Python for Finance

More documents

Recommendations

Info

Chapter 7• http://canisius.edu/~yany/python/ffMonthly.pklffMonthly.pklffcMonthly.pklffMonthly5.pklyanMonthly.pklyanMonthly.pklusGDPannual.pklusCPImonthly.pklFama-French monthly three factorsFama-French-Carhart monthly four factorsFama-French monthly five factorsFama-French daily three factorsA monthly dataset generated by the authorUS GDP annualConsumer Price Index (CPI) monthlySeveral questions:• Which criterion?• Is the performance time-period independent?• In-sample estimation versus out-sample predictionReferencesPlease refer to the following articles:• Carhart, Mark M., 1997, On Persistence in Mutual Fund Performance, Journal ofFinance 52, 57-82• Fama, Eugene and Kenneth R. French, 2015, A five-factor asset pricing model,Journal of Financial Economics 116, 1, 1-22• Fama, Eugene and Kenneth R. French, 1993, Common risk factors in the returns onstocks and bonds, Journal of Financial Economics 33, 3056• Fama, Eugene and Kenneth R. French, 1992, The cross-section of expected stockreturns, Journal of Finance 47, 427-465• Jegadeesh, N., & Titman, S., 1993, Returns to buying winners and selling losers:Implications for stock market efficiency, Journal of Finance 48(1): 65–91• Sharpe, W. F., 1966, Mutual Fund Performance, Journal of Business 39 (S1),119–138• Sharpe, William F., 1994, The Sharpe Ratio, The Journal of Portfolio Management21 (1), 49–58• Sortino, F.A., Price, L.N.,1994, Performance measurement in a downside riskframework, Journal of Investing 3, 50–8• Treynor, Jack L., 1965, How to Rate Management of Investment Funds, HarvardBusiness Review 43, pp. 63–75[ 239 ]
Multifactor Models and Performance MeasuresExercises1. What are the differences between the CAPM and Fama-French 3three-factormodels?2. What are the meanings of SMB and HML in the Fama-French three-factormodel?3. What is the meaning of MOM in the Fama-French-Carhart four-factor model?4. What are the meanings of RMW and CMA in the Fama-French five-factormodel?5. What is the difference between R2 and adjusted R2 when runningmultifactor models?6. How many OLS functions we could use? Please offer at least two functionsfrom different Python modules.7. Which module contains the function called rolling_kurt? How can you usethe function?8. Based on daily data downloaded from Yahoo! Finance, find the results forIBM based on the last 5 years by running both the CAPM and Fama-Frenchthree-factor models. Which model is better?9. What is the momentum factor? How do you run a Fama-French-Carhartfour-factor model? Please use a few tickers as an illustration.10. What is the definition of the Fama-French 5 factor model? How do you run itfor Citi Group? The ticker of the financial institution is C.11. For the following stock tickers, IBM, DELL, WMT, ^GSPC, C, A, AA, andMOFT, run regression based on CAPM, FF3, FFC4, and FF5. Which one is thebest? Discuss your benchmark or criteria to compare.12. Write a Python program to estimate rolling beta on a yearly basis based onthe Fama-French three-factor model. Use it to show the annual beta for IBMfrom 1962 to 2016.13. Update the following Python datasets. The original datasets can bedownloaded from the author's web page. For example, in order to downloadthe first dataset, called ffMonthly.pkl, go to http://canisius.edu/~yany/python/ffMonthly.pkl:ffMonthly.pklffcMonthly.pklffMonthly5.pklFama-French monthly three factorsFama-French-Carhart monthly four factorsFama-French monthly five factors[ 240 ]
Page 2 and 3:
Python for FinanceSecond EditionFin
Page 4 and 5:
CreditsAuthorYuxing YanProject Coor
Page 6 and 7:
I would like to thank Ben Amoako-Ad
Page 8 and 9:
www.PacktPub.comeBooks, discount of
Page 10 and 11:
Table of ContentsPrefaceixChapter 1
Page 12 and 13:
Table of ContentsAppendix F -how to
Page 14 and 15:
Table of ContentsAppendix C - data
Page 16 and 17:
Table of ContentsCapital budgeting
Page 18 and 19:
PrefaceIt is our firm belief that a
Page 20 and 21:
PrefaceChapter 2, Introduction to P
Page 22 and 23:
PrefaceChapter 13, Credit Risk Anal
Page 24 and 25:
ConventionsIn this book, you will f
Page 26:
PrefaceErrataAlthough we have taken
Page 29 and 30:
Python BasicsThere are several reas
Page 31 and 32:
Python BasicsThe top-left panel (wi
Page 33 and 34:
Python Basics4. From the IPython sh
Page 35 and 36:
Python BasicsIn the preceding codes
Page 37 and 38:
Python Basics(1+r)^nfv: fture value
Page 39 and 40:
Python BasicsAssuming that we inves
Page 41 and 42:
Python BasicsBy using a loop, we ca
Page 43 and 44:
Python BasicsThe print() function c
Page 45 and 46:
Python BasicsWe could retrieve data
Page 47 and 48:
Python BasicsData typesDescriptioni
Page 49 and 50:
Python BasicsFor example, from the
Page 51 and 52:
Python BasicsIn Chapter 3, Time Val
Page 53 and 54:
Python BasicsTo retrieve the file,
Page 55 and 56:
Python BasicsHere PV is the present
Page 58 and 59:
Introduction toPython ModulesIn thi
Page 60 and 61:
Chapter 2In addition, when a module
Page 62 and 63:
Chapter 2Recall that in Chapter 1,
Page 64 and 65:
Chapter 2To find a specific module,
Page 66 and 67:
Chapter 2In the previous example, d
Page 68 and 69:
Chapter 2----------a : array_likeCa
Page 70 and 71:
Chapter 2If n returns are given, th
Page 72 and 73:
Chapter 2Similarly, we could save a
Page 74 and 75:
Chapter 2The next example presents
Page 76 and 77:
Chapter 2Introduction to statsmodel
Page 78 and 79:
Chapter 2From the preceding output,
Page 80 and 81:
Chapter 2'core', 'crosstab', 'cut',
Page 82 and 83:
x[bad] = np.nan# missing code is np
Page 84 and 85:
[ 57 ]Chapter 2For finance, time se
Page 86 and 87:
Chapter 2Stock A Stock B0 0.452820
Page 88 and 89:
Chapter 22010-01-08 110.929466>> >d
Page 90 and 91:
Chapter 2After clicking the logo sh
Page 92 and 93:
For a Mac, we have the following co
Page 94 and 95:
Chapter 2After answering y, the fol
Page 96:
Chapter 213. What is wrong with the
Page 99 and 100:
Time Value of MoneyIntroduction to
Page 101 and 102:
Time Value of MoneyLater in this ch
Page 103 and 104:
Time Value of MoneyNote that in the
Page 105 and 106:
Time Value of MoneyFor perpetuity,
Page 107 and 108:
Time Value of MoneyExample 1: >>>pv
Page 109 and 110:
Time Value of MoneyThe following ta
Page 111 and 112:
Time Value of MoneyExample 1: >>>pv
Page 113 and 114:
Time Value of MoneyIf using the sam
Page 115 and 116:
Time Value of MoneyWhile using just
Page 117 and 118:
Time Value of MoneyThe related grap
Page 119 and 120:
Time Value of MoneyNow, let's write
Page 121 and 122:
Time Value of MoneyFor annuity due,
Page 123 and 124:
Time Value of MoneyThe total future
Page 125 and 126:
Time Value of MoneyNormally, if a s
Page 127 and 128:
Time Value of MoneyAn annuity is de
Page 129 and 130:
Time Value of MoneyTo find out the
Page 131 and 132:
Time Value of MoneyThe output is sh
Page 133 and 134:
Time Value of Moneydef pvGrowingAnn
Page 135 and 136:
Time Value of Money15. We know that
Page 137 and 138:
Sources of DataDiving into deeper c
Page 139 and 140:
Sources of DataTo find the usage of
Page 141 and 142:
Sources of DataTo view the first an
Page 143 and 144:
Sources of DataIn the preceding pro
Page 145 and 146:
Sources of Data>>>ret[0:3]array([-0
Page 147 and 148:
Sources of DataAssume that two pric
Page 149 and 150:
Sources of Datad0=x.datefor i in ra
Page 151 and 152:
Sources of DataThe so-called candle
Page 153 and 154:
Sources of DataRetrieving data from
Page 155 and 156:
Sources of DataAssume that the data
Page 157 and 158:
Sources of DataGenerating two dozen
Page 159 and 160:
Sources of Data15 spreadBBB Moody's
Page 161 and 162:
Sources of Data0 201512311 20130228
Page 163 and 164:
Sources of DataAssume that TORQcq.p
Page 165 and 166:
Sources of Datasigma = np.std(ret)x
Page 167 and 168:
Sources of DataAppendix C - Python
Page 169 and 170:
Sources of DataThe corresponding gr
Page 171 and 172:
Sources of DataAppendix F -how to g
Page 173 and 174:
Sources of Data3. Manually download
Page 176 and 177:
Bond and Stock ValuationBond or fix
Page 178 and 179:
Chapter 5In Chapter 3, Time Value o
Page 180 and 181:
Chapter 5On the bottom, we have the
Page 182 and 183:
Chapter 5By applying equation (8),
Page 184 and 185:
Chapter 5Note that Rm=APR/m is from
Page 186 and 187:
Chapter 5Interest rate quotationEff
Page 188 and 189:
Chapter 5The related graph is given
Page 190 and 191:
Chapter 5For this reason, a Python
Page 192 and 193:
Chapter 5>>>w20.4878049>>>w1*1 + w2
Page 194 and 195:
Chapter 5If a bond holder could con
Page 196 and 197:
Based on this result, the YTM is 7.
