▪ HPRt = (Pt – Pt-1 + Dt) / Pt-1, Note: When returns stable, AAR and GAR gives the ▪ Sharpe

arpe Ratio =
(𝑁(𝑁𝑁)−𝑁𝑁 )
, measures risk premium per unit of risk ▪ Assumptions: 1) Individual investors are price takers, 2) Investments are
𝑁𝑁
same result. Bias of AAR relative to GAR increases w return volatility. Both limited to traded financial assets, 3) No taxes and no transaction costs, 4)
AAR and GAR ignore the effects of trading on portfolio returns Diversification with Many Risky Assets People only care about mean and variances of returns, 5) People have same
Asset Classes and Financial Risk and Risk Premiums ▪ As more assets are included, the efficient frontier shifts to the NW expectations, mean and variance of returns known (homogeneous expectns)
(↑North=higher return, West=lower risk), where efficient frontier (risky ▪ Implications: Under CAPM, all investors will hold some combination of the
Instruments ▪ Ex ante (before the event): forecast scenarios based on probability
▪ E(r) = ∑p(s) x rs, σ2 = ∑p(s) x [rs – E(r)]2 assets) is the set of portfolios that have highest return for given level of risk market portfolio (well-diversified, no idiosyncratic risk) & the risk-free rate.
Bid = you SELL, Ask = you BUY, Ask-Bid = Dealer’s profit ▪ Ex post (after, use average returns to appx expected returns) Optimal Complete Portfolio (Risky Portfolio + Risk-free Asset) ▪ Market portfolio is 1) market-value weighted of all asset, 2) optimal risky
Money Market portfolio on efficient frontier of all risky asset w max Sharpe ratio
▪ ST debt instruments (bills): no default risk, zero-coupon, not callable, highly Moving along CML
Understanding Beta: Measures sensitivity of security’s return to systmtc risk
liquid, dominated by institutional investors for large denominated instru is a passive
▪ ▪ Pricing of individual securities depends on risk that individual securities
strategy since it
▪ LIBOR: Determined by group banks, an avg of i/r that banks expect to charge contribute to the market portfolio. Only systematic risk (𝛽) matter since
only involves
: Optimistic  charge lower LIBOR, Pessimistic  charge higher LIBOR changing weights, idiosyncratic risk is diversified away, and
Bond Market: Treasury Inflation Protected Securities (TIPS): have principal Normal Distribution: risk: possibility of realized returns differing from and not combi of 𝑁𝑁𝑁(𝑁𝑁 , 𝑁𝑁 )
𝑁

adjusted by consumer price index. expected returns, σ measures deviation above or below mean (Appx 95% occur stocks 𝑁 = 2 , 𝑁 = ∑ 𝑁𝑁 𝑁𝑁
▪ Nominal YTM = Real YTM + Expected Inflation within E(r) ± 2 σ, 99% within E(R) ± 3 σ) 𝑁 𝑁
Everyone, 𝑁=1
Stock and Bond Indexes Ex-ante risk premium and risk aversion: Risk Premium = E(r) - rf regardless of risk ▪ β > 1: greater sensitivity to economy than avg stock, β < 1: below-average
▪ Track average returns, compare performance of money managers, and base ▪ As investors are risk averse, they require compensation for bearing risk aversion, holds a sensitivity to economy, β = 1: market portfolio has a beta of 1.
of derivatives ▪ rreal = rnom x (1-tax rate) – inflation rate portfolio along CAPM: Securities Market Line
▪ Mutual fund: Investment fund that pools funds from investors to buy Asset allocation across risky and risk-free portfolios CML
securities, Index fund: A mutual fund which buys securities w the goal of ▪ Amount invested depends on risk, risk premium and risk aversion
tracking an index, Exchange-traded funds: an index fund which trades on ex ▪ Complete portfolio (risky portfolio + rf asset): E(rc) = yE(rp) + (1-y)rf, σc = y
Construction of Indexes σrp + (1-y) σrf = yσrp
𝑁𝑁𝑁 (𝑁𝑁𝑁𝑁𝑁1 × 𝑁𝑁𝑁𝑁𝑁𝑁1 ) + (𝑁𝑁𝑁𝑁𝑁2 × 𝑁𝑁𝑁𝑁𝑁𝑁2 ) + (𝑁𝑁𝑁𝑁𝑁3 × 𝑁𝑁𝑁𝑁𝑁𝑁3▪) Capital Allocation Line: y-intercept: rf, gradient = risk premium per unit of
𝑁𝑁𝑁 (𝑁𝑁𝑁𝑁𝑁1 × 𝑁𝑁𝑁𝑁𝑁𝑁1 ) + (𝑁𝑁𝑁𝑁𝑁2 × 𝑁𝑁𝑁𝑁𝑁𝑁2 ) + (𝑁𝑁𝑁𝑁𝑁3 × 𝑁𝑁𝑁𝑁𝑁𝑁3 ) risk (E(rp)-rf)/σp | Investors can borrow at rf rate  y>1 | Ignore y<0 (i.e.
