THEORETICAL AND EMPIRICAL STUDIES ON INTERNATIONAL FINANCIAL MARKETS AND FINANCIAL REGULATION

在这项研究计划中取得的成果来自四个相互关联的项目，这些项目考察了与利率平价重新调整、银行部门的羊群行为以及股票市场波动与外汇波动和期权市场隐含波动率的关系相关的重要问题。综合结果揭示了国际金融市场和机构的情况。从经济一体化的角度来看，有证据表明亚洲实际利差的行为是非线性的，因为实际利率平价的大冲击比小冲击更有可能导致更快的重新调整(Holmes和Maghrebi，IREF，2004)。结果还提供了不对称货币政策的证据，这表明对于许多国家来说，在名义利率上升和通胀下降的制度下，与美国的联系可能不那么普遍(Holmes和Maghrebi，JIFMIM…更多，即将推出)。根据一种评估G7国家之间经济和金融一体化程度的新方法，也有证据表明，美国和加拿大之间的一体化程度更强，而与日本的一体化程度要弱得多。没有明确证据表明与欧洲汇率机制危机或引入欧元等重大事件相关的条件相关性结构性崩溃(Holmes和Maghrebi，EREF，2005年)。从行为金融学的角度来看，从城市银行到地区性银行，从长期信贷银行和信托银行到城市银行，从城市银行到地区性银行，从行为金融学的角度来看，研究结果表明日本贷款市场存在羊群行为。来自日本银行的一组数据的证据表明，存在一种周期性的羊群效应。仅在金融泡沫期间观察到的非理性羊群行为的迹象表明，非理性的银行行为是导致20世纪80年代末不良贷款问题的原因(Nakagawa和Uchida，WP 2003&Uchida和Nakagawa，WP 2005)。就金融市场之间的关系而言，来自动态协方差模型的新证据表明，市场波动性往往是同等的，对股市的坏消息比好消息更敏感，对货币贬值的反应比升值更快。结果表明，升值比贬值更有利于降低货币市场的波动性。此外，将坏消息与货币贬值联系在一起，可能会增加货币市场的波动性，并影响股市杠杆效应的程度。这些实证结果可能在一定程度上解释了亚洲国家对以邻为壑的汇率政策的厌恶(Maghrebi，Holmes和Pentecost RPBFMP，即将出版)。最后，这项研究计划的另一个成果是开发了日本股票市场预期波动率的新基准，以便综合对期权价格未来波动率的预期。日经225隐含波动率在金融数据库中不可用，它是按照芝加哥期权交易所发布的新波动率指数的基本方法构建的。有证据表明，尽管隐含波动率存在上升倾向，但它比替代GARCH模型更能反映已实现波动率的变化，后者解释了波动率的持续性和新闻的不对称影响。鉴于隐含波动率的高出样本表现，证据支持了从期权价格推断未来股市波动性的基本原理，以及新的隐含波动率指数的有用性(Nishina，Maghrebi和Kim，DP 2006)。较少

The results achieved during this research program derive from four interrelated projects, which examine important issues related to the realignment towards interest rate parity, the herding behavior in the banking sector, and stock market volatility in its relationship with foreign exchange fluctuations and with implied volatility from options markets. The aggregate results shed light on international financial markets and institutions. From the perspective of economic integration, there is evidence of nonlinearities in the behavior of Asian real interest differentials in the sense that large shocks to real interest parity are more likely to lead to faster readjustments than smaller shocks (Holmes and Maghrebi, IREF, 2004). The results also provide evidence of asymmetric monetary-policy, which suggests that for many countries, linkages with respect to the U.S. are likely to be less prevalent in a regime of rising nominal interest rates and falling inflation (Holmes and Maghrebi, JIFMIM … More , forthcoming). Based on a new approach for assessing the degree of economic and financial integration between G7 countries, there is also evidence of stronger integration between the US and Canada and much weaker integration with Japan. There is no clear evidence of structural breakdown in conditional correlation associated with significant events such as the ERM crisis or the introduction of the Euro (Holmes and Maghrebi, EREF, 2005). From the perspective of behavioral finance, the results suggest the existence of herding behavior in the Japanese loan market with evidence of causality from city banks to regional banks, from long-term credit banks and trust banks to city banks. The evidence from a panel data of Japanese banks suggests the existence of a cyclical pattern of herding. The signs of irrational herding observed only in association with the financial bubble, suggest that the irrational bank behavior was contributive to the bad loan problems in the late 1980s (Nakagawa and Uchida, WP 2003 & Uchida and Nakagawa, WP 2005). As far as the relationship between financial markets is concerned, there is new evidence from dynamic covariance modelling that market volatility tends to be ceteris paribus, more responsive to bad news about equity than good news and to currency depreciation than appreciations. The results indicate that appreciations are conducive to lower volatility in currency markets than depreciation. Also, the association of bad news with currency depreciation is likely to increase the level of volatility in currency markets and affect the extent of leverage effects in equity markets. These empirical results may partly explain the aversion of Asian countries toward beggar-thy-neighbor policies of exchange rate (Maghrebi, Holmes and Pentecost RPBFMP, forthcoming). Finally, another, achievement of this research program is the development of a new benchmark of expected volatility in the Japanese stock market in order to aggregate anticipation of future volatility across options prices. The Nikkei 225 implied volatility is not available in financial databases and it is constructed following the methodology underlying the new VIX index disseminated by Chicago Board of Options Exchange. There is evidence that despite its upward bias, implied volatility is more reflective of changes in realized volatility than alternative GARCH models, which account for volatility persistence and the asymmetric impact of news. Given the higher out-of sample performance of implied volatility the evidence lend support to the rationale behind drawing inference about future stock market volatility from option prices and to the usefulness of the new implied volatility index (Nishina, Maghrebi and Kim, DP 2006). Less