Working Papers

THE CHOICE OF DIRECT DEALING OR ELECTRONIC BROKERAGE IN FOREIGN EXCHANGE TRADING

by

Michael Melvin & Lin Wen

Abstract

Central bank surveys indicate that the use of electronic brokerage systems account for the great majority of inter-dealer spot foreign exchange market trade execution.  This share has grown from zero in the early 1990s and is up sharply from that reported in the surveys taken in 1998.  While the surveys point out the rapid growth of electronic brokers as an important FX institution, there has been no research on the microstructure issues that lead traders to choose electronic brokerage (EB) over the historically dominant, and still quite relevant, institution of direct dealing where bilateral conversations (either telephone or electronic) occur between two FX traders and a deal is struck.  We provide theory and empirical analysis to further our understanding of the choice of trading venue in foreign exchange. 

Our theoretical model analyzes the choice of trading venue for “large” and “small” traders.  The theory illustrates the importance of asymmetric information, transaction costs, and speed of execution. The most likely outcome has direct dealing used for large trades while the EB is used for small trades.   

The empirical analysis utilizes data on orders submitted to the Reuters 2000-2 EB system. We focus on the duration of time between order submission and finding a match for trade execution.  An autoregressive conditional duration (ACD) model is specified using the Burr distribution. Given the price competitiveness of an order, duration is increasing in order size.  Because of this longer duration for large orders on the EB, large traders will prefer the direct dealing market to the brokerage.  We also find that the greater the depth of the market, the shorter the duration of orders of all sizes.  This result is consistent with traders clustering in time to submit orders so as to increase the probability of finding a match.

• click for pdf file of paper

THE ROLE OF U.S. TRADING IN PRICING INTERNATIONALLY CROSS-LISTED STOCKS

 by

 Joachim Grammig^a, Michael Melvin^b, and Christian Schlag^c

Abstract:  This paper addresses two issues: 1) where does price discovery occur for firms that are traded simultaneously in the U.S. and in their home markets and 2) what explains the differences across firms in the share of price discovery that occurs in the U.S?  The answer to the first question is that the home market is typically where the majority of price discovery occurs, but there are significant exceptions to this rule and the nature of price discovery across international markets during the time of trading overlap is richer and more complex that previously realized.  For the second question, the results provide strong support that liquidity is an important factor. For a particular firm, the greater the liquidity of U.S. trading relative to the home market, the greater the role for U.S. price discovery.

• click for pdf file of paper

THE EFFECTS OF INTERNATIONAL CROSS-LISTING ON RIVAL FIRMS

Michael Melvin and Magali Valero-Tonone

Abstract

The analysis focuses on the stock price impact of firms’ U.S. cross-listing on home-market rival firms. A theoretical model is presented that indicates the effect is negative, and that the effect is stronger for emerging market firms than for developed market firms. The empirical work uses both listing dates and announcement dates of forthcoming ADR programs. An event study approach is employed to analyze the impact on the home market price of the rival firm around the dates of  listing and announcement of listing. We find negative cumulative average abnormal returns for the rival firms around the announcement and listing dates, consistent with rival firms being hurt by the listing.  The evidence suggests that investors see rivals as less transparent and less informative relative to the listing firm. We also find evidence that the effect on the rival firm is stronger for firms from emerging market countries than for firms from developed market countries. 

• click for pdf file of paper

Explaining the Early Years of the Euro Exchange Rate: an episode of learning about a new central bank

  Manuel Gómez and Michael Melvin*

Abstract:

