The primary assumptions of the efficient market hypothesis (EMH) are that information is universally shared and that stock prices follow a random walk, meaning that they're determined by today's news rather than yesterday's trends. The strength of these assumptions, however, depends on the form of EMH under consideration.
The weak form of the theory states that public market information is fully reflected in prices and that past performance has no relationship to future returns—in other words, trends don't matter. The semi-strong form says that stock prices are updated to reflect both market and non-market public information. The strong form states that all public and private information is fully and immediately factored into prices.
The assumptions about information underlying EMH vary depending on the form, with the weak form of the hypothesis assuming that only public market information is known to all market participants and the strong form assuming perfect information transparency. In all forms, future stock price movements are assumed to be independent of past stock price movements—the random walk.
The Implication of the Efficient Market Hypothesis
The implication of EMH is that the market can't be beaten because all information that could predict performance is already built into the stock price. The concept has fallen out of favor in the last couple of decades with research advances in behavioral finance and, to a lesser extent, with the success of quantitative trading algorithms. High-frequency trading is one example. Over time, it's been shown to contribute to market efficiency, implying that markets weren't efficient before.