漫谈金融全球化 魏尚进 纯属个人观点, 言责自负

文献小结: 金融全球化, 理论上好处一条接一条; 费尽吃奶尽, 实证的效果不妙也不巧. 综述文章: Prasad, Rogoff, Wei, and Kose, IMF, 2003 Kose, Prasad, Rogoff, and Wei, 2005 (JEL?)

Turkey India U.S.A. China

Benefits of Financial Integration: Theory Direct Channels Efficient international allocation of capital For capital-poor economies—lower cost of capital, higher investment Spillovers of technological and managerial expertise

Benefits of Financial Integration: Theory Indirect Channels Domestic financial sector development Specialization Disciplining device for macro policies

Benefits of Financial Integration: Theory Effects on Volatility No strong implication for output volatility Consumption smoothing via international risk sharing => Lower consumption volatility

Growth Benefits of Financial Integration: Evidence About 20 studies of growth effects No effect: 4 Mixed: 12 Positive: 4

閾值效应 (Threshold Effect): 政府缺良治, 金融开放则危机增发百姓更吃苦; 政府有良治, 金融开放则投资增加生活有出路. 例子: 新加坡, 香港 阿根廷, 刚果 文献: Wei, 2000; Alfaro, Kalemli-Ozcan and Volosovych, 2003; Klein, 2005; Chinn and Ito, 2005.

组合效应 (Composition Effect): 国际银行贷款两刃刀, 危机增加效率也难保; 国际直接投资效果好, 资本增加风险又减少. 例子: 中国 vs. 阿根廷 文献: Frankel and Rose; Kose, Prasad, Rogoff, and Wei; Frankel and Wei.

The composition hypothesis Not all capital flows are equal FDI and maybe portfolio inflow are more beneficial to growth than debt Desoto and Reisen 2001; Bekaert, Harvey, and Lundblad, 2005, JFE FDI is also less volatile than international bank loans -> More reliance on bank loans increases vulnerability to currency crashes Frankel and Rose, 1996; Frankel and Wei, 2005

Volatility of (FDI/GDP) and (Loan/GDP) (1980-2003, Measured by Standard Deviation) FDI/GDP Loan/GDP

Volatility of (FDI/GDP) and (Loan/GDP) (1980-2003, Measured by Coefficient of Variation) FDI/GDP Loan/GDP

閾值效应 与组合效应: 两者是否可能是同一硬币的两面? 文献中提出为两个独立的解释 魏尚进 (2000, 2001); 魏尚进与武屹 (2002) Corruption is linked to a composition of capital inflows away from FDI and towards bank loans. Feira and Mauro (2005): Higher institutional quality -> higher share of “equity-like” capital flows in total foreign liabilities.

反对意见: Albuqurque (2004 JIE): worse credit rating related to more (not less) FDI/total inflows. 即使成立, 怎么知道是因果关系?

迎击挑战: 金融体制与其他政府良治指标 对外来资本组合的影响可能不一. 寻找工具变量 来确认因果关系.

金融体制与其他政府良治指标 对外来资本组合的影响可能不一. 鞠建东 与 魏尚进 (2005), Caballero, Farhi, and Gourinchas (2005): Low financial development + Low K/L ratio -> inward FDI + capital outflow

寻找工具变量 来确认因果关系. 需要 IV for corruption IV for financial institution

殖民过程代价高, 政权掠夺又残暴; 殖民容易象过桥, 民众财产有担保. 腐败的工具变量: 殖民成本 initial mortality of European colonizers 殖民过程代价高, 政权掠夺又残暴; 殖民容易象过桥, 民众财产有担保. 谁先想出? Acemoglu, Johnson & Robinson 例子: 非洲 与 澳洲 墨西哥 与 加拿大

英美common law, 制约百姓也限当权; 法西civil law, 表面公平实偏特权. 金融体制的工具变量: 法律体系的渊源 (legal origin) 英美式的法律体系更多保护股权人债权人利益 英美common law, 制约百姓也限当权; 法西civil law, 表面公平实偏特权. 谁先想出? La Porta, De-silanes, Shleifer & Vishny 例子: 美国 与 法国 墨西哥 与 加拿大

注意: Settler mortality and legal origin are not highly correlated.

实证结果: 腐败越高, FDI 的比重就越低 (外国银行贷款的比重就越高). 我2000的假说的确反映的是因果关系. 金融体制优劣与外国银行贷款的比重无关系; 但金融体制质量越低, FDI 的比重可能更高.

今后的研究工作 刚才讨论了政府的良治程度怎么通过影响金融全球化的閾值效应 来影响金融全球化的组合效应. 也需知道金融全球化怎么影响政府的行为, 包括向良治转变的改革.

欲知结果, 且听下回分解