Dynamics of exchange rate determination and ...

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/317836587

Dynamics of exchange rate determination and currency order flow in the Thailand foreign exchange market: An empirical analysis Article in Journal of Chinese Economic and Foreign Trade Studies · June 2017 DOI: 10.1108/JCEFTS-11-2016-0031

CITATION

READS

1

73

3 authors, including: Anifowose Daniel Abolaji

Izlin Ismail

University of Malaya

University of Malaya

6 PUBLICATIONS 2 CITATIONS

18 PUBLICATIONS 15 CITATIONS

SEE PROFILE

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Palestine Project View project

The Post Consolidation of Banks: Human Resource Management Challenges and Prospects in Nigeria Banking Sector View project

All content following this page was uploaded by Anifowose Daniel Abolaji on 09 August 2017.

The user has requested enhancement of the downloaded file.

Journal of Chinese Economic and Foreign Trade Studies Dynamics of exchange rate determination and currency order flow in the Thailand foreign exchange market: An empirical analysis Abolaji Daniel Anifowose, Izlin Ismail, Mohd Edil Abd Sukor,

Downloaded by University of Malaya At 20:42 01 August 2017 (PT)

Article information: To cite this document: Abolaji Daniel Anifowose, Izlin Ismail, Mohd Edil Abd Sukor, (2017) "Dynamics of exchange rate determination and currency order flow in the Thailand foreign exchange market: An empirical analysis", Journal of Chinese Economic and Foreign Trade Studies, Vol. 10 Issue: 2, pp.143-161, https://doi.org/10.1108/JCEFTS-11-2016-0031 Permanent link to this document: https://doi.org/10.1108/JCEFTS-11-2016-0031 Downloaded on: 01 August 2017, At: 20:42 (PT) References: this document contains references to 52 other documents. To copy this document: [email protected] The fulltext of this document has been downloaded 18 times since 2017* Access to this document was granted through an Emerald subscription provided by Token:Eprints:28JP7FIVVIGHCWAVN2I4:

For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services. Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download.

The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/1754-4408.htm

Dynamics of exchange rate determination and currency order flow in the Thailand foreign exchange market

Thailand foreign exchange market 143

An empirical analysis


Abolaji Daniel Anifowose, Izlin Ismail and Mohd Edil Abd Sukor Department of Finance and Banking, University of Malaya, Kuala Lumpur, Malaysia

Abstract Purpose – The purpose of this paper is to present the essential role that currency order flow plays in the foreign exchange markets of emerging economies in the determination of their currencies in the short and the long-run against major currencies of the world, which cannot be over emphasized, most especially against the US dollar. Insomuch that, if some of these emerging economies can be successfully transmitted into full development, it would be a good model for other emerging economies and the world at large. Design/methodology/approach – A hybrid model (portfolio shift model) proposed by Evans and Lyons (2002a, 2002b) is extended to analyze a data set of every quarter of an hour currency order flow and currency exchange rate fluctuations of Thai Baht (THB) against the US$ for the period of six years (January 2010 to December 2015). To reflect the pressure of currency excess demand, the authors construct a measure of currency order flow in the Thailand currency exchange market. Vector autoregression model is applied to estimate the effectual role of currency order flow in the determination of exchange rate for the THB against the US$. Findings – Currency order flow indeed accounted for a sizeable and significant portion of the fluctuations in the THB and the US$ exchange rate. Originality/value – Insomuch that, the results show that currency order flow has significant explanatory power in the emerging markets economy to capture the THB exchange rate variability, and it then brings to the attention of the Thailand Monetary Authority the importance that should be attached to the market microstructure.

Keywords Currency exchange market, Currency exchange rate, Currency order flow, Market microstructure Paper type Research paper

Introduction The experiential study of the asset market method reveals the manner in which traditional models of currency exchange rate determination completely go on the blink to elucidate exchange rate movements in the short term. In the other words, it can only indicate longterm trends (Vitale, 2003). Hence, both the financial economist and international finance researchers now give more consideration to the organization of currency exchange markets. Furthermore, with the knowledge that the market microstructure theory actually investigates dealers’ behavior in the securities market has led the academics in the field of economics to suggest that, such a theory may guide in the foreign exchange markets. O’Hara (1995) defines market microstructure “as the study of the process and outcomes of exchanging assets under explicit trading rules”. Therefore, dealing in securities market

Journal of Chinese Economic and Foreign Trade Studies Vol. 10 No. 2, 2017 pp. 143-161 © Emerald Publishing Limited 1754-4408 DOI 10.1108/JCEFTS-11-2016-0031

JCEFTS 10,2


144

has essential consequences while determining the price at which deals are consummated. It can be rightly said that, this understanding is real for currency exchange markets in a sense. The traditional models of exchange rate determination actually based on two basic fundamental principles: (1) currency exchange rate determination is mainly macroeconomic variable occurrence, i.e. changes in macroeconomic variables aggregates exclusively determine exchange rate movements; and (2) currency exchange rates instantaneously react to changes in macroeconomic variables aggregates. However, inadequate explanatory capability of these traditional models together with the experiential proof revealing the significance of micro-structure parts of the stock market while given explanation to short-term fluctuations in stock prices have made researchers to concentrate on order flow in the foreign exchange markets. Order flow is defined as the net of the buyer-initiated and seller-initiated orders in the foreign exchange market (Evans and Lyons, 2002a, 2002b). Currency order flow represents the measure of net buying pressure. Insomuch that, currency order flow may as well be inferred to be a communication nexus that links information and currency exchange rates in the foreign exchange market, as it sends information meaningful enough on the determination of currency exchange rates, for markets aggregation and subsequently impounding this information in the value of currencies. These currency order flows are the driving force behind the turnover in the currency exchange market (Evans and Lyons, 2002a, 2002b). Researchers in this field of international finance concentrated majorly on matured economies and the world currency pairs, but a small number of studies have investigated the essential role currency order flow plays in the foreign exchange markets in the emerging markets. Thailand, an industrialized country, also heavily export-dependent, with exports accounting for more than two-thirds of the country’s gross domestic product, has made the country to be one of the most robust in Asia. In the recent years, international trade has become increasingly important to this nation, having the main trading partners like the USA, ASEAN, EU and Japan. It is interesting that the exports to the USA are the highest compared to other trading partners to this one of the fastest growing countries in the world. Also, being the second largest economy in the Southeast Asia, Thailand has been a frontrunner in the region in relations to trade liberalization and facilitation with the rest of the world. With Thailand, enormous trade transactions with the USA, her currency ought to appreciate in value and to a reasonable degree, achieve exchange rate stability in the international market. However, the continuous reduction in the foreign reserves of the country and subsequent currency depreciation in the international market, especially, against US dollar (US$) is worrisome. With the country daily turnover of inter-bank and customer transactions in the spot and forward market averages around US$3bn, this is considerably large compared to other financial markets. Therefore, it is paramount that, we should know how the value of Thai Baht (THB) against the US$ is determined in the international currency market, in the short run and long run, respectively. More so, if this emerging economy can be successfully transmitted into full development, it would be a good model for other emerging economies and the world economy at large. A data set for every quarter of an hour currency order flow and exchange rate fluctuations for the period of six years (January 2010 to December 2015) is analyzed using hybrid model of order flow and exchange rate dynamics proposed by Evans and Lyons (2002a, 2002b).

