Financial Market Experiments

长期以来，金融产品及其交易市场一直是所有经济互动中最具创新性的。然而，2008年的全球金融危机以及随后出台的监管金融市场的新法律，特别是2010年的《多德·弗兰克法案》，完全有理由引起经济学家的极大关注，既要找出金融机构脆弱性的原因，又要更好地阐明如何改变金融市场的架构，以便在不过分限制这些市场提供的产品和服务的好处或限制创新的情况下提高稳定性。与设计新法规相关的一个固有问题是，缺乏关于从未被遵守的政策的有效性的证据。正是在这种情况下，实验室方法可以非常有用。这个项目包括一系列的三个实验，应用于金融市场监管中的新问题。每一次实验都代表着一种洞察替代政策潜在后果的方法。实验结果将有益地让人们对替代政策选择的潜在后果有一些了解。第一个实验评估将或有资本债券转换为股权的触发机制。或有资本债券(“COCO”)已被确定为促进具有系统重要性银行稳定性的重要新工具。成功实施COCO的关键是有效的转换触发机制。我们的实验关注的是两种领先候选机制的相对表现，一种是固定价格触发机制，其中转换在超过预先宣布的价格门槛后自动发生；另一种是可自由支配的监管机构，后者也看到价格信息，但根据不精确的非价格信号做出决定，并面临干预或不干预的偏见激励。第二个实验与银行挤兑有关。在许多情况下，暂停流动性而不是存款保险是阻止陷入困境的银行挤兑的最佳(或唯一)政策工具。在这种情况下，银行监管机构很少能够简单地暂停支付，直到资产到期--这是最理想的。相反，他们暂时暂停流动性，然后重新为剩余储户设定合同条款。在重新设定合同条款时，银行当局必须选择是否将付款条款偏向于患者的利益。把存款存在银行的储户，还是倾向于不耐烦？有意外流动性需求的储户。与重新签约条件密切相关的是公布储户取款行为信息的(去)稳定效应。我们的实验将重新签约条件和信息制度相互作用，以确定最有可能促进稳定的那种事后重新签约和信息计划。第三个实验评估了一个简单的银行间同业拆借市场模型。在2008年金融危机中，银行间市场？被锁定，？尽管央行积极干预，但这引发了人们的疑问，即为什么市场会崩溃，央行可以做些什么来改善这个市场在压力时期的运作。我们建议对Allen，Carletti和Gale(2009)的模型进行简单的实验实施，以更好地理解银行在这个市场上面临的协调问题，并获得一些关于何时以及如何进行央行持续干预的洞察。这项实验还将作为评估最近更复杂的模型中提出的囤积流动性的替代动机的基线环境。

Financial products and the markets in which they are traded have long been among the most highly innovative of all economic interactions. However, the worldwide financial crisis of 2008 and the ensuing new laws regulating financial markets, including in particular the 2010 Dodd Frank Act, have quite justifiably stimulated immense attention among economists, both to identify the causes of fragility in financial institutions, and to better articulate ways to alter the architecture of financial markets so as to improve stability without unduly limiting the benefits of the products and services offered in these markets or restricting innovation. An inherent problem associated with designing new regulations is an absence of evidence regarding the effectiveness of policies that have never been observed. In exactly these circumstances laboratory methods can be extremely useful. This project consists of a series of three experiments with applications to emerging issues in financial market regulation. Each experiment represents a way to gain some insight into potential consequences of alternative policies. Experimental results will usefully allow some insight into potential consequences of alternative policy options. The first experiment evaluates triggering mechanisms for converting contingent capital bonds into equity. Contingent capital bonds ("CoCo's") have been identified as an important new instrument for promoting the stability of systemically important banks. Critical to the successful implementation of CoCo's is an efficient conversion triggering mechanism. Our experiment focuses on the relative performance of the two leading candidate mechanisms, a fixed price trigger in which conversion occurs automatically after crossing a preannounced price threshold and a discretionary regulator, who also sees price information, but who makes decisions in light of imprecise non-price signals, and also face biased incentives to intervene or not. The second experiment pertains to bank runs. In many instances liquidity suspension rather than deposit insurance is the best (or only) policy tool available for discouraging runs on distressed banks. In such instances, bank regulators are rarely able to simply suspend payments until asset maturation, as would be ideal. Rather, they temporarily suspend liquidity, and then re-set contract terms for remaining depositors. In re-setting contract terms, a banking authority must choose whether to skew the terms of payment toward the interests of ?patient? depositors who maintain their deposits in the bank, or toward ?impatient? depositors with unanticipated liquidity needs. Intimately related to re-contracting conditions are the (de)stabilizing effects of publicizing information about depositor withdrawal behavior. Our experiment interacts re-contracting conditions and information regimes to identify the sort of ex post re-contracting and information scheme most likely to promote stability. The third experiment evaluates a simple model of the interbank loan market. In the 2008 financial crisis interbank markets ?locked up,? despite active central bank intervention, raising questions as to why the market failed and what a central bank can do to improve the functioning of this market in times of stress. We propose a simple experimental implementation of a model by Allen, Carletti and Gale (2009) to both better understand the coordination problem faced by banks in this market and to gain some insight as to when and how persistent central bank intervention may be desirable. This experiment will also serve as a baseline environment for evaluating alternative motivations for liquidity hoarding proposed in recent, more complex models.