The information coefficient is a number used to evaluate the quality of predictions about the value of stocks. It describes the difference between predicted and actual returns on stock. Companies use the information coefficient to determine the effectiveness of particular financial analysts; the higher their information coefficients, the better their predictions.
The information coefficient describes the ultimate success of an analyst's predictions, regardless of how they were obtained. The number is obtained by calculating the correlation coefficient between the predicted and actual stock prices. An information coefficient of 1 suggests that the analyst has predicted stock prices perfectly. A coefficient of 0 suggests that the analyst was no more effective than your average monkey with a typewriter. A significantly negative information coefficient suggests that the analyst's predictions tend to be the opposite of correct.
An analyst whose predictions have a high information coefficient has a very high value to her employer. If an investment bank has an analyst who can predict changes in the stock price with a high level of confidence, it can act on this information to gain a large profit. If the star analyst predicts that a company's stock will double during the next year, the company can buy a large amount of this stock and sell it when its value increases. The higher the average information coefficient of the analyst recommending the purchase, the safer a bet it probably is.
It is the job of financial analysts to get information before other people do. They study publicly available information about companies in order to form assessments about their value. The idea is that the analysts can use this public data to produce high-quality predictions that constitute better information about the companies they study. Of course, it is the value of information that also incentivizes illegal behaviors such as insider trading.
When companies seek to make money from investments, they rely on having information that puts them ahead of competitors. According to the efficient market hypothesis—which has been proved somewhat true—all publicly available information is encoded into market prices as soon as it becomes available. For example, if a company is releasing a new product, the primary effect on its stock price occurs when information about the release becomes available. Secondary changes in the price only emerge as a consequence of new information, such as consumer reaction or the appearance of a defect.