× 100 short risky portfolio) cos y<0 imply complete portfolio has -ve risk premium
▪ Index is driven more by changes in large-cap than small-cap stocks Risk aversion and allocation
▪ Investors w high level of risk aversion will choose lower % of risky portfolio
▪ Risk aversion ↑ as age ↑ and wealth  (investor cares about the complete
Securities Markets portfolio risk premium vs complete portfolio risk)
Types of Orders 𝑁(𝑁𝑁 ) − 𝑁𝑁
𝑁= > 0 𝑁𝑁 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁𝑁, 𝑁%𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁
▪ Market order: execute immediately at the best price (buy/sell now) 𝑁𝑁 2
▪ Limit order: buy/sell at a specified price or better (If buy limit at $15, only (𝑁(𝑁𝑁 ) − 𝑁𝑁)/𝑁𝑁 2
buy if <$15), (if sell limit at $16, only sell if >$16) = ▪ FALSE: CAPM implies that investors require a higher return to hold highly
𝑁 volatile securities (because investors require higher return for bearing
▪ Stop loss (stop-sell) order: If own stock trading at $40, and place stop loss Implications: A = 0  risk neutral (if E(r) exceeds rf rate, since a risk neutral
at $38, becomes mkt order to sell if stock hits <$38. ▪ Because combinations of P & rf have an equal or higher Sharpe ratio than systematic risk, not volatility σ)
investor only focuses on returns & is indifferent to risk, 100% in mkt portfolio other portfolios on the efficient frontier. Planner varies the asset allocation
▪ Stop buy order: If short stocks at $40 and place stop buy at $42, becomes mkt ↑variance, less likely to invest in risky portfolio ▪ Consider Sharpe ratio to determine if CAPM holds, i.e. no portfolio should
order to buy if stock hits >$42. between risky and rf according to risk aversion to create complete portfolio have higher Sharpe ratio than the mkt portfolio
▪ Discretionary order: Gives the broker the power to buy/sell at discretion Understanding Alpha (α): measures how much expected returns differ from
▪ Spread: cost of trading with dealer, i.e. ask-bid Efficient Diversification Index Model & Capital Asset Pricing CAPM-implied expected returns
Alpha: = 𝑁(𝑁𝑁 ) − 𝑁𝑁 − 𝑁𝑁 [𝑁(𝑁𝑁 ) − 𝑁𝑁 ] → Excess returns adjusted for
Types of Risks Index Models (Y = α + βx + ϵ)
Margin Trading compensation for systematic risk
▪ Systematic risk: Macro, “mkt risk”, cannot be diversified away ▪ Estimate the two components of risk by using historical HPR to separate Disequilibrium example: According to SML, E(r) = 13%, but actual E(r)=15%, α =
▪ Buying on margin: Borrow money from broker, buy stocks, int paid on loan, ▪ Unsystematic risk: Micro, “idiosyncratic risk”, can be diversified away
either for speculation or magnify exposure to risk/expected return profile systematic and firm-specific components 2% → UNDERPRICED, offering too high expected returns for its level of risk
Two Risky Assets ▪ Advantages: Reduces no of inputs req to diversification in a portfolio
▪ Initial margin requirement (IMR): min % of initial investor equity  (1- ▪ Expected Return: 𝑁(𝑁𝑁 ) = 𝑁1 𝑁(𝑁1 ) + 𝑁2 𝑁(𝑁2 ) ▪ Required rate > actual return  overpriced  -ve α
IMR) = max % amount investor can borrow ▪ Convenient: easy starting point for understanding risk using the measure β ▪ Required rate < actual return  underpriced  +ve α
▪ Variance: 𝜎 2𝑁 = 𝑁𝑁2 𝑁2𝑁 + 𝑁𝑁2 𝑁2𝑁 + 2𝑁𝑁 𝑁𝑁 𝑁𝑁𝑁(𝑁𝑁 , 𝑁𝑁 ) Single Factor Index Model [ R i – Rf = βi(Rm – Rf) + αi + ϵi,t ], uses histor. data
▪ Maintenance margin req (MMR): min % of equity before additional funds ▪ Note: Variance: Cov(rA, rA) = σ2A, Covariance w rf asset = 0
