The impact of herding on the sustainability of pegged exchange rates

This thesis presents a theoretical analysis of the impact of herding on the sustainability of pegged exchange rates. The first chapter outlines the motivations for this analysis and provides a comprehensive review of existing literature, relating both to currency crises and herding. It also highlights the following questions left unanswered by existing research: i. How is the likelihood of a currency crisis affected by the timing of traders’ moves? In other words, is the likelihood of coordination among traders on an attack strategy higher if traders move simultaneously or if past actions are observable and traders either move sequentially or can choose when to move? ii. How does the presence of a visible large trader affect the probability of coordination among smaller traders on an attack strategy when these smaller traders are heterogeneous and are able to take discrete rather than infinitesimal positions against the currency? iii. In a currency crisis scenario, how are the predictions of sequential move herding models, characterised by information updating over time, affected if gains to coordination exist and these gains are incurred given a majority rather than unanimity requirement? iv. Given the existence of many different sources of information on the underlying fundamentals in currency markets, in a sequential move setting in which this information is costly to collect and evaluate but past actions are freely observable, how large must the cost of information acquisition be in order that traders are dissuaded from investing in information and choose instead simply to follow the actions of preceding traders? Each of these questions forms the basis of the theoretical models presented in chapters 2 - 5. Chapter 2 presents an analysis of the bandwagon behaviour possible when traders take their actions over time and compares this with the outcome when actions are taken simultaneously. Chapter 3 presents an analysis of the coordinating effects of a large trader and the associated impact his presence has on the likelihood of currency crises. Chapter 4 presents a model closer in spirit to traditional herding models, but incorporating both information updating and gains to coordinated actions, and compares the predictions of this model with those from seminal herding research and chapter 5 presents an analysis of the effects of costs of information acquisition on traders’ incentives to herd. All references referred to in the text and footnotes of chapters 1- 5 of the thesis are provided in the bibliography in chapter 6.