Week 3 - Economic and financial theories Flashcards
(39 cards)
What is the efficient market hypothesis?
The EMH suggests that security prices reflect information (that information is absorbed into prices) - this is the concept of market efficiency.
What is the difference between market efficiency and portfolio efficiency?
Market efficiency - information is reflected in the prices of stocks
Portfolio efficiency - minimum volatility for a fixed rate of return
Why are price changes random?
Price changes are random because prices react to information and the flow of information is random
What is the concept of ‘random walk’?
Random walk is the concept that stock prices are random, and that they are random about an expected positive trend. It is a statistically naive hypothesis
Explain how to test random walk and the results of analysis performed on the Australian market.
Volatility of log returns can be calculated at different data frequencies (e.g. annual, quarterly, monthly), and if independent the ratio between them should theoretically be the same (i.e. volatility ratio of 1).
When this test is applied to the Australian market, split across the total market vs industrials vs resources. Across the whole market there is mild statistical contradiction of random walk, particularly in the industrials sector, which has been clouded by the resources sector (an ever-smaller proportion of the market).
Define the three levels of the efficient market hypothesis.
Weak form - Information derived from market trading data includes past prices, volumes and interest
Semi-strong form - Publicly available information, e.g. financial reports, earnings forecasts, included in pricing
Strong form - all information, including insider information, is available
What are the three basic patterns recognised in research into the momentum of stock prices?
1) Longer term reversal over multiple years (winners become losers)
2) Intermediate momentum where trailing 3-12 mth performance continues (winners continue to win)
3) Short term reversal effect where trailing returns quickly reverse because of liquidity effects
How are price momentum strategies able to outperform the market?
Changes occur in a company, which are not picked up by the market. Then, the market slowly realises these changes and begins to adjust price according to the impact that these changes have, until it outperforms expectations and what really should happen. Then it reverses back to a fair estimate.
What are some of the global conclusions about how to exploit price momentum?
1) Large stock momentum occurs at an industry level
2) If sector allocation is controlled, a portfolio will not be exposed to momentum effect (otherwise sectoral risk is present)
3) Small stock momentum is stock specific
4) Large cap momentum is best exploited by overweighting winners, while small stock momentum is best exploited by underweighting losers (no help in a long-only market)
What is the relationship between high risk and low risk stocks across different market conditions?
1) High risk stocks outperform low risk stocks in times of high growth, e.g. after recessions
2) Low risk stocks outperform high risk stocks when prices of low risk have been discounted relative to high risk stocks (e.g. at the end of the 2000’s tech bubble). Historically this has occurred because volatile stocks underperformed, not because stable stocks outperformed (esp. prior to 1990)
Explain the value anomaly.
The value anomaly occurs because investors wrongly assume past growth patterns will continue into the future. Thus stocks which are momentarily experiencing a downturn and are for the moment underpriced, will eventually revert and grow at a faster rate than the market, outperforming glamour stocks, particularly in times of distress. It continues to exist because it is difficult to arbitrage away.
How is it possible to exploit the value anomaly?
It is best to buy undervalued stocks rather than selling overvalued stocks. Institutional investors are unable to take advantage of this, as it primarily occurs in sectors where it cannot be utilised and is almost always restricted to small stocks. Further, there is also a tendency of companies which have ‘boosted’ earnings to experience profit growth reversal (investors are surprised by this reversal so high accrual stocks underperform).
Does strong form market efficiency exist? Reference three event studies.
No. Earnings surprises have very little leakage in the build up to the event, while dividend surprises have a substantial amount of leakage up to the day of the event (and the pattern continues afterwards also). Further, there is some anticipation of profit warnings, particularly as the downward pattern occurs before and after the drop at t=0.
Leakage occurs when the pattern is noticeable before the drop. If the pattern is not noticeable before the drop then it is likely that information is not publicly available,
What questions should be asked in anomaly investigations?
1) Is the sample representative of the investment universe? Must treat unweighted analysis with caution
2) Is the anomaly asymmetric?