must be put into the account 𝑁𝑁𝑁 (𝑁1 ,𝑁2)
▪ Ri – Rf : excess return of individual stock Arbitrage Pricing Theory
▪ Correlation coefficient (𝑁): 𝜌(1,2) = ▪ Rm – Rf : excess return of well-diversified portfolio (mkt/benchmark) Review Qn: When ϵ<0 → bad news, if stock px still ↑, cos +ve systematic shock
▪ Equity = Current value of trade = Mkt value – loan – int on loan + add. cash 𝑁1 𝑁2
▪ Margin call: notification from broker to add additional funds or have your ▪ p=1  σ(1,2) = w1σ1+w2σ2 | p=-1  σ(1,2) = ±(w1σ1-w2σ2) | -1<p(1,2)<1 by ▪ βi : sensitivity of a security’s excess return to the benchmark index APT’s difference from CAPM: CAPM only has one source of systematic risk (i.e.
position liquidated. A margin call occurs when: combining stocks 1 & 2, there is some  in risk due to diversification ▪ ϵ i,t : unanticipated firm-specific events unexpected Δ in mkt portfolio. In reality, many other sources (unexpected Δi/r,
𝑁𝑁𝑁𝑁𝑁𝑁 3-stock portfolio: σ2p = w12σ21 + w22σ22 + w32σ23 + 2w1w2Cov(r1,r2) + Δ inflation, Δ aggregate corporate default risk, Δ industrial pdn)
≤ 𝑁𝑁𝑁, 𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 ▪ Describes the relation between an expected returns and risk when there are
𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 2w1w3Cov(r1,r3) + 2w2w3Cov(r2,r3)
𝑁𝑁𝑁𝑁 + 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 − 𝑁𝑁𝑁. 𝑁𝑁𝑁𝑁 Tradeoff between risk and return: The positive effects of diversification one or more sources of systematic risk, systematic risk must be pervasive and
= undiversifiable in a large portfolio
1 − 𝑁𝑁𝑁 increase as the correlation between assets decrease since the returns of the two
▪ When margin call occurs, need to restore equity to IMR%. assets move in opposite directions and can offset each other, lower risk. ▪ Assumptions: 1) No taxes, no txn costs, 2) investors form well-diversified
▪ Ignore int if considering equity immediately, include int if equity over 1yr ▪ Minimum variance portfolio: portfolio w lowest risk portfolios (only systematic risk), 3) No arbitrage opportunities exist (two
▪ Rate of returns = (Gain/loss – int on loan)/Initial equity securities that always have the same payoff must always have the same
When p=+1, E(r) price), arb (α≠0)is the act of exploiting mispricing to achieve risk-free profit
Short Selling increases linearly, no Factor Loadings & Risk Premiums
▪ Borrow stocks from broker, requires initial margin (investor’s equity), broker effects of diversification ▪ E(rp) = rf + βP,1 [E(r1) - rf ] + βP,2 [E(r2) - rf ]
sells stock, deposit proceeds and margin into margin account. Close out the When p=-1, can find ▪ Loading on factor portfolios: βP,1, βP,2  how a security’s returns covary w a
position by buying the stock and returning stock title. Liable for div weights such that σc=0 risk factor portfolio’s returns
▪ Equity = Total margin acct – market value – div paid after opening trade portfolio is risk-free ▪ Factor risk premium: E(r1) - rf & E(r2) - rf
▪ Margin call occurs when: Optimal portfolio is the ▪ Risk factor portfolio: a well-diversified portfolio with zero idiosyn risk, a
𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁 − 𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 one that maximizes beta of 1 related to one risk factor, a beta of 0 w.r.t. the other risk factors
≤ 𝑁𝑁𝑁, 𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 To formulate arbitrage strategy: Calculations
𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 Sharpe ratio
𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁 Two Components of Stock Risk ▪ Step 1: Form an arbitrage portfolio (only rf and risk factor) so βARB = βA,IR
=
1 + 𝑁𝑁𝑁 𝑁𝑁𝑁(𝑁𝑁 ) = 𝑁𝑁𝑁(𝑁𝑁 (𝑁𝑁 − 𝑁𝑁 )) + 𝑁𝑁𝑁(𝑁𝑁 ) + 𝑁𝑁𝑁(𝑁𝑁 ) 𝑁(𝑁𝑁𝑁𝑁 ) = 𝑁𝑁 + 𝑁𝑁,𝑁𝑁𝑁 [𝑁(𝑁𝑁𝑁𝑁 ) − 𝑁𝑁 ] + 𝑁𝑁,𝑁𝑁 [𝑁(𝑁𝑁𝑁) − 𝑁𝑁 ]
▪ Find market value x 0.5 = net equity req, current equity = total margin – mkt = 𝑁2 𝑁 𝑁2 𝑁 + 𝑁2 (𝑁𝑁 ) = ▪ Step 2: Compare arbitrage portfolio to portfolio A
value, diff = equity to top up 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁 + 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁 𝑁(𝑁𝑁𝑁𝑁 ) = 15.5% and 𝑁(𝑁𝑁 ) = 16%
▪ A round trip is a purchase and a sale (2 x commission), Long (buy first, sell ▪ Note: Var(αi) = 0 because αi is a constant, Cov(Rm, ϵi) = 0 because firm-specific ▪ Step 3: Buy low (higher returns), sell high (lower returns)
later, bullish), Short (sell first, buy later, bearish) risk is the part that is not linked to market risk ▪ Lend = buy t-bill (ppl pay you back at rf rate), Borrow = short sell t-bill
▪ ϵi has an expected value of 0 as the impact of unanticipated events avg to 0 Example: There is one risk factor, i/r risk factor. 𝑁(𝑁𝑁 ) = 9%, E(rARB) = 8.8%.
Risk and Return: Past and Prologue 2 𝑁2 𝑁 𝑁2 𝑁
▪ 𝑁 = 𝑁2  For index model, R measures how much of the variation of R Short sell i/r factor: 0.8x$1m, borrow rf: 0.2x$1m, buy portfolio A: -$1m. Risk-
𝑁
2 2

free profit = (0.09-0.088)x$1m = $2,000

Arithmetic Avg (AAR) Geometric Avg (GAR) is explained by variation in market returns, Rm.
For two risk factors, wp,M = 0.5 = βA,M, wp,IR = 0.75 = βA,IR
Capital Asset Pricing Model (CAPM)
▪ CAPM shows how risk and expected returns relate in equilibrium. In stable
state (no mispricing), α = 0  E(Ri) = Rf + βi[E(Rm)-Rf] The Efficient Market Hypothesis
Multi-factor Model: Fama-French 3 Factor Model
1) Market Index Excess Return: rM-rf 𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁 − 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁
2) SMB: Small firm returns minus big firm returns (rS-rB): α > 0 for smaller
𝑁𝑁𝑁 =
𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 If coupon > IR, bond price Option Markets
firms since CAPM assumptions truer for bigger companies (no transaction 1) FRONT-END LOAD: Commission charged as % of offering price, one-off, > par premium bond ▪ A contract that gives the buyer right (but not obligation) to buy and sell a
costs relative to size of coy, homogeneous expectations since more public e.g. NAV of Fund A = $10.00, 8% front-end load  Investor pays 10/(1-0.08) = designated instrument at specific price (ex/strike price) over specified time
news, price takers (many buyers for big firms) $10.87 per share to buy shares. It does not affect NAV. If coupon < IR, bond price ▪ Call option: option to buy, purchase call when expect value to increase
3) HML: High B/M firm returns minus Low B/M firm returns (rH-rL): 2) BACK-END LOAD: Commission charged as % of withdrawal amount, e.g. < par discount bond ▪ Put option: option to sell, purchase put when expect value to decrease
Value firms (High B/M since low E(FCF) means the firm did well in the past, NAV of Fund A = $12.70, 8% back-end load. Investor receives $12.70 x (1- ▪ European (exercise only on expiry date), American (any time)
but poor future perf. Growth firms (low B/M since high E(FCF)). α > 0 for high If coupon = IR, bond price ▪ In the money (+ve CF), out of the money (-ve CF), at the money (exercise
0.08) = $11.684 per share when they redeem the shares. Does not affect NAV