Covering this extensive period, and the quality of our data set, and that of its precise high frequency, these data sets are unique. To reflect the pressure of currency excess demand, we therefore construct a measure of currency order flow in the Thailand foreign exchange market context. Vector autoregression (VAR) model is applied to estimate the cointegrating relations between cumulative currency order flow and exchange rate fluctuations in the Thailand currency exchange market. Our major concern is to proffer answers to the following questions: Q1. In the international currency market, does currency order flow capture the movements of THB exchange rate against the US$?


Q2. In the international currency market, does the long-term and short-term elements impact on the estimation of the THB? Our results show that, there exists bidirectional causality between the currency order flow and exchange rate. Meaning that, currency order flow Granger causes exchange and vice-versa. While testing the potency of the relationship at longer horizons, we consider 6 weeks as 30 trading days, 4 weeks as 20 trading days and 2 weeks as 10 trading days. Therefore, we test for 30 trading days’ time horizon using Cholesky decomposition. The result shows that, there is a strong relationship between cumulative currency order flow and currency exchange rate at 30 trading days. Thus, even at longer horizon, there is a positive and strong relation between cumulative currency order flow and exchange rates in the Thailand foreign exchange market. From the results, it appears that currency order flow is the most exogenous variable relative to other variables in our specification, evidencing that, currency order flow can explain up to 15 per cent of the fluctuations in exchange rates for every US$10m/THB purchase. The motivation for this study comes on the premise that currency exchange rate determination using market microstructure approach requires further understanding and light shedding, most especially, in the emerging market of this nature. Therefore, our daily cumulative currency order flow is computed from the Thailand foreign exchange market using every quarter of an hour tick-by-tick trading data. Given our span of data, we are able to shed more light on the usefulness and appreciativeness of currency order flows in the emerging markets. More so, while concentrating on the currency order flow and determination of the exchange rate in the international market, our research contributes to the market microstructure of the exchange rate theory in the emerging markets economy. Furthermore, it will help scholars to have profound grasp of currency order flow as one of the major microeconomic factors to be considered in the currency exchange market, most especially, in the emerging markets economy. In addition, the policy makers and practitioners will have a deeper understanding of the explanatory power of currency order flow on how this influential variable drives the exchange rate movements in the foreign exchange market, not only developed but also emerging markets. This research paper is structured as follows: the next section reviews literature in brief on exchange rate dynamics with reference to market microstructure. Then the paper discusses the data and methodology. Finally, the paper presents the empirical results and provides the conclusion. Review of literature Market microstructure of exchange rate stresses on the role trading in foreign currencies play in price formation via a concept known as order flow. Evans and Lyons (2007) defined


JCEFTS 10,2


146

currency order flow to be the difference between the buyer-initiated and the seller-initiated trading interest in a given market and thus relates largely to what practitioners in the market might refer to as aggressive buying and selling of foreign currencies in the foreign exchange markets. Although, in the models of the following researchers, Lyons (1995), Perraudin and Vitale (1996) and Evans and Lyons (2002a, 2002b), currency order flow gives explanation on concomitant exchange rate fluctuations, as it includes information, either about fundamentals or long-run risk premia, which was hitherto circulated among foreign exchange market dealers and participants. Hence, the uniqueness of the microstructure level analysis when compared to the traditional exchange rate framework is that even though the same information is made available to all market participants but interpreted differently. Following the research work of Meese and Rogoff (1983) and Frankel and Rose (1995), other researchers (Evans and Lyons, 2002a, 2002b; Osler, 2006; Cheung et al., 2005) follow suit to explain currency exchange rate fluctuations via the process and procedure of technical trading approaches, currency order flows and price formations. Therefore, financial economists and international finance academia are at ease with an information perception in financial markets, thereby depending on a number of analytical models involving market microstructure and economic fundamentals for an enhanced and appreciativeness of the financial markets. As a measure of the sum of the signed seller-initiated order and that of the buyer-initiated orders in the experiential stipulation, currency order flow is deemed to be an essential information transmission device connecting price fluctuations and diffuse information (Evans, 2011). In fact, market microstructure research works have focused on the explanatory role of currency order flow in exchange rate models with two basic classifications of data: customer order flow data and interdealer order flow data (Evans and Lyons, 2007). The work of Evans and Lyons (2002a, 2002b), using interdealer order flow of four months exchange rate transaction data to analyze the daily fluctuations between DM/US$ and JPY/ US$ shows that, order flow actually accounts for more than 60 per cent of daily fluctuations in the DM/US$. Further research study by Evans and Lyons (2002a, 2002b), focusing on seven different currencies against the US$ shows that, currency order flow can generate an R2 of 78 per cent daily. Furthermore, Berger et al. (2008) examine the relationship between order flow and exchange rate of the EUR/USD, using interdealer transaction data over a time period of six-year (1999-2004). The results show that, a substantial relationship exists between interdealer order flow and exchange rate returns at short horizons. The simple description of inventory effect, information effect and liquidity effect with how currency order flow drives the movements of the exchange rate is summarized by Osler (2006). There exists an unwarranted risk that dealers are exposed to when anticipated currency position is not achieved. To guide against this risk when their inventory positions are not in conformity with their desired levels, they therefore adjust the price by sliding it or increase it to attract more buying or selling orders to maintain and retain their desired currency positions. Consequently, inventory models cannot be used to explain permanent exchange rate movements but momentarily. However, market prices should be permanently affected by order flow using information models. Therefore, there should be cointegrating relationship between currency order flow and exchange rate (Zhang et al., 2013). Hasbrouck (1991) proposed microstructure VAR model to investigate New York stock exchange. Payne (2003) apply the same model to examine US$/DM for a period of one week (October 6 to 10, 1997), and the result shows that trading with an informed dealers, currency order flow can generate up to 60 per cent fluctuations on exchange returns. Froot and Ramadorai (2005) investigate the interaction between permanent shock and transitory shock on exchange rate earnings applying order flow as a main factor of exchange rate fluctuations.


