5. PORTFOLIO THEORIES Flashcards
MPT - Modern Portfolio Theory
Prior to MPT - ports constructed by looking at risks/rewards of indiv secs
Optimal portfolio = comprised of secs offering highest return and lowest risk
Markowitz - said assets can’t be looked at in isolation, investor exposed to more risks if secs are all positively correlated
- overall risk reward characteristics of portfolio are what’s important
MPT aims to prove?
More diversified portfolio = greater risk reduction
- optimum mix of assets known as efficient frontier
- main limitation is reliance on historic data
Efficient frontier
Set of investments that provide highest return for given level of risk
Rational investors = want max return for lowest risk so would want portfolio on efficient frontier
M - optimal portfolio = tanget to RFR
MVP = min variance portfolio = lowest SD
Effect of correlation on efficient frontier
Perf positive corr - return increases but risk also increases in a linear way since no diversification benefits
CAPM
Calcs min expected return investor should receive for given level of risk
Helps identify overvalued/undervalued stocks
Assumes diversification (no specific/unsystematic risk)
CAPM = rfr + B(Rm - rfr)
Rm -rfr = asset risk premium
CAPM assumptions
- investors are rational and risk averse (BF says not true)
- investors make decisions on risk and return alone (unrealistic)
- All secs are fairly valued
- All secs are transparently valued @ market prices
- all investors have same holding period (unrealistic)
- market has many buyers and sellers (sometimes can be v illiquid)
- non one individual can affect market price (unrealistic)
- no taxes, costs, restrictions on shorting (unrealistic, reg restrictions on naked shorts etc)
- info is free and simultaneously avail to all (unrealistic, research behind paywall)
- unlimited funds can be borrowed or lent by all investors at rfr (not possible)
- no dealing costs and perfect liquidity (less liquid stocks have higher sppreads in reality)
- homogeneity of investor expectations
- No limitations on short selling
- Normal distro of returns (kurtosys, skew, fat tails)
- single period model - assumes single investment period for all investors
Main limitations of CAPM
- most of the assumptions are unrealistic
- single period (1y) model)
- studies have shown that some unsystematic risk present even in diversified portfolios
- only suitable for diversified portfolios (as uses B only)
- Fama and French found CAPM was not a good predictor of return in US markets in certain types of share - so they added 2 more fundamental factors to the CAPM formula
- single factor model
- Difficult to estimate B - relies on historical data and B subject to huge variations
Two main methods of stock analysis
Technical analysis = forecasts price through study of market data, usually price and volume
- Chartism = price movements in a security are not random but can be predicted through a study of past trends and other forms of technical analysis
- predict prices from past patterns
Fundamental analysis
- try to ascertain the intrinsic values of stocks and shares
- Study anything that could impact sec value, macro, financials, management etc
- looking for over/under valued secs
Random walk theory
- Changes in asset prices are random
- so stock prices move unpredictably + past prices cant be used to predict future ones
- Implies market is efficient + reflects all available info
- challenges idea that technical analysis can be used to predict/profit from trends in stock price movements
-suggests that investment advisors add little or no value to an investor’s portfolio
Efficient Market Hypothesis (EMH)
Hypothesis that considers extent to which information is priced into a share by the market
- more efficient market = more quickly info is priced into the share and less op to find undervalued stocks
- EMH hypothesizes that stocks trade at their fair market value on exchanges
Believers = benefit from investing in low cost passive portfolio as stocks trade @ fair val
Opponents = possible to beat market (consistently) and that stocks deviate from fair val
Weak form efficiency
All historical info related to price movements and patterns fully reflected in share price
Nothing in previous share price movements that can be used to predict future price
Technical analysis cannot be used to generate excess returns as it uses historical info
Low valuations sec approach will not work because uses historical data to derive metric
BUT valuation metrics based on past data do have some predictive power
Small comp shares and EM - less research and investment. Less liquidity.
EMH assumptions
Large no. of profit maximising participants who analyse and value secs independently
Investors are rational and risk averse
New info regarding secs arrives randomly
New info is not predictable and random
Profit maximising investors adjust sec prices rapidly to reflect effect of new info
Dealing costs not too high
No market participants have sufficient wealth that they can dominate the market
Semi strong info
All historical info and all publicly available info is fully reflected in current price
New info priced in immediately and investors cant make excess returns from release of new info
Fundamental analysis cant be used to generate excess returns (as they use only public info)
Technical analysis also cant be used as historical info priced in
Only inside info can generate excess returns (if you were allowed to trade on it)
Copying director trades could produce alpha as mimicking behaviour of insider and inside info not priced
Following recs of top analysts may work - they construct ratings with non material private and grey info that may lead to similar conclusion as if private info were held and private info not priced
Only a few pricing anomalies for active managers to exploit - average analyst recs wouldnt work
Strong form efficiency
All info, historic, public and private, is priced into a stock
Markets move so quickly to price info in that noone can consistently produce excess returns
Investor cannot beat market except via luck - even most corrupt investor could not find info the price doesnt already reflect
Suggests inside info cannot produce excess returns which evidence suggests is untrue
Also long term director’s trades generally t profitable (not ST) - disproves
Consistent strong form efficient market = passive tracker
Could copy LT director trades or hold market
What makes market inefficient?
- asset prices dont accurately reflect true value
- lack of liquidity
-high transaction costs or delays - market psychology
- human emotion
- behavioral biases
EMH weak, semi strong, strong
What’s in price?
How to beat?
Fama 1991 test?
Semi strong vs strong = is there a big chnage in price when new info is released to market? If yes the prob semi strong and if no then semi strong
Fama and French 3 Factor model
Back tested CAPM and saw it was a good predictor except for
- small caps
- high book to price (more value)
1) Beta
2) Size = Big minus small (more big than small = ‘positive big minus small’, more small than big = ‘negative big minus small’)
3) Value = High Minus Low (Book to Value) Positive High Minus Low = , value, high book to price.
Negative high minus low = growth, low book to price)
Formulas look like CAPM + factor sensitivity x risk premium for factor
for each factor
Carhart Four Factor Model
1)Beta
2) Size - Big Minus Small
3) Value - High Minus Low (Book to Value)
4) Momentum
Refinement of Fama and French 3 factor model to include momentum
Formulas look like CAPM + factor sensitivity x risk premium for factor
APT
Uses unspecified number of factors - mainly macroeconomic factors (systematic risk only)
S = sensitivity to factor (+ or -)
Rpf = XS return over rfr (risk premium for factor)
APT assumes markets sometimes misprice secs unlike CAPM which assumes secs always correctly priced
CAPM and SML
Overpriced? Underpriced?
Overpriced security = has expected return below that predicted by CAPM
Underprices sec = expected return is higher than that predicted by CAPM
CAPM = sec mispriced if not on the SML
