Handbook of Exchange Rates E-Book
147,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Serie: Wiley Handbooks in Financial Engineering and Econometrics
- Sprache: Englisch
Praise for Handbook of Exchange Rates "This book is remarkable. I expect it to become the anchor reference for people working in the foreign exchange field." --Richard K. Lyons, Dean and Professor of Finance, Haas School of Business, University of California Berkeley "It is quite easily the most wide ranging treaty of expertise on the forex market I have ever come across. I will be keeping a copy close to my fingertips." --Jim O'Neill, Chairman, Goldman Sachs Asset Management How should we evaluate the forecasting power of models? What are appropriate loss functions for major market participants? Is the exchange rate the only means of adjustment? Handbook of Exchange Rates answers these questions and many more, equipping readers with the relevant concepts and policies for working in today's international economic climate. Featuring contributions written by leading specialists from the global financial arena, this handbook provides a collection of original ideas on foreign exchange (FX) rates in four succinct sections: * Overview introduces the history of the FX market and exchange rate regimes, discussing key instruments in the trading environment as well as macro and micro approaches to FX determination. * Exchange Rate Models and Methods focuses on forecasting exchange rates, featuring methodological contributions on the statistical methods for evaluating forecast performance, parity relationships, fair value models, and flow-based models. * FX Markets and Products outlines active currency management, currency hedging, hedge accounting; high frequency and algorithmic trading in FX; and FX strategy-based products. * FX Markets and Policy explores the current policies in place in global markets and presents a framework for analyzing financial crises. Throughout the book, topics are explored in-depth alongside their founding principles. Each chapter uses real-world examples from the financial industry and concludes with a summary that outlines key points and concepts. Handbook of Exchange Rates is an essential reference for fund managers and investors as well as practitioners and researchers working in finance, banking, business, and econometrics. The book also serves as a valuable supplement for courses on economics, business, and international finance at the upper-undergraduate and graduate levels.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 1433
Veröffentlichungsjahr: 2012
Ähnliche
Table of Contents
Series Page
Title Page
Copyright
Dedication Page
Preface
Contributors
Part I: Overview
Chapter 1: Foreign Exchange Market Structure, Players, and Evolution
1.1 Introduction
1.2 Geography and Composition of Currency Trading
1.3 Players and Information in FX Market
1.4 Electronic Trading Revolution in FX Market
1.5 Survey of Multibank FX Platforms
1.6 Summary
1.7 Glossary
1.8 Acknowledgments
References
Chapter 2: Macro Approaches to Foreign Exchange Determination
2.1 Introduction
2.2 Models of the Nominal Exchange Rate
2.3 Real Models of the Real Exchange Rate
2.4 New Directions in Exchange-Rate Modeling
2.5 Conclusions
2.6 Acknowledgments
References
Chapter 3: Micro Approaches to Foreign Exchange Determination
3.1 Introduction
3.2 Perspectives on Spot-Rate Dynamics
3.3 Currency Trading Models and their Implications
3.4 Exchange Rates, Order Flows, and the Macro Economy
3.5 Conclusion
3.6 Appendix
3.7 Acknowledgment
References
Chapter 4: The Exchange Rate in a Behavioral Finance Framework
4.1 Introduction
4.2 Exchange Rate Puzzles
4.3 A Prototype Behavioral Model of the Foreign Exchange Market
4.4 Conclusion
References
Chapter 5: The Evolution of Exchange Rate Regimes and Some Future Perspectives
5.1 Introduction
5.2 A Brief History of Currency Regimes
5.3 Performance of the Laisser-Faire Exchange Rate System, 1973–2010
5.4 Trends in Currency Use
5.5 Prospects for the Future
5.6 Concluding Comments
5.7 Appendix A: A Formal Test of Hollowing Out
References
Part II: Exchange Rate Models and Methods
Chapter 6: Purchasing Power Parity in Economic History
6.1 Introduction
6.2 Categorization of Purchasing-Power-Parity Theories
6.3 Historical Application of PPP: Premodern Periods
6.4 Techniques of Testing PPP Theory in Economic-History Literature
6.5 Price Variable in PPP Computations
6.6 Modern Period: Testing of PPP
6.7 Analysis of U.S. Return to Gold Standard in 1879
6.8 Establishment and Assessment of a Fixed Exchange Rate in Interwar Period
6.9 Conclusions
References
Chapter 7: Purchasing Power Parity in Tradable Goods
7.1 Introduction
7.2 The LOP and Price Indices
7.3 Empirical Evidence on the LOP
7.4 Purchasing Power Parity
7.5 Aggregating from the LOP to PPP: What Can We Infer?
7.6 Conclusion and Implications
7.7 Appendix: TAR modeling
7.8 Acknowledgments
References
Chapter 8: Statistical and Economic Methods for Evaluating Exchange Rate Predictability
8.1 Introduction
8.2 Models for Exchange Rate Predictability
8.3 Statistical Evaluation of Exchange Rate Predictability
8.4 Economic Evaluation of Exchange Rate Predictability
8.5 Combined Forecasts
8.6 Empirical Results
8.7 Conclusion
8.8 Appendix A: The Bootstrap Algorithm
8.9 Acknowledgments
References
Chapter 9: When Are Pooled Panel-Data Regression Forecasts of Exchange Rates More Accurate than the Time-Series Regression Forecasts?