B/M. B/M is related to systematic financial distress risk. Book value (ex-post = par par bond price = current share price)
3) EXPENSE RATIO/12B-1: % of NAV, recurring
measure), market value (ex-ante measure) Payoff vs Profit at Expiration
FF 3 Factor Model vs CAPM Model 4) Operating Expenses: % of NAV, recurring. If OE ↑, NAV ↓ Breakeven = Exercise Price + Premium Limited downside
Mutual Fund Returns Long Call
𝑁𝑁3: 𝑁(𝑁𝑁 ) − 𝑁𝑁 = 𝑁𝑁,𝑁 [𝑁(𝑁𝑁 ) − 𝑁𝑁 ] + 𝑁𝑁,𝑁𝑁𝑁 𝑁(𝑁𝑁𝑁𝑁 ) + 𝑁𝑁,𝑁𝑁𝑁 𝑁(𝑁𝑁𝑁𝑁) Overall Profit = Max(ST –X, 0) – Premium loss, unlimited
𝑁𝑁𝑁𝑁 𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁 = (buyer)
𝑁𝑁𝑁𝑁: 𝑁(𝑁𝑁 ) − 𝑁𝑁 = 𝑁𝑁,𝑁 [𝑁(𝑁𝑁 ) − 𝑁𝑁 ] Payoff for buyer = Max (ST-X, 0) upside gain
𝑁𝑁𝑁𝑁=1 −𝑁𝑁𝑁𝑁=0 + 𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 Breakeven = Exercise Price + Premium Unlimited downside
Diff = α under CAPM = 𝛽𝑁,𝑁𝑁𝑁 𝑁(𝑁𝑁𝑁𝑁 ) + 𝑁𝑁,𝑁𝑁𝑁 𝑁(𝑁𝑁𝑁𝑁 ) Short Call
𝑁𝑁𝑁𝑁=0 Overall Profit = -Max(ST – X, 0) + Premium loss, limited upside
▪ FF3 is better than single-index CAPM at explaining stock returns because it (seller)
includes important factors. Payoff for seller = – Max (ST-X, 0) gain
Efficient Market Hypothesis (EMH)
▪ Security prices accurately reflect all available information. If markets are
Portfolio Performance Evaluation Long Put
Breakeven = Exercise Price – Premium
Overall Profit: Max(X-ST, 0) – Premium
Limited downside
loss and upside gain
Evaluating Performance: Single Fund (buyer)
efficient, on average, investors cannot earn risk-adjusted positive profit If coupon rate > IR; bond price > par → premium Payoff for buyer = Max (X-ST, 0)
1) Benchmark portfolio: Compare directly with an established, passive index If coupon rate < IR; bond price < par → discount Breakeven = Exercise Price – Premium Limited downside
using publically available info. If not efficient, active strategies should earn If coupon rate = IR; bond price = par → par Short Put
2) Peer Group: Rank perf of competitor funds within a given category Overall Profit = -Max(X-ST, 0) + Premium loss and upside gain
risk-adjusted +ve profits and outperform passive strategies (on CML) Term Structure of I/R: Yield Curve (rs btw yields and time to maturity) (seller)
3) Risk Model: Estimate risk-to-reward measures (alpha, Sharpe, Treynor) Payoff for seller = – Max (X-ST, 0)
▪ Investor competition causes stock prices to fully and accurately reflect ▪ X-axis: Time to maturity, Y-axis: yield (3 types  rising, falling, flat)
▪ Jensen’s alpha: measures risk-adjusted average returns
relevant, available information very quickly. Low trading cost assures that 1) Expectations Hypothesis (Rising / Falling / Flat)
price reflects available info (otherwise unexploited profit opportunities) 𝛼𝑁 = (𝑁𝑁 − 𝑁𝑁 ) − 𝑁𝑁 (𝑁𝑁 − 𝑁𝑁 ) ▪ Long term rates equal cumulative expected future short-term rates
▪ Competition ≠ info efficiency when there is unequal access to info, structural ▪ At 95% confidence level, use t-test of statistic to test whether alpha is ▪ Upward slope  expecting higher future short-term rates, Downward slope
market frictions and investor psychology statistically different from zero. If null hypothesis is not rejected  alpha is  expecting lower future short-term rates
▪ Random Walk: Stock prices should exhibit a random walk if price changes zero  should invest in passive mutual funds instead (1 + 𝑁𝑁 )𝑁 = (1 + 𝑁𝑁−1 )𝑁−1 (1 + 𝑁𝑁 )
are unpredictable and random [ Pi,t = Bi x Pi,t-1 + ei,t, where Bi = 1+E(ri) Mutual Fund Performance
Where f = forward rate
Forms of EMH ▪ If markets efficient, then before expenses, an average mutual fund has α = n0.