They find out that although macroeconomic fundamentals can be used to explain currency return in the long term, but order flow, a microeconomic variable is more appropriate to explain currency return in the short term. Therefore, going by these findings, currency order flow is of great importance to research on in the foreign exchange market, by examining its role in the determination of exchange rate both in the long- and short-term dynamics. Although, most of the researchers in this field truly concentrated on the major currency pairs. For example, Rime (2000) employ microstructure approach, investigates the influence of order flow on exchange rate determination on deutschmark, British pound sterling, Canadian dollar, Swiss franc and Japanese yen, all against US dollar, for the period July 1995 to September 1999. The results show that there is a cointegrating relationship between exchange rate and order flow for deutschmark/US$, British pound sterling/US$ and Swiss franc/US$. It implies that, there is an explanatory power of exchange rate fluctuations when order flow is lagged. In addition, Andersen et al. (2003), Evans and Lyons (2005) and Berger et al. (2008) investigate the explanatory power of order flow in their empirical studies. In the studies of currency order flow and exchange rate in the emerging markets, Zhang et al. (2013) examine the influential role of currency order flow on exchange rate fluctuations between Chinese RMB and US$, and they find out that order flow explains significantly exchange rate fluctuations in the Chinese foreign exchange market. More so, research work of Duffuor et al. (2012) reveals that in the Ghanaian foreign exchange market, the end-user order flow does not have much influence on the exchange rate fluctuations. In essence, there is a weak performance. In the Brazilian foreign exchange market, Wu (2010) investigates the interactions between the commercial and financial customer order flow and finds that positive relationship exists between the financial customer order flow and intervention flows, whereas a negative relationship exists between the commercial customer order flow and exchange rate. The inconclusiveness of these research studies and their findings inspired us to investigate further, the emerging market currency of Thailand (THB) and that of developed market, US$, to examine the strength at which currency order flow can explain exchange rate movements in the Thailand foreign exchange market. Thailand foreign exchange market: market-microstructure perspective The foreign exchange market is the ambit for a country’s currency in exchange for another. This market can be described as the leading financial market in the world, in the sense that it accommodates a daily trading volume of an equivalent of over US$4tn. This is three times over and above the total aggregate amount of transactions on the US equity and Treasury market combined. A spot-on 24-h market opens each trading in Sydney and then shifts as the business day commences in other financial center, i.e. from Sydney to Tokyo, London, New York and Frankfurt. Although, a time comes where two trading sessions are open at the same time. This is described as overlapping trading sessions. In this situation, there is a tendency for more volume to be traded, as all the market participants are “wheel-in” and “deal-in”, meaning that more money is transferring hands among the market participants in the foreign exchange market. In Thailand, in relative terms, it is the forces of demand and supply that do determine the exchange rate to an extent. Even though, such forces of demand for currency and supply of currency are derived from international trade value, international capital flows and market expectations among other factors. On July 2, 1997, the country adopted a managed-float exchange rate regime which made the Bank to implement foreign exchange rate management structure that aims to maintain currency stability. In the foreign exchange market, monitoring and supervision of Thai-Baht (THB) exchange rate against other


JCEFTS 10,2


148

currencies is the responsibility of the Bank of Thailand (BOT). The Bank does this to guide against excessive volatility in the foreign exchange markets. One of the major reasons for adopting a managed floating regime is for the monetary policy to implement flexibility and efficiency and also to increase the confidence of the investors in the domestic and international markets. The Ministry of Finance actually assigned the responsibility of foreign exchange administration to the Bank of Thailand. Foreign exchange transactions in Thailand must be carried-out through authorized commercial banks and authorized non-banks, which include authorized money changers, authorized money transfer agents and authorized companies that are granted licenses by the Ministry of Finance to officially carry-out foreign exchange transaction (BOT, 2016). Currently, only few major currencies, for example US$, Euro and Japanese yen are normally used for international trade and service settlement. As international trade relationship with China plays an important role at global and regional level, which made the Chinese authority announced the Reminbi (RMB) internationalization policy in 2009 to promote the use of RMB as the international currency. The Chinese authority has allowed the use of onshore RMB for international trade settlement, thereby appointing the first offshore RMB business in Hong Kong. However, the supportiveness of this policy which is in line with the Bank of Thailand to use local currencies for the settlement of international trade as made it possible to reduce foreign exchange risks. Importantly, to support the use of Reminbi as a means of payment for international trade transactions, the Bank of Thailand collaborate with three Thai commercial banks for ease of foreign transactions as it relates to Chinese Reminbi (RMB) and THB exchange rate. Figure 1 shows the correlation between US$/THB and currency order flow. Spotted from Figure 1, currency order flows are constant between January 2012 and July 2013 and September 2013 and July 2015, respectively. This strange occurrence made us to investigate further what could have been the major cause. Although, most of the emerging markets economy do not operate free floating rather managed floating which may lead to frequent occurrence of currency intervention by the monetary authority. Our findings show that Central Bank of Thailand consistently intervene to curtail the depreciation of THB against the US$ in the foreign exchange market during these periods. This may be one of the major reasons for the currency order flows to remain constant during these periods. [The results for the Central Bank of Thailand (CBT) intervention and currency order flows in the foreign exchange markets can be made available on request]. Data and methodology Data sources We source our data from Reuters and Bloomberg. Spot foreign exchange market and trade transactions on the Thailand FX market is our focus. THB against the US$ for the period January 4, 2010 to December 31, 2015 is applied to analyze a new data set for every quarter of an hour currency order flow and exchange rate movements. We explore further on the influential role of currency order flow in the determination of exchange rate in the emerging markets using a high frequency of every-fifteen minute currency order flow on exchange rate. Therefore, it examines the impact that currency order flows may have on exchange rate using a unique high frequency data set of every 15-min currency order flow over a six-year period. While some data sets have been for a relatively short period in earlier studies (Froot and Ramadorai, 2005; Bjønnes and Rime, 2005; Evans and Lyons, 2007; Rime et al., 2010; Zhang et al., 2013). To the best our knowledge, it is the largest data set and more recent period ever used in the literature to examine the impact of currency order flow on exchange rate determination in the emerging markets. Covering this extensive period, and the quality

Thailand foreign exchange market

USD/THB 37 36 35 34 33

149

32 31 30 29


28 I II III IV I II III IV I II III IV I II III IV I II III IV I II III IV 2010

2011

2012

2013

2014

2015

Exch. rate (USD/THB) and order flow 37

50,000

36

40,000

35

30,000

34

20,000

33

10,000

32

0

31

–10,000

30

–20,000

29

–30,000

28

–40,000 I

II III IV 2010

I

II III IV I

II III IV

2011

2012 PRICE

I

II III IV I

II III IV I

II III IV

2013

2014

2015

ORDER FL

of our data set, and that of its precise high frequency, these data sets are unique. A total sample of 1564 trading days excluding weekends and public holidays. We exclude Saturdays and Sundays from our sample as much that, foreign currency trading activity during this period is slight. Also, general public holidays were excluded from our sample, for these days are of unusually insubstantial trading volume: New Year, January 1 and, Christmas day, December 25. Spot currency exchange trading usually opens for business on Monday morning and closes on Friday evening. Even though, trading in the spot foreign exchange market in Thailand is conducted on a 24-h basis (i.e. from 1700 h to 1659 h). Currently, foreign currency transaction settlement period in Thailand is set at T þ 2. (i.e. two days after the transaction day). Measurement of variables Our variables are measure in this order; Pt represent closing exchange rate transaction price for each working day, Xt represent accumulated currency order flow for each working day, (hi – fi) represent the differential in interest rate for short-term period, (hr – fr) represent the differential in the interest rate for long-term period and (Rh – Rf) represent the difference in

