Several papers examined the ability of price and return models along with some alternative forms to accommodate the return-earnings relationship. An important ratio for the more low-risk, defensive investor was introduced by Graham and Dodd in 1933. They introduced the Earnings/Price ratio which is simply the Price/Earnings Ratio but flipped around as a benchmark for equity valuation. After the 1929 stock market crash, they recommend the investors that rather by trying to guess what the future bring, they should concentrate on other factors such as the company’s past earnings or the value of its assets. According to Graham and Dodd, a company with strong profits and a relatively low stock price was probably undervalued. Also the fact that each share is value a number of times its current earnings became commonly satisfactory as a specific P/E level enables financial investors to make their buy/sell decision. The authors specified that P/E ratio, which is calculated by current fundamentals, never provide an exact appraisal for stocks. As a conclusion, P/E ratio was first regarded as a rough benchmark for selective stock investment and a tool for applying specific financial strategies so that in the long term, above-market returns can be generated. One of the first works showing the effect of the Price Earnings ratio was done by Nicholson (1960). The first study was based on a sample of 100 stocks which were mainly from industrial issues of trust investment quality and the stock was taken from the period 1939-1959. The stocks were arranged into groups of five according to their P/E ration in ascending order and were rebalanced every five years. The author found out that the twenty lowest multiple stocks had larger price gains as compared to the twenty highest multiple stocks. Those with the lowest P/E would deliver 14.7 times its original investment after the 20-year period, whereas the portfolio with the highest P/E stocks only earned 4.7 times its initial investment. Eight years later, Nicholson (1968) conducted another study where he looked at the earnings of 189 companies between 1937 and 1962. By dividing companies into groups of five, he found out that the average return for companies with a P/E ratio below ten was 12.7% per annum as compare to companies with a P/E ratio above twenty which had an average return of 7.97% per annum. Another studies done by Basu’s papers (1977) confirmed the results of Nicholson. The author tried to find the relationship between the investment performance of common stocks and their P/E ratio. He studied the price performance of NYSE industrial firms from 1957 to 1971. Two or more portfolios will be computed whereby risk-return relationship is weighted against each other and their performance is measured in pre-determined terms. Price to earnings ratio for every sample was calculated and they were ranked. Five portfolios were formed according to their P/E ratio. Considering the inter-quartile range, dispersion of the P/E ratio over the 14 years period can be noted where the low portfolio earned a return of 16.3% per annum compared to 9.3% for the high portfolio. Later researches (Jaffe, Keim and westerfield (1989) and Fama and French (1992)) supported the effectiveness that stocks with low P/E ratios produce higher returns. However a possible rejection of Nicholson and Basu’s studies on the Price Earnings ratio was made by Ball (1978). He conceded the apparent of such effects and argued that abnormal returns could not be produced on the basis of information available in the public area as they are of little or zero costs. Other reasons that could account for this irregularity are the systematic experimental error, transaction and processing costs and failure of Sharpe’s two parameter CAPM model. Beaver and Morse (1978) found out that when combining stocks into portfolios based on their price to earnings ratio, the differences among the portfolio continued up to the 14 years and that growth is not able to explain the existence of these little differences. I the years in which the portfolios are created, the price earnings are negatively correlated with earnings growth but positively correlated with earning growth in the next year implying that investors are considering only short-live distortions. In the study, the correlation of earnings growth in 1957 is negative 0.28 and the median correlation over the 19 years is negative 0.28. This is due to investors’ belief that earnings have been affected by temporary, random events or accounting management policies (rate of inflation, change in accounting treatment), firms which have low earning growth tend to have a high P/E ratio in the same year. As the portfolio are formed on the basis of ratio of price to realised earnings, stocks with transitory earnings will be grouped together meaning portfolio with the highest P/E ratio will be likely to include firms with negative transitory elements, that is, realized earnings are lower than the expected earnings. In the next year, while investors’ expectations are confirmed and earning growth increases, there is a positive correlation between P/E ration and earning growth. The author concluded that differences in accounting methods are the most evident