بازار های سرایت در طول بحران مالی: یک رویکرد رژیم سوئیچینگ

The recent financial crisis provides us with an opportune backdrop to investigate the contagion effects among the stock market, credit default market, real estate market, and the energy market. In this paper, we employ a Markov regime-switching VAR framework to delve into the process and the magnitude of the impacts of certain economic shocks on different markets and how the effects spill over to the other segments of the economy. In the early 2000s, after the high-tech bubble rocked the US stock market, the government decided to pump more funds into the real estate market to keep prices stable. As a house is usually the largest single asset of most households and its value represents an important component of the aggregate portfolio of financial intermediaries, housing policies usually have pervasive economic and social effects and attentions have been focused on the real estate price stability and mortgage affordability. Quasi-government agencies such as Fannie Mae and Freddie Mac purchased a significant amount of subprime-laden securities. As the US housing bubble peaked, a highly competitive and liquid market had made mortgage money easily available to households which otherwise would not qualify for underwriting. These low credit mortgages planted the seeds of the eventual subprime mortgage crisis. The onset of financial crisis started after the real estate market peaked in 2006. Mortgage delinquencies and foreclosures rose sharply, which in turn caused huge losses for banks and financial institutions that held a large amount of mortgages and mortgage-backed securities. During the first half of 2007, several mortgage companies, including ResMae and New Century Financial, filed for bankruptcy protection. By August 2007, Countrywide drew $11.5 billion from credit lines and Bank of America injected $2 billion of equity capital into Countrywide. Such events began to spark a widespread loss of confidence in the banking system in the investors' mind. As a result, banks tightened their lending standards, hence ignited another round of credit crisis. The crisis reached a critical point in September 2008 during which the Federal Housing Finance Agency placed Fannie Mae and Freddie Mac in government conservatorship, Bank of America purchased Merrill Lynch, Lehman Brothers filed for Chapter 11 protection, and the Federal Reserve Board lent $85 billion to the American International Group. Companies and financial institutions hastened to deleverage to reduce their risk exposures, hence selling massive assets at discount. The tightening of the credit default market and liquidity, the decline of the real estate value, the rapid rising energy prices in conjunction with the tremendous loss of wealth in the stock market finally cut into economic growth, which started a domestic recession and the recession propagated into a global one. Therefore we are granted with this rare opportunity to study the dynamics of market contagion effects arising from various economic shocks. Not only this financial crisis involves the credit default swap market which was almost non-existent in the past recessions, the simultaneous collapse of the housing market and the stock market also differs from past experiences. Our analysis centers on market contagion during the financial crisis. Specifically, in this setting we study the magnitude and the process whereby certain economic shocks impact different segments of the economy. This issue not only warrants a timely study but also could be paramount to scholars and market participants. Toward this end, we utilize a multivariate vector autoregressive (VAR) model with a Markov regime-switching feature in order to accommodate any potential regime shifts. As Dungey and Zhumabekova (2001) have shown, tests of contagion effects can be seriously affected by the size of the “crisis” and “non-crisis” periods. Our empirical results demonstrate that the watershed of regimes occurs when the subprime crisis starts in 2007, after which the “risky” regime (larger mean and high volatility) dominates the evolution of market chaos. The technical virtue of our model makes it accommodate the data and actual events sufficiently well. Subject to different market shocks, the responses of all markets exhibit regime-dependent patterns. These effects tend to be more dramatic in the “risky” regime, which corresponds to the financial crisis period. In the low volatility or “stable” regime, impulse response functions and variance decomposition analysis suggest that economic factors' own shocks explain the lion's share of the variations. On the other hand, the high volatility or “risky” regime finds that excluding their own shocks, stock market shock and oil price shock are the main driving forces behind the evolution of the credit default market and stock market, respectively. Contrary to general perception, however, the impacts from the credit default market and stock market on the real estate market are not significant. The paper proceeds as follows. The next section reviews the background and related studies on the four markets in question. In Section 3, we present our econometric methodologies. Data source and preliminary analysis are provided in Section 4. We provide in detail the empirical results and their implications in Section 5. The last section summarizes the main findings.