Many observers were surprised by the depreciation of the euro after its launch in 1999.  Handicapped by a short sample, explanations tended to appeal to anecdotes and lessons learned from the experiences of other currencies.  Now sample sizes are just becoming large enough to permit reasonable empirical analyses. This paper begins with a theoretical model of pre- and post-euro foreign exchange trading that generates three possible causes of euro depreciation: a reduction in hedging opportunities due to the elimination of the legacy currencies, asymmetric information due to some traders having superior information regarding shocks to the euro exchange rate, and policy uncertainty on the part of the ECB.  One empirical implication of the model is that higher transaction costs associated with the euro than the German mark may have contributed to euro depreciation.  However, empirical evidence on percentage spreads tends to reject the hypothesis that percentage spreads were larger on the euro than the mark for all but the first few months.  This seems like an unlikely candidate to explain euro depreciation over the prolonged period observed.  Reviewing evidence on market dynamics around ECB, Bundesbank, and Federal Reserve meetings, there is no evidence suggesting that the market is “front running” in a different manner than the other central banks.  However, we do find empirical support for the euro exchange rate to be affected by learning. By focusing on euro-area inflation as the key fundamental, the model is structured toward the dynamics of learning about ECB policy with regard to inflation. While a stated target inflation rate of 2 percent existed, it may be that market participants had to be convinced that the ECB would, indeed, generate low and stable inflation.  The theory motivates an empirical model of Bayesian updating related to market participants learning about the underlying inflation process under the ECB regime.  With a prior distribution drawn from the pre-euro EMS experience and updating based upon the realized experience each month following the introduction of the euro, the evidence suggests that it was not until the fall of 2000 that the market assessed a greater than 50 percent probability that the inflation process had changed to a new regime.  From this point on, trend depreciation of the euro ends and further increases in the probability of the new inflation process are associated with euro appreciation. 

• click for pdf file of paper

A Stock Market Boom During a Financial Crisis?

ADRs and capital outflows in Argentina

Michael Melvin

Abstract

Beginning in late 2001, Argentina experienced a tumultuous economic and social crisis including the end of the decade-long peso peg to the dollar, drastic foreign exchange and capital controls, violent anti-government demonstrations, social unrest, and the largest debt default in history.  Yet the Argentine stock market experienced a boom during the early period of the crisis.  This is in contrast to the experience of other countries undergoing financial crises, where the domestic stock market experiences sharp declines in value.  This paper explains the surprising Argentine experience as a result of investors using the stock market to shift funds out of Argentina and into the United States.  This was accomplished via purchases in Argentina of shares of firms listed in the United States and traded as American Depositary Receipts (ADRs).  These Argentine shares were converted into ADRs and sold in the U.S. to shift out of pesos in Argentina into dollars in the United States.  While ADRs and underlying share prices typically trade in a very narrow range, during the time when ADR conversions were permitted in Argentina, a large premium on share prices in Argentina relative to ADR prices existed.  This premium reflected the capital loss expected on peso investments in Argentina and the value of capital control avoidance.  On March 25, 2002, the conversion of Argentine shares into ADRs was prohibited and the premium of Argentine share prices over ADR prices once again returned to fluctuate about zero.

• click for pdf file of paper

Exchange Rates and FOMC Days

 Michael Melvin

 Abstract:  

FOMC meeting days provide a natural laboratory for exploring the effects of policy uncertainty and learning on exchange rate determination. Intradaily mark/dollar exchange rates are employed for 10 FOMC meetings. The meetings examined are the first 10 following the February 1994 change in policy where the meeting outcome is announced after meetings end.  A reasonable hypothesis is that the meeting outcomes are price-relevant public information associated with a switch to an “informed-trading state.”

 A markov-switching model is used to estimate the time of informed trading. The data suggest that on most days, there is a switch to the informed-trading state during the time of the meeting, well before the end of the meeting. An extensive search of public news indicates that the informed trading cannot be explained as the response to public information.

 An ordered probit model of the bid-ask spread is estimated as a function of the probability of being in the informed trading state.  The estimation results indicate that the greater the probability of being in the informed trading state, the wider spreads.  This is consistent with dealers protecting against adverse selection in quoting.

 The evidence indicates that traders are adjusting positions on more than just public news during FOMC meetings. In this sense, the realization of the meeting outcome is often not news.  An interesting question is why such positioning does not occur earlier in the morning prior to the meeting or on a prior day. A remaining puzzle is what causes the switch to the informed–trading state during meeting time.

• Click for pdf file of paper

Recently Published Papers

Internationally Cross-Listed Stock Prices During Overlapping Trading Hours:

price discovery and exchange rate effects

 Joachim Grammig, Michael Melvin, and Christian Schlag

Abstract

This study addresses two questions:  where does price discovery occur for internationally-traded firms and how do international stock prices adjust to an exchange rate shock?  These questions are answered by analyzing quotes originating in New York and Frankfurt for three large German firms, DaimlerChrysler, Deutsche Telekom, and SAP, during overlapping trading hours.  A high-frequency sample of quotes from both locations along with the dollar/euro exchange rate yields evidence of one cointegrating relation among the 3 variables.  Vector error correction models are estimated for each firm and the associated vector moving average representations are utilized to infer the share of price discovery coming from the exchange rate, New York, and Frankfurt quotes.  The evidence suggests a structure of the international equity market that has the home-market largely determining the random walk component of the international value of a firm along with an independent role for exchange rate shocks to affect prices in the U.S. markets.  However, there is a significant information share for New York in the case of DaimlerChrysler and an even bigger role for New York with respect to SAP.  Following a shock to the exchange rate, we find that almost all of the adjustment comes through the New York price.