Page 198 and 199:
Chapter 5Stock valuationThere are s
Page 200 and 201:
Chapter 5……(23)The selling pric
Page 202 and 203:
Chapter 5Here,Equity is the market
Page 204 and 205:
Here, PV is the present value, R is
Page 206 and 207:
Chapter 5Appendix C - Python progra
Page 208 and 209:
Chapter 5sellingPrice/(1+R)**nselli
Page 210 and 211:
Chapter 5A3/A- 32.00 49.00 59.00 72
Page 212 and 213:
Capital Asset Pricing ModelCapital
Page 214 and 215:
Chapter 6Here is the covariance bet
Page 216 and 217: Chapter 6Here, yi is the ith observ
Page 218 and 219: default to the md5 hash or the url
Page 220 and 221: Chapter 6(beta,alpha,r_value,p_valu
Page 222 and 223: Chapter 60 1.0 NaN 1.51 1.5 1.0 2.0
Page 224 and 225: Chapter 6The first 10 lines are sho
Page 226 and 227: Chapter 6Saving our data to an Exce
Page 228 and 229: Simple string manipulationFor Pytho
Page 230 and 231: Chapter 6The output is shown here:T
Page 232 and 233: Chapter 6The two most used panels a
Page 234 and 235: Chapter 6ReferencesPlease refer to
Page 236 and 237: Chapter 6Note 2 - how to download a
Page 238 and 239: Chapter 616. From this chapter, we
Page 240 and 241: Multifactor Models andPerformance M
Page 242 and 243: The pandas DataFrame is used to con
Page 244 and 245: Chapter 7import statsmodels.api as
Page 246 and 247: Chapter 7When ratios of equity book
Page 248 and 249: Chapter 7To save space, we will not
Page 250 and 251: Chapter 7The corresponding output i
Page 252 and 253: Chapter 7The output is shown here:P
Page 254 and 255: Chapter 7The only modification is t
Page 256 and 257: Chapter 7y = pd.DataFrame({'key': [
Page 258 and 259: Chapter 7'SP500': [0.1,0.17, -0.05,
Page 260 and 261: Chapter 7Sometimes, we need to merg
Page 262 and 263: Appendix A - list of related Python
Page 264 and 265: Chapter 7The first and last several
Page 268 and 269: 14. Data source: http://mba.tuck.da
Page 270 and 271: Time-Series AnalysisIn finance and
Page 272 and 273: There is one end of chapter problem
Page 274 and 275: Chapter 8In the preceding program,
Page 276 and 277: Function Description Examplesrelati
Page 278 and 279: Chapter 8The output is shown here:P
Page 280 and 281: ticker='IBM'begdate=(2013,1,1)endda
Page 282 and 283: Chapter 8Here, v2(v1) is the second
Page 284 and 285: 2016-09-01 0.0025 0.0200 -0.0134 0.
Page 286 and 287: Chapter 8import numpy as npticker='
Page 288 and 289: Chapter 8Some books distinguish the
Page 290 and 291: Chapter 8ticker='ibm'begdate=(2013,
Page 292 and 293: Chapter 852-week high and low tradi
Page 294 and 295: Chapter 8The covariance between the
Page 296 and 297: ff=pd.read_pickle('c:/temp/ffDaily.
Page 298 and 299: Chapter 8The following Python progr
Page 300 and 301: Chapter 8Here, n is the number of o
Page 302 and 303: Chapter 8x=np.array(pd.read_csv(pat
Page 304 and 305: Chapter 8The output is shown here:[
Page 306 and 307: Chapter 8In the preceding example,
Page 308 and 309: Chapter 8• George, T.J., and Hwan
Page 310 and 311: The key part of the program used to
Page 312 and 313: Chapter 8For readers from schools w
Page 314 and 315: 