Physica A 391 (2012) 31703179

Contents lists available at SciVerse ScienceDirect

Physica A
journal homepage: www.elsevier.com/locate/physa

An investigation of Forex market efficiency based on detrended

fluctuation analysis: A case study for Iran
Esmaiel Abounoori a, , Mahdi Shahrazi b , Saeed Rasekhi b
a

Department of Economics, University of Semnan, Semnan, Iran

Department of Economics, University of Mazandaran, Babolsar, Iran

article

info

Article history:
Received 1 March 2011
Received in revised form 15 November
2011
Available online 5 January 2012
Keywords:
Efficient market hypothesis (EMH)
Forex market
RialDollar exchange rates
Detrended fluctuation analysis (DFA)
Iran

abstract
The efficient market hypothesis (EMH) states that asset prices fully reflect all available
information. As a result, speculators cannot predict the future behavior of asset prices and
earn excess profits at least after adjusting for risk. Although initial tests of the EMH were
performed on stock market data, the EMH was soon applied to other markets including
foreign exchange (FX). This study uses the detrended fluctuation analysis (DFA) technique
to test 01:12:200518:04:2010 Iranian Rial/US Dollar exchange rate time series data to
see if it can be explained by the weak form of the EMH. Moreover, to determine changes
in the degree of inefficiency over time, the whole period has been divided into four
subperiods. The study shows that the Iranian Forex market (the Rial/Dollar case) is weakform inefficient over the whole period and in each of the subperiods. However, the degree
of inefficiency is not constant over time. The findings suggest that profitable risk-adjusted
trades could be made using past data.
2011 Elsevier B.V. All rights reserved.

1. Introduction
The efficient market hypothesis (EMH) originally introduced by Fama [1] is concerned with the behavior of prices in
asset markets, and it states that prices fully reflect all available information. To be more precise, this hypothesis focuses on
informational efficiency [2]. In an informationally efficient market, changes in asset prices will occur in response to news
which cannot be predicted in any systematic manner. In other words, asset prices respond only to the unexpected part of
any news, since the expected part of the news is already embedded in prices, and so more profit without higher risk is
impossible [3].
The analysis of foreign exchange market efficiency has been considered in the international finance literature over
three decades. The question of currency market efficiency is important. Exchange rates have a penetrating effect on all
other prices. If the currency market is inefficient, currencies are often incorrectly priced. This distortion will spread over
all other markets, and it causes misallocation of resources that leads eventually to welfare losses. On the other hand,
inefficiency in currency markets can lead to excessive exchange rate volatility. Exchange rate volatility is inevitable when
rates float, but excess exchange rate volatility increases the exchange rate risk and may decrease the flow of trade and
investment [4].
The structure of this paper is as follows. Section 2 reviews the literature. Section 3 provides the data and the methodology
of the research. Section 4 presents an estimation along with some discussion. Finally, the paper ends with the conclusions
in Section 5.

Corresponding author. Tel.: +98 911 111 2176; fax: +98 112 534 2502.
E-mail addresses: esmaiel.abounoori@gmail.com, e.abounoori@profs.semnan.ac.ir, e.abounoori@umz.ac.ir (E. Abounoori), mehdishahrazi@yahoo.com
(M. Shahrazi), srasekhi@umz.ac.ir (S. Rasekhi).
0378-4371/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.physa.2011.12.045

E. Abounoori et al. / Physica A 391 (2012) 31703179

3171

2. Literature review
Although the idea of the EMH may be traced to the 1990 work of Louis Bachelier, the organized modern presentation of
the EMH is due to Eugene Fama. In 1970, Fama published his now-famous paper, Efficient Capital Markets: A Review of
Theory and Empirical Work. He helps with the focus and direction of future research by defining three different forms of
market efficiency depending on the specification of the information set: weak form, semi-strong form, and strong form; see
Refs. [5,6].
Weak-form efficiency.
The weak form of market efficiency states that the information content of all past prices is reflected in the current market
price. This implies that, in contrast to the methods of technical analysts, it is impossible to obtain a market advantage by
analyzing only past prices.1 For a foreign exchange market this means that no profitable information about future exchange
rate movements can be obtained by analyzing the past exchange rates. However, fundamental analysis can be successfully
applied [8].2 The present paper examines the validity of this sample version of EMH. Test of weak-form foreign exchange
market efficiency can be categorized into three groups: uncovered interest rate parity (UIP), forward market efficiency, and
spot market efficiency.
Uncovered interest parity (UIP) postulates that the expected foreign exchange gain from holding one currency rather than
another must be just offset by the interest rate differential. The UIP condition can be expressed as
Et (rt +k ) rt = it ,k it ,k ,