3) Have confounding effects been controlled, e.g. size, sector?
4) Have transaction costs been allowed for?
As a final note, we can also ask the question - “Is it relevant?” - as the majority of the Australian market index is accounted for by a select few stocks
Define the following terms:
1) Prospect theory
2) Overconfidence
3) Attribution bias
4) Hindsight bias
1) Prospect theory assumes investors are loss averse, and that they are willing to sell their winners early but hold their losers to postpone the regret of incurring a loss
2) Overconfidence is when investors over-estimate the accuracy of their knowledge and overrate their own abilities
3) Attribution bias is the theory that investors attribute cause to the wrong event (e.g. good results are because of skill, bad ones because of uncontrollable factors)
4) Hindsight bias is the theory that investors are likely to overestimate the probability they assigned to an event that has already occurred
Define the following terms:
1) Representative heuristic
2) Underreaction/confirmation bias
3) Overreaction
1) Representative heuristic occurs when investors wrongly assume small samples are highly representative of the parent population from which they are drawn
2) Confirmation bias assigns more importance to information that confirms prior views (e.g. momentum)
3) Over-reaction is when investors exaggerate optimism and pessimism over the medium term (e.g. value anomaly, where stocks that underperform eventually reverse and appreciate)
How should we invest on the basis of anomaly research?
1) Avoid small stocks with negative momentum
2) Overweight sectors with positive momentum (large cap) but beware of associated risks (sectors)
3) If you expect a market crash avoid volatile stocks, if you expect a market run, favour volatile stocks
4) Buy cheap small stocks, not with negative momentum, but with strong cash backing of earnings
5) There are very few anomalies for large stocks - no real opportunities for a typical manager
What is the major issue with superannuation funds and equity managers?
Super funds typically imagine that managers’ past performance is an indicator of future performance. This is not true, particularly when performance is assessed pre-tax/transaction costs. As such, the selection decision of managers has a large impact. Managers are typically fired after a bad performance but hired after a good performance, but managers perform in a cycle which means that they are often hired when their performance is about to worsen and fired when performance is likely to improve.
What are some of the benefits of passive management?
Passive management involves simply matching the index, and is not just about whether some managers can reliably outperform the index portfolio but whether it is consistent. Typical survey results indicate that super funds would be better off if they avoided active management (due to selection issue).
What are the 3 hurdles for active management?
1) Are there anomalies to efficient markets
2) If so, can they be effectively exploited
3) Can funds identify and employ effective active managers without spoiling by ill-timed hiring and firing
Even if 1) and 2) are answered with yes, it is unlikely that super funds are able to do 3) effectively, despite managers often outperforming the market (esp. in Australia, not so much US)
What is the Black-Litterman model used for?
The Black-Litterman model is one which uses risk estimation as a basis for estimating returns. It is particularly useful as it provides a framework to blend investor views with implied market consensus.
What are the BL model formulas for:
1) Covariance matrix
2) Unscaled risk premia
3) Scaled risk premia/implied returns
1) Covariance matrix is defined as the correlation matrix pre and post multiplied by a diagonal matrix of volatilities (all size n x n)
2) Unscaled risk premia is given by covariance matrix * weight vectors (n x 1 outcome)
3) Scaled risk premia/Implied returns is the unscaled risk premia * scaling factor (n x 1 outcome). The formula in Greek letters is delta (risk aversion coefficient) * sigma (covariance matrix) * w (weights of stocks). Delta can be found by making an assumption on one return and scaling everything to be in line with that
What is the information ratio? What components does it have?
InfRatio = alpha/omega, where alpha is the excess return over benchmark and omega is the standard deviation of active return.
Alternatively it can be defined as root(M)*c where M is the breadth of the strategy and c is the skill involved.
Which level of information ratio is preferred - higher or lower? Why?
Higher - indicates more return and/or less volatility. This has implications for concentrated portfolios, thematic investment (less breadth) and unskilled managers (less skill)