Figure 1. The daily exchange rate of USD/THB THB and order flow (04/01/2010-31/12/ 2015)

JCEFTS 10,2


150

the country risks premium. The research work of Evans and Lyons (2002a, 2002b) reveals that, the currency order flow Xt trading transactions for the day, represent the net position of order flow initiated between the buyer and the seller. The interest rate differential for shortterm period (hi – fi) represent Thailand interest rate daily overnight minus the US interest rate daily overnight. The differential in interest rate for long-term period (hr – fr) represents Thailand daily inter-bank lending rate for 12-month minus the US daily inter-bank lending rate for 12 months. Also, the country’s daily risk premium Rh actually represent the difference between the prime lending rate (PLR) and three months Treasury bill rate (TB). Hence, the difference between the two countries risk premium (Rh – Rf) that is, Thailand risk premium minus that of the US risk premium. Meanwhile, we express on an annual basis, our interest rate data. From the definition of currency order flow, there are two main important things to take note of the trade direction and the sum of transaction volume. We therefore determine the trade direction and subsequently sum-up the tick-by-tick trading direction of our every quarter of an hour intraday data. We construct a measure of spot currency order flow by assigning values to every single buying trade þ1 and selling trade 1, respectively. Hence, one-day spot currency order flow over the entire trading period is equal to the summation of these trade signs. The stationarity of our data is checked, and Table I presents the results of the unit root test. The tests show that the null hypothesis cannot be rejected with level data. But at first difference, the null hypothesis can be rejected. This confirms that our variables are stationary as 1(1) process. Table II presents the summary of descriptive statistics and then correlation matrix of the major items; Pt transaction price, Xt daily accumulated currency order flow, (hi – fi) differential in interest rate for short-term period, (hr – fr) differential in interest rate for longterm period and (Rh – Rf) difference in the country risk premium. Our findings indicate that, all the variables fail the Jarque–Bera (JB) test. Meaning that, all the variables depart from normality. The skewness for all the variables is less than 1. The correlation matrix results show that, currency order flow and other three key variables have a positive relationship with the exchange rate. Going by these results, the extent to which interaction exists amongst these variables needs further investigation. Transaction price (pt) and cumulative currency order flow (xt). Evans and Lyons (2002a, 2002b) propose a model based on a portfolio shift model. This model can be stated as:

Variables Exchange rate (Pt) Order flow (Xt) Short-term IRD (hi – fi) Long-term IRD Count. risk prem. diff. (Rh – Rf)

Table I. Unit root test

Intercept

At level Trend and intercept

At first difference Intercept Trend and intercept

0.0782 (0.9499) 2.9202 (0.2433) 1.5643 (0.5007) 1.1220 (0.7091)

1.8516 (0.6789) 4.1959 (0.3246) 1.6129 (0.7878) 1.0455 (0.9359)

37.6773 (0.0000)*** 37.8059 (0.0000)*** 18.7022 (0.0000)*** 18.6964 (0.0000)*** 31.4174 (0.0000)*** 31.5652 (0.0000)*** 25.9245 (0.0000)*** 26.2823 (0.0000)***

1.5831 (0.4910)

1.6745 (0.7622)

40.5452 (0.0000)*** 40.5344 (0.0000)***

Notes: This table reports the unit root test of the five key variables, i.e. Pt Transaction price, Xt Daily accumulated currency order flow, (hi – fi) differential in interest rate for short-term period, (hr – fr) differential in interest rate for long-term period and (Rh – Rf) difference in the country risk premium. 1% level denoted by *** represent the level of statistical significance

Variables

(Pt)


Stratum I: summary statistics Observations 1,564 Mean 0.0315 Std. Dev. 0.0015 Skewness 0.6319 Kurtosis 3.2438 Jarque–Bera 07.9754 Normality test (0.00000)*** Stratum II: correlation 1.00000 (Pt) 0.2185 (Xt) 0.5721 (hi – fi) (hr – fr) 0.6193 0.3843 (Rh – Rf)

(Xt)

(hi – fi)

(hr – fr)

(Rh – Rf)

1,564 5968.590 6707.127 0.5803 2.5968 98.3938 (0.00000)***

1,564 2.0283 0.6896 0.1782 2.0901 62.2237 (0.00000)***

1,564 1.8890 0.5895 0.6090 2.6745 103.5845 (0.00000)***

1,564 0.5710 0.3620 0.2129 1.6933 123.0695 (0.00000)***

0.2185 1.00000 0.0926 0.1395 0.0538

0.5721 0.0926 1.00000 0.9341 0.7641

0.6193 0.1395 0.9341 1.00000 0.6531


0.3843 0.0538 0.7641 0.6531 1.00000

Table II.

Notes: The table presents the summary of descriptive statistics then correlation matrix of the major items Summary statistics Pt Transaction price, Xt Daily accumulated order flow, (hi – fi) differential in interest rate for short-term period, (hr – fr) differential in interest rate for long-term period and (Rh – Rf) difference in the country risk and the correlation of the main variables premium. 1% level denoted by *** represent the level of statistical significance

DPt ¼ Dmt þ lDXt

(1)

where DPt represent changes in spot exchange rate; Dmt represent macroeconomic information innovations (for example, changes in interest rate differential); l represent positive constant; and DXt is daily accumulated signed order flows. Transaction price (pt) and interest rate (hr – fr). As a result of public information innovations Dmt and the change in the log of the spot exchange rate DPt, our equation (1) needs modification to be comparable to the standard macroeconomic models. The estimation specification can be expressed as: DPt ¼ a Dðhr fr Þ þ b DXt þ et

(2)

where DPt represent change in log of the spot exchange rate, Dmt in equation (1) is the change in interest rate differential, i.e.Dmt = D(hr – fr), we substitute Dmt for change in longterm interest rate differential D(hr – fr). Interest rate is considered to be an important variable that causes exchange rate movements in macroeconomic models, also available on a daily basis. Hence, it is considered suitable for experiential research. DXt represent the daily cumulative order flow, while a and b represent regression parameters, and et is the error term. Term spread and country risk premium (rh – rf). Country risk premium is a variable considered in the literature to have a positive and strong significance in the studies of emerging markets (De Medeiros, 2004, Wu, 2010, Duffuor et al., 2012; Zhang et al., 2013). Country’s daily risk premium Rh represent the difference between the PLR and three months TB. Therefore, the difference between the countries risk premium is given as (Rh – Rf) the Thailand risk premium minus that of the US risk premium. The research work of Evans (2011) states that currency transaction spot rate Pt of a pair currency with their interest rate for short-term period is practically determined according to

JCEFTS 10,2

the standard of the monetary policy of the central banks concerned. Therefore, we consider Bank of Thailand and the Federal Reserve as the central banks concerned in this study. Quote for all dealers is at US$ to THB and is given as: Pt ¼ EPt þ ðhi fi Þ Rh

152

(3)

Pt is the transaction price; (hi – fi) represent differential in interest rate for short-term period; Rh represent country’s daily risk premium. That is, the difference between the PLR and three months TB. The long-term (hr) and short-term (hi) difference represent term spread are given as: ðhr fr Þ ðhi fi Þ ¼ Rh Rf (4)


