Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Eco525: Financial Economics I Lecture 06: Factor Pricing Prof. Markus K. Brunnermeier 09:55 Lecture 06 Factor Pricing Slide 06-1 Eco525: Financial Economics I Overview • Theory of Factor Pricing (APT) ¾ Merits of Factor Pricing ¾ Exact Factor Pricing and Factor Pricing Errors ¾ Factor Structure and Pricing Error Bounds ¾ Single Factor and Beta Pricing (and CAPM) ¾ (Factor) Mimicking Portfolios ¾ Unobserved Factor Models ¾ Multi-period outlook • Empirical Factor Pricing Models ¾ Arbitrage Pricing Theory (APT) Factors ¾ The Fama-French Factor Model + Momentum ¾ Factor Models from the Street • Salomon Smith Barney’s and Morgan Stanley’s Model 09:55 Lecture 06 Factor Pricing Slide 06-2 Eco525: Financial Economics I The Merits of Factor Models • Without any structure one has to estimate ¾ J expected returns E[Rj] (for each asset j) ¾ J standard deviations ¾ J(J-1)/2 co-variances • Assume that the correlation between any two assets is explained by systematic components/factors, one can restrict attention to only K (non-diversifiable) factors ¾ Advantages: d Drastically reduces number of input variables d Models expected returns (priced risk) d Allows to estimate systematic risk (even if it is not priced, i.e. uncorrelated with SDF) d Analysts can specialize along factors ¾ Drawbacks: d Purely statistical model (no theory) (does not explain why factor deserves compensation: risk vs mispricing) d relies on past data and assumes stationarity 09:55 Lecture 06 Factor Pricing Slide 06-3 Eco525: Financial Economics I Factor Pricing Setup … • K factors f1, f2, …, fK ¾ E[fk]=0 ¾ K is small relative to dimension of M ¾ fk are not necessarily in M • F space spanned by f1,…,fK,e • in payoffs ¾ bj,k factor loading of payoff xj 09:55 Lecture 06 Factor Pricing Slide 06-4 Eco525: Financial Economics I …Factor Pricing Setup • in returns • Remarks: ¾One can always choose orthogonal factors Cov[fk, fk’]=0 ¾Factors can be observable or unobservable 09:55 Lecture 06 Factor Pricing Slide 06-5 Eco525: Financial Economics I Factor Structure • Definition of “factor structure:” • ⇒ risk can be split in systematic risk and idiosyncratic (diversifiable) risk 09:55 Lecture 06 Factor Pricing Slide 06-6 Eco525: Financial Economics I Exact vs. Approximate Factor Pricing • Multiplying (1) by kq and taking expectations • Rearranging • Exact factor pricing: ¾ error: ψj = 0 (i.e. εj s orthogonal to kq ) ¾ e.g. if kq ∈ F 09:55 Lecture 06 Factor Pricing Slide 06-7 Eco525: Financial Economics I Bound on Factor Pricing Error… • Recall error ¾ Note, if ∃ risk-free asset and all fk ∈ M, then … • If kq ∈ F, then factor pricing is exact • If kq ∉ F, then ¾ Let’s make use of the Cauchy-Schwarz inequality (which holds for any two random variables z1 and z2) ¾ Error-bound 09:55 Lecture 06 Factor Pricing Slide 06-8 Eco525: Financial Economics I Error-Bound if Factor Structure Holds • Factor structure ⇒ split idiosyncratic from systematic risk • ⇒ all idiosyncratic risk εj are linearly independent and span space orthogonal to F. Hence, • Note • Error • Pythagorean Thm: If {z1, …, zn} is orthogonal system in Hilbert space, then ¾ Follows from def. of inner product and orthogonality 09:55 Lecture 06 Factor Pricing Slide 06-9 Eco525: Financial Economics I Error-Bound if Factor Structure Holds Applying