TA: Announcement Wei Cen Won't hold office hours and TA sections this week due to some family issues. He will make up the missed section.
Review: Short Interest
What is short interest?
What do we know short interest in general?
Short Interest on the rise The typical firm has very little short interest
Review: Short Interest
What do we know market for borrowing?
Most stocks are easy to borrow
Majority (91%) stocks are "general collateral" 9% stocks are "special" (with 1% "very special") Recalls are rare
Review: Short Interest
What do we know market for borrowing?
What do we know the "special"?
High investor disagreement, turnover, and dispersion in analyst forecasts, increased message board activity, and low cash flows.
Review: Short Interest
How smart are bears?
Large
short interest is bad news for a firm's future returns (incremental to 3-factor model) Heavily shorted stocks look like late-stage glamour stocks (high V/P ratios, high turnover) But profitability of this strategy is mainly from heavily shorted stocks with high volume and volatility
Desai et al. (JF 2002)
Calendar
abnormal returns for heavily shorted firms from month EN=1 to Ex+1
Market Risk
Required Return = Risk-free return + Risk Premium
Models: CAPM (Assuming market risk only)
Rit Rft = i + imt (Rmt - Rft) + it Ri - Rf = + im x ( Rm - Rf ) + i
Multiple Risks
Required Return (i) = Risk-free return + Beta1 (i) x Risk Premium for factor 1 + Beta2 (i) x Risk Premium for factor 2 + ... + Betak (i) x Risk Premium for factor k
Models: Multifactor model
Ri - Rf = + 1 x ( Rm - Rf ) + 2 x ( Risk premium factor 2 ) + ... + k x ( Risk premium factor k ) + i
Data Characteristics
(Fama and French, 1992)
Returns don't vary across portfolios sorted on Why when 's are strongly correlated with size MVE decreases as increases (but small variation) Over the whole period doesn't explain average returns
Portfolio average returns
(Fama and French, 1992)
Returns increase as portfolio size decreases Returns across a given size decile invariant to changes in Variation in unrelated to size is not compensated in average returns
Portfolio returns (Size and BE/MVE) (Fama and French, 1992)
Returns decrease as size increases Returns increase as Book to Market increases after controlling for size What does Book to Market mean after controlling for size
Book Value would have given the same results
Sub-period results
(Fama and French, 1992)
insignificant even across sub-periods Coefficient for size is negative and significant Coefficient for Book to Market is positive and significant
Results are economically significant
Three-factor Model(Fama and
French, 1992) Ri - Rf = + im x ( Km - Rf ) + bs x SMB + bv x HML + i
Where SMB=(Rsmall-Rbig), HML=(Rvalue-Rgrowth)
FF-four Factor Model
Four-factor Model (Jegadeesh and Titman,
1993 ; Carhart, 1997)
Ri - Rft = + im x ( Rm - Rf ) + bis x SMB + biv x HML + bmi x UMD + Where SMB=(Rsmall-Rbig), HML=(Rvalue-Rgrowth) and UMD=(Rwinner-Rlose).
http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/d
http://www.efficientfrontier.com/ef/701/sg.htm William J. Bernstein
Valuation Class 7 Dividend & Valuation
AEM 428 Manny Dong
Readings: Penman
Summary
Introduction to Valuation Model: Dividend Discount Model Dividend Irrelevancy Principles: Problems of Dividend Discount Model Dividend and Free Cash Flow
Present value based Valuation Models
Dividend Discount Model Free Cash Flow Model Residual Income Model
Payoffs to Investing in an Equity
The investments are made at time zero and held for T periods when they terminate or are liquidated.
Initial price
Investment horizon When stock is sold
1
2
3
T-1
T
0 d1 d2
Io
d3
dT-1
Dividends
TVT + dT
Selling price at T + Dividend (if sold at T)
Dividend Discount Model
Dividend discount model says:
P0 = E d t + TVT E
-t t =1
T
T
In other words....
d1 d 2 d3 dT TVT P0 = + 2 + 3 + ..... + T + T E E E E E
Terminal dividend is assumed to follow perpetuity with growth
d T +1 TVT PT = = E - g G
Two small points
E
refers to gross discount rate e.g. 1.12, while r refers to net discount rate e.g. 0.12 In terminal value calculation, gross discount rate should be used with gross growth rate e.g. 1.05, and net discount rate should be used with net growth rate e.g. 0.05.
G refers one plus growth rate i.e. G = 1+g
The Dividend Discount Model (DDM)
E.g.: Suppose the required rate of return is 10%. You also know that next year's dividend is 220, and then dividends are 110 per year after that. What is the value of the project?
Example 2
Now suppose you know the price of the company which is: 10200. The required rate of return is still 10%. You also know that next year's dividend is 220, and then dividends are 110 the year after that. What is the perpetual growth rate implied by the price?
Example 2: Backing out the growth rate
Understanding dividends in the grand scheme of things
What
Are Some Financing Activities? they create value for the investors?
Do
The Question for Forecasting: What Creates Value in a Firm
You
can get value from a firm from getting the dividends. How else can you get money from a firm?
Would you prefer to have dividends, have capital gain, have stock repurchases or have the firms do nothing?
Modigliani-Miller: (M&M) Dividend Irrelevance
Reason: If
a company issues dividend, then shareholders get the dividend. They can reinvest these dividends in project of the same risk class and get the same expected returns.
Dividend Irrelevance
If a company doesn't issue dividend, then shareholders don't get the dividend . But the dividends remain in the firm, resulting in higher capital gain. ==> same thing for investors (They can sell the stocks for the capital gain: "homemade dividend")
Merton Miller
The Question for Forecasting:
What Creates Value in a Firm Equity Financing Activities ?
Share
Issues ? Share Repurchases ? Dividends ?
Debt Financing Activities ? Investing and Operating Activities?
Value
is usually created when you sell products and when you lower costs, not when you play around with distributing value!
The Dividend Discount Model:Targeting Dividends
DDM:
d1 d 2 d 3 V0 = + 2 + 3 + ..... E E E
Problems: How far does one project? Does d d d d V0T = 1 + 2 + 3 + ..... + T 2 3 T E E E E provide a good estimate of V0?
The Dividend Discount Model:Problems
More Problems: Take our e.g.
E
220 110 V0 = 1.1 = + 10 ,200 1.10 1.10 - 1.09
(i) If the firm can borrow a lot to pay the dividends ... does this create value? (ii) Suppose a firm earns twice the amount as another firm, and they both pay the 110 per year of dividend. Does the formula say anything? (iii) If a firm does not pay dividend, is it worth nothing?
Do you see any contradiction for these two statements?
1) M&M dividend irrelevancy concept 2) We forecast dividends over finite horizons to get the value.
The Dividend Discount Model:Targeting Dividends
1) M&M dividend irrelevancy concept leads to the dividend conundrum: Equity price is based on future dividends, but forecasting dividends over finite horizons does not give an indication of this price Conclusion: Focus on creation of wealth rather than distribution of wealth.
Does Dividend Discount Analysis make sense?
YES Dividends are what shareholders get, so forecast them Dividends are usually fairly stable in the short run and hence easy to forecast NO Dividends payout is not related to value, at least in the short run; dividend forecasts ignore the capital component of payoffs You have to forecast for much longer periods Terminal values for shorter periods are hard to calculate with any reliability. Bottom Line: DDM works best when payout is permanently tied to the value generation in the firm. For example, when a firm has a fixed payout ratio (dividends/earnings).
We want another measure that captures the value generation activity better
Free Cash Flow The firm is earning money for both the equity holders and the debt holders. Free cash flow is the amount available to both.