Therefore, we can equate country’s daily risk premium difference to the term spread (Figure 2). Order flow

Exchange rate

30,000

38

20,000

36

10,000

34 1.5

30,000

0

20,000

–10,000

10,000

–20,000

0

–30,000

32

1.0

30

0.5 28 0.0 –0.5

–10,000

–1.0

–20,000 –30,000

–1.5 I

II III IV I

II III IV I

II III IV I

II III IV I

II III IV I

II III IV

2010

2011

2012

2013

2014

2015

PRICE

Trend

I

II III IV I

II III IV I

II III IV I

II III IV I

II III IV I

II III IV

2010

2011

2012

2013

2014

2015

ORDER FL

Trend

Cycle

Cycle

Short-term interest Long-term interest

3.5

4

3.0 3

2.5

2

2.0

1

1.5 0.4

1.0

0.5

0

0.2

0.5

0.0

-1

–0.5

0.0

–1.0

–0.2

–1.5 –0.4

–2.0

–0.6

–2.5 I

II III IV

I

2010

II III IV

I

2011

II III IV

I

2012 STIRD

II III IV

I

2013 Trend

II III IV

I

2014

II III IV

I

2015

1.6 1.2 0.8 0.4

0.4

0.0 0.2 -0.4 0.0

–0.2 –0.4 I

II III IV 2010

I

II III IV 2011

I

II III IV 2012

CRPD

I

II III IV 2013

Trend

I

II III IV 2014

Cycle

I

II III IV 2011

I

II III IV 2012

LTIRD

Country risk premium

Figure 2. Hodrick–Prescott filter for the major items of measurement (Thailand)

II III IV 2010

Cycle

I

II III IV 2015

I

II III IV 2013

Trend

I

II III IV 2014

Cycle

I

II III IV 2015


Methodology The portfolio shift model (Evans and Lyons, 2002a, 2002b) is extended in this study, and we apply a VAR model proposed by Hasbrouck’s (1991) to examine the market microstructure elements of Thailand currency exchange rate fluctuations. From the literature reviewed, VAR model is one of the preferred econometric methods to investigate both the long-run as well as short-run relation between currency order flow and exchange rate fluctuations with their feedback effects. (Froot and Ramadorai, 2005; Wu, 2010, Duffuor et al., 2012, Zhang et al., 2013). Importantly, it considers the currency order flow coefficient on the ordinary least square regression with their likely feedback effect (Zhang et al., 2013). Johansen’s (1995) cointegration is applied to run our analysis with particular reference to the setting of VAR. With cointegration analysis, we are more precise about the long-run relation between cumulative currency order flow and exchange rate fluctuations. Cointegration is said to exist between two time series if they are individually nonstationary, even though there exists a linear combination of them with stationarity (Evans and Lyons, 2007). By interpretation, we can say that a stable long-run equilibrium relation exists. Therefore, VAR framework is extended in our analysis to calculate approximately the explanatory power of currency order flow on exchange rate movements in Thailand. The vector autoregression model The VAR model assumes that quotes from the market are immediately reflected based on public information available to the traders; hence, the informed traders take advantage of this to earn returns via their currency market orders. Therefore, let Nt denote attribute vector, Ct is the log of each transaction attribute, t is the time event. The model: Nt ¼ BCt þ Et

(5)

and 0

Pt

1

B C B C Xt B C B C B ð Þ h f Nt ¼ B i i C C; B C B ðhr fr Þ C @ A Rh Rf 51

2 6 6 6 6 6 B ¼6 6 6 6 6 4

b 1;1

L

b 1;5l

.. .

.. .

.. .

M .. .

T .. .

M .. .

b 5;1

S

b 5;5l

3 7 7 7 7 7 7; 7 7 7 7 5

5 5l

0

Pt1 B . B . B . B B Ct ¼ B M B B . B .. @

1

0

«t

1

C C C C C C; C C C A

B C B «t C B C B C C Et ¼ B B «t C B C B «t C @ A «t

5l 1

51

Rh1

(6) where Pt represent transaction price, Xt represent daily accumulated currency order flow, (hi – fi) represent differential in interest rate for short-term period, (hr – fr) represent differential in interest rate for long-term period and (Rh – Rf) represent the difference in the country risk premium. B represents matrices of coefficients to be estimated and random/realization variables ( b , L, M, T and S). Ordinary least square with heteroskedasticity robust standard errors is applied to estimate each VAR equation.


JCEFTS 10,2

VAR terms: Nt ¼ CCt1 þ « t

(7)

0 Nt ¼ f Pt ; Xt ; ðhi fi Þ; ðhr fr Þ; Rh Rf

(8)

hence:

154


Nt represent the transaction attributes vector, Pt represent the transaction price, Xt represent daily accumulated currency order flow, (hi – fi) represent differential in interest rate for shortterm period, (hr – fr) represent differential in interest rate for long-term period and (Rh – Rf) represent the difference in the country risk premium. The companion matrix U and variable Pt are let on uniform crosswise the currencies and also the lags. Table III reports the result of Johansen cointegration tests. The cointegration rank test (Trace and maximum eigenvalue statistics) analyzes the propositions at maximum g number of cointegrating relations of the key variables. The subscript g denotes the number of significant cointegrating vectors. The results show that, two cointegrating relationships exist, based on our full sample. At the 1 per cent significance level, the null hypothesis L0: g # II cannot be rejected. Table IV shows the results of the uniqueness of the cointegrating relationships of our 0 variable space tested in the VAR specification, i.e. Nt ¼ f ½Pt ; Xt ; ðhi fi Þ; ðhr fr Þ; Rh Rf ; Trend. We tested whether there exists any trend among the variables from the model using H1 to tests the cointegrating relationships. The results show that, there exists a trend among the variables from the model. Furthermore, using H2 to H4 to test the long-run cointegrating relationships between exchange rate and currency order flow, i.e. Pt = –Xt, interest rate spread and country risk difference, i.e. (hi – fi) = – (hr – fr), Rh = – Rf . However, the p-value of 0.4930 (H4(a)) is accepted from the test results. This shows that relationship exists between exchange rate and country risk premium. An automatic specification based on SC and AIC with maximum lag of 23 is used for the selection of our optimal lag length (Table V). Table IV presents the results of Granger causality tests and long-run weak exogeneity test of the key variables. The results show that exchange rate Granger causes order flow Levels (ranks) Eigenvalue Log likelihood Trace test Crit. value (0.05) Probability Max-Eigen Crit. value (0.05) Probability

Table III. Cointegration analyses with levels (ranks)

L0: g # NIL

L0 : g # I

L0: g # II

L0: g # III

L0: g # IV

0.0599 4118.685 153.6605 69.8188 (0.0000)*** 96.4488 33.8768 (0.0000)***

0.0180 4132.637 57.2116 47.8561 (0.0052)*** 28.4651 27.5843 (0.0385)**

0.0097 4146.870 28.7464 29.7970 (0.0657) 15.3575 21.1316 (0.2646)

