A solution manual to the econometrics of financial markets

Consider a stock with a virtual price process that follows a continuous geometricBrownian motion, with a net expected annual return and standard deviation of returnnot continuously compo

The Econometrics of Financial Markets

Petr Adamek John Y Campbell Andrew W Lo

A Craig MacKinlay Luis M ViceiraAuthor address:

MIT Sloan School, 50 Memorial Drive, Cambridge, MA 02142{1347 Department of Economics, Harvard University, Littauer Center, Cam- bridge, MA 02138

MIT Sloan School, 50 Memorial Drive, Cambridge, MA 02142{1347 Wharton School, University of Pennsylvania, 3620 Locust Walk, Philadel- phia, PA 19104{6367

Department of Economics, Harvard University, Littauer Center, bridge, MA 02138

List of Figures

3.3 Histogram of IBM Price Changes Falling on Odd or Even Eighth 15

12.1 Kernel Regression of IBM Returns on S&P 500 Returns 63

List of Tables

2.3 Statistics for Daily and Monthly Simple and Continuously Compounded

3.4 Unconditional and Conditional Distributions of Bid/Ask Spreads 20

PrefaceThe problems inThe Econometrics of Financial Marketshave been tested in PhD courses

at Harvard, MIT, Princeton, and Wharton over a number of years We are grateful to thestudents in these courses who served as guinea pigs for early versions of these problems,and to our teaching assistants who helped to prepare versions of the solutions We alsothank Leonid Kogan for assistance with some of the more challenging problems in Chapter9

Problems in Chapter 2

Solution 2.1 2.1.1Recall the martingale property given by (2.1.2) and observe that the mean-squarederror of the time-tforecastXt of pricePt +1is

(S2.1.2)

= E[0(Pt ; k ; l ;Pt ;2 k ; l)] = 0:

Solution 2.2Denote the martingale property (2.1.2) by M Then

nand2 n  jn j ;1=2 satis es RW3 but not M; (ii)n nn ;1 satis es Mbut not RW2; (iii)n nnsatis es RW2 but not RW1

Solution 2.3

A necessary condition for the log-price processptin (2.2.9) to satisfy RW1 is+ = 1 Let

c+ and consider the set of all non-RW1 Markov processes (2.2.9), i.e.,c6= 1 Therestriction CJ = 1 is equivalent to =c=4 The constraints 0; 1 are satis edexactly forc2[1;4=3] and therefore the set of all two-state Markov chains represented bythe pair (; ) that cannot support any RW1 process but still yields CJ = 1 is simply

f(1 p

Var[Zt(q)] = q;1

X

k =0Var[Zt ; k] + 2q;1

X

k =1(q;k)Cov[Zt;Zt ; k](S2.4.1)

which yields (2.4.19) The coecients of Cov[Zt;Zt ; k] are simply the number of k-thorder autocovariance terms in the variance of the multiperiod returnZt(q) (recall that thismultiperiod return is the sum ofq one-period returns) The coecients decline linearly

Ten individual stocksused for problem 2.5, identi rstNassets and the market portfolio,



 2 4

(S5.4.4)

 2 4

0;1

1

 2 m +

0;1

(S5.4.6)

;1 :

(S5.4.9)

From (S5.4.2) and (S5.4.9) we have

0;1 =2

arithmeticaverage ofdailyreturns, and the estimator ^ of the volatility of annual returns

dened as a rescaled standard deviation ofdaily... of discretization of prices to< /p>

a $1/8 grid on naive estimates of annual mean and standard deviation based on dailyreturns data is simulated for a hypothetical stock following a continuous... changes to volume, according to the Gauss-Markov theorem

Histogram of IBM’s Bid/Ask