(1)

where Et (rt +k ) denotes the logarithm of the expected spot exchange rate at time t + k, when the expectation is made at time
t; rt denotes the logarithm of the spot exchange rate at time t; it ,k and it ,k are the nominal interest rates at time t available on
similar domestic and foreign securities, respectively, with k periods to maturity. A test of market efficiency can therefore be
designed to examine whether the interest rate differential is an unbiased estimator of the future spot rate [4]. Unbiasedness
has been tested by regressing the change in exchange rate on the interest rate differential using the following equation:

1rt +k = 1 + 2 (it ,k it ,k ) + t +k .

(2)

This is known as the uncovered interest parity condition test for market efficiency. 1rt +k is the change in the logarithm of
the spot price over period k, and t +k is a disturbance term. The null hypotheses are 1 = 0 and 2 = 1.
According to forward market efficiency, the forward exchange rate should be an unbiased predictor of the future spot
exchange rate. By replacing the interest rate differential in the above equation with the difference between the forward and
spot rate such that

1rt +k = 1 + 2 (ft ,k rt ) + t +k ,

(3)

the forward rate unbiasedness can be tested. ft ,k is the logarithm of the k-period forward rate, i.e., the rate agreed now for
an exchange of currencies k periods ahead. Similar to the UIP test, foreign exchange market efficiency is considered to hold
if the estimates of and are not significantly different from zero and unity, respectively [4].
Since a forward market has not yet been developed in Iran and as a result the forward exchange rate is absent, so this
category cannot be used for testing the efficiency of the foreign exchange market of Iran.
Spot market efficiency can easily be tested using historical data and studying the predictability in asset returns through
correlations [9]. The null hypothesis is that changes in spot exchange rates are serially uncorrelated. In other words, the
question is whether the spot exchange rate behaves as a random walk. In this paper, the focus is on the spot market efficiency.
Semi strong-form efficiency.
In a semi-strong efficient market, prices reflect all publicly available information about economic fundamentals [8].
Therefore, it is not possible for any investor to earn abnormal returns using any publicly available information [6]. In this
level of efficiency, neither fundamental nor technical analysis can systematically produce abnormal returns [7].
Strong-form efficiency.
In the highest level of efficiency, no investor can earn excess return by using any information available, whether it is
private or public information. In another words, while the semi-strong form precludes the profitability of both technical
and fundamental analyses, strong-form efficiency implies that even those with privileged information cannot expect to
earn excess returns. This extreme form serves mainly as a limiting case [6]. Grossman and Stiglitz [10] identified a major
theoretical problem with the hypothesis termed the paradox of efficient markets, which they developed in the context of
equity markets. The argument starts by noting that all available information is costly, and therefore there must be a financial
incentive for traders to gather and analyze it. But if markets were perfectly efficient, traders would not be able to make
excess returns on any available information. Partly for these reason, Campbell et al. [11] suggest that the debate about
perfect efficiency is pointless, and that it is more sensible to evaluate the degree of inefficiency than to test for absolute
efficiency [12].
1 Technical analysis is the practice of identifying recurring patterns in historical prices in order to forecast future price trends [7].
2 In fundamental analysis, one tries to forecast future price trends by studying macroeconomic variables such as inflation, interest rate, public debt levels,
GDP, and money supply.