Pythagorean Thm to implies Multiply by use of …… and making RHS is constant for constant max[σ2(εj)]. ⇒ For large J, most securities must have small pricing error • Intuition for Approximate Factor Pricing: Idiosyncratic risk can be diversified away 09:55 Lecture 06 Factor Pricing Slide 06-10 Eco525: Financial Economics I One Factor Beta Model… • Let r be a risky frontier return and set f = r – E[r] (i.e. f has zero mean) ¾ q(f) = q(r) – q(E[r]) _ • Risk free asset exists with gross return of r _ ¾ q(f)_= 1 – E[r]/r • f and r span E and hence kq ∈ F ⇒ Exact Factor Pricing 09:55 Lecture 06 Factor Pricing Slide 06-11 Eco525: Financial Economics I …One Factor Beta Model • Recall _ _ ¾ E[rj] = _r - βj r q(f) _ ¾ E[rj] = r - βj {E[r] - r} • Recall ¾ βj = Cov[rj, f] / Var[f] = Cov[rj, r] / Var[r] • If rm ∈ E then CAPM 09:55 Lecture 06 Factor Pricing Slide 06-12 Eco525: Financial Economics I Mimicking Portfolios… • Regress on factor directly or on portfolio that mimics factor ¾ Theoretical justification: project factor on M ¾ Advantage: portfolios have smaller measurement error • Suppose portfolio contains shares α1, …, αJ with ∑jJ αj = 1. • Sensitivity of portfolio w.r.t. to factor fk is γk = ∑j αj βjk • Idiosyncratic risk of portfolio is ν = ∑j α εj ¾ σ2 (ν) = ∑j α2 σ(εj) ¾ diversification 09:55 Lecture 06 Factor Pricing Slide 06-13 Eco525: Financial Economics I …Mimicking Portfolios • Portfolio is only sensitive to factor k0 (and idiosyncratic risks) if for each k ≠ k0 γk=∑ αj βjk=0, and γk0=∑ αj βjk0≠ 0. • The dimension of the space of portfolios sensitive to a particular factor is J-(K-1). • A portfolio mimics factor k0 if it is the portfolio with smallest idiosyncratic risk among portfolios that are sensitive only to k0. 09:55 Lecture 06 Factor Pricing Slide 06-14 Eco525: Financial Economics I Observable vs. Unobservable Factors… • Observable factors: GDP, inflation etc. • Unobservable factors: ¾ Let data determine “abstract” factors ¾ Mimic these factors with “mimicking portfolios” ¾ Can always choose factors such that • factors are orthogonal, Cov[fk, fk’]=0 for all k ≠ k’ • Factors satisfy “factor structure” (systemic & idiosyncratic risk) • Normalize variance of each factor to ONE ⇒ pins down factor sensitivity (but not sign, - one can always change sign of factor) 09:55 Lecture 06 Factor Pricing Slide 06-15 Eco525: Financial Economics I …Unobservable Factors… • Empirical content of factors ¾Cov[ri,rj] = ∑k βikβjkσ2(fk) ¾ σ2(rj) =∑k βjkβjkσ2(fk)+σ2(εj) ¾ σ(fk)=1 for k=1,L,K. (normalization) ¾In matrix notation • Cov[r,r‘] = ∑kβk’βkσ2(fk) + D, – where βk = (β1k,…,βJk). • Ω= B B’ + D, – where Bjk=βjk, and D diagonal. – For PRINCIPAL COMPONENT ANALYSIS assume D=0 (if D contains the same value along the diagonal it does affect eigenvalues but not eigenvectors – which we are after) 09:55 Lecture 06 Factor Pricing Slide 06-16 Eco525: Financial Economics I …Unobservable Factors… • For any symmetric JxJ matrix A (like BB’), which is semipositive definite, i.e. y’Ay ≥ 0, there exist numbers λ1 ≥ λ2 ≥…≥ lambdaJ ≥ 0 and non-zero vectors y1, …, yJ such that ¾ yj is an eigenvector of A assoc. w/ eigenvalue λj, that is A yj = λj yj ¾ ∑jJ yij yij’ = 0 for j ≠ j’ ¾ ∑jJ yij yij = 1 ¾ rank (A) = number of non-zero λ ‘s ¾ The yj ‘s are unique (except for sign) if the λi ‘s are distinct • Let Y be the matrix with columns (y1,…,yJ), and let Λ the diagonal matrix with entries λi then 09:55 Lecture 06 Factor Pricing Slide 06-17 Eco525: Financial Economics I …Unobservable Factors • If K-factor model is true, BB' is a symmetric positive semi-definite matrix of rank $K.