0.0084 4154.549 13.3888 15.4947 (0.1013) 13.3076 14.2646 (0.0704)

0.0005 4161.202 0.0812 3.8414 (0.7756) 0.0812 3.8414 (0.7756)

Notes: The table reports the result of Johansen cointegration analyses. The cointegration rank test (trace and maximum eigenvalue statistics), analyze the propositions at maximum g number of cointegrating relations of the key variables. g denotes the cointegrating vectors number of significance. 5 and 1% level denoted by ** and ***respectively, represent the level of statistical significance Source: Authors’ calculations


Variables

Pt

Unrestricted: b1 b2

0.0293 1.0000

Xt 0.5517 0.5149

(hi – fi)

(hr – fr)

Rh

Rf

Trend

1.0000 0.0917

9.1388 0.0132

0.4277 0.0927

0.7567 0.0163

0.0069 0.0023

H1. Trend = 0, x 2 (2) = 14.92294 [0.000575]*** b1 1.8075 1.1938 1.0000 b2 1.0000 1.1185 9.9121

19.0354 0.2688

1.1606 2.7837

2.0776 7.9653

0.00 0.00

H2. Pt = – Xt, x 2 (2) = 105.0275 [0.0000]*** b1 0.5517 0.5517 1.0000 b2 1.0000 1.0000 941.0245

19.1294 33.1633

1.8543 1328.486

0.8444 1805.02

0.0092 0.0305

H3(a). (hi – fi) = – (hr – fr), x 2 (2) = 14.67810 [0.000650]*** b1 73.0535 0.0693 1.0000 1.0000 b2 1.0000 0.0599 429.4696 429.4696

28.0038 35.0770

3.5418 125.170

0.0035 0.0019

H3(b). (hi – fi) = – (hr – fr), Trend = 0 x 2(4) = 45.42680 [0.0000]*** b1 5.3147 0.0200 1.0000 1.0000 b2 1.0000 0.0116 7.2101 7.2101

5.1972 0.0259

1.2204 4.3515

0.00 0.00

H4(a). Rh = – Rf, x 2(2) = 1.414421 [0.493018] b1 6.7824 1.1185 11.7625 b2 1.0000 0.5517 0.6731

48.0946 0.7561

1.0000 0.7898

1.0000 0.7898

0.0019 0.0040

H4(b). Rh = – Rf, Trend = 0 x 2(4) = 17.48644 [0.001554]*** b1 5.3786 0.0019 17.2705 0.6696 b2 1.0000 0.5517 0.7262 6.7386

1.0000 0.7469

1.0000 0.7469

0.00 0.00

Notes: This table reports the result of cointegrating relationships among of key variables with and without trends. *** indicate statistical significance at the 1% level Source: Authors’ calculations

Variables 2

X (4) Probability

Pt

Xt

(hi – fi)

(hr – fr)

(Rh – Rf)

48.2478 (0.0000)***

82.4138 (0.0000)***

36.3023 (0.0026)***

25.2877 (0.0649)*

37.0779 (0.0020)***

Notes: This table present the results of Granger causality tests and long-run weak exogeneity test of the key variables. 10 and 1% level denoted by * and *** represent the level of statistical significance Source: Authors’ calculations

and order flow Granger cause exchange rate. In essence, there exists bidirectional causality. Table VI presents the results of hypotheses test on the cointegrating relationship among the key variables with their cointegration coefficients b and adjustment coefficients a and their standard errors. It appears that none of the variables in the model is weak, based on the results of the p-values for the long-run beta. Therefore, the following cointegrating equations can be formulated with level data:


Table IV. Cointegrating equations restriction tests

Table V. Granger causality/ long-run weak exogeneity test

JCEFTS 10,2

Variables

Pt

Cointegrating vector. b Feedback coefficients (a) with 2 ranks

156


Table VI. Long-run formation

10.1159 1.0000 0.0007 (0.0012) 0.0002 (0.0002)

Xt 0.0169 0.0027 14122.66 (8381.94) 1251.773 (109.258)

(hi – fi)

(hr – fr)

(Rh – Rf)

1.0000 2.3142 0.0041 (0.0016) 0.0015 (0.0013)

5.1350 1.0000 0.0077 (0.0034) 0.0132 (0.0034)

2.1321 1.0000 0.0030 (0.0007) 0.0091 (0.0029)

Notes: The table reports the outcomes of hypotheses test on the cointegrating relationship amongst the variables. The cointegration coefficients b and adjustment coefficients a with their standard errors in (). We consider 1 to 4 lag interval Source: Authors’ calculations

Ki ¼ Pt þ 0:0027 Xt þ 2:3142 ðhi fi Þ ðhr fr Þ þ Rh Rf Kii ¼ 10:1159 Pt 0:0169 Xt þ ðhi fi Þ 5:1350 ðhr fr Þ 2:1321 Rh Rf

(9)

(10)

The interest rate differential is appropriately signed and significant. When there is a higher imbalance currency position in the net buying activities, there is a tendency that there will be higher THB price against US$, in as much that the currency order flow is positively significant. In addition, US$/THB exchange rate calculation gives a beta coefficient of 0.0027. This connotes that, within the day transactions, for every currency order flow increasing at 1 per cent, there will be a corresponding increase of 27 basis points of the THB price against the US$. More so, the long-run coefficient b for the country risk premium is significant: DPt ¼ K þ a1 DPt1 þ b 1 DXt1 þ b 2 DXt2 þ u Kiit1 þ « p;t DXt ¼ K þ a1 DPt1 þ a2 DPt2 þ u Kit1 þ « X;t D Rh Rf ¼ a1 DPt1 þ a3 DPt3 þ w 3 Dðhi3 fi3 Þ þ l 1 Dðhr1 fr1 Þ þ l 2 Dðhr2 fr2 Þ þ l 3 Dðhr3 fr3 Þ þ d 1 D Rh1 Rf 1 þ d 2 DðRh2 Rf2 Þ þ d 3 DðRh3 Rf3 Þ þ « R;t From Table VII, we present the results of the short-run vector error correction model estimates for the variables DPt, DXt and D(Rh – Rf), respectively. However, insignificant variables were removed from our model, thereby reducing it to a partial vector error correction model. The short-term correction result with a coefficient of 0.0330 is negatively significant at 5 per cent level. This implies that in the Thailand currency exchange market, currency order flow Granger causes exchange rate fluctuations in the short term. The long run relation for order flow speed of adjustment is negatively significant. Also, the relationship between the country risk premium and exchange rate fluctuations is negative. While comparing our results, the coefficients of ours and that


Variables

DPt

DXt

D(Rh – Rf)

Constant a1 a2 a3 b1 b2 H w3 l1 l2 l3 d1 d2 d3 R2

0.0425 (0.0262) 0.0564*** (0.0264) – – 0.4585 (0.0693) 0.9998 (0.2544) 0.0330 (0.0013) – – – – – – – 0.1280