3172

E. Abounoori et al. / Physica A 391 (2012) 31703179

The term Efficient Market was initially applied to the stock market, but was soon generalized to other asset markets
such as foreign exchange. For some empirical studies about stock market efficiency, see Refs. [5,1321].
In most of the empirical studies, developed markets show stronger evidence of informational efficiency than emerging
markets [22]. It should be mentioned that emerging markets differ from developed markets in some important categories
such as size, liquidity, transparency, trading costs, and operational efficiency [23].
The fact that the market efficiency (especially in the weak form) usually holds in many highly developed economies
but usually has been violated in transition (developing) economics shows that a number of universality laws discovered in
physics do not extend to economics [24]. Even if an emerging market is efficient, the reason for this efficiency is likely to
be different than for an efficient developed market. Many suppose there to be two types of trader in the market: informed
traders and noise traders. Informed traders are rational agents that decide by their expectations about future underlying
fundamentals and so keep the exchange rate in line with macroeconomic fundamentals and help stabilize markets around a
new equilibrium. Noise traders make their decisions on the basis of trends and/or fads unrelated to fundamentals. Noise
traders are not able to distinguish between real and pseudo signals, and therefore they are destabilizing speculators.
Currency market efficiency in developed markets may result from the dominant position of informed traders. In developed
markets, informed traders are usually more numerous relative to noise traders. Consequently, the noise traders lose money
to informed traders and so eventually disappear from the market [25]. In contrast to developed markets, efficiency in
emerging currency markets (if it holds) is likely due to official intervention. When the noise traders are more active than the
informed traders, the currency market cannot function efficiently. In this case, a central bank may play the role of informed
traders, act as a stabilizing speculator, earn profits from such deviations, and lead the market to efficiency.
Many empirical studies have been carried out for testing currency market efficiency using different methods. A few of
these studies are presented below.
In a seminal study [26], Fama examined the efficiency in nine exchange markets (nine currencies against the US dollar),
using monthly data from 8:1973 to 12:1982. Ordinary least-squares (OLS) estimation showed that the EMH is not accepted
because of a time-varying risk premium.
The foreign exchange market efficiency for the British pound, Japanese Yen, and Canadian dollar against the US dollar
from 1:1976 to 6:1996 (monthly observations) was tested [27]. The efficiency hypothesis in all exchange rates using the
vector error correction model (VECM) was strongly rejected.
The applicability of the EMH to the Australian foreign exchange market (US dollar, British Pound, Japanese Yen, and
Swiss Franc against the Australian dollar) has been analyzed [28]. The filter rule technique on the daily exchange rates for
the period from 1984 to 2003 has been applied. For the total period, the filter rules were, on average, profitable for the British
Pound, Japanese Yen, and Swiss Franc against the Australian dollar. In contrast, the filter rule for the US dollar against the
Australian dollar, on average, generated losses. Under these results, the EMH was accepted for Australian dollar/US dollar
exchange rate time series only.
Using rescaled range analysis during the period 19952006 (daily RealDollar data) the efficiency of the Brazilian foreign
exchange market has been examined [29]. The authors conclude that the market was inefficient before the currency crisis
in 1999, but after the crisis it has tended to efficiency.
The relative market efficiency in foreign exchange markets, applying the approximate entropy (ApEn) method, has been
investigated [30]. The authors used global foreign exchange market indices for 17 countries during two periods, from 1984
to 1998 and from 1999 to 2004 (daily observations). The results suggested that more liquid European, North American, and
Japanese markets are more efficient than smaller, less liquid, African and non-Japanese Asian markets.
Using monthly return series for each of these markets over a period of 21 years (19852005), the weak-form efficiency of
the foreign exchange markets in seven SAARC (The South Asian Association for Regional Cooperation) countries3 has been
analyzed [31]. The authors applied a battery of unit root tests and variance ratio tests to see whether the return series follow
a random walk process. The results suggested that the increments of the return series are not serially correlated and follow
a random walk process. Therefore, they concluded that the foreign exchange markets in SAARC countries are weak-form
efficient.
Some crucial properties of the spot foreign exchange market, efficiency, liquidity, and volatility, were studied in Ref. [8]
using minute-by-minute exchange rates on seven currency pairs with the US dollar for the period 03:200905:2009. The
exchange rates considered were US dollar versus seven major currencies. The results of the augmented DickeyFuller and
the PhillipsPerron tests showed that the market is efficient in weak form.
3. Data and methodology
3.1. Data description
Forex is a market that functions 24 h a day. Unlike other financial markets, the Forex market has neither a physical location
nor a central exchange, and it operates through an electronic network of banks, corporations, and individuals trading one

3 SAARC is an economic and political organization of eight countries, namely India, Pakistan, Bangladesh, Sri Lanka, Nepal, Bhutan, Maldives, and
Afghanistan.

E. Abounoori et al. / Physica A 391 (2012) 31703179

3173

Fig. 1a. Rial/Dollar Forex exchange rates from 01:12:2005 to 18:04:2010.