▪ Weak form: Prices reflect info in historical price data and trading data  Most actively managed funds charge fees. After fees, the average mutual fund 2) Liquidity Preference Theory (Rising)
▪ Longer term bonds  less liquid  require a higher liquidity premium on
cannot earn using historical price and trading data will underperform the market
top of expected short-term rates
▪ Semi-strong: Prices reflect all publicly available info (past + current)  ▪ To be better off w actively managed funds, markets are NOT semi-strong
cannot earn using growth forecasts, acc statements, past price, volume etc efficient, can identify managers that outperform benchmark after expenses
Selecting Funds/Portfolios in Practice
Managing Bond Portfolios
▪ Strong: Prices reflect ALL info (both public and private)  cannot earn Interest Rate Sensitivity (Price Sensitivity/Volatility)
Approaches to Stock Analysis 1) Entire-wealth portfolio: small investor selects one portfolio, cannot
1) Maturity: When maturity ↑, interest rate sensitivity ↑ because long term
▪ Technical Analysis: uses historical stock prices and volume info to predict diversify across funds, consider total risk, including unsystematic risk. Select
bonds are compounded/discounted more
future price changes  assumes weak form violated portfolios w highest Sharpe ratio
2) Coupon: When coupon ↑, interest rate sensitivity ↓, since cash flow is
▪ Fundamental Analysis: uses publicly available info (fin statements & future 2) Fund of Funds: large investor holds many funds, well diversified, select
delivered sooner
prospects to identify mispriced securities  assumes semi-strong violated funds using Treynor Ratio. Compares avg excess return to systematic risk
3) Yield to maturity: When YTM ↑, interest rate sensitivity ↓ because bond
Implications of EMH for Active/Passive Management (rather than total risk) because idiosyncratic diversified away
price decreases at a decreasing rate, when YTM↑
▪ Active management: 1) Security analysis (violates weak/semi-strong), 2) 3) Portfolio added to benchmark: when evaluating adding an actively-
 Bonds that pay cash flows sooner have lower interest rate sensitivity
Timing Strategies, 3) Investment newsletters (violate semi-strong) managed portfolio, can deliver the benefit of α, but adds idiosyncratic risk
 When expect YTM to fall, buy zero coupon LT bond to experience largest
▪ Passive management: Consistent w semi-strong, e.g. buy and hold portfolio, →Select using info ratio
price appreciation (can sell at high price)
index funds (benchmark, focus on minimizing costs) Macaulay’s Duration
Abnormal Returns: Index Model ▪ Effective maturity of a bond: since the bond pays cash (coupon) prior to
Use regression to estimate a and b coefficients: 𝑁𝑁 = 𝑁 + 𝑁(𝑁𝑁𝑁𝑁𝑁𝑁.𝑁 ) + 𝑁𝑁 maturity, effective maturity < actual maturity.
Then calculate Abnormal Returns: 𝑁𝑁𝑁 = 𝑁 + 𝑁𝑁 = 𝑁𝑁 − 𝑁(𝑁𝑁𝑁𝑁𝑁𝑁.𝑁 ) where 𝑁(𝑁𝑁 ) is the measure of unsystematic risk, deviation from index model 𝑁
Abnormal Returns: FF 3 Factor 𝑁𝑁𝑁 /(1 + 𝑁𝑁𝑁)𝑁
Use a multivariate linear regression to estimate a, b M, bHML, bSMB
Questions: Small cap stocks have larger betas. Thus, the returns, on average, 𝑁𝑁 = 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 = ∑ 𝑁𝑁 𝑁 𝑁
should be larger, and the cumulative returns will be > S&P 500. | Alpha is an 𝑁𝑁𝑁𝑁𝑁
𝑁𝑁 = 𝑁 + 𝑁𝑁 𝑁𝑁,𝑁 + 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁,𝑁 + 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁,𝑁 + 𝑁𝑁 𝑁=1
expected return and realized returns may be diff from expected. Also, alpha may ▪ Duration is equal to maturity for zero coupon bonds
𝑁𝑁𝑁 = 𝑁 + 𝑁𝑁 = 𝑁𝑁 − 𝑁𝑁 𝑁𝑁,𝑁 + 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁,𝑁 + 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁,𝑁
be capturing omitted risk factors. ▪ If semi-annual coupon payments, duration = sum (weighted % x T x 0.5)
Find the idiosyncratic return and subtract total stock return from part of stock’s
return due to the market Bond Prices and Yields ▪ As maturity ↑, duration ↑ | As coupon ↑, duration ↓ | As YTM↓, duration ↓
▪ Duration of a level payment perpetuity: 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 = 𝑁𝑁𝑁
1+𝑁𝑁𝑁 Protective Put Strategy (Long stock + long put option)
Testing Market Efficiency Bond Characteristics and Pricing ▪ Guarantees minimum floor on the long stock position equal to strike price of
▪ Event studies: Test for semi-strong (how quickly info integrates into prices) ▪ A bond is a LT debt instru in which a borrower agrees to make payments of Duration/Price Relationship & Modified Duration the put
▪ Testing a trading rule: Test for weak form (based on past trading info) principal and interest, on specific dates to the holder of the bond ▪ Price change is proportional to duration, not maturity
▪ Assessing perf of prof managers: Test for strong form 𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁 ∆𝑁 ∆ 𝑁𝑁𝑁
Issues in Testing Market Efficiency ▪ Bond pricing: 𝑁 = ∑𝑁
𝑁=1 + = −𝑁 ×
(1+𝑁)𝑁 (1+𝑁)𝑁 𝑁 1 + 𝑁𝑁𝑁
▪ Model mis-specification: Use of wrong model to measure risk & exp return ▪ Holding period yield (HPY)  uses actual reinvestment rate on coupons (Not → Negative sign: price and yield have an inverse relationship
▪ Lucky event issue: Outperform by luck/chance, may not violate strong form YTM), and considers change in price if bond is sold prior to maturity ▪ Modified Duration (D*)
▪ Selection bias issue: Successful investors may not reveal strategies Bond Yields 𝑁 ∆𝑁
Questions: 1) Since upward trend is due to a positive risk premium, it is ▪ Yield to maturity (YTM): Discount rate that makes PV(bond) = Price, 𝑁∗ = ;∴ = −𝑁∗ × ∆ 𝑁𝑁𝑁
consistent with market efficiency, 2) CAPM alpha is an expected value, and (1 + 𝑁𝑁𝑁) 𝑁
Assumptions: bond holders hold bond till maturity, if bond bought halfway,
realized risk-adjusted excess returns can differ from what was expected. The Duration Pricing Error & Convexity
will not receive YTM, coupons reinvested at YTM
APT alpha is the diff in expected returns between a mispriced portfolio w no ▪ Duration is a linear approximation, but the price-yield curve is non-linear (If
idiosyncratic risk and an arbitrage portfolio w the same systematic risk as the use duration, when I/R ↑, overestimate loss. When I/R ↓, underestimate gain
mispriced portfolio. Since the mispriced portfolio and arbitrage portfolio move 1 𝑁𝑁𝑁
in tandem the diff in realized returns is known w certainty and realized risk- 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 = ∑ [ (𝑁2 + 𝑁)]
adjusted excess returns will always equal alpha. 𝑁 × (1 + 𝑁𝑁𝑁)2 (1 + 𝑁𝑁𝑁)𝑁
Prediction Model Including Convexity
1
Mutual Funds and Other Investment ∆𝑁
𝑁
= −𝑁 ×
∆𝑁𝑁𝑁
(1 + 𝑁𝑁𝑁)
+ [ × 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 × ∆𝑁𝑁𝑁2 ]
2
Companies  In practice, convexity is important when interest rate changes are large Payoff ST ≤ X ST > X
Mutual Funds Interest Rate Risk Stock ST ST
▪ Portfolio of financial securities, with many investors providing capital 1) Price risk: Risk that an investor cannot sell bond at the expected price Put X – ST 0
▪ Benefits: Professional mgt, diversification & divisibility for small investors (when I/R ↑, bond price ↓  ↑price risk) Total X ST
(low funds but well diversified portfolio), reduced txn costs 2) Reinvestment risk: Risk that an investor cannot reinvest coupon at YTM Covered Call Strategy (long stock + write/sell call option)
Net Asset Value (NAV): price per share of a mutual fund (when I/R ↑, FV of reinvested coupons ↑  ↓reinvestment risk
Note: these two risks are potentially offsetting.