$ ¾ Exactly K non-zero eigenvalues λ1,…,λk and associated eigenvectors y1,…,yK ¾ YK the matrix with columns given by y1,…,yK ΛK the diagonal matrix with entries λj, j=1,…, K. ¾ BB'= K Hence, • Factors are not identified but sensitivities are (except for sign.) • In practice choose K so that λk is small for k>K. 09:55 Lecture 06 Factor Pricing Slide 06-18 Eco525: Financial Economics I Why more than ONE mimicking portfolio? • Mimic (un)observable factors with portfolios [Projection of factor on asset span] • Isn’t a single portfolio which mimics pricing kernel sufficient ⇒ ONE factor • So why multiple factors? ¾ Not all assets are included (real estate, human capital …) ¾ Other factors capture dynamic effects [since e.g. conditional ≠ unconditional. CAPM] (more later to this topic) 09:55 Lecture 06 Factor Pricing Slide 06-19 Eco525: Financial Economics I Overview • Theory of Factor Pricing (APT) ¾ Merits of Factor Pricing ¾ Exact Factor Pricing and Factor Pricing Errors ¾ Factor Structure and Pricing Error Bounds ¾ Single Factor and Beta Pricing (and CAPM) ¾ (Factor) Mimicking Portfolios ¾ Unobserved Factor Models ¾ Multi-period outlook • Empirical Factor Pricing Models ¾ Arbitrage Pricing Theory (APT) Factors ¾ The Fama-French Factor Model + Momentum ¾ Factor Models from the Street • Salomon Smith Barney’s and Morgan Stanley’s Model 09:55 Lecture 06 Factor Pricing Slide 06-20 Eco525: Financial Economics I APT Factors of Chen, Roll and Ross (1986) 1. Industrial production (reflects changes in cash flow expectations) 2. Yield spread btw high risk and low risk corporate bonds (reflects changes in risk preferences) 3. Difference between short- and long-term interest rate (reflects shifts in time preferences) 4. Unanticipated inflation 5. Expected inflation (less important) Note: The factors replicate market portfolio. 09:55 Lecture 06 Factor Pricing Slide 06-21 Eco525: Financial Economics I Fama-MacBeth 2 Stage Method • Stage 1: Use time series data to obtain estimates for each individual stock’s βj (e.g. use monthly data for last 5 years) Note: is just an estimate [around true βj ] • Stage 2: Use cross sectional data and estimated βjs to estimate SML b=market risk premium 09:55 Lecture 06 Factor Pricing Slide 06-22 Eco525: Financial Economics I CAPM β−Τesting Fama French (1992) • Using newer data slope of SML b is not significant (adding size and B/M) • Dealing with econometrics problem: ¾ s are only noisy estimates, hence estimate of b is biased ¾ Solution: – Group stocks in 10 x 10 groups sorted to size and estimated βj – Conduct Stage 1 of Fama-MacBeth for portfolios – Assign all stocks in same portfolio same β – Problem: Does not resolve insignificance size • Standard Answer: Find instrumental variable • Answer in Finance: Derive estimates for portfolios Portfolio • CAPM predictions: b is significant, all other variables insignificant • Regressions: size and B/M are significant, b becomes insignificant ¾ Rejects CAPM 09:55 Lecture 06 Factor Pricing Slide 06-23 Eco525: Financial Economics I Book to Market and Size 09:55 Lecture 06 Factor Pricing Slide 06-24 Eco525: Financial Economics I Fama French Three Factor Model • Form 2x3 portfolios book/market ¾Size factor (SMB) ¾Book/Market factor (HML) size • Return of small minus big • Return of high minus low • For … αs are big and βs do not vary much • For … (for each portfolio p using time series data) αs are zero, coefficients significant, high R2. 