9.1 Introduction
9.2 Panel Data Exchange Rate Determination Studies
9.3 Asymptotic Consequences of Pooling
9.4 Monte Carlo Study
9.5 An Illustration with Data
9.6 Conclusions
References
Chapter 10: Carry Trades and Risk
10.1 Introduction
10.2 The Carry Trade: Basic Facts
10.3 Pricing the Returns to the Carry Trade
10.4 Empirical Findings
10.5 Time-Varying Risk and Rare Events
10.6 Conclusion
10.7 Acknowledgments
References
Chapter 11: Currency Fair Value Models
11.1 Introduction
11.2 Models/Taxonomy
11.3 Implementation Choices and Model Characteristics
11.4 Conclusion
11.5 Acknowledgments
References
Chapter 12: Technical Analysis in the Foreign Exchange Market
12.1 Introduction
12.2 The Practice of Technical Analysis
12.3 Studies of Technical Analysis in the Foreign Exchange Market
12.4 Explaining The Success of Technical Analysis
12.5 The Future of Research on Technical Analysis
12.6 Conclusion
12.7 Acknowledgments
References
Chapter 13: Modeling Exchange Rates with Incomplete Information
13.1 Introduction
13.2 Basic Monetary Model
13.3 Information Heterogeneity
13.4 Model Uncertainty
13.5 Infrequent Decision Making
13.6 Conclusion
13.7 Acknowledgments
References
Chapter 14: Exchange Rates in a Stochastic Discount Factor Framework
14.1 Introduction
14.2 Exchange Rates and Stochastic Discount Factors
14.3 Empirical Evidence
14.4 Models
14.5 Conclusion
References
Chapter 15: Volatility and Correlation Timing in Active Currency Management
15.1 Introduction
15.2 Dynamic Models for Volatility and Correlation
15.3 The Economic Value of Volatility and Correlation Timing
15.4 Parameter Uncertainty in Bayesian Asset Allocation
15.5 Model Uncertainty
15.6 Empirical Results
15.7 Conclusion
15.8 Appendix A: Univariate Models for Volatility Timing
15.9 Appendix B: Parameter Uncertainty and the Predictive Density
15.10 Acknowledgments
References
Part III: FX Markets and Products
Chapter 16: Active Currency Management Part I: Is There a Premium for Currency Investing (Beta)
16.1 Introduction
16.2 Beta in the Foreign Exchange Markets
16.3 Multiple Forms of FX Beta
16.4 Carry FX Indices from Banks
16.5 Trend-Following FX Indices from Banks
16.6 Conclusion
References
Chapter 17: Active Currency Management Part II: Is There Skill or Alpha in Currency Investing?
17.1 Introduction
17.2 Alternative Currency Management Mandates
17.3 Benchmarks for Currency Fund Management
17.4 Empirical Evidence with the Barclay Currency Traders Index and Individual Fund Managers
17.5 Empirical Evidence: Fund Managers on the DB FX Select Platform
17.6 Conclusions and Investment Implications
References
Chapter 18: Currency Hedging for International Bond and Equity Investors
18.1 Introduction
18.2 Overview of Empirical Hedging Studies
18.3 Return and Volatility Impact of Currency Hedging
18.4 Hedge Instruments—Currency Forwards versus Options
18.5 Managing Tracking Error in Forward Hedges
18.6 Conclusions
References
Chapter 19: FX Reserve Management
19.1 FX Reserve Management
19.2 FX Reserve Uses
19.3 FX Reserve Sources
19.4 Objectives of Reserves Management
19.5 Techniques of Reserve Management
19.6 Historical Perspective
19.7 What Assets Do Central Banks Hold?
19.8 Constraints
19.9 External Managers
19.10 Costs of Accumulation and Holding of Reserves
19.11 Diversification
19.12 Challenges to Diversification and Size of Reserves
19.13 Changing Role of the Dollar as the International Reserve Currency
19.14 Reserve Management if the Dollar is Replaced as the Reserve Currency
19.15 Conclusion
19.16 Acknowledgments
References
Chapter 20: High Frequency Finance: Using Scaling Laws to Build Trading Models
20.1 Introduction
20.2 The Intrinsic Time Framework
20.3 Scaling Laws
20.4 The Scale of Market Quakes
20.5 Trading Models
20.6 Conclusion
20.7 Acknowledgments
References
Chapter 21: Algorithmic Execution in Foreign Exchange
21.1 Introduction
21.2 Key Components of an Algorithmic Execution Framework
21.3 Types of Algorithms
21.4 What Execution Strategies are Most Effective?
21.5 Looking Forward
21.6 Appendix A
References
Chapter 22: Foreign Exchange Strategy Based Products
22.1 Introduction
22.2 Evolution of the Foreign Exchange Market
22.3 Foreign Exchange Investable Indices and Strategy-Based Products
22.4 Conclusion
References
Chapter 23: Foreign Exchange Futures, Forwards, and Swaps
23.1 Introduction
23.2 Market Basics and Size
23.3 Dislocations of the FX and Cross-Currency Swap Markets under Financial Crises
23.4 Conclusion
23.5 Acknowledgments
References
Chapter 24: FX Options and Volatility Derivatives: An Overview from the Buy-Side Perspective
24.1 Introduction
24.2 Why Would One Bother With an Option?
24.3 Market for FX Options
24.4 Volatility
24.5 FX Options from the Buy-Side Perspective
24.6 Acknowledgment
References
Part IV: FX Markets and Policy
Chapter 25: A Common Framework for Thinking about Currency Crises
25.1 Introduction
25.2 The KFG Model
25.3 Extensions
25.4 Empirical Work
25.5 Conclusion
References
Chapter 26: Official Intervention in the Foreign Exchange Market
26.1 Introduction
26.2 Official FX Interventions and Reserve Accumulation: Stylized Facts, Motives, and Effects
26.3 Empirical Evidence on the Effectiveness of Official FX Interventions
26.4 Conclusions
26.5 Acknowledgements
References
Chapter 27: Exchange Rate Misalignment—The Case of the Chinese Renminbi
27.1 Introduction
27.2 Background
27.3 Undervalued or Overvalued
27.4 Concluding Remarks
27.5 Acknowledgments
References
Chapter 28: Choosing an Exchange Rate Regime
28.1 Five Advantages of Fixed Exchange Rates
28.2 Econometric Evidence on the Bilateral Trade Effects of Currency Regimes
28.3 Five Advantages of Floating Exchange Rates
28.4 How to Weigh Up the Advantages of Fixing Versus Floating
28.5 Country Characteristics That Should Help Determine the Choice of Regime
28.6 Alternative Nominal Anchors
References
Index
Wiley Series
Copyright 2012 by John Wiley & Sons, Inc. All rights reserved
Published by John Wiley & Sons, Inc., Hoboken, New Jersey
Published simultaneously in Canada
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.
Library of Congress Cataloging-in-Publication Data:
Handbook of exchange rates / [edited by] Jessica James, Ian W. Marsh, Lucio Sarno.
pages cm. – (Wiley handbooks in financial engineering and econometrics ; 2)
Includes bibliographical references and index.
ISBN 978-0-470-76883-9
1. Foreign exchange rates. 2. Foreign exchange. I. Sarno, Lucio.
II. James, Jessica, 1968– III. Marsh, Ian W., 1966–
HG3810.H35 2012
332.4′56–dc23
2011052323
For my children Zach and Miriam, my constant sources
of inspiration!
Jessica James
To my wonderful wife Michela
Ian W. Marsh
To my fantastic wife Julia
Lucio Sarno
Preface
The Foreign Exchange market, as we know, was born in the 1970s. To us in the second decade of the third millennium, mainly accustomed to freely floating and convertible currencies, the events that led to its birth would seem almost incredible. It is huge news when a government, via its central bank, intervenes to shift its currency or keep it stable. But from the end of World War II to the mid-1970s, pegs and interventions were the norm for almost all currencies.