The recent financial crisis provides us with an opportune backdrop to investigate the contagion effects among the stock market, credit default market, real estate market, and the energy market. In this paper, we employ a Markov regime-switching VAR framework to delve into the process and the magnitude of the impacts of certain economic shocks on different markets and how the effects spill over to the other segments of the economy. In the early 2000s, after the high-tech bubble rocked the US stock market, the government decided to pump more funds into the real estate market to keep prices stable. As a house is usually the largest single asset of most households and its value represents an important component of the aggregate portfolio of financial intermediaries, housing policies usually have pervasive economic and social effects and attentions have been focused on the real estate price stability and mortgage affordability. Quasi-government agencies such as Fannie Mae and Freddie Mac purchased a significant amount of subprime-laden securities. As the US housing bubble peaked, a highly competitive and liquid market had made mortgage money easily available to households which otherwise would not qualify for underwriting. These low credit mortgages planted the seeds of the eventual subprime mortgage crisis. The onset of financial crisis started after the real estate market peaked in 2006. Mortgage delinquencies and foreclosures rose sharply, which in turn caused huge losses for banks and financial institutions that held a large amount of mortgages and mortgage-backed securities. During the first half of 2007, several mortgage companies, including ResMae and New Century Financial, filed for bankruptcy protection. By August 2007, Countrywide drew $11.5 billion from credit lines and Bank of America injected $2 billion of equity capital into Countrywide. Such events began to spark a widespread loss of confidence in the banking system in the investors' mind. As a result, banks tightened their lending standards, hence ignited another round of credit crisis. The crisis reached a critical point in September 2008 during which the Federal Housing Finance Agency placed Fannie Mae and Freddie Mac in government conservatorship, Bank of America purchased Merrill Lynch, Lehman Brothers filed for Chapter 11 protection, and the Federal Reserve Board lent $85 billion to the American International Group. Companies and financial institutions hastened to deleverage to reduce their risk exposures, hence selling massive assets at discount. The tightening of the credit default market and liquidity, the decline of the real estate value, the rapid rising energy prices in conjunction with the tremendous loss of wealth in the stock market finally cut into economic growth, which started a domestic recession and the recession propagated into a global one. Therefore we are granted with this rare opportunity to study the dynamics of market contagion effects arising from various economic shocks. Not only this financial crisis involves the credit default swap market which was almost non-existent in the past recessions, the simultaneous collapse of the housing market and the stock market also differs from past experiences. Our analysis centers on market contagion during the financial crisis. Specifically, in this setting we study the magnitude and the process whereby certain economic shocks impact different segments of the economy. This issue not only warrants a timely study but also could be paramount to scholars and market participants. Toward this end, we utilize a multivariate vector autoregressive (VAR) model with a Markov regime-switching feature in order to accommodate any potential regime shifts. As Dungey and Zhumabekova (2001) have shown, tests of contagion effects can be seriously affected by the size of the “crisis” and “non-crisis” periods. Our empirical results demonstrate that the watershed of regimes occurs when the subprime crisis starts in 2007, after which the “risky” regime (larger mean and high volatility) dominates the evolution of market chaos. The technical virtue of our model makes it accommodate the data and actual events sufficiently well. Subject to different market shocks, the responses of all markets exhibit regime-dependent patterns. These effects tend to be more dramatic in the “risky” regime, which corresponds to the financial crisis period. In the low volatility or “stable” regime, impulse response functions and variance decomposition analysis suggest that economic factors' own shocks explain the lion's share of the variations. On the other hand, the high volatility or “risky” regime finds that excluding their own shocks, stock market shock and oil price shock are the main driving forces behind the evolution of the credit default market and stock market, respectively. Contrary to general perception, however, the impacts from the credit default market and stock market on the real estate market are not significant. The paper proceeds as follows. The next section reviews the background and related studies on the four markets in question. In Section 3, we present our econometric methodologies. Data source and preliminary analysis are provided in Section 4. We provide in detail the empirical results and their implications in Section 5. The last section summarizes the main findings.

The subprime mortgage meltdown, the massive default in the credit default swap market, the crash of the stock market and the skyrocketing oil prices are factors believed to have played complex roles in the recent financial turmoil and recession. However, relatively little information is known as to what extent the four markets interact and contribute to the catastrophic economic meltdown. In this paper, we model the nonlinear relationships of these risk factors within a four-dimensional Markov switching VAR model. Using weekly data of oil price, stock index, CDS index and housing price index from October 2003 through March 2009, we investigate the key issue mentioned above and several important conclusions have been derived from our analysis. The results from a Markov switching specification reveal that the contagion effects among these markets are characterized by nonlinearity with two distinct regimes. The watershed of the two regimes within the MS-VAR framework occurs around the start of the subprime crisis in 2007, after which the “risky” regime (larger mean and high volatility) dominates the evolution of the market chaos. In addition, the duration of a “stable” market is found to be about twice longer than that of a “risky” market. This finding is consistent with the empirical observation that the durations of economic booms tend to be longer than those of economic slumps. The technical virtue of our model makes it accommodate the data and actual events sufficiently well. Assessment of the relations of the economic variables utilizing regime-dependent impulse response functions reveals that all market indicators respond more significantly to various economic shocks when the “risky” regime is dominant. That is, the contagion effects among markets are more prominent during the financial crisis. Specifically, we observe that the stock market activities generate higher volatilities in the CDS market. The contagion effects between stock and energy markets also appear to be larger in the “risky” regime. However, the impacts from the credit default market on the real estate market are not significant as expected. Similar conclusions emerge from the variance decomposition analysis. This study also uncovers a few findings that may differ from the general public postulations but are worth noting. First, credit default market shows strong reactions to the stock market fluctuations, but obviously is much less responsive to the housing market shock in the “risky” regime. Second, although the real estate market shock is an important factor in explaining the variation in the stock market in the “stable” regime, its influence diminishes during the financial crisis period, after the oil and CDS market shocks. Finally, we find that the stock market ranks as the second largest contributor in the fluctuations of the variation in the energy market —bonly after its own shock during the financial crisis. Interestingly, the impact on the stock market from itself subsides rapidly in the “risky” regime, while the other factors gain momentum in their influences over the longer horizon. Our findings have important implications for policymakers regarding the amplified contagion effects among the markets during financial crisis, so as to avoid any reckless lack of oversight. In particular, they should carefully examine and uncover the underlying primary driving forces behind the crisis, and take precautions against the potential risk factors in making future policy decisions. It is critical for policymakers to guide investors to pay special attention to those unexpected factors arising from various markets.