explanation in differences in the P/E ratio rather than risk and growth. Studies that relate to accounting and price data normally derived the accounting measures from the COMPUSTAT database and for the quality returns they use CRSP data. However some difficulties may arise when using the COMPUSTAT database and Branz and Breen (1986) explained on the two possible problems that may crop up, that is, the ex-post-selection bias and the look-ahead bias. The ex-post-selection meant that companies which have merged, gone bankrupt or otherwise disappeared are no more included in the COMPUSTAT database and also new companies appeared with a full accounting history which does not exist before. The look-ahead bias resulted because of a dating problem where investors would not have access to portfolios that were formed at the end of the year and they had to wait several months before having access to it. Branz and Breen eliminated these factors by collecting certain COMPUSTAT items on a monthly basis that contain information on companies that was available to the investors and which also include all companies that had gone bankrupt, merged or disappeared on the COMPUSTAT. They concluded that even though the size effect was present, the Price Earning ratio was no more important as it is the data biases that had generated the evident P/E effect. Alford (1992) studied the accuracy of the valuation of the price to earning ratio when comparable firms are selected on the basis of industry, risk and earning growth. Alford (1992) used a sample of NYSE, ASE and OTC firms for the years 1978, 1982 and 1986 to analyse the accuracy of the price earning valuation. Each of the selected comparable firm’s predicted stock price is compared to its actual price and the author found that the price to earning ratio is an accurate measure of equity valuation. The findings of his research concluded that much of the diversity of P/E is attributed to the variety in the level of risk and earning growth of the individual firm. In addition, the industry factor appears to be a good proxy for risk and earning growth realed to the P/E ratios. Alford (1992) showed that the use of the industry P/E multiplied by the firm’s earnings per share (EPS) was proved to be an accurate estimator of its equity. The assessment of the accuracy of the P/E estimator was made under the efficient market hypothesis. In an efficient market, the market price changes randomly to reflect all new information. Thus, it can be used to test the accuracy of the theoretical price that Alford (1992) calculated, using P/E ratio. However this condition might not hold for different market. Value strategies have been defined by lakonishok, Schleifer and Vishny (1994) as the buying of stocks whose price are low as compared to other indicators of fundamental values such as earnings, dividends, historical prices, book asset or other measures of value in a comprehensive treatment of the issue of value strategies versus glamour stocks. They collected and studied stock prices between the periods 1963 to 1990. Firms are then classified into ‘value’ or ‘glamour’ stocks based on their past growth in sales and expected future growth as implied by the then-current P/E ratio. Differences in the expected future growth rated between the ‘value’ and ‘glamour’ stocks were found and as shown by the P/E ratio, investors were always overestimating them. For the first couple of years, the glamour stocks grew more quickly but afterwards the growth rates for the two types of stocks were almost identical. Glamour strategies were outperformed by 10-11% per year by the value strategies which used both past low growth and low multiples. Thus, glamour stock became overestimated, failed to meet investor’s expectation and were gradually abandoned. “Stocks with temporarily depressed earnings are lumped together with well-performing glamour stocks in the high expected growth/low E/P category. These stocks with depressed earnings do not experience the same degree of poor future stock performance as the glamour stocks, perhaps because they are less overpriced by the market” is the possible reason why the P/E did not produce a large effect as he other measures of fundamental value such as price-to book value or price-to-cash flow. Lakonishok, Schleifer and Vishny (1994) argued that such strategies offer higher return because they take advantage of investors’ sub-optimal behaviour. They came across little, if any, support that the value strategies were riskier and also found that the value stocks performed better than the glamour stocks. As a conclusion we can say that there has been much research that has been done on the price earnings ratio. Also many studies have been done throughout the world on different stock exchange market such as the Athens Stock Exchange (ASE). These studies concentrate on the impact of the price earning ration on the stock returns and it has been seen that price earning ratio do affect the stock returns, for example, Basu (1977) confirmed that stocks with low P/E ratio produce higher returns. However these researches had focus mainly on the empirical review rather than the theoretical review and this is the reason why we concentrate more on the empirical review.