The post war era gave us the Bretton Woods agreement that set the United States as the world's reserve currency and pegged most others to it. The United States was backed by gold, and the stability that this system brought enabled trade flows to enormously grow worldwide. This was deliberate; in the preceding decades, the architects of the system had seen how economic stresses could lead countries to war and were driven by a desire to promote trade and allow the world economy to grow.
However, the Vietnam war brought a growing US trade deficit and a drift of the alignment of the United States and the US dollar with the rest of the world. Pressure grew on the exchange rates and one by one, led by the Japanese yen, the currency became freely convertible and floating. By 1976, the gold standard was no more present and the majority of the world's currencies had the form we know today.
From that not-so-auspicious beginning grew the world's largest market. Today, according to the 2010 BIS report, currencies to the value of over US$4 trillion are traded each working day in global markets. The total value of global equity is a fraction of this amount. In most countries, the traded volume of the currency vastly exceeds the total amount of government debt. Foreign exchange has indeed come of age.
But what is it like, compared to other rates? It is not an interest rate. An interest rate is highly predictable and stable between close to zero, and perhaps exceptionally 15%. Outside these ranges, interest rates historically do not last long; high rates are often followed by spiralling hyperinflations, and negative rates do not occur without anomalous market stresses. It is not an equity, which has a floor at zero and is at least expected to appreciate with inflation.
There is a symmetry to foreign exchange, which is lacking in other market rates. One is “long” or “short” of an equity; in foreign exchange to be long of one currency is to be short of another. It is a rate of exchange and not an ownership of an asset. It has no floor or ceiling, no particular “stable region.” As befits the largest global market, it is also the easiest and cheapest to trade in, with transactions often costing just a single basis point—1% of 1%—in bid-offer spread.
How and why has it grown to such a size? Partly, it is the nature of the beast and the way it is traded. Most foreign exchange transactions are done on a margined basis, with the principal amounts being notional only. Thus, for a simple forward trade of 1 month's duration, where a rate of exchange is locked in for 1 month ahead, no capital is exchanged at the start. At the end, the difference between the agreed rate and the market rate is paid out, scaled by the notional amount of the deal. This is usually at most a few percent of this notional amount. So one can see that it is unnecessary to actually own a sum of money equal to the notional amount—it is more of a scaling factor to the deal. Hence, the notional flows in the market can be very large relative to a smaller capital base.
Another reason for the size of the foreign exchange flows is that speculative trades, which have grown vastly in size and number over the last few years, are usually leveraged. The foreign exchange market is somewhat less volatile than the equity or commodity market, and so to generate comparable profits, profit seeking traders usually apply a multiplier to the notional amount. The last decade has seen the development of foreign exchange as an asset class, where structures suitable for investors are created around foreign exchange transactions. These often utilize large notional amounts to generate returns.
These additions increase an already considerable bulk of foreign exchange trades in the market. One critical activity made possible by the advent of floating exchange rates is hedging. Overseas investments can have their value drastically altered by fluctuations in the foreign exchange rates; hedges allow investors and corporates to immunize their portfolios and cashflows against these variations. Thus the growth of overseas investment has itself driven the volume of foreign exchange transactions higher; the investments, particularly less volatile fixed income instruments, will be significantly affected by foreign exchange movements unless they are hedged.
What are the instruments that make up this flow? The vast majority are spot and forward transactions, but the option market is also deep and liquid. This market came into its own in 1983 when Garman and Kolhagen published the formula for an option on a foreign exchange rate. Additionally, a vast variety of structures may be constructed, which allow investors or corporates to eliminate or transform risks, or gain exposure to specific areas and events.
Why do we need a Handbook of Exchange Rates? In part because it is so important, and in part because it is evolving so rapidly. In the last 5 years, we have seen the growth of foreign exchange as an asset class, which 10 years ago was the exclusive property of large trading floors and a few fund managers. Most large investors will now have a portion of their risk allocated to active strategies, which deliver fundlike returns derived from positioning themselves in the foreign exchange market. Another significant change is the growth of correlation and volatility derivatives. The correlation of rates has become a traded quantity, allowing a remarkably subtle set of risks and rewards to be accessed. The fascinating dynamics of high frequency data are now largely understood via scaling laws, to the extent that co-location—the siting of trading hubs close to exchanges—is becoming an issue; the speed of light has become a limiting factor in foreign exchange trading.
Another fascinating development is strongly connected with the correlation properties of foreign exchange rates. In floating emerging markets, the majority of foreign exchange rates are closely connected with their equity markets for obvious reasons. So high is this correlation that one can replicate emerging market equity indices to good accuracy using only foreign exchange rates—with their greater liquidity and low trading costs. Other interesting correlations are those between the Japanese yen or Australian dollar to the world's equity markets. So strong is the negative (in the case of the Yen) or positive (the Australian dollar) correlation that currency positions are often used to hedge equity drawdowns. So the foreign exchange market is being used to proxy for other rates. Its liquidity and depth make this attractive, though market stresses could bring an end to some of these useful correlations.
The Handbook is designed to span this extensive subject, with experts in the different areas contributing to each section. In planning the book, we drew up a list of the key subjects deserving of a chapter and against each we wrote the name of the key people in that topic. In almost every case, to our delight, those people accepted our invitation to contribute. As a result, the chapters have been written by leading specialists in their fields, often with extensive experience in academia and/or professional practice. Each chapter then benefitted from the feedback of at least one anonymous referee and at least one coeditor, and most authors kindly also acted as referees.
The initial section “Overview” covers structure, regimes, and general underlying behaviors. Part 2 covers models and methods, focussing on the predictability—or lack of it—of foreign exchange rates. In Part 3, we go to the practitioner side of the subject to cover hedging, active management, high frequency trading, and products. Part 4 wraps up with a look at policy and a framework for analyzing currency crises. The chapters are mainly in the form of self-contained surveys, and trace the key developments in a well-defined topic with specific reference to the relevant research frontier. Some contributions also present new empirical findings, especially where competing paradigms are evaluated. So rapid has the evolution of this market been, that it would not be surprising if in a few years there are many new chapters to add! Alternatively, market stresses and political crises can influence foreign exchange rates to an enormous extent. It is easy to forget that our current regime of floating market determined rates is recent, and by no means the norm in a historical context. The foreign exchange market is a fascinating subject, which never stops evolving.
