they explain interrelated portfolios and payment commitments forge financial linkages among agents and thus make two related types of the financial crisis can occur in response.
Trade links is another type of shock that has its similarities to common shocks and financial links. These types of shocks are more focused on its integration causing local impacts. "Any major trading partner of a country in which a financial crisis has induced a sharp current depreciation could experience declining asset prices and large capital outflows or could become the target of a speculative attack as investors anticipate a decline in exports to the crisis country and hence a deterioration in the trade account." document the evidence that trade links in goods and services and exposure to a common creditor can explain earlier crises clusters, not only the debt crisis of the early 1980s and 1990s, but also the observed historical pattern of contagion.
Competitive devaluation is also associated with financial contagion. Competitive devaluation, which is also known as a currency war, is when multiple countries compete against one another to gain a competitive advantage by having low exchange rates for their currency. "Devaluation in a country hit by a crisis reduces the export competitiveness of the countries with which it competes in third markets, putting pressure on the currencies of other countries; especially when those currencies do not float freely." Also, Calvo (2004) argues for correlated liquidity shock channel meaning that when some market participants need to liquidate and withdraw some of their assets to obtain cash, perhaps after experiencing an unexpected loss in another country and need to restore capital adequacy ratios. This behavior will effectively transmit the shock.
Out of the four agents, an investor's behavior seems to be one of the biggest one that can impact a country's financial system. and classical early models of disease diffusion were applied to financial markets by Shiller (1984). Also, Kirman (1993) analyses a simple model of influence that is motivated by the foraging behavior of ants, but applicable, he argues, to the behaviour of stock market investors. Faced with a choice between two identical piles of food, ants switch periodically from one pile to the other. Kirman supposes that there are N ants and that each switch randomly between piles with probability ε (this prevents the system getting stuck with all at one pile or the other), and imitates a randomly chosen other ant with probability δ. Eichengreen, Hale and Mody (2001) focus on the transmission of recent crises through the market for developing country debt. They find the impact of changes in market sentiment tends to be limited to the original region. They also find market sentiments can more influence prices but less on quantities in Latin America, compared with Asian countries.
Besides, there are some researches on geographic factors driving the contagion. De Gregorio and Valdes (2001) examine how the 1982 debt crisis, the 1994 Mexican crisis, and the 1997 Asian crisis spread to a sample of twenty other countries. They find that a neighborhood effect is the strongest determinant of which countries suffer from contagion. Trade links and pre-crisis growth similarities are also important, although to a lesser extent than the neighborhood effect.
History
The term "contagion" was first introduced in July 1997, when the currency crisis in Thailand quickly spread throughout East Asia and then on to Russia and Brazil. Even developed markets in North America and Europe were affected, as the relative prices of financial instruments shifted and caused the collapse of Long-Term Capital Management (LTCM), a large U.S. hedge fund. The financial crisis beginning from Thailand with the collapse of the Thai baht spread to Indonesia, the Philippines, Malaysia, South Korea and Hong Kong in less than two months. This caused economists to realize the importance of financial contagion and produced a large volume of researches on it. Yet, there were episodes of international financial crisis that occurred before the introduction of the term contagion.
Some analysts, including Bordo and Murshid, identify the crisis that happened in 1825 as the first international financial crisis. "The liberation of Latin American in the early 1820s led to a massive inflow of capital from Britain to finance the exploitation of gold and silver mines and of sovereign loans to the newly independent republics." Between new industries beginning to grow, an increase in foreign influence, and a liberal monetary expansion after the Napoleonic Wars, there was an increase in irrationality on the London Stock Exchange. As a result, the bank decided to increase its discount rate. The stock market crashed in October, which triggered a banking crisis around December. This crisis spread throughout the continent. "This crisis spread to Latin America as the overseas loans were cut off, a decline in investment and exports reduced tax revenues and led to sovereign debt defaults across the region." The first currency that faced problems was the Thai baht. With the Thai baht having issues, it doubled the debt of Thai organizations, which started the spread of the crisis to other countries. As this was happening, investors started reevaluating their investments in this region. This caused the flow of money to disappear rapidly, resulting in the growth of this crisis.
The crisis of 2007–08 has been identified as the most severe since the 1930 Great Depression. Major financial institutions around the world were greatly affected. The history of the 2007–08 crisis traces back to the bursting of the housing bubble in the United States, and the increase in mortgage defaults. This came about as a result of the mandate by the U.S. Congress for the Federal National Mortgage to increase access to low-income housing. As a result of the high default rates, many financial institutions across the U.S. were affected. Although the U.S. government had attempted to salvage the situation through liquidity doses, the crisis further deepened. By March 2008, Bear Sterns, a U.S. investment bank, required the efforts of the government to be rescued. At this stage, it was clear that the crisis had deepened. Other financial institutions, such as the Lehman bank and American International Group (AIG), started to feel the effects of the crisis. It proposes a concrete definition, a significant increase in cross-market linkages after a shock, and suggests using the term "interdependence" in order to differentiate this explicit definition from the existing literature. It shows the elementary weakness of simple correlation tests: with an unchanged regression coefficient, a rise in the variance of the explanatory variable reduces the coefficient standard error, causing a rise in the correlation of a regression.