Fig. 1b. Log returns of Rial/Dollar Forex exchange rates from 01:12:2005 to 18:04:2010.

currency for another. Low barriers to entry, generally no commission, high leverage,and free trading tools make the Forex
market attractive for traders. They may reply to a currency fluctuation caused by economic, social, or political events. In
the stock market and the futures market or any market with a central exchange, there are transaction costs, including
commission charges, which lower the profit margins. The commission charges in the Forex market are almost negligible.
For the most part, Forex brokers do not charge commission. Another transaction cost is the slippage cost caused by the fact
that buying pushes prices up and selling pushes them down. In a liquid market like the Euro/US dollar (USD) market, this
slippage cost is also very small except for the largest of traders, but, in a smaller market like the Rial/USD market, this slippage
could become substantial. Therefore, even a strong pattern in the data is not necessarily going to lead to a profitable trading
rule.
Data used in this paper includes daily spot Forex data between the US Dollar (USD) and the Iranian Rial (IRR) during the
period from 01:12:2005 to 18:04:2010. Before 01:12:2005, the Forex exchange rates were fixed. The number of data points
is 1600. During this period, Iran adopted a managed floating exchange rate regime and followed the policy of leaning against
the wind. The data was obtained via the OANDA electronic foreign exchange service [31]. This paper represents the first
study on the efficiency of the Iranian Forex market.
Let yt denote the Rial/Dollar exchange rate at time t. The return rt at time t is given by the logarithmic difference between
consecutive daily exchange rates:

rt = ln

yt
yt 1

(4)

Figs. 1a and 1b respectively exhibit the Rial/Dollar exchange rate and log returns of the Rial/Dollar daily Forex exchange
rates from 01:12:2005 to 18:04:2010. According to Fig. 1a, the value of the Dollar against the Rial has remained in a narrow
range during most of this period, except for a decreasing trend during the 2 months between 15:02:2008 and 17:04:2008
and for an increasing trend during the 8 months between 17:04:2008 and 15:12:2008. Nevertheless, these observations do
not allow us to draw conclusions about the efficiency of this Forex market.
By looking at Fig. 1b, it seems that there are recurring patterns in the log returns of the Rial/Dollar Forex exchange rates.
One may not be able to identify this from Fig. 1b, but a careful examination of the exchange rate returns may reveal the
validity of this statement. This may indicate that the Iranian Forex market is inefficient. However, more exact judgment is
deferred to Section 4.

3174

E. Abounoori et al. / Physica A 391 (2012) 31703179

3.2. Detrended fluctuation analysis (DFA)

In the empirical literature, test of foreign exchange market efficiency in weak form has been carried out using different
econometric tests such as augmented DickeyFuller (ADF) and PhillipsPerron (PP) unit root tests [6,32]; and various
econophysics approaches such as Hurst analysis [33,34], detrended cross-correlation analysis (DCCA) [35], detrended
moving average (DMA) [36], detrended fluctuation analysis (DFA), and algorithmic complexity theory [37].
In this article, we have applied the detrended fluctuation analysis (DFA) proposed in Ref. [39], to determine whether
Rial/Dollar daily Forex exchange rates behave as a random walk. DFA provides a simple quantitative parameter to represent
the correlation properties of a time series. The advantages of DFA over older methods (e.g., Hurst analysis) are that it removes
local trends through the least-squares regression fit and is relatively immune to non-stationarity [37]. Therefore, it permits
the detection of long-range correlations embedded in non-stationary series, avoiding the spurious detection of apparent
long-range correlations that are artifacts of non-stationarity [38].
In the DFA technique, one divides a time series y(t ) of length N into N /n no overlapping boxes of equal size [39]. The
variable t is discrete, t = 1, . . . , N; N /n is an integer, and each box contains n points. Let z (t ) be the local trend function in
each n-size box:
z (t ) = at + b.

(5)

z (t ) in each box can be obtained by local an ordinary least-squares fit of the data points in that box. Then detrended
fluctuation function F (n) is calculated from
F 2 ( n) =

(k+1)n
1

n t =kn+1

|y(t ) z (t )|2

k = 0, 1, 2, . . . ,

N
n

1 .

