Portfolio Management |
[ An Introduction To Multi-factor Models ]
1. Arbitrage Pricing Theory
- unsystematic risk can be diversified away in a portfolio
- returns are generated using a factor model
- no arbitrage opportunities exist
2. Multi-factor Model
1) Macroeconomic factor model
- Ri = E(Ri) + bi1FGDP + bi2FQS + ei
- Ri = return for Asset i
- E(Ri) = expected return for Asset i (in the absence of any surprises)
- FGDP = surprise in GDP
- FQS = surprise in Quality Spread (BB rate bond yield - treasury bond yield)
- sensitivities : regression slope estimates (historical)
2) Fundamental factor model
- Ri = ai + bi1FP/E + bi2FSIZE + ei
- ai = intercept, no economic interpretation
- factor : P/E and size (multiple regression)
- sensitivities : calculated from attribute data
3) statistical factor model
3. Active P/F Management & Performance measurement
- active return = RP - RB
- active risk = TE = s(RP - RB)
- information ratio = (RP - RB) / s(RP - RB)
4. multi-factor model application
1) return attribution = factor return + security selection return
2) risk attribution = factor risk + security selection risk
3) portfolio construction = passive, active, rules-based or algorithmic active management
4) carhart model
[ Analysis of Active Portfolio Management ]
○ Return
- excess return = Rp - Rf
- active return = Rp - Rb
- Jensen's alpha = Rp - CAPM
- alpha = Rp - βp × RB/M
○ Sharpe Ratio and Information Ratio
- SR = (RP - RF) / σP
- IR = (RP - RB) / s(RP - RB)
|
SR |
IR |
± Cash |
× |
change |
± B/M |
change |
× |
- unconstrained active portfolio, optimal active risk σA = ( IR / SRB ) × σP
- active + BM portfolio, optimal active risk
- active + BM portfolio, total risk
- highest IR = highest SR = optimal P/F for all investors regardless of risk tolerance
○ Fundamental law of active portfolio management
- information coefficient(IC) = manager's skill = 2 × (%correct) - 1
- transfer coefficient(TC) = correlation between actual active weights and optimal active weights
- breadth(BR) = independent active bets per year = N / 1 + (N-1)r,
N = number of decisions, r = correlation between the decision
-
-
○ Correlation triangle
[ Economics and Investment Markets ]
○ Required return
- R = real risk-free discount rate
- π = expected inflation
- θ = uncertainty about inflation
- γ = credit spread
- κ = additional RP relative to risky debt for an investment in equities
- λ = equity risk premium = γ +κ
- φ = risk premium for illiquidity
- Risk premium = depends on asset classes and investor's perception of risk
- nominal RP short term = R + π
- nominal RP long term = R + π + θ
- credit bond = R + π + θ + γ
- equity = R + π + θ + γ + κ
- real estate = R + π + θ + γ + κ + φ
○ Inter-temporal rate of substitution
|
|
marginal utility of consuming 1 unit in the future |
|
|
marginal utility of current consumption of 1 unit |
- investor's trade-off between real consumption now and in the future
○ Taylor rule
- r = Rn +π + 0.5(π - π*) + 0.5(y -y*)
- Rn = neural real policy interest rate
- π = current inflation rate
- y = current level of output
○ Break-even inflation (BEI)
- BEI = yield on non-inflation indexed bond - yield on inflation indexed bond
- BEI = expected inflation + RP(uncertainty about actual inflation)
○ Credit spread
- economic expansion → spread narrow → risky bonds outperform
- economic downturn → spread widen → higher rate bonds outperform
[ The Portfolio Management Process and the Investment Policy Statement ]
○ Objectives
1. return objectives
2. risk objectives
- ability to take risk (portfolio's ability) : above, average, below
- willingness to take risk (investor's behavior) : above, average, below
⇒ ability define the maximum risk tolerance
○ Constraints
- time horizon
- taxes
- liquidity : expected cash outflow
- legal & regulatory issue
- unique circumstances
'CFA > CFA lv2' 카테고리의 다른 글
Quantatitive (0) | 2016.05.27 |
---|---|
Derivatives (0) | 2016.05.27 |
Fixed Income (0) | 2016.05.23 |
Equity (0) | 2016.05.23 |
Economics (0) | 2016.05.23 |