[MUSIC] Okay, so let me recall now the second notion. So we saw the first notion, which is a weak form, so now let us be a little bit more stronger. And this is called the semi strong notion of efficiency. And in top of past prices, we'd also look at what is called public information. And what attorney we'll talk a little bit now is about particular type of study, and he will mention that. >> So question of semi strong form efficiency relates to the way information, publicly available information, is reflected in to prices. When we observe new information pertaining to a particular security, we would like to know how it impacts the price. And we can think of two ways, at least, two general ways in which it can affect the price. It could impact it immediately, and be incorporated in a single jump. Or it could incorporate information gradually, and it could be reflected as a slowly moving or fast moving increase or decrease in the price. The way we address this type of issue statistically and economically is by constructing what we call event studies. So what we do is take a particular security, a particular company, look at a date where a particular event that will or that is supposed to impact it's prices. For example, the announcement of its quarterly earnings or information about a new big contract signed by the company. Which could be a good news and should be reflected in a positive way on the price. And we look at that particular date and we call that the event date. Then we take a bunch of comparable companies that are maybe active in the same industry, that have similar capitalization, similar type of activity. And we compare the return of our candidate company that just received the news to the other company. The difference in return between these two groups, so, one company that is affected by the information and the peer group. The difference between these two returns is called the abnormal return. And we look at the cumulative abnormal returns. So, if there is an excess return of our company versus the other today of 1%, this is 1% in the bank for that company. And tomorrow another 1%. Now we have 2% cumulative abnormal returns, and so forth and so on. If there are no significant differences between the two groups, we should observer something that fluctuates around zero in terms of cumulative abnormal return. Around the event date we would like to see whether there is a particular reaction of that company's price vis-a-vis the others. So let's look at how market efficiency would imply the price would react to the introduction of news by looking at the abnormal return evolution over time. So, in the slide that you're looking at, you have, in red, the behavior we expect to see in an efficient market, and in blue, in an inefficient market. Recall here that we're talking about the semi strong form of efficiency. In the efficient market, the cumulative of normal return, prior to the event date, which is the point denoted with 0 in the middle of the graph, it's fluctuating around 0. At the event date the price jumps immediately. Here we would expect that this is related to a positive piece of information that affects the company, and its return is larger on that date than the other peer companies that have similar characteristics. After the event date, nothing more. The cumulative abnormal return stays at the level it has reached on the event date, and there are no more fluctuations. In the blue graph, there we see that at the event date, there is a reaction of the security but the reaction is actually gradual. It increases from date 0 to date 50. So, this form of behavior we would expect to see in a market that is not efficient in the semi strong form. And looking at this, what we could do is build an investment strategy that would be profitable in the blue case but not in the red case. Immediately after hearing about the new piece of information related to the company, we could invest in the company in the blue market and expect a profitable abnormal return. This would not be possible in the red market because right after the revelation of the information, the price jumps. Coming back to the poll that we asked earlier. We said if markets are more efficient are they going to be more or less predictable? This is exactly what we had in mind. In the red market, the most efficient one, you can't predict anything. The news is a surprise for you. The price reacts immediately, but you cannot benefit from it. In the blue market, the market is more predictable because it is less efficient, and the information affects the price gradually. Market can become efficient simply because of competition among traders. Let's say we start in the blue case, where the market is initially inefficient. Everybody would like to benefit from the availability of such a simple strategy, and it would push prices up because everybody would purchase the security on the event date and not gradually after. This would have an impact on the price and would be reflected by an abnormal return for that security. So the blue pattern would actually transform into the red one, just because trader in the market are actively seeking profit. If they can benefit from such a simple strategy, they probably would. [MUSIC]