(6)

Averaging F 2 (n) over all N /n box sizes centered on time n gives the fluctuations F 2 (n) as a function of n. The calculation is

12

repeated for all possible different values of n [39]. According to Ref. [36], a relationship between F 2 (n)
as

F 2 (n)

12

n .

and n is expected

(7)

1/2
is called the scaling exponent [39]. By performing a least-squares regression with log F 2 (n)
as the dependent variable

and log n as the independent one, where a straight line is observed, we find the slope of the regression, which is the estimate
of the scaling exponent (). can be interpreted as follows.

If = 0.5, the time series is independent, there is no correlation at all, and the time series represent a random walk. This
is evidence for market efficiency.

If 0.5 < < 1, persistent long-range correlations are present in the time series. This implies that if the series has been
up or down in the last period then the chances are that it will continue to be up or down, respectively.

If 0 < < 0.5, anti-persistent long-range correlations are present in the time series. This means that whenever the time
series have been up in the last period, it is more likely that it will be down in the next period.
The presence of long-range correlations indicates evidence of a predictable component for future returns and invalidates
the weak form of the EMH [5].
4. Empirical results
In this study, y(t ) is the daily spot Forex data between the USD and the IRR from 01:12:2005 to 18:04:2010. The series
contain N = 1600 data points; hence the variable t runs between t = 1 and t = 1600. Also, each n is a factor of 1600;, for

example, n = 10, n = 16 and n = 20. For each n, we have calculated F 2 (n) 2 as shown in Table 1. Using linear regression,
we have obtained the following estimate for the scaling exponent: = 0.303 0.01 (Fig. 2).
We have tested the null hypothesis of = 0.5 using the Wald test to understand whether or not the scaling
0.5 2
exponent obtained is statistically different from 0.5. The Wald statistic is given by W = (
) , and it follows a chi
squared distribution with one degree of freedom. is the estimated scaling exponent and is its standard error. The
critical values at the 1% and 5% significance levels are 3.8 and 6.6, respectively [33].
The calculated Wald statistic and its associated p-value were 251.71 and 0.000, respectively. This result show that
= 0.303 is significantly different from 0.5, and the null hypothesis is strongly rejected.
According to Fig. 2, the scaling exponent from the DFA technique shows a significant amount of anti-persistence or
negative long-range correlations ( = 0.30). This means that, whenever the time series have been up in the last period,
it is more likely that they will be down in the next period. Therefore, the DFA technique provides evidence of inefficiency

E. Abounoori et al. / Physica A 391 (2012) 31703179

3175

Table 1
The values of log F 2 (n)

1/2

for the period 01:12:200518:04:2010.

log n

log F 2 (n)

1
1.20412
1.30103
1.39794
1.50515
1.60206
1.69897
1.80618
1.90309
2
2.20412
2.30103
2.50515
2.60206
2.90309
3.20412

1.71453
1.74776
1.76113
1.77305
1.7924
1.81949
1.86065
1.86511
1.91734
1.93562
1.97363
2.03156
2.12052
2.12811
2.27723
2.31837

1/2

Fig. 2. Scaling exponent from the DFA technique for the period 01:12:200518:04:2010.

in the Iranian Forex market.4 However, by looking carefully at Fig. 1b, one can see very noisy signals with a strong negative
correlation at the start of the selected period. As mentioned above, before 01:12:2005 the Iranial Rial/USD Forex rate was
fixed. It is possible that the negative long-range correlation and inefficiency discovered over the entire period is just an
artifact of the beginning of the exchange rate flexibility. To examine this issue, and also to determine changes in the degree
of inefficiency over time, we divide the whole period into four subperiods: 01:12:200504:1:2007, 04:01:200708:02:2008,
08:02:200814:03:2009, and 14:03:200918:04:2010. The number of data points in each subperiod is 400. Tables 25 and

1/2

Figs. 36 show the values of log F 2 (n)

and the estimation of the scaling exponent for each subperiod, respectively.
The estimated scaling exponent for the subperiods as well as for the whole period and their associated Wald tests are
presented in Table 6. According to this table, in all four subperiods the scaling exponent is statistically smaller than 0.5.
Therefore, there is negative long-range correlation in each subperiod as evidence against the weak form of market efficiency.
However, the degree of inefficiency is not constant. In the first subperiod, the inefficiency is very high ( = 0.07).
The market became less inefficient in the second subperiod ( = 0.34). The inefficiency in the third subperiod is small
( = 0.45), but it increases again in the last subperiod ( = 0.21). Hence, one can conclude that the Iranian Forex market
(the Rial/Dollar case) is weak-form inefficient not only in the first subperiod but also in the other three subperiods as well as
over the whole period. As a result, it should be possible for speculators to extract risk-free profits by studying past Rial/Dollar
series. In this case, one may suggest that a central bank can play the role of informed traders (especially in an oil exporter
country such as Iran in which bulks of foreign currency are held in a central bank), act as a stabilizing speculator, earn profits
from such deviations, and lead the market to efficiency.