0.1711 (0.3506) 0.0963*** (0.0208) 0.1666** (0.0319) – – – 0.4104*** (0.0264) – – – – – – – 0.2478

– 1.9240 (0.3181) – 1.5819 (0.3400) – – – 0.6683*** (0.0294) 12.0587** (3.3127) 11.6286*** (8.8983) 51.4493*** (20.0920) 0.5989*** (0.7286) 0.3147*** (0.0287) 0.1774*** (0.0232) 0.3044

Notes: The table reports the estimating outcomes for DPt, DXt and D(Rh – Rf) of the short-run vector error correction model. 5 and 1% level denoted by ** and ***respectively, represent the level of statistical significance Source: Authors’ calculations


Table VII. Error correction modeling estimates

of Evans and Lyons (2002a, 2002b) are significant. However, our R2 is relatively low at 0.128 compared to 0.64 and 0.46 from the research work of Evans and Lyons (2002a, 2002b). This is not amazing in the sense that, the level at which the currencies of emerging markets economy being traded in the international market is relatively low compared with world major currencies of the developed markets. In addition, most of the emerging markets economy do not operate free floating rather managed floating which may lead to frequent occurrence of currency interventions by the government. In view of this, we can reasonably say that this may be one of the main rationales for the difference in our results with that of Evans and Lyons (2002a, 2002b). The results our findings are in consistent with that of De Medeiros (2004), while analyzing order flow in the Brazilian foreign exchange markets, Cerrato et al. (2011) when they investigate the extent to which customer order flow analysis can help to explain exchange rate movements over and above the influence of macroeconomic variables and Zhang et al. (2013) when they examine the influential role of currency order flow on exchange rate fluctuations between Chinese RMB and US$. Other empirical studies in line with our findings include Evans and Lyons (2005), Marsh and O’Rourke (2005), Sager and Taylor (2008), Evans (2010) and Rime et al. (2010) (Table VIII).

Period 10 20 30

Standard error

Pt

Xt

(hi – fi)

(hr – fr)

(Rh – Rf)

0.0031 0.0044 0.0054

97.8298 95.2908 92.3822

7.0724 12.8567 15.0243

0.0427 0.1102 0.2045

0.1174 0.0764 0.0415

3.6732 3.4356 3.2023

Notes: The table reports the outcomes of decomposition of each item forecast error variance in our specification. We use Cholesky decomposition to test for a period of 30 trading days Source: Authors’ calculations

Table VIII. Variance decomposition of exchange rate

JCEFTS 10,2


158

While testing the potency of the relationship at longer horizons, we consider 6 weeks as 30 trading days, 4 weeks as 20 trading days and 2 weeks as 10 trading days. Therefore, we test for 30 trading days’ time horizon using Cholesky decomposition. The results of decomposition of each variable forecast error variance is reported in Table VII. However, it appears from the results that currency order flow is the most exogenous variable relative to other variables in our specification. That is, from the results, up to 15 per cent of the changes in the exchange rate fluctuations are caused by currency order flow. It is noteworthy to say that, per day trading period in the Thailand foreign exchange markets (THB/US$), currency order flow may account for 15 per cent of exchange rate fluctuations. In addition, 3.2 per cent of exchange rate fluctuations is brought about by the country risk premium difference. However, less than 1 per cent of exchange rate movement is explained by short-term and long-term interest differentials. In Thailand foreign exchange market, it appears that currency order flow and country risk variables are two influential determinant components of exchange rate fluctuations. Conclusion The determination of THB exchange rate against the US$ in the long term as well as short term are hereby investigated by this paper, taken into consideration, the influential role of cumulative currency order flow. To reflect the pressure of currency excess demand, we construct a measure of currency order flow in the Thailand foreign exchange market context. VAR is applied to estimate the long-run components and short-run dynamics, and the results show that between the cumulative currency order flow and exchange rate of US$ and THB, there exists cointegrating relationship. Therefore, the major fluctuations in the exchange rate of the THB/US$ is actually due to currency order flow. The explanatory power of the currency order flow is positively strong. With a positive beta coefficient of currency order flow (0.0027) in the US$/THB exchange rate, it means that within the day transaction, for every currency order flow increasing at 1 per cent, there will be a corresponding increase of 27 basis points of the THB price against the US$. Insomuch that, the results show that currency order flow, a microeconomic variable, has significant explanatory power to capture the THB exchange rate movements in the foreign exchange market, it then brings to the attention of the Monetary Authority of Thailand the importance that should be attached to the market microstructure. In addition, while comparing our results, the coefficients of ours and that of Evans and Lyons (2002a, 2002b) are significant. Even though, our R2s relatively low at 0.128 compared to 0.64 and 0.46 from the research work of Evans and Lyons (2002a, 2002b). Although, this is not amazing, in the sense that the level at which the currencies of emerging markets economy being traded in the international market are relatively low compared with that of world major currencies of the developed markets. In addition, most of the emerging markets economy do not operate free floating rather managed floating which may lead to frequent occurrence of currency interventions by the government. In view of these, we can reasonably say that this may be one of the main rationales for the difference in our results with that of Evans and Lyons (2002a, 2002b). The results our findings are in consistent with other empirical studies, such as that of De Medeiros (2004), Marsh and O’Rourke (2005), Evans and Lyons (2005), Sager and Taylor (2008), Evans (2010), Rime et al. (2010), Cerrato et al. (2011) and Zhang et al. (2013).