09:55 Lecture 06 Factor Pricing Slide 06-25 Eco525: Financial Economics I Fama French Three Factor Model • Form 2x3 portfolios book/market ¾Size factor (SMB) ¾Book/Market factor (HML) size • Return of small minus big • Return of high minus low • For … αs are big and βs do not vary much • For … (for each portfolio p using time series data) αps are zero, coefficients significant, high R2. 09:55 Lecture 06 Factor Pricing Slide 06-26 Book to Market as a Predictor of Return Annualized Rate of Return 25% 20% 15% 10% Value 5% Growth 0% 1 2 High Book/Market 3 4 5 6 7 8 9 10 Low Book/Market Book to Market Equity of Portfolios Ranked by Beta Book to Market Equity 1 0.9 0.8 0.7 0.6 0.5 0.6 0.8 1 1.2 Beta 1.4 1.6 1.8 Eco525: Financial Economics I Adding Momentum Factor • 5x5x5 portfolios • Jegadeesh & Titman 1993 JF rank stocks according to performance to past 6 months ¾Momentum Factor Top Winner minus Bottom Losers Portfolios 09:55 Lecture 06 Factor Pricing Slide 06-29 Monthly Difference Between Winner and Loser Portfolios Monthly Difference Between Winner and Loser Portfolios at Announcement Dates 1.0% 0.5% 0.0% 1 3 5 7 9 11 13 15 17 19 21 23 25 27 -0.5% -1.0% -1.5% Months Following 6 Month Performance Period 29 31 33 35 Cumulative Difference Between Winner and Loser Portfolios Cumulative Difference Between Winner and Loser Portfolios at Announcement Dates 5% 4% 3% 2% 1% 0% -1% 1 3 5 7 9 11 13 15 17 19 21 23 25 27 -2% -3% -4% -5% Months Following 6 Month Performance Period 29 31 33 35 Eco525: Financial Economics I Morgan Stanley’s Macro Proxy Model • Factors ¾ GDP growth ¾ Long-term interest rates ¾ Foreign exchange (Yen, Euro, Pound basket) ¾ Market Factor ¾ Commodities or oil price index • Factor-mimicking portfolios (“Macro Proxy”) ¾ Stage 1: Regress individual stocks on macro factors ¾ Stage 2: Create long-short portfolios of most and least sensitive stocks [5 quintiles] • Macro Proxy return predicts macro factor 09:55 Lecture 06 Factor Pricing Slide 06-32 Eco525: Financial Economics I Salomon Smith Barney Factor Model • Factors ¾Market trend (drift) ¾Economic growth ¾Credit quality ¾Interest rates ¾Inflation shocks ¾Small cap premium 09:55 Lecture 06 Factor Pricing Slide 06-33 Eco525: Financial Economics I Haugen’s view: The Evolution of Academic Finance The Old Finance 1930’s 40’s 50’s 60’s 70’s 80’s 90’s beyond Modern Finance Modern Finance Theme: Paradigms: Valuation Based on Rational Economic Behavior Optimization Irrelevance CAPM (Markowitz) (Modigliani & Miller) EMH (Sharpe, Lintner & Mossen) (Fama) Foundation: Financial Economics 09:55 Lecture 06 Factor Pricing Slide 06-34 Eco525: Financial Economics I Haugen’s view: The Evolution of Academic Finance The Old Finance 1930’s 40’s 50’s The New Finance 60’s 70’s 80’s 90’s beyond Modern Finance The New Finance Theme: Paradigms: Inefficient Markets Inductive ad hoc Factor Models Expected Return Risk Behavioral Models Foundation: Statistics, Econometrics, and Psychology 09:55 Lecture 06 Factor Pricing Slide 06-35