We owe a debt of gratitude to many people for helping us bring this project to a successful conclusion: the contributing authors, for the high quality of their chapters; the many anonymous referees who provided, without exception, valuable feedback to the authors; and Chiara Banti who collated the chapters and worked very hard to copy edit the first draft of the Handbook. Needless to say, without their help, we would never have completed this project.
Jessica James
Ian W. Marsh
Contributors
Part I
OVERVIEW
Chapter 1
Foreign Exchange Market Structure, Players, and Evolution
Michael R. King
Richard Ivey School of Business, University of Western Ontario
Carol Osler
Brandeis International Business School, Brandeis University
Dagfinn Rime
Norges Bank, Norwegian University of Science and Technology
1.1 Introduction
It would be hard to overstate the importance of exchange rates for the world economy. They affect output and employment through real exchange rates. They affect inflation through the cost of imports and commodity prices. They affect international capital flows through the risks and returns of different assets. Exchange rates are justifiably a major focus for policymakers, the public, and, of course, the media.
To understand exchange rates, it is essential to know how these prices are determined. This chapter describes the FX market and presents new evidence on recent trends, thereby setting the stage for the rest of the handbook. It presents stylized facts on the market's size and composition. It then describes more closely the motives, incentives, and constraints of the major players. Trading is a search problem, and the constraints and costs related to this search are affected by the structure of the market. Our starting point is that the exchange rates are primarily driven by new information about economic fundamentals. In this light, we review which agents bring information to the market and exactly how their information becomes embedded in the market price.
The chapter describes the momentous changes in trading practices and market structure which have taken place over the recent decades. It concludes by presenting new evidence on some of the most recent technological advances. Twenty-five years ago, most FX trading involved the telephone and all trading involved institutions: individuals were essentially shut out. Trading was opaque, there was a sharp division between interdealer trading and dealer–customer trading, and market concentration among dealers was low.
Today, only the least liquid corners of the FX market can still be described this way. In the liquid markets, telephones are obsolete. New electronic trading platforms have streamlined trade processing and settlement, reduced operational risks, and lowered trading costs. Lower trading costs have enabled the participation of retail traders and the adoption of new strategies such as high frequency trading—a form of computer-automated trading that relies on high execution speeds to make profits from small price movements. Since streaming real-time prices are now available to virtually all participants, these markets are now regarded as transparent. On some of the new platforms, any trader can provide liquidity, so the division between dealers and their most sophisticated customers is, at times, ambiguous. To remain competitive, the major dealing banks have made heavy investments in software and hardware, which has led to sharply higher concentration among market makers.
1.2 Geography and Composition of Currency Trading
Given the pervasive influence of exchange rates, it is no surprise that the dollar value of trading activity in spot and forward FX market dwarfs most other economic measures (BIS, 2010). With daily average turnover estimated at $2.0 trillion, the market is 36 times larger than the combined exports and imports for the world's 35 largest economies, 16 times their combined GDP, and roughly 10 times exchange-traded equity turnover.
FX trading volume has exploded reflecting an electronic revolution that has lowered trading costs, attracted new groups of market participants, and enabled aggressive new trading strategies. Between 1998 and 2010, turnover in the FX market grew by over 250% (BIS, 2010). The associated 8.4% average annual growth rate far exceeds the contemporary 5.5% annual expansion of global real GDP (Table 1.1).
Table 1.1 FX Turnover and Growth: Comparison with Trade, GDP, and Equity Trading Volume
Many aspects of the FX market remain constant despite the electronic revolution. As has been true for decades, the markets remain decentralized with high liquidity and continuous trading (Lyons, 2001; Rime, 2003; Osler, 2009). As ever, the trading day begins when dealers arrive for work in Australia and Asia. Activity then moves to Europe when markets open in Frankfurt, London, and Paris, and finishes late in the afternoon in New York. As always, there is no time during the day when the market formally closes, although there is a brief lull in activity between about 19:00 and 22:00 GMT, when most New York traders have gone home and most Sydney traders are still on their way to work (Figure 1.1). As ever, overall market liquidity is deepest when both London and New York are open, though liquidity for most individual currencies tends to be deepest during their local trading hours.
Figure 1.1 Average daily interdealer trading activity by the hour across different currencies. Note: The horizontal axis shows hour of the day (GMT) and the vertical axis shows the average number of trades. The five lines are for 1997, 4-year averages for 1998–2001, 2002–2005, and 2006–2009, respectively, and 2010. From example GBP/USD (c), we see the growth in number of trades since 1997. The exchange rates EUR/USD and USD/JPY are now primarily traded on the competing platform EBS, hence the decrease in number of trades from 1997 to 2010 for these two exchange rates. Source: Thomson Reuters Matching.
FX trading remains heavily concentrated in London, which captures over one-third of global trading, and New York, which captures almost one-fifth of trading (Table 1.2). London's traditional dominance in FX grew out of the United Kingdom's worldwide economic dominance in the nineteenth century. It remains secure at the beginning of the twenty-first century because of its geographic location: London's morning session overlaps with Asian trading and its afternoon session overlaps with New York's trading. Trading in the Asia-Pacific region, which in aggregate accounts for about one-quarter of global trading, is dispersed among Tokyo, Hong Kong, Singapore, and Sydney. Latin America, Africa, and the Middle East each account for less than 1% of global turnover.
Table 1.2 Geographical Distribution of Global Foreign Exchange Market Turnover (%)
Despite the continued dominance of London and New York, there have been some subtle shifts in the global distribution of currency trading. The advent of the euro (EUR) brought a decline in the share of European trading outside of London. Meanwhile, rapid economic growth in Asia has supported a surge in trading in the Asian regional centers. Hong Kong and Singapore now vie in importance with traditional European centers such as Switzerland and France.
1.2.1 Which Currencies Are Traded?
Another unchanging aspect of the FX market is the dominance of the US dollar (USD), which is still involved on one side of roughly three-quarters of all spot transactions (Table 1.3).1 The dollar's dominance reflects the market's practice of trading minor currencies via a major currency (called the vehicle currency). A trade from Mexican pesos (MXP) to Australian dollars (AUD), for example, would typically involve two trades, one from MXP to USD and the second from USD to AUD. This “vehicle” trading through the major crosses concentrates liquidity in a narrower range of currency pairs, reducing overall transaction costs.