General models
Let <math>\mathcal{V}</math> is the set of financial assets and <math>p_v(t)</math> be the price of asset <math>v \in \mathcal{V}</math> at time <math>t</math>. A network with contagion is defined in matrix form as <math>\Gamma(t) \in \{0,1\}^{n \times n}</math>, whose <math>(v, \, v^{\prime})</math> component represents the connection between two stocks <math>v</math> and <math>v^{\prime}</math>. In vector notation, the standard model for contagion tests can be written as a VAR (vector autoregression) model of order <math>\tau</math>:
:<math> \ln\mathbf{p}(t) \,=\, \Gamma(t-1) \ln \mathbf{p}(t-1) \,+\, \ldots \,+\, \Gamma(t-\tau) \ln \mathbf{p}(t-\tau) \,+\, \epsilon_v(t),</math>
where <math>\epsilon_v(t_i)</math> is a random term. In their specific application, Forbes and Rigobon (2002) estimated a variant of this model to study contagion between countries. They first estimated the variance-covariance matrices for each pair of countries during the stable period, turmoil period, and full period. Then, they use the estimated variance-covariance matrices to calculate the cross-market correlation coefficients (and their asymptotic distributions) for each set of markets and periods.
As Pesaran and Pick (2007) observe, however, financial contagion is a difficult system to estimate econometrically. To disentangle contagion from interaction effects, county-specific variables have to be used to instrument foreign returns. Choosing the crisis period introduces sample selection bias, and it has to be assumed that crisis periods are sufficiently long to allow correlations to be reliably estimated. In consequence, there appears to be no strong consensus in the empirical literature as to whether contagion occurs between markets, or how strong it is.
The financial and economic literature presents ample evidence that in time of crisis co-movements between the returns of assets increase. This increase in correlation between the returns of the loans' collateral causes an increase in the volatility of bank assets and, therefore an increase in the value of the bank's stock and its cost of default, while decreasing the value of its debt. The increase in correlation can be explained by a procyclical forbearance policy of regulators. Since regulators have greater forbearance during systemic crises, the increase in correlation creates incentives for banks to herd and become interconnected so that when they fail, they fail together, increasing their chances of being bailed-out. Peleg and Raviv (2018) shows that as the correlation between the returns of bank's borrowers' increases, asset risk increases as well. Thus an increase in co-movement of a bank's loan portfolio increases the bank's cost of default through a second channel: an increase in risk shifting.
Multi-channel models
Recently, Nasini and Erdemlioglu have proposed a model to study how the effects on stock price dynamics of different network propagation channels vary according to the state of the economy. Drawing on the view that decisions and outcomes of financial firms are influenced by multiple network channels, they studied the stock price dynamics of listed enterprises connected by supply-chain relationships, competition linkages and business partnerships.
Let <math>s_{v}(t)</math> be the market value of asset <math>v \in \mathcal{V}</math>, defined as the share price times the number of shares outstanding: <math>s_v(t) = p_v(t)\nu_v(t)</math>. At each moment in time <math>t</math>, a network with connection of type <math>c \in \mathcal{C}</math> is defined in matrix form as <math>\Gamma(t,c) \in \{0,1\}^{n \times n}</math>, whose <math>(v, \, v^{\prime})</math> component represents the <math>c^{th}</math> connection between two stocks <math>v</math> and <math>v^{\prime}</math>. Define <math>\hat{s}_{v,v^{\prime(\ell) = \sum_{c} \beta_{c \ell} s^{\Delta}_{v, v^{\prime(t-\ell) \gamma_{v,v^{\prime(c)</math>, where <math>s^{\Delta}_{v, v^{\prime(t-\ell)</math> quantifies the difference between the market value of <math>v</math> and <math>v^{\prime}</math>. The financial econometric model of Nasini and Erdemlioglu can be written as
:<math> p_v(t) ~=~ \eta_v(t) \prod_{\ell = 1}^{\tau} \prod_{v^{\prime} \in \mathcal{V p_{v^{\prime(t_{i-\ell})^{ \hat{s}_{v,v^{\prime(\ell) },</math>
where <math>\eta_v(t)</math> is a random term. They derived an important relationship between this model and the classical Fama–French three-factor model. Let <math>q_{v,v^{\prime(c, t-\ell) = hat{\gamma}_{v,v^{\prime(c) \ln p_{v^{\prime(t-\ell)</math>, and <math>s_{\max}(t)</math> and <math>s_{\min}(t)</math> be the maximum and the minimum market capitalization among the <math>n</math> listed enterprises at time <math>t</math> and consider <math>SMB(t) = s_{\min}(t) - s_{\max}(t)</math> (small minus big) and <math>SOB(t) = s_{\min}(t)/s_{\max}(t)</math> (small over big). When <math>s^{\Delta}_{v, v^{\prime(t-\ell) = s_v^{\prime}(t-\ell) - s_{v}(t-\ell)</math>, well known properties of the log-normal distribution implies
:<math> \mathbb{E}\left[ p_v(t) ~|~ \boldsymbol{\beta}_{\tau_{\min \ldots \boldsymbol{\beta}_{\tau_{\max \right] \leq \mathbb{E}\left[ p_v(t) ~|~ \boldsymbol{\beta}_{\tau_{\min^{SMB} \ldots \boldsymbol{\beta}_{\tau_{\max^{SMB} \right],</math>
where <math>\beta_{c,\ell}^{SMB} = \beta_{c,\ell} SMB(t-\ell)</math>. Similarly, when <math>s^{\Delta}_{v, v^{\prime(t-\ell) = s_v^{\prime}(t-\ell) - s_{v}(t-\ell)</math>,
:<math> \mathbb{E}\left[ p_v(t) ~|~ \boldsymbol{\beta}_{\tau_{\min \ldots \boldsymbol{\beta}_{\tau_{\max \right] \leq \mathbb{E}\left[ p_v(t) ~|~ \boldsymbol{\beta}_{\tau_{\min^{SOB} \ldots \boldsymbol{\beta}_{\tau_{\max^{SMO} \right],</math>
Next to it, their approach allows decomposing the financial dynamics into networks propagation and firms structural positions effects.
See also
- Financial crisis
- Flight-to-liquidity
- Stock market crash
- Currency crisis
- Systemic risk
- Viral phenomenon