Emprical tests/Findings of CAPM, APT and Diversification Flashcards
(17 cards)
Describe Evidence of the CAPM - Rolls Critique of CAPM tests
-Roll argued all empirical tests of CAPM are flawed because systematic risk is measured against a proxy, not the true market port.
-Common proxies used typically levae out: some tradeable shares, untradable assets and other asset classes (bonds, property).
-this problem has got worse over time due to increase in international diversification.
-Therefore Since the true market portfolio is unobservable, validity of CAPM cannot be tested.
Empirical evidence of the CAPM - Fama and French
-FAMA and French studied 2000+ stocks (1941-1990) and conclide beta/systematic risk alone cannot explain a stocks performance/returns over time.
-Find small cap stocks earn higher avg returns and stocks with high book-market rartios (value stocks) earn higher returns.
-Introducing additional factors, specifcally size and value improves relationship between returns and risk.
Strengths and limitations of the CAPM
Stengths:
-Highlights impact of un diversifiable (systematic) risk on expected returns.
-Model is simple and intuitive because its based on the idea only systematic risk matters.
Limitations:
-Potentially to simplistic to explain real world returns.
-Other factors, such as company size and industry may influence returns.
Describe the process of estimating the APT + results
Process:
-Identify macro risk facrots and using historical data estimtate the Lamda’s for each factor (market price of risk for each factor.)
-Estimate factor sensitivities (β) for each stock using historical data. Determine how sensitive stock returns are to each factor.
-Calculate expected return into the multi factor model.
CAPM and APT give similar results in some industries (e.g., banking, oil & gas), but differ in others (e.g., machinery).
Both models emphasize the importance of diversification to avoid uncompensated unsystematic risk.
Higher risk requires higher expected return.
Define the APT
-Arbitrage pricing theory model is an alternative approach to risk and return, thus the capm, debeloped by Ross (1976).
-More generalized and more complex version of the CAPM.
-assumes returns depend on several macroeconomic factors, known as “factors,” rather than a single market portfolio.
Main difference between CAPM and APT
APT does not assume investors construct efficient portfolios and does not require strong utility assumptions.
-CAPM uses the market portfolio return as the single systematic risk factor, APT uses multiple systematic risk factors.
Explain the key emprical findinfs from early CAPM studies and name 3 studies
-Jacob
(1971), Black et al. (1972) and Fama and Macbeth (1973).
Returns are linearly and positively related to beta (↑β ⇒ ↑𝑅̄𝑖).
Adding a squared beta term offers no extra explanatory power.
Including unsystematic risk controls does not improve model fit.
These findings support CAPM: beta is a valid measure of systematic risk.
Explain the criticisms of empirical CAPM tests?
Most tests are based on portfolios, not individual stocks.
Findings are sensitive to sample periods and data frequency.
Results depend heavily on the market portfolio proxy used.
These issues raise questions about the robustness and generalizability of CAPM results.
How do real world results differ from capm predictions
(SML) is flatter than theory predicts.
Low-beta stocks earn higher returns than predicted. and High-beta stocks earn lower returns than predicted.
The intercept of the regression is greater than the risk-free rate (𝑅𝑓), meaning γ₀ ≠ 0.
These differences may suggest measurement issues or limitations .
General equation for multi factor APT model:
E(R) = Rf + Bi1λ1 + Bi2λ2 + … + BinλN + ei
where: 𝛽𝑖 is the systematic risk of the ith risk factor
λ the market price of risk for the factor; and 𝜀𝑖 is error term .
Describe the factors included in the APT (MACRO ECONMOIC BASED MODEL ) - Chen, Roll, Ross.
-Many different multi factor models have been used in practice.
-This multi factor APT model takes into account Return on NYSE index, industrial production growth, expected and unexpected inflation, credit spread changes and shifts in the term structure of interest rates to explain exp returns.
Describe the factors included in the APT (Micro ECONMOIC BASED MODEL ) - Fama and french
Developed by Fama and Frejch The 3-factor model explains returns using:
Market risk
Size effect (SMB: small minus big stocks),
Value effect (HML: high minus low book-to-market stocks).
fama and french developed model to The 5-factor model adds profiliatbilty and investment. Aiming to explain returns better.
Investors expect higher returns for exposure to these factors because such stocks have more volatile market values.
Describe some emprical findings on diversifitcation + 2 references
-Stock holding puzzle, few households hold stocks despite high exp returns.
-Impact of under-divsersification, limits investment and consumption, can slow econ growth and reduce wellbeing.
-Florentsen et al 2019 found in denmark only 2% of investors hold more than 20 stocks, and estimated a annual 400 mil usd loss due to under diversifcation. 60% of port risk can be reduced by diversifying further.
Calvet et al. (2007) found many sweedish households are well diversified, with many investors outperforming the odmestic stock index’s sharpe ratio. Success may be due to international diversidication in mutual funds.
Benefits to international diversdification
-Reduces portfolio risk, specifically domestic shocks/systemtic risk and improves returns.
-Minimieses/clears country specific risk
-Gain access to more industries globally, more potenital for returns in markets not accessible domesitcally.
-Offers currency diversification
Implications of International Diversification
-Higher returns, lower risk (unsystematic)
-Shifts in the efficinet frontier, higher frontier compared to domestic portfolios only. Higher return for same or lower risk.
-Tends to be bias, behavioural bias, for domesitc stocls. Familiarity heuristic why? Foregin exchange risk, unfamiliar, bias.
how and which graphs are used to test the capm validity to predict expected returns.
SML graps
-Assets on SML = correctly priced,
-Above SML = underpriced, below SML = overpriced.
-Tests if returns align with CAPM predictions.
Regression graph:
Y-intercept = alpha (should be zero if CAPM holds).
-Significant alpha means CAPM doesnt fully explain returns.
-R² shows how well market returns explain asset returns.
Steps to emprically test the capm + validity criteia
-Run regression: Ri - Rf = a + b(Rm - Rf) + e (same as jensens alpha). Alpha is the intercept measureing return unemplained by beta,
-l Use cross-sectional regressions to check if beta alone explains returns.Assess R² to measure explanatory power of beta
Criteia for validity of capm:
-Alpha should be zero a = 0,
-Beta must be only statitically significant factor explaining returns
-Relationship between return and beta should be linearr.
-Slope should = market risk premium