References Andersen, T.G., Bollerslev, T., Diebold, F.X. and Vega, C. (2003), “Micro effects of macro announcements: real-time price discovery in foreign exchange”, The American Economic Review, Vol. 93 No. 1, pp. 38-62. Berger, D.W., Chaboud, A.P., Chernenko, S.V., Howorka, E. and Wright, J.H. (2008), “Order flow and exchange rate dynamics in electronic brokerage system data”, Journal of International Economics, Vol. 75 No. 1, pp. 93-109. Bjønnes, G.H. and Rime, D. (2005), “Dealer behavior and trading systems in foreign exchange markets”, Journal of Financial Economics, Vol. 75 No. 3, pp. 571-605. BOT (2016), available at: www.bot.or.th/English/monetarypolicy/default.aspx (accessed 27 October 2016). Cerrato, M., Sarantis, N. and Saunders, A. (2011), “An investigation of customer order flow in the foreign exchange market”, Journal of Banking and Finance, Vol. 35 No. 8, pp. 1892-1906. Cheung, Y.W., Chinn, M.D. and Marsh, I.W. (2005), “How do UK-based foreign exchange dealers think their market operates?”, International Journal of Finance and Economics, Vol. 9 No. 1, pp. 289-306. De Medeiros, O.R. (2004), Order Flow and Exchange Rate Dynamics in Brazil, SSRN eLibrary. Duffuor, K., Marsh, I.W. and Phylaktis, K. (2012), “Order flow and exchange rate dynamics: an application to emerging markets”, International Journal of Finance & Economics, Vol. 17 No. 3, pp. 290-304. Evans, M. (2010), “Order flows and the exchange rate disconnect puzzle”, Journal of International Economics, Vol. 80 No. 1, pp. 58-71. Evans, M. (2011), Exchange-Rate Dynamics, Princeton University Press, Princeton, NJ. Evans, M. and Lyons, R. (2002a), “Informational integration and FX trading”, Journal of International Money and Finance, Vol. 21 No. 6, pp. 807-831. Evans, M. and Lyons, R. (2002b), “Order flow and exchange rate dynamics”, Journal of Political Economy, Vol. 110 No. 1, pp. 170-180. Evans, M. and Lyons, R. (2005), “Do currency markets absorb news quickly?”, Journal of International Money and Finance, Vol. 24 No. 2, pp. 197-217. Evans, M. and Lyons, R. (2007), “Exchange rate fundamentals and order flow”, NBER Working Paper (w13151). Frankel, J.A. and Rose, A.K. (1995), “Empirical research on nominal exchange rates”, Handbook of International Economics, Princeton, North Holland, Vol. 3, pp. 1689-1729. Froot, K.A. and Ramadorai, T. (2005), “Currency returns, intrinsic value, and institutional-investor flows”, The Journal of Finance, Vol. 60 No. 3, pp. 1535-1566. Hasbrouck, J. (1991), “Measuring the information content of stock trades”, Journal of Finance, Vol. 46 No. 1, pp. 179-207. Johansen, S. (1995), “Likelihood-based inference in cointegrated vector autoregressive models”, Advanced Texts in Econometrics, Oxford University Press, Oxford; New York, NY. Lyons, R.K. (1995), “Tests of microstructural hypotheses in the foreign exchange market”, Journal of Financial Economics, Vol. 39 No. 2, pp. 321-351. Marsh, I.W. and O’Rourke, C. (2005), “Customer order flow and exchange rate movements: is there really information content?”, Cass Business School Research Paper. Meese, R.A. and Rogoff, K. (1983), “Empirical exchange rate models of the seventies: do they fit out of sample?”, Journal of International Economics, Vol. 14 No. 1, pp. 3-24. O’Hara, M. (1995), Market Microstructure Theory, Vol. 108, Blackwell Cambridge, Cambridge, MA. Osler, C.L. (2006), “Macro lessons from microstructure”, International Journal of Finance & Economics, Vol. 11 No. 1, pp. 55-80.


JCEFTS 10,2


160

Payne, R. (2003), “Informed trade in spot foreign exchange markets: an empirical investigation”, Journal of International Economics, Vol. 61 No. 2, pp. 307-329. Perraudin, W. and Vitale, P. (1996), “Interdealer trade and information flows in a decentralized foreign exchange market”, The Microstructure of Foreign Exchange Markets, University of Chicago Press, Chicago, IL, pp. 73-106. Rime, D. (2000), “Private or public information in foreign exchange markets? An empirical analysis”, Working Paper No. 14/2000 Available at SSRN. Rime, D., Sarno, L. and Sojli, E. (2010), “Exchange rate forecasting, order flow and macroeconomic information”, Journal of International Economics, Vol. 80 No. 1, pp. 72-88. Sager, M. and Taylor, M.P. (2008), “Commercially available order flow data and exchange rate movements: caveat emptor”, Journal of Money, Credit and Banking, Vol. 40 No. 4, pp. 583-625. Vitale, P. (2003), New Exchange Rate Economics, Dottorato Di Ricerca, Università Di Tor Vergata, Fall. Wu, T. (2010), “Order flow in the South: anatomy of the Brazilian FX market”, The North American Journal of Economics and Finance, Vol. 23 No. 3, pp. 310-324. Zhang, Z., Chau, F. and Zhang, W. (2013), “Exchange rate determination and dynamics in China: a market microstructure analysis”, International Review of Financial Analysis, Vol. 29, pp. 303-316. Further reading Ariff, M. (1991), The Pacific Economy: Growth and External Stability, Allen & Unwin Pty, North Sydney. Asian Development Bank (2015), Asian Development Bank Bulletin. Bacchetta, P. and Wincoop, E. (2006), “Can information heterogeneity explain the exchange rate determination puzzle?”, American Economic Review, Vol. 96 No. 3, pp. 552-576. Bank of Thailand (2015), Annual Report. Bank of Thailand (2016), Monetary Policy Report. Boyer, M.M. and Norden, S.V. (2006), “Exchange rates and order flow in the long run”, FinanceResearch Letters, Vol. 3 No. 4, pp. 235-243. Danielsson, J. and Love, R. (2006), “Feedback trading”, International Journal of Finance & Economics, Vol. 11 No. 1, pp. 35-53. Danielsson, J., Luo, J. and Payne, P. (2012), “Exchange rate determination and inter-market order flow effects”, The European Journal of Finance, Vol. 18 No. 9, pp. 823-840. Effiong, E.L. (2014), “Exchange rate dynamics and monetary fundamentals: a cointegrated SVAR approach for Nigeria”, Global Business Review, Vol. 15 No. 2, pp. 205-221. Evans, M. (2002), “FX trading and exchange rate dynamics”, The Journal of Finance, Vol. 57 No. 6, pp. 2405-2447. Evans, M. and Lyons, R. (2006), “Understanding order flow”, International Journal of Finance& Economics, Vol. 11 No. 1, pp. 3-23. Evans, M. and Lyons, R. (2008), “How is macro news transmitted to exchange rates?”, Journal of Financial Economics, Vol. 88 No. 1, pp. 26-50. Fisher, P. and Hillman, R. (2003), “Comments on Richard K. Lyons, Explaining and forecasting exchange rates with order flows”, Economic Policy Forum. Galati, G. (2000), Trading Volumes, Volatility and Spreads in Foreign Exchange Markets: Evidence from Emerging Market Countries, BIS, Monetary and Economic Department. IMF (2015), World Economic Outlook Database-2015. Ito, T., Lyons, R.K. and Melvin, M.T. (1998), “Is there private information in the FX market? The Tokyo experiment”, The Journal of Finance, Vol. 53 No. 3, pp. 1111-1130.

Lyons, R.K. (2001a), The Microstructure Approach to Exchange Rates, MIT Press, Cambridge, MA. Lyons, R.K. (2001b), “New perspective on FX markets: order-flow analysis”, International Finance, Vol. 4 No. 2, pp. 303-320. Martin, A.D. (2001), “Technical trading rules in the spot foreign exchange markets of developing countries”, Journal of Multinational Financial Management, Vol. 11, pp. 59-68. Stephen, G. and Takatoshi, I. (2011), “An independent evaluation of the bank of Thailand’s monetary policy under the inflation targeting framework, 2000-2010”, available at: www.bot.or.th/English/ MonetaryPolicy/Pages/Assessment.aspx World Currency Yearbook (WCY) (2008), IMF Annual Report on Exchange Arrangement and Exchange Restriction.


Corresponding author Abolaji Daniel Anifowose can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website: www.emeraldgrouppublishing.com/licensing/reprints.htm Or contact us for further details: [email protected]

View publication stats