4 According to data available in the International Financial Statistics (IFS) data base, the mean and standard deviation of the gap between interest rates
in Iran versus those in the US has been about 0.086 and 0.027, while the mean and standard deviation of the currency exchange rate has been about 0.002
and 0.008, respectively. Using the UIP method and quarterly data, we have estimated the simple regression (2) and then tested 1 = 0 and 2 = 1 applying
the Wald test. According to the Wald statistic, the null hypothesis of 1 = 0 and 2 = 1 is rejected at less than 0.0001 significance level, supporting the
results obtained from the DFA.

3176

E. Abounoori et al. / Physica A 391 (2012) 31703179

Fig. 3. Scaling exponent from the DFA technique for the first subperiod, from 01:12:2005 to 04:01:2007.

Fig. 4. Scaling exponent from the DFA technique for the second subperiod, from 04:01:2007 to 08:02:20.

Fig. 5. Scaling exponent from the DFA technique for the third subperiod, from 08:02:2008 to 14:03:2009.

Fig. 6. Scaling exponent from the DFA technique for the fourth subperiod, from 14:03:2009 to 18:04:2010.

E. Abounoori et al. / Physica A 391 (2012) 31703179

Table 2

1/2

The values of log F 2 (n)

for the first subperiod, from
01:12:2005 to 04:01:2007.

log n

log F 2 (n)

0.90309
1
1.20412
1.30103
1.39794
1.60206
1.69897
1.90309
2
2.30103
2.60206

1.904304
1.915711
1.930148
1.933246
1.939398
1.960426
1.964784
1.978883
1.989787
1.997618
2.092274

1/2

Table 3

1/2

for the second subperiod, from

The values of log F 2 (n)
04:01:2007 to 08:02:2008.

log n

log F 2 (n)

0.90309
1
1.20412
1.30103
1.39794
1.60206
1.69897
1.90309
2
2.30103
2.60206

1.178904
1.119025
1.275975
1.239009
1.343343
1.398969
1.436598
1.48487
1.49441
1.610586
1.719279

1/2

Table 4

1/2

The values of log F 2 (n)

for the third subperiod, from
08:02:2008 to 14:03:2009.

log n

log F 2 (n)

0.90309
1
1.20412
1.30103
1.39794
1.60206
1.69897
1.90309
2
2.30103
2.60206

1.597106
1.613496
1.702932
1.727691
1.75312
1.871121
1.947857
2.016253
2.055753
2.220107
2.335098

1/2

Table 5

1/2

The values of log F 2 (n)

for the fourth subperiod, from
14:03:2009 to 18.04.2010.

log n

log F 2 (n)

0.90309
1
1.20412
1.30103
1.39794
1.60206
1.69897
1.90309
2
2.30103
2.60206

1.627659
1.661136
1.687137
1.696034
1.724092
1.733712
1.797772
1.908227
1.899506
1.927058
1.932364

1/2

3177

3178

E. Abounoori et al. / Physica A 391 (2012) 31703179

Table 6
Results of the DFA technique and the associated Wald test for the null hypothesis of
H = 0.5 for the Rial/Dollar exchange rate series from 01:12:2005 to 18:04:2010.
Time interval

Scaling exponent from the DFA

Wald test (p-value)

01:12:200504:01:2007
04:01:200708:02:2008
08:02:200814:03:2009
14:03:200918:04:2010
01:12:200518:04:2010

0.069 0.00
0.336 0.02
0.453 0.01
0.206 0.02
0.303 0.01

19896.22 (0.000)
74.12 (0.000)
14.84 (0.000)
193.99 (0.000)
251.71 (0.000)

5. Conclusion
The efficient market hypothesis (EMH) states that prices fully reflect all available information. Examination of foreign
exchange market efficiency has been a debated issue within international finance literature over three decades. Whether
currency markets are efficient or not is an important question, since exchange rates have a penetrating effect on all other
prices.
In this paper, we have investigated the validity of the weak-form EMH for Iranian Rial/US Dollar Daily Forex exchange
rate time series from 01:12:2005 to 18:04:2010 using the DFA technique. Low barriers to entry, generally no commission,
high leverage, and free trading tools make the Forex market attractive for traders.
Our results suggest negative long-range dependence (anti-persistence). This means that, whenever the exchange rate
has been up, it is more likely that it will be down in the close future. Therefore, the Rial/Dollar daily Forex exchange rate is
inefficient, implying that it should be possible for speculators to extract risk-free profits by studying past Rial/Dollar series.
In this case, one may suggest that a central bank can play the role of informed traders, especially in an oil-exporting country
like Iran in which the bulk of foreign currency reserves are held in a central bank. In this way, the Iranian central bank could
act as a stabilizing speculator, earn profits from such deviations, and lead the market to efficiency.
Finally, we have divided the whole period into four subperiods (01:12:200504:1:2007, 04:01:200708:02:2008,
08:02:200814:03:2009, and 14:03:200918:04:2010) to identify changes in inefficiency over time. According to our
calculations, the inefficiency was very high in the first subperiod ( = 0.07). The market became less inefficient in the
second subperiod ( = 0.34). The inefficiency in the third subperiod was small ( = 0.45), but it increased again in the last
subperiod ( = 0.21). Hence, one can conclude that the Iranian Forex market (the Rial/Dollar case) is weak-form inefficient
not only in the first subperiod but also in the other three subperiods as well as over the whole period.
Acknowledgments
We are grateful for the valuable suggestions, and major and minor corrections which have been recommended by the
reviewers, as well as the main editor H.E. Stanley, which significantly improved the quality of this paper.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]

E. Fama, Efficient capital market: a review of theory and empirical work, Journal of Finance 25 (1970) 383417.
A. Timmermann, Clive W.J. Granger, Efficient market hypothesis and forecasting, International Journal of Forecasting 20 (2004) 1527.
F. Shostak, In defense of financial analysis: a critique of the efficient market hypothesis, The Review of Austrian Economics 10 (1997) 2745.
T. Boulter, The efficiency of currency markets: studies of volatility and speed of adjustment, Queensland University of Technology, 2007.
J.A. Skjeltorp, Scaling in the Norwegian stock market, Physica A 283 (2000) 486528.
L. Jones, M. Netter, Efficient capital markets, 2008. Available at: http://www.econlib.org.
M. Beechey, D. Gruen, J. Vickery, The efficient market hypothesis: a survey, Research Discussion Paper, Reserve Bank of Australia, 2000.
A. Serbinenko, S.T. Rachev, Intraday spot foreign exchange market. Analysis of efficiency, liquidity and volatility, Journal of Investment Management
and Financial Innovations 4 (2009) 3545.
M. Couillard, M. Davison, A comment on measuring the Hurst exponent of financial time series, Physica A 348 (2005) 404418.
S.J. Grossman, J.E. Stiglitz, On the impossibility of informationally efficient markets, American Economic Review 70 (1980) 393408.
J.Y. Campbell, A.W. Lo, A.C. MacKinlay, The Econometrics of Financial Markets, Princeton University Press, Princeton, 1997.
C.J. Neely, Technical analysis in the foreign exchange market: a Laymans guide, Federal Reserve Bank of St. Louis Review 79 (1997) 2338.
J.T. Barkoulas, C.F. Baum, Long-term dependence in stock returns, Economics Letters 53 (1996) 253259.
D.O. Cajueiro, B.M. Tabak, Evidence of long range dependence in Asian equity markets: the role of liquidity and market restrictions, Physica A 342
(2004) 656664.
E. Panas, Estimating fractal dimension using stable distributions and exploring long memory through ARFIMA models in Athens stock exchange,
Applied Financial Economics 11 (2001) 395402.
L. Zunino, M. Zanin, B.M. Tabak, D.G. Perez, O.A. Rosso, Forbidden patterns, permutation entropy and stock market inefficiency, Physica A 388 (2009)
28542864.
C. Eom, S. Choi, G. Oh, W.-S. Jung, Hurst exponent and prediction based on weak-form efficient market hypothesis of stock markets, Physica A 387
(2008) 46304636.
H.F. Coronel-Brizio, A.R. Hernandez-Montoya, R. Huerta-Quintanilla, M. Rodriguez-Achach, Evidence of increment of efficiency of the Mexican stock
market through the analysis of its variations, Physica A 380 (2007) 391398.
B. Toth, J. Kertesz, Increasing market efficiency: evolution of cross-correlations of stock returns, Physica A 360 (2006) 505515.
B. Podobnik, Do. Fu, T. Jagric, I. Grosse, H.E. Stanley, Fractionally integrated process for transition economics, Physica A 362 (2006) 465470.
R. Costa, G. Vasconcelos, Long-range correlations and nonstationarity in the Brazilian stock market, Physica A 329 (2003) 231248.
J.H. Wright, Long memory in emerging market stock returns, Emerging Markets Quarterly 5 (2001) 5055.
R. Levich, The importance of emerging capital markets, Brookings-Wharton Papers on Financial Services (2001).

E. Abounoori et al. / Physica A 391 (2012) 31703179

3179

[24] J. Tenenbaum, D. Horvatic, S.C. Bajic, B. Pehlivanovic, B. Podobnik, H.E. Stanley, Comparison between response dynamics in transition economies and
developed economies, Physical Review E 82 (2010) 046104.
[25] A. Kaltenbrunner, M. Nissanke, The case for an intermediate exchange rate regime with endogenizing market structures and capital mobility, Working
Papers UNU-WIDER Research Paper, World Institute for Development Economic Research, 2009.
[26] E. Fama, Forward and spot exchange rates, Journal of Monetary Economics 14 (1984) 319338.
[27] E. Zivot, Cointegration and forward and spot exchange rate regressions, Journal of International Money and Finance 19 (2000) 785812.
[28] J. Nguyen, The efficient market hypothesis: is it applicable to the foreign exchange market?, Economics Working Papers, University of Wollongong,
Australia, 2004.
[29] S. Da Silva, R. Matsushita, I. Gleria, A. Figueiredo, Hurst exponents, power laws, and efficiency in Brazilian foreign exchange market, Economics Bulletin
7 (2007) 111.
[30] G. Oh, S. Kim, Market efficiency in foreign exchange markets, Physica A 382 (2007) 209212.
[31] http://www.oanda.com/.
[32] G.B. Wickremasinghe, Weak-form efficiency of foreign exchange markets of developing economies, Journal of Emerging Market Finance 7 (2008)
169196.
[33] S.R.S. De Souza, B.M. Tabak, D.O. Cajueiro, Long-range dependence in exchange rates: the case of the European monetary system, International Journal
of Theoretical and Applied Finance 11 (2008) 199223.
[34] S.V. Muniandy, S.C. Lim, R. Murugan, Inhomogeneous scaling behaviors in Malaysian foreign currency exchange rates, Physica A 301 (2001) 407428.
[35] B. Podobnik, H.E. Stanley, Detrended cross-correlation analysis: a new method for analyzing two nonstationary time series, Physical Review Letters
100 (2008) 084102.
[36] L. Xu, P.Ch. Ivanov, K. Hu, Z. Chen, A. Carbone, H.E. Stanley, Quantifying signals with power-law correlations: a comparative study of detrended
fluctuation analysis and detrended moving average techniques, Physical Review E 71 (2005) 051101.
[37] C. Taufemback, R. Giglio, S. Da Silva, Algorithmic complexity theory detects decreases in the relative efficiency of stock markets in the aftermath of
the 2008 financial crisis, Economics Bulletin 31 (2011) 16311647.
[38] C.-K. Peng, S.V. Buldyrev, S. Havlin, M. Simons, H.E. Stanley, A.L. Goldberger, Mosaic organization of DNA nucleotides, Physical Review 49 (1994)
16851689.
[39] C.-K. Peng, S. Havlin, H.E. Stanley, A.L. Goldberger, Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time
series, Chaos 5 (1995) 8287.

Further reading
[1] A. Noman, M.U. Ahmed, Efficiency of the foreign exchange markets in South Asia, Afro-Asian Journal of Finance and Accounting 1 (2009) 295305.
[2] K. Hu, P.C. Ivanov, Z. Chen, P. Carpena, H.E. Stanley, Effect of trends on detrended fluctuation analysis, Physical Review 64 (2001) 119.
[3] M. Ausloos, Statistical physics in foreign exchange currency and stock markets, Physica A 285 (2000) 4865.