Table 1.3 Currency Distribution of Spot Turnover (%)
The EUR is involved in 46% of trades, in part because it serves as the vehicle currency within the eurozone.2 The next most actively traded currencies are the JPY (20%) and the UK pound (GBP, 14%). Together, these four currencies are known as “the majors” (or G4).3
The next tier below the majors comprises the AUD (7.5%), the Swiss franc (CHF, 6.2%), and the Canadian dollar (CAD, 5.2%). A notable recent shift is the rising share of the so-called commodity currencies, specifically the AUD, CAD, NOK, and the New Zealand dollar (NZD). These currencies' combined share rose from 7% in 1998 to 16% in 2010.
The share of emerging-market currencies rose sharply in the 1990s but has been fairly stable around 18% since then. Nonetheless, currencies from the most advanced emerging markets, such as the South Korean won (KRW) and Hong Kong dollar (HKD), have more than doubled their market share since 1998 and now rival the Swedish krona (SEK). Turnover in more recently emerging countries, such as Turkey, Thailand, Brazil, and India, has grown even faster.
1.2.2 What Instruments Are Traded?
The dominance of spot FX trading is another area of relative stability.5 Daily spot turnover in 2010 was $1.5 trillion while turnover in outright forwards was far lower, at $0.5 trillion (BIS, 2010). A number of other currency-related instruments—FX futures, currency options, FX swaps, and currency swaps—swell total daily turnover in FX market beyond $4.0 trillion (Table 1.4).6 These assets are traded entirely separately from spot and forward contracts and for entirely different purposes, so they generally have little influence on exchange rates and are not discussed in this chapter.
Table 1.4 Instruments Traded in Global FX Market
FX swaps deserve some discussion, however, because of their immense average daily turnover of $1.8 trillion. Like repos in the fixed income markets, FX swaps are primarily used for overnight position management by banks. Collapsing swap volumes following the Lehman Brothers bankruptcy in late 2008 triggered a rapid expansion of central bank swap activity, as authorities tried to stabilize the world banking system (Baba and Packer, 2009; Melvin and Taylor, 2009). In 2010, FX swap trading activity remained below its previous peak.
1.2.3 How Is Trading Regulated?
The vast majority of FX trading is essentially unregulated, in striking contrast to the extensive regulations in most equity and bond markets.7 Governments have learned through experience that dealers will simply move elsewhere if they are regulated. In the 1960s, for example, bond dealers moved offshore when the US government attempted to regulate the foreign issuance of USD denominated bonds in the domestic market.
Some well-known regulations in other asset markets are missing in FX market. Their absence is not a problem because of unique features of the FX market. Short-sales restrictions, for example, though severe in most developed equity and bond markets, cannot even be defined in this asset class because the sale of one currency is simply the purchase of another. Other practices that are illegal on most organized exchanges are discouraged in FX by market conventions and best practices. For example, front running of customer orders is widely considered bad practice even though it is not illegal.8 Fortunately, the FX market is sufficiently liquid that significant manipulation by any single actor is all but impossible during active trading hours for the major currencies.
Since FX market are subject to minimal regulation, they are also subject to minimal reporting requirements, which explains the scarcity of aggregate data on FX trading. Although equity trading volume is a staple on the evening news on any given day, no one knows how much was traded in FX market—not the regulators, not the monetary authorities, and not even the major FX dealers. The only comprehensive information source is the Triennial Central Bank Survey of FX Market Activity, a joint effort of central banks around the world, coordinated every third year in April by the Bank for International Settlements (BIS).9 In the absence of official sources of high frequency data, most research on currency trading relies on proprietary data from banks and brokers.
1.3 Players and Information in FX Market
A key goal of exchange rate economics is to understand currency returns. Exchange rates, like asset prices, more generally, move in response to new information about their fundamental value. Over the past decade, microstructure research has revealed that this “price discovery” process involves different categories of market participants. Each participant's distinct role is determined by (i) whether the agent is a liquidity maker or taker and (ii) the extent to which the agent is informed.
The original FX market participants were traders in goods and services. Currencies came into existence because they solved the problem of the coincidence of wants with respect to goods. Most countries have their own currencies, so international trade in goods requires trade in currencies. The motives for currency exchange have expanded over the centuries to include speculation, hedging, and arbitrage, with the list of key players expanding accordingly. Beyond importers and exporters, the major categories of market participants now include asset managers, dealers, central banks, small individual (retail) traders, and, most recently, high frequency traders.
“Dealers” or “market makers” emerged naturally to fulfill the search function among trading counterparties. Dealers stand ready to trade with anyone needing FX at a moment's notice. To initiate an FX trade, an agent calls a dealer indicating the currency and quantity she/he wishes to trade and asking for the price. The dealer states a price at which she/he is willing to buy (the “bid”) and a price at which she/he is willing to sell (the “ask”). Finally, the customer decides whether to buy, sell, or pass. The dealer is compensated for the burdens of liquidity provision, such as bearing inventory risk and screening agents for credit quality, by a favorable gap between the quoted buy and sell prices, the “bid–ask spread.” Markets of this structure, known as “over-the-counter” (or OTC) markets, have arisen naturally in contexts including municipal and corporate bonds, derivatives, and equities. Although OTC dealers are under no formal obligation to provide liquidity, they tend to be reliable because otherwise their reputation—and potentially their market share—will suffer.
Existing theory indicates that information is brought to the market by liquidity takers rather than market makers. Among liquidity takers in FX, the evidence indicates that information relevant to short horizons comes from financial customers, especially leveraged investors. Information does not come from firms involved in international trade, small individual traders, or governments/central banks (Bjønnes et al., 2005, 2011; Evans and Lyons, 2006; Nolte and Nolte, 2009; Osler and Vandrovych, 2009; King et al., 2010). To identify whether some category of participant is typically informed, it is standard to examine whether their trades anticipate FX returns. If an agent consistently tends to buy (sell) before prices rise (fall) and the subsequent price change tends to be at least partially permanent, researchers infer that the agent was trading based on information about the asset's fundamental value. Academic studies have long indicated that FX dealers are informed. But, until recently, both theory and practice assumed that this information originated entirely with end-customers (Evans and Lyons, 2002). It is now recognized that dealers bring their own independent information to the market (Bjønnes et al., 2011; Moore and Payne, 2011).
1.3.1 Who Needs Liquidity?
Traditionally, the end-customers routinely needing FX liquidity were non-dealer financial institutions on one hand, and corporations and governments on the other. Over the last decade, the set of active FX end-users has been augmented by retail investors and by computer-automated traders known as algorithmic traders.