Published  in the Journal of Empirical Finance, January 2005

• Click for pdf file of paper

DOES CORPORATE GOVERNANCE MATTER IN THE MARKET RESPONSE TO MERGER ANNOUNCEMENTS?  Evidence from the U.S. and Germany

 Paul Lowengrub, Torsten Luedecke, and Michael Melvin

 ABSTRACT: This paper attempts to determine if differences in corporate governance standards between Germany and the U.S would cause differences in the market response to merger announcements. German executives can deliberately make misleading statements regarding merger activities while U.S. executives must either state “no comment” or else provide a truthful statement. A sample of German and U.S. firms that announced acquisition plans in the 1995-1999 period was collected to determine if such a difference exists. Results over all firms in the sample suggest that merger news has no significant impact on cumulative abnormal returns for German firms but a significant positive impact for U.S. firms.  Such results are consistent with the market having discounted the German price impact prior to announcements as may occur with insider trading while U.S. announcements are truly price-relevant news. A finer breakdown of the sample reveals that German-U.S. firm differences are not as simple as the aggregate results suggest. Specifically, larger firms in Germany and the U.S. have similar experiences.  In addition, controlling for size it is found that German firms listed on a U.S. stock exchange also have similar experiences as U.S. firms. This is likely due to the greater disclosure requirements for such firms.  So differences only exist between smaller firms in the U.S. and smaller German firms that are not traded on a U.S. exchange. Aside from this smaller-firm effect, the evidence is consistent with no price-relevant differences arising from the differences in corporate governance rules.

Published in Advances in Financial Economics, Corporate Governance, Hirschey, John, and Makhija, eds, Elsevier, 2004.

• click for pdf file of paper

A YEN IS NOT A YEN:

TIBOR/LIBOR and the determinants of the 'Japan Premium'

Vincentiu Covrig, Buen Sin Low, and Michael Melvin

Abstract

Pricing in the Euroyen market is based on LIBOR, the London Interbank Offered Rate, set at 11am London time or TIBOR, the Tokyo Interbank Offered Rate, set at 11am Tokyo time.  Since the TIBOR panel is dominated by Tokyo city banks while the LIBOR panel is dominated by non-Japanese banks, the changing TIBOR-LIBOR spread reflects the credit risk associated with Japanese banks or the “Japan premium.”  In this paper, we investigate the determinants of this "Japan premium."  The spread is modeled as a function of determinants of bank default and firm value suggested by a theory of credit spreads.  Our results suggest that systematic variation in the spread can be explained by interest rate and stock price effects along with public information flows of good and bad news regarding Japanese banking, with a separate individual role for Japanese bank credit downgrades and upgrades.

Published  in the Journal of Financial and Quantitative Analysis, March 2004.

• Click for pdf file of paper

A Stock Market Boom During a Financial Crisis?

ADRs and capital outflows in Argentina

Michael Melvin

Abstract

 The surprising Argentine stock market boom during a financial crisis is explained as a result of investors using the stock market to circumvent cash withdrawal restrictions and capital controls and shift funds out of Argentina and into the United States. 

Published in Economics Letters, October 2003

• click for pdf file of paper

THE GLOBAL TRANSMISSION OF VOLATILITY IN THE

FOREIGN EXCHANGE MARKET

 Michael Melvin and Bettina Peiers

 Abstract: 

Volatility spillovers of the DM/$ and ¥/$ exchange rate across regional markets are examined using the integrated volatility of high-frequency data.  An analysis of quoting patterns reveals 5 distinct regions: Asia, Asia/Europe overlap, Europe, Europe/America overlap, and America.  After reviewing theoretical foundations for persistence of volatility in dealership markets, regional volatility models are constructed   where volatility in one region is a function of  yesterday’s volatility in that region (“heat-wave effect”) and volatility in other regions (“meteor-shower effect”). While evidence of statistically significant effects is found for both own-region spillovers as well as inter-regional spillovers, the economic significance of own-region spillovers indicates that heat wave effects are more important than meteor showers.