Market efficiency Flashcards
How will we define the efficient market hypothesis?
1) Prices reflect all available information
2) The market responds only to new information
3) It is not possible to forecast price changes
Why should markets be efficient?
Sophisticated arbitrageurs with unlimited capital and risk capacity trade on information to eliminate inefficiencies, competing intensely for profitable opportunities.
What is weak form, semi-strong and strong form efficiency?
Also if the market is strong form efficient what does it mean?
1) Weak form -where current asset prices fully reflect all available historical trading information, including past price patterns and trading volumes.
2) Semi-strong where current asset prices incorporate not only historical trading information but also all publicly available information. This includes financial statements, earnings reports, economic indicators, and other public news.
3) Strong form - Strong form efficiency is the highest level of market efficiency, where current asset prices fully reflect all available information, both public and private (inside information). . Note that insider trading is illegal.
Strong form => Semi-srong form => Weak form.
What are 3 implications of weak form efficiency and what are 2 tests?
3) Technical analysis is useless.
What is technical analysis ( test for weak form efficiency - current prices reflect all information constrained in past prices and price movements)
Technical analysis uses historical price patterns and trading volumes to forecast future prices and pinpoint profitable trades, testing weak form efficiency by implying that past trading data isn’t fully incorporated in current asset prices.
Give example of a trading rule by Lakonishok and LeBaron (1992) in regards to technical analysis?
moving average ( average of returns over a certain period of time) , (buy when the short-period MA rises above the long period MA and vice versa);
What is another type of trading strategy used in technical analysis?
Filter rules are simple strategies that tell you when to buy or sell based on a set percentage change in price, helping you take advantage of price trends. For instance, a 5% rule means you buy when the price goes up 5% from a low point and sell when it drops 5% from a high point
Now moving on to Semi-strong efficiency, what are 2 implications of them? and what are 2 tests of this?
1) Prices react immediately to economic events
2) No systematic under/overshooting after announcements of macroeconomic news.
The test is fundamental analysis
event studies of reaction to news
What is fundamental analysis?
Is the practice of using firm earnings, dividends, and other sources of accounting and financial data to forecast future returns of the stock.
The success of fundemental analysis would be a violation of semi-strong efficiency.
What is an event study?
to assess the impact of a specific event on a company’s stock price or on the overall stock market.
Suppose that we want to study the effect of analyst downgrades on stock returns of Cisco. Is this an evidence that the downgrade affected returns? ( BTW it happened on september 28, 2000.
It looks like it but remember correlation doesn’t imply causation, it might be that on that day the whole stock market went down, so it might not be linked to the announcement, just by looking at that photo.
So essentially with event studies, we want to see if there are any abnormal returns following an announcement of news. How do we show this formally?
To calculate expected return we are going to rely on an Asset pricing model which here is CAPM, we could use fama and french 3 factor model.
So its realised return - expected return.
( for example bit) We run a regression on the left hand side ( realised excess return) on a constant and market portfolio, then estimate coefficients of alpha and beta. We do this before announcement date to not capture the effect of the announcement.
So what is the event study actual methodology ? Also draw a timeline of event studients methodology?
1) estimating alpha and betas through regression in the estimation window which excludes sample relating to the event.
2) we will then calculate abnormal returns in the event window ( the reason we observe abnormal returns a period before the event day is to see whether there was some prior insider information before the event causing abnormal return)
3) we will construct the following variable CAR ( the sum of all returns over all datas in our event window
What is the abnormal return at time t? can we say the downgrade conclusively caused the negative abnormal returns?
The SD is within 1 SD of the return, suggesting that it this is quite frequent
So as we have seen in the previous example our test has lower power because the return includes other idiosracritic shocks, so how would we diversify away those idioscraytic shocks which are not related to the news we are interested in ?
We construct a porttolfio of firms that have downgrades, then the idioscraytic risk related to noise will be diversified away. So we can construct CAR for each company in portfolio and take average, eliminating this risk.
This helps us test if the null hypothesis H0 : CAR = 0.
So how does this show a jump?
At the event date, there is a significant downgrade impact on stock prices
How does this show underreaction?
Firstly, prices jump on event date consistent with semi-strong efficiency, but continue to trend down 50 subsequent days, not consistent with EMH. This means that the price of the stock may not fully reflect the true impact of the event, and may continue to adjust in the days or weeks following the event as investors gradually incorporate the new information, HENCE UNDERREACTION - it should have fallen more.
How does this show overreaction?
Firstly, prices jump on event date consistent with semi-strong efficiency, but continue to trend upwards 50 subsequent days, not consistent with EMH. the stock market pushed prices down far too much, then price gradually goes back to some equilibrium level.
How does this show information leakage?
We see here that CAR start trending down before the announcement date . Some investors realise their will be a downgrade before date 0 and start trading on this before announcement date