Until the mid-1980s, non-dealer financial institutions, corporations, and governments each traded roughly equal amounts with their FX dealers. Since then trading by corporate customers and governments has maintained an overall market share of 17% on an average, though this share naturally rises and falls with economic activity. During the recessions of 2001 and 2010, for example, their share of activity fell to 15%. The share of financial trading in total trading, by contrast, rose steadily from 20% in 1998 to over 50% in 2010 (Figure 1.2). This trend partly reflects the rapid growth of trading on retail trading platforms,10 which reached an estimated $125 to 150 billion per day in 2010, equivalent to 8–10% of global spot turnover (King and Rime, 2010). It also reflects rapid growth in algorithmic trading, especially, high frequency trading. Although data on the extent of algorithmic trading are limited, the survey reported in Section 1.1 suggest that it now accounts for between one-third and one-half of trading in the most liquid currencies.
Figure 1.2 Rising share of financial turnover. Note: Figure shows the share of financial customers (left axis) and non-financial customers (right axis, dot symbols) out of total spot trading. Third group not shown in graph is dealers. G4-currencies (solid lines) are USD, EUR (DEM before 1999), JPY, and GBP; emerging-market currencies (dashed lines) are MXN, KRW, RUB, PLN, TRL, TWD, INR, HUF, ZAR, and BRL.
Financial Institutions
Financial institutions are a diverse category that includes hedge funds and other asset managers, regional and local banks, broker–dealers, and central banks. Relative to corporate customers, financial institutions trade larger amounts and hold FX positions for far longer. Financial institutions tend to be better informed than other end-users, because they have strong incentives to invest in information acquisition. Since financial institutions use currencies primarily as a store of value, they gain or lose according to changes in the currency's value.
Among financial institutions, leveraged institutional investors—meaning mostly hedge funds and their close cousins the commodity trading advisors (CTAs)—appear to be best informed. This finding seems logical since leveraged institutional investors face particularly intense incentives to acquire information. Hedge-fund managers are often paid 2% of underlying assets as a flat annual fee and 20% of investment returns. Leveraged currency funds, which grew dramatically during the late 1990s, are known to favor well-defined speculative strategies focused on four factors: fundamentals, interest differentials (i.e., the carry trade), momentum, and volatility.
Unleveraged asset managers (“real-money investors”) include mutual funds, pension funds, endowments, and insurance firms. Perhaps surprisingly, such funds often pay little attention to the exchange rate component of returns when choosing asset allocations (Taylor and Farstrup, 2006). Instead, they concentrate on maximizing expected returns to foreign assets measured in the asset's home currency. This approach may be rational given ample evidence that major exchange rates are well approximated by a random walk. Some real-money investors outsource the management of FX exposures to currency overlay managers, who focus on risk reduction, return maximization, or some combination of the two.
FX hedging has become more important among portfolio managers since the financial crisis (Melvin and Prins, 2010). Market participants report that it is common to adopt a 50% hedge ratio, with the hedge reset periodically (e.g., once a month). A 50% ratio minimizes “embarrassment risk,” meaning the risk that a firm incurs either an absolute loss (when the rate moves adversely on an unhedged position) or an opportunity loss (when the rate moves favorably on a fully hedged position).
Private financial institutions dominate financial trading on a day-to-day basis, but central banks are noteworthy participants nonetheless. When these public sector institutions intervene to influence exchange rates, their trades are considered informed. Major dealing banks ensure that they know of such trades by cultivating their relationships with central banks. For example, dealers may share market intelligence on a daily basis with these valued clients. Central banks also trade FX as part of the regular procurement process for military and other government functions. Such trades are not considered informative, and central banks often ensure that such trades are not confused with intervention by announcing them in advance.11
Corporate Customers
Corporate customers use FX market to support treasury operations associated with their core business activities such as mining, shipping, and manufacturing. As such, corporations primarily use foreign currencies as a medium of exchange, trade relatively small amounts, and hold these positions only briefly. Most corporate customers choose not to engage in speculative FX trading; indeed some firms explicitly prohibit it. Given their institutional goals, this restriction seems logical. FX forecasting is not among a corporation's “core competencies,” so cultivating in-house speculative expertise can be ill-advised (Goodhart, 1988). Further, creating a trading operation is expensive. Not only it is costly to hire currency analysts and traders but also it is expensive to hire the extra staff required to protect against “rogue trader risk,” meaning the risk that a single trader brings down the firm (Osler, 2009).12 Even corporate firms that hedge their foreign cash flows pay little attention to future exchange rate movements. A survey by Bodnar et al. (1998) finds that among corporations that hedge their exposures—as most do—they typically choose hedge ratios between 40% and 50% and favor maturities below 6 months. They also review their currency hedge ratios at most a few times each year. Since corporate customers generally choose not to engage in speculative trading, it is not surprising that their trades do not anticipate short-term returns and are, therefore, not considered informative.
Corporations typically only use the FX market for one side of each exposure. A US multinational needing EUR to pay taxes in Germany, for example, sells its USD to buy the EUR in the FX market but then delivers the currency directly to the German government, bypassing the FX market entirely. Similarly, a Japanese exporter of manufactured goods to the United States receives USD from the American importer and then sells those USD in the FX market.
Retail Investors
Historically, few private individuals have had sufficient net worth to qualify for a credit line at a FX dealing bank. This barrier to entry effectively made the FX market an entirely wholesale market. Trading by small investors was also discouraged by the relatively high bid–ask spreads on small trades, meaning those below $1 million. Retail investors gained access to FX market around the year 2000 with the arrival of internet-based trading platforms tailored to their unique needs, so-called “retail aggregators” (RAs, described below).
Retail investors primarily trade FX spot in the major currency pairs, although the number of emerging-market currencies offered is growing. These individuals or small institutions tend to focus on just one or two currencies and to hold positions for very short time horizons, typically under a day. According to a recent survey (CitiFX Pro, 2010), these traders find the FX market attractive in part because of its low correlation with other markets, its high liquidity, and its 24-h market.
Retail traders should have strong incentives to be informed, since they trade for speculative purposes and employ substantial leverage. The evidence indicates, however, that retail trades are not informed. Their trades do not generally anticipate exchange rate returns (Nolte and Nolte, 2009) and the retail traders themselves are generally unprofitable (Heimer and Simon, 2011). In 2011, Oanda.com claimed that 48% of their retail customers were profitable. A systematic lack of trading acumen also appears to characterize retail traders in equities (Barber and Odean, 2000, 2002; Linnainmaa, 2010). Well-documented forces that might drive traders to stay active even when losing money include wishful thinking and overconfidence (Oberlechner and Osler, 2012).
Algorithmic and High Frequency Traders
Algorithmic trading is a form of electronic trading where a computer algorithm (or program) determines an order-submission strategy and executes trades without human intervention (Chaboud et al., 2009). Human involvement is limited to designing the algorithm (or algo), monitoring it, and occasionally adjusting the trading parameters. Some algorithms simply automate existing strategies—for example, they break up large trades to minimize transaction costs—while others take advantage of superior execution speeds such as high frequency trading.
High frequency trading relies on a technological advantage to exploit small price discrepancies across different online trading platforms. The time to execute a high frequency trade is measured in milliseconds (where it takes 100 ms to blink). Speed is so essential that high frequency traders co-locate their computer servers as geographically close to the pricing engines of leading electronic trading platforms as possible. Popular high frequency strategies include triangular arbitrage and covered interest rate arbitrage. More commonly, high frequency traders simply pick off dealers' posted quotes that are briefly out of line with the market due to slight time delays between trade instructions and execution (known as price-latency arbitrage).
1.3.2 Who Provides Liquidity?
Historically, liquidity in currency markets was provided exclusively by the bigger commercial and investment banks. Over the past decade the spectrum of formal liquidity providers has expanded to include global custodial banks, retail aggregators, and high frequency traders. The rest of this section discusses each agent type in detail.
FX Dealers
As they have for decades, FX dealers earn income by taking speculative positions and by providing liquidity to customers. Speculation is typically the responsibility of interbank traders, whose positions are only held open a few minutes or hours, and proprietary traders, whose positions are held open for longer periods (Bjønnes and Rime, 2005). Salespeople, who are responsible for maintaining good relationships with customers, are the third major group on a trading floor. To align the interests of FX dealers with those of bank shareholders, dealers receive bonuses tied to their individual profits and the profits of the entire trading floor while their individual risk-taking is constrained by position and loss limits.
Historically, the largest dealing banks earned substantial revenue from both speculative trading and liquidity provision for customers; the smaller dealing banks, by contrast, primarily earned income from customer service (Mende and Menkhoff, 2006). Since the 2007–2009 global financial crisis, proprietary FX trading activity has shrunk dramatically, discouraged by intensified regulatory scrutiny and a smaller appetite for risk. A few top dealers now report that liquidity provision for customers is their biggest source of FX revenue.
When not servicing customers, dealers have historically preferred to trade on the fast and inexpensive interdealer market. In the 1980s and early 1990s, interdealer trading represented over 60% of spot FX trades. In recent years, this fraction (the residual of the two shares presented in Figure 1.2) has declined steadily, and it was most recently estimated at only 35% (BIS, 2010). Reasons for this decline are discussed at length in Section 1.3 and include greater market transparency and heightened efficiency in matching trades electronically.
Interdealer trading can be carried out either directly in the OTC market or indirectly via FX limit-order markets run by FX brokers (Lyons, 1995). In OTC markets, all trades take place through designated market makers. In limit-order markets, no agents are specifically tasked with providing liquidity. Every agent can either supply (“make”) liquidity by placing a limit order or demand (“take”) liquidity by entering a market order. Limit orders indicate the prices at which an agent is willing to buy (bid) or sell (offer/ask) a specific quantity. The set of outstanding limit orders, which represents the existing liquidity supply, is known as the limit-order book. Orders remain in the limit-order book until executed or cancelled. The best bid or ask prices are known as the quotes, and the quantity available at the quotes is the market's “depth.” Trades occur when another dealer indicates that she/he wishes to trade immediately by placing a “market order.” Market orders are executed against limit orders in the book, beginning with the best-priced limit order and, for large quantities, moving to limit orders with successively less attractive prices.
Interdealer bid–ask spreads in the major currencies are typically 0.5–2.0 pips; those on less liquid currencies, particularly those with exchange controls, can reach 40 pip (Osler et al., 2011). These spreads, like spreads in equity and bond markets, are influenced by inventory risk, as indicated by the positive cross-sectional relation between average spreads and average market volatility (Bollerslev and Melvin, 1994). Nonetheless, the forces that drive FX interdealer spreads vary in striking ways from those that drive equity and bond spreads. FX interdealer spreads are highest during the short overnight period when trading is light and volatility is low. By contrast, spreads in equity and bond markets are typically highest at the market open, when trading and volatility are highest. This difference has been traced to the absence of regulation—and specifically the absence of formal opening and closing hours—in FX (Osler and Yusim, 2009).
Historically, dealers preferred to have zero inventories, so after trading with a customer, the dealer typically passed any accumulated inventory quickly onto other dealers (Lyons, 1995; Bjønnes and Rime, 2005). On the basis of pre-2000 data, estimated inventory half-lives range from a minute for dealers at large banks (Bjønnes and Rime, 2005) to 12 minutes for dealers at small banks (Osler et al., 2011). In recent years, these speeds have come down sharply because of the advancement in electronic trading technologies. If a dealer passed inventory to another dealer, that second dealer typically laid the position off on yet another dealer in a process known as hot-potato trading (Lyons, 1997). In other words, the position continued to move until some dealer laid the risk off onto its own customers. In this respect, the FX market differed from some other OTC markets, such as the US municipal bond market or the European government bond market, where dealers “warehouse the risk” and eliminate the inventory via later customer trades, shading their quoted prices to attract trades in their preferred direction (Dunne et al., 2008).
Dealers are perhaps the best-informed agents in FX market. Not only does their order flow anticipate returns (Rime et al., 2010) but it does so better than the trades of any individual group, including leveraged investors (Osler and Vandrovych, 2009). This advantage is thought to reflect, at least in part, the dealers' extensive networks of financial customers. If so, then larger dealers should be better informed than smaller dealers, a prediction that is supported by evidence (Bjønnes et al., 2011).
To increase the amount of customer information coming their way, dealers have historically quoted narrower spreads for larger trades and for financial-customer trades (Ding, 2009; Osler et al., 2011). This pattern is the opposite of that observed in prominent equity markets such as the NYSE. As discussed in Osler et al. (2011), these dealer choices could also be influenced by fixed operating costs and the dealers' stronger bargaining power relative to their least informed customers.
Recent evidence suggests that dealers generate some market-relevant information independently of their customers (Bjønnes et al., 2011). Consistent with this, Moore and Payne (2011) find that dealers specializing in liquid dollar rates can forecast both order flow, which is known to drive returns, and the component of returns that is uncorrelated with flow.
Global Custodian Banks
Large asset managers typically hire administrators or “custodians” who track their assets, calculate portfolio values, process dividend and interest payments, buy and sell assets, and settle trades. When they need to trade foreign currencies, real-money investors typically do not contact the major banks. Instead, they trade with their custodian, motivated largely by administrative efficiency (DuCharme, 2007).
Consistent with standard practice in the broader FX market, custodial FX trades are handled on a principal basis. Global custodians thus provide a second layer of FX market-making services. The custodian provides liquidity to its clients and, in turn, counts on receiving liquidity from its regular dealing banks. The custodian charges its clients a mark-up over the prices it pays in the interbank market.
Bid–ask spreads on custodial FX trades average a striking 30–40 basis points more than interbank spreads (Osler et al., 2011). This has been traced to the relative opacity of these trades. Typically, when a fund manager instructs its “fund accountant” at the custodian to buy or sell a foreign asset, the order includes instructions to carry out any required FX transactions. As a consequence, the custodian's client gets very little information about the trade; it learns the price actually paid with a delay of days or even weeks and it learns no information at all about the time of the trade or the effective bid–ask spread. This opacity makes it difficult for the fund to monitor execution quality.
In recent years, institutional investors have focused much more on the costs of their FX trades. Transaction cost analysis (TCA) is a growing business and increasingly a requirement for pension funds and other fiduciaries.
Research has not yet examined whether the interbank trades of global custodians are informed, but this could well vary across custodians. Small and mid-sized custodians typically focus exclusively on customer service and their customer trades may not be highly informative since they serve few leveraged investors (Osler et al., 2011). The custodial business has become increasingly concentrated, however, and some large custodians intentionally gather market-relevant information from customers to support speculative position-taking (Ramadorai, 2008).
Retail Aggregators
In recent years, the emergence of retail FX trading was enabled by the development of a new type of financial intermediary, the retail aggregator (RA). Retail aggregators exclusively operate over internet-trading platforms, bundling small retail trades into larger trades that can be conveniently handled by dealing banks. Some retail aggregators act purely as FX brokers, matching retail trades with quotes from banks. Other retail aggregators combine a broker model with a dealer model; they match some trades but strategically act as the counterparty for others. Retail aggregators typically provide their customers with leverage, which can range up to 200%. They protect themselves against default by insisting that each retail customer posts an initial cash deposit (“margin”). When a trade is executed, the retail aggregator settles it against the margin in the customer's account. When the margin in the retail customer's account is exhausted, the retail aggregator liquidates the retail customer's positions and closes the account immediately. This behavior allows retail aggregators to avoid taking on credit risk from their customers, who must therefore monitor their own trading activity closely.
1.3.3 Asymmetric Information And Exchange Rate Determination
For exchange rate modeling, it is important to know not only just which agents bring information to the market but also how their information becomes embedded in the market price. In FX market, this process appears to have at least three steps. First, end-customers reveal their information to market makers indirectly by trading with them. Second, the information becomes embedded in interdealer prices. Third, the information is disseminated widely as quotes in the customer segment of the FX market are adjusted to reflect the new interdealer prices.
The process through which dealers learn information from their customers has already been described. To examine the next step in this process, consider how a dealer adjusts her position after providing liquidity to a customer. As discussed earlier, a dealer who has bought currency from a customer then sells that currency in the interdealer market. The dealer will be especially anxious to trade quickly and aggressively after trading with informed customers (Bjønnes et al., 2011; Osler et al., 2011). An informed customer's purchase implies that the price is likely to rise. If the dealer is short the appreciating currency as a result of the trade, the inventory position represents a bad risk. If the customer is uninformed, however, the dealer has no strong reason for trading aggressively, as there is a 50–50 chance that the associated inventory position will be a winner.
Consistent with this hypothetical dealer behavior, Osler et al. (2011) provide evidence that, after customer trades, dealers are more likely to trade aggressively and in the same direction as the customer if the customer is informed. Similarly, Bjønnes et al. (2011) provide evidence that dealers with larger networks of financial customers tend to trade more aggressively, on average. By contrast, the average aggressiveness of dealer trades is not influenced by the extent of their trading with corporations or governments.
To connect this dealer behavior to price discovery, note that when a dealer sells aggressively, she trades at the lower, less attractive bid price. The downward movement of the interdealer price is consistent with the information implied by the initial informed-customer sale that the price is likely to decline. The interdealer price thus naturally moves to embed the information brought to the market by end-customers. The price shift can be sustained, because other dealers move their own quotes in parallel when they observe the interdealer trade (Goodhart et al., 1996). The interdealer price shift will immediately influence the prices quoted to customers, because these are typically set as a mark-up relative to the interdealer price.
The contemporaneous relationship between daily FX returns and order flow provides evidence consistent with this view of price discovery. (Order flow is defined as the net of aggressive buy trades and aggressive sell trades.) This relationship is positive for financial customers, as one would expect since their individual buy (sell) trades are typically followed by price rises (declines). Dealer order flow also has a positive contemporaneous relationship with daily FX returns.
Since corporate customers are generally uninformed, the price discovery process just described should imply that FX returns are unrelated to corporate order flow. But the contemporaneous relationship between FX returns and corporate order flow is actually negative. That is, on days when corporate customers are net buyers (sellers) of a currency, FX returns are negative (positive) (Bjønnes et al., 2005; Evans and Lyons, 2006; Lyons, 2001; Marsh and O′Rourke, 2005).
The intraday response of corporate customers to changing currency values explains the negative contemporaneous relationship with daily FX returns. Corporate customers are motivated to buy imports inexpensively or to sell exports lucratively. Once a currency has declined in value, imports are less expensive; after a currency gains in value, the effective mark-up on exports is higher. Thus, changes in exchange rates elicit trading responses from corporate customers. Corporate customers can respond nimbly, even though they do not normally monitor the market closely, by using “take-profit” orders, effectively hiring the dealer to monitor the market for them (Osler, 2003, 2005).13
Because of their different trading motivations, financial and corporate customers appear to play different roles in exchange rate determination. Sager and Taylor (2006) use a “push-pull” metaphor. When push agents sell a currency to FX dealers, the currency typically depreciates. The depreciation serves to “pull” other agents into the market as buyers. This process has a similar flavor to the model suggested by Evans and Lyons (2002), though their “push” agents are uninformed financial customers and their “pull” agents are risk-averse investors. The evidence suggests, instead, that the push agents tend to be informed financial customers and the pull agents tend to be non-speculative traders.