 ▪ Parameters: S0 (current stock price), X (strike price), T (time to maturity), r
(risk-free i/r), σ (volatility of stock), δ (stock dividend)  all known except
for volatility of the stock.
If this variable increases Value of Call Option Value of Put Option
Stock Price, S Increase Decrease
Exercise Price, X Decrease Increase
Volatility, σ Increase Increase
Time to expiration, T Increase Increase
Interest Rate Increase Decrease
Dividend Payouts Decrease Increase

Payoff ST ≤ X ST > X
Stock ST ST
Call 0 ST – X
Total ST X

Implied volatility: Standard deviation of stock returns consistent with

Option Valuation option’s market value
1) Binomial Option Pricing Model / Risk-Neutral Valuation
▪ Consider setting up a riskless portfolio by buying H shares of XYZ and short
one call option. Current share price = $100, exercise price = $110, upside
price = $120, downside price = $90.
𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁
▪ 𝑁 |𝑁= , u = 1.2, d= 0.9, Cu = option payout = future
𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁
stock price – exercise price = 120 – 110 = 10 (sell at $120, buy at $110), Cd =
$0 (since X > stock price)
▪ Upside payout = $120H - $10 (since sell call option), downside payout =
$90H – 0. Find H shares so that portfolio has identical payout in either state
▪ Since H = 1/3, Today = Present value (payout in 3 months), rf = 10%,
100(1/3) – C = $30/1.1  C = $6.06
* Can use risk-free rate since the payout is riskless

Risk-Neutral Shortcut

where p represents a risk-neutral probability

2) Put Call Parity: defines the relation btw a call and put premium
▪ The call and put must have the same (1) underlying asset, (2) strike price,
(3) maturity date
At T = 0, At T = Expiration,

Where C/P = call/put premium, S = stock price, X = strike price, r = annual i/r,
T = time to maturity (in annual terms)
▪ Portfolio A (Long call, short put), Portfolio B(long stock, short bond/borrow)
▪ In no-arbitrage equilibrium, the cost of portfolio A must equal the cost of
portfolio B (given same X, same T, same S0)

Calculation of Arbitrage Profit

▪ No-arbitrage principle: if two portfolios have the same payout in the future,
then in no arbitrage equilibrium, the portfolios will have the same cost today
Scenario 1: 𝑁0 − 𝑁0 > 𝑁0 − 𝑁𝑁−𝑁𝑁 → there is an arbitrage opportunity
▪ Sell overvalued portfolio A: sell call, buy put (Net CF today= C – P)
▪ Buy portfolio B: buy stock, borrow 𝑁𝑁−𝑁𝑁 (Net CF today = 𝑁𝑁−𝑁𝑁 – S0)
▪ Net CF today = C – P – (S0 - 𝑁𝑁−𝑁𝑁 ) > 0, Net CF at expiration = ST – X – (ST – X)
= 0  therefore, there is an arbitrage profit
Scenario 2: 𝑁0 − 𝑁0 < 𝑁0 − 𝑁𝑁−𝑁𝑁 → there is an arbitrage opportunity
▪ Buy portfolio A: buy call, sell put (Net CF today= P - C)
▪ Sell portfolio B: sell stock, lend/buy 𝑁𝑁−𝑁𝑁 (Net CF today = S0 - 𝑁𝑁−𝑁𝑁 )
▪ Net CF today = S0 - 𝑁𝑁−𝑁𝑁 – (C – P) > 0, Net CF at expiration = ST – X – (ST – X)
= 0  therefore, there is an arbitrage profit
Black-Scholes Pricing Model
▪ Increasing the number of time intervals until expiration decreases time btw
each interval. At the limit, investors trade in continuous time.