The surveies on market efficiency so far have reached different decisions because different ways have been used to depict efficiency and assorted methods have been applied to prove the being of efficiency conditions. The earlier surveies were chiefly focused on mature and well-known US markets. Most of these surveies were based on the entropy of the stock market monetary values or returns and tested the correlativities of them.
In a really early survey, Working ( 1934, p.12 ) states that for capriciousness, the first difference of the monetary value series should be every bit random as possible and suggests look intoing random walk behavior of the monetary value alterations instead than that of the monetary value series. He examines the alterations in wheat monetary values and compares the frequence distribution of these alterations with the frequence distribution of an unnaturally created random-difference series. He concludes that the wheat monetary value returns are acting as a random difference series.
After about 30 old ages, Fama ( 1965 ) asks the inquiry whether the past monetary values can be used to foretell future monetary values and investigates an reply to this inquiry by proving the day-to-day alterations in the monetary values of 30 stocks of the Dow Jones Industrial Average. He uses the independency premise of the random walk theoretical account and investigates dependance by utilizing consecutive correlativities and runs trials. He finds strong grounds for random walk theory in favor of Chartist theories and concludes that the past monetary values can non be used to foretell future monetary values as their behavior is random and independent ( Fama, 1965, p.80 ) .
Most of the surveies in the sixtiess ( such as Working, 1960 ; Godfrey, et al. , 1963 ; Samuelson, 1965 ) suggest that the US stock market monetary value or monetary value returns fluctuate randomly, whereas merely a few surveies find non-random motions ( such as Niderhoffer & A ; Osborne, 1966 ) . In 1970, Fama ( 1970 ) refers to some surveies on market efficiency until that clip and defines three different signifiers of market efficiency by utilizing the consequences of those surveies. In his survey, the efficient market hypothesis is evolved and Fama ( 1970, p.383 ) concludes that the empirical surveies provide strong grounds that this hypothesis, with few exclusions, works good.
After the early 1970s, the surveies started to happen strong grounds against efficient market hypothesis ( such as Watts, 1978 ; Galai, 1978 ) , unlike the findings of the surveies in the sixtiess. Jensen ( 1978, p.95 ) provinces: “I believe there is no other proposition in economic science which has more solid empirical grounds back uping it than the Efficient Market Hypothesis.” and explains that as more complicated methods are used to look into efficient market hypothesis and as the quality of informations improves, the consequences of the new surveies start beliing with the older 1s and, therefore, the consequences are non parallel with efficient market hypothesis every bit by and large as they were earlier. He besides claims that a new stage in which the consequences are anomalous is get downing.
In 1980, it is shown that perfect informational efficiency is impossible ( Grossman & A ; Stiglitz, 1980 ) . Grosmann and Stiglitz ( 1980, p.404 ) province that a absolutely informational efficient market can non be since the information is non complimentary besides claim that costless information is non merely a sufficient but besides a necessary status for market efficiency in the context of efficient market hypothesis.
After the 1980s, besides the methods used by earlier surveies started to be questioned and the power and dependability of those techniques ( i.e. runs trials ) were criticised. Summers ( 1986 ) showed that the statistical trials used in market efficiency surveies until that clip hold really low power and can non observe plausible signifiers of inefficiency.
Poterba and Summers ( 1988 ) look into the mean-reversion feature of 18 markets around the universe ( including USA ) . They find that the monetary value returns show positive auto-correlation in short periods, whereas for long periods the auto-correlation is negative. They besides conclude that these consequences provide strong grounds against the cogency of random walk hypothesis although they can non happen strong grounds to reject it ( Poterba & A ; Summers, 1988, p.53 ) .
After strong groundss against the power and dependability of old statistical methods to prove efficient market hypothesis, Lo and Mackinlay ( 1988 ) proposed a new method to prove random walk behaviors and they showed that the hebdomadal US stock market returns are non following a random walk by utilizing this trial. Huge Numberss of surveies use this new method to prove random walk hypothesis. This trial has become a normally used trial for entropy of monetary value returns.
Although the earlier surveies were focused on mature and well-known markets, the efficiency of emerging markets in developing states has besides been investigated after 1990s. There have been assorted consequences ( non surprisingly ) from the efficiency trials of the emerging markets and the relationship between the adulthood of a market and its efficiency has been debated.
Harvey ( 1993 ) investigates 41 different markets in the universe, 20 of which are developed, whereas the others are emerging markets of developing states. He finds that the monetary values in emerging markets are predictable which implies that these markets are weak form inefficient. Besides Claessens, et al. , ( 1995 ) look into 20 emerging stock markets and happen important predictability of returns in those markets by utilizing the first and 2nd autocorrelations of monthly returns.
On the other manus, in a more recent survey, Karemera, et al. , ( 1999 ) find grounds for the weak signifier efficiency of most of the 20 emerging markets, 12 of which are the same markets in Claessens, et al. , ( 1995 ) , by utilizing assorted discrepancy ratio trials. Furthermore, Kawakatsu and Morrey ( 1999 ) tries to happen out the impact of liberalisation policies on the degree of efficiency for 16 emerging markets, 13 of which are the same markets of Claessens, et Al. ‘s ( 1995 ) and 11 of which are the same markets in Karemera, et Al. ‘s ( 1999 ) , and they show that the markets were already efficient before liberalisation policies and still are. They find grounds for random walk behaviour in most of the markets by utilizing unit root and discrepancy ratio trials.
Ayadi and Pyunn ( 1994 ) applied the Lo & A ; Mackinlay discrepancy ratio trial to the emerging stock market of South Korea that is shown to be weak signifier efficient after this survey ( Karemera, et al. , 1999 ; Kawakatsu & A ; Morrey, 1999 ) . They find that the random walk is rejected under homoskedasticity premise of the error term for short period returns. However, test consequences can non reject the entropy of the monetary value returns when the trial applied under heteroskedasticity premise of the error term for shorter period returns. Besides, for longer period returns, they province that there is no grounds to reject the random walk hypothesis under either homoskedasticity or heteroskedasticity premise.
In 1999, Ojah and Karemera ( 1999 ) show that the equity indices of major emerging Latin American markets-Argentina, Brazil, Chile and Mexico are weak form efficient. They use multiple discrepancy ratio and auto-regression trials and supply grounds for capriciousness of future monetary values by utilizing past monetary values. Their consequence for the stock market of Chile contradicts the consequence of the survey by Claessens, et al. , ( 1995 ) . Besides Grieb and Reyes ( 1999 ) look into the efficiencies of Argentina and Brazil stock markets. They find strong grounds for weak signifier efficiency for the Brazilian stock market which is parallel to the consequences of Ojah and Karemera ( 1999 ) for that market, whereas the consequences reject the random walk behavior for Mexican stock monetary values which contradicts the findings of Ojah and Karemera ( 1999 ) for that market.
Antoniou and Ergul ( 1997 ) look into the stock market of Turkey and compare the degree of efficiency after considerable regulative alterations in 1990. They provide grounds for the random walk behavior of the stock market monetary values after 1990 by utilizing arrested development analysis and province that the market is weak signifier efficient after the regulative alterations. However, they showed that the monetary values do non follow a random walk before 1990. Besides, Antoniou and Ergul ( 1997, p.187 ) conclude that the adulthood of a market is related with efficient regulative alterations which improve engagement and the quality of information. Besides, Buguk and Brorsen ( 2003 ) provide back uping grounds for the weak signifier efficiency of the Turkish stock exchange market.
Although most of the surveies on market efficiency have been focused on the stock markets, options and hereafters markets have besides been investigated. The efficiency surveies on options and hereafters markets are largely based on two different methods: I ) Investigating arbitrage chances and two ) proving random walk behavior of the monetary value returns. The intuition behind the first method is that the being of chances to gain higher than normal net incomes is a misdemeanor of market efficiency while, the latter method is used to look into weak form informational efficiency in the context of efficient market hypothesis.
Stevenson and Bear ( 1970 ) investigates the cogency of random walk hypothesis for two trade good hereafters contracts, wheat and maize, on the Chicago Board of Trade. They used several mechanical trading techniques ( filters ) to look into the being of chances to do higher than normal net incomes. Besides they test the independence of the consecutive monetary value alterations by utilizing correlativity coefficients and runs analysis. They find that some mechanical trading techniques provide profitableness in long intervals which shows the being of long term tendencies. Although they province that the being of long term tendencies do non belie the random walk hypothesis, their consequences from other trials provide grounds against random walk behavior.
In 1984, Elton et al. , ( 1984 ) tests different schemes to supply an empirical probe of the efficiency of the US Treasury measures hereafters contracts by utilizing intraday informations. They provide grounds for arbitrage chances and, as a consequence, they conclude that the market is non efficient.
Figlewski ( 1984 ) surveies S & A ; P500 stock index hereafters and the public presentation of the hereafters market from 1982 to 1983. He finds differences between the existent and theoretical monetary values which may connote that there exist arbitrage chances. Figlewski ( 1984, p.667 ) states that if the mispricing is eliminated rapidly, this suggests that there is an arbitrage chance and the minutess of arbitragers push the existent monetary value to its theoretical degree. He shows that 70 % of the divergences from the theoretical monetary value are eliminated within one twenty-four hours and explains that there are arbitrage chances if it can be realized. Sing the fact that the stock indices contain big Numberss of stocks, coincident minutess for a proper portfolio to use an arbitrage chance is rather hard. Because of this, he states that the increasing velocity of extinguishing the difference between the existent and theoretical monetary values shows betterment of market efficiency. His consequences show that the market is going more efficient with clip.
Furthermore, in 1988, Mackinlay and Ramaswamy ( 1988 ) besides investigate S & A ; P500 stock index hereafters and they use intraday minutess informations from 1982 to 1987. They besides report mispricing and they find that the magnitude of mispricing additions with clip to adulthood of the S & A ; P500 hereafters contract. In add-on, they examine the auto-correlation of the monetary value alterations and the consequences show that the monetary value alterations are uncorrelated. They besides reported that the correlativity decreases when the interval between the monetary value alterations increases.
Chung ( 1991 ) uses Major Market Index ( MMI ) of the American Stock Exchange and MMI stock index hereafters contract traded in the Chicago Board of Trade from 1984 to 1986. By utilizing dealing costs in an index arbitrage scheme he shows that the old surveies overestimated the size and frequence of arbitrage chances. He besides concludes that the frequence of such chances declined dramatically as the market matures.
Clare and Miffre ( 1995 ) examine the MATIF CAC40 stock index hereafters contract traded in International French Futures and Option Exchange between 1990 and 1993 and Frankfurt Stock Exchange DAX hereafters contract from 1991 to 1993. They test trading theoretical accounts and they find a profitable trading theoretical account for Gallic stock index hereafters contract, nevertheless, they conclude that they can non happen such a regulation for German stock index hereafters. They besides use these consequences to claim that the Gallic Futures and Options Exchange is weak form inefficient.
In 1996, Antoniou and Holmes ( 1996 ) look into the efficiency of the FTSE100 stock index hereafters and its function on monetary value find from 1984 to 1993. They province that the efficiency of hereafters market has great importance for the monetary value find function of those markets ( Antoniou & A ; Holmes, 1996, p.116 ) . They use co-integration theoretical accounts and discrepancy edge trials for both short and long-term. The consequences suggest that the market is efficient, particularly, 1 or 2 months before adulthood. Besides they conclude that the market provides monetary value find for the investors.
Evans ( 2006 ) surveies the weak signifier efficiency of three UK fiscal hereafters contracts which are FTSE100 stock index hereafters, Long Gilt bond hereafters and Short Sterling involvement rate hereafters between 1990 and 2002. Besides he investigates the impact of the electronic trading system implemented in 1999 by utilizing the consequences of Lo and Mackinlay discrepancy ratio trial. Augmented Dickey-Fuller and KPSS trials are besides used as unit root trials to analyze the random walk behavior of the monetary values. The consequences show that the monetary value returns of all three fiscal derivative contracts are random and their monetary value series follow a random walk. Therefore, these hereafters markets are weak form informational efficient. It is shown that the largest impact of the electronic trading system on the degree of efficiency occurred for the FTSE100 hereafters contract and the consequences suggest that this hereafters contract became the most weak signifier efficient one among the 3 hereafters contracts under probe after electronic trading system, whereas the most weak signifier efficient contract was the Long Gilt bond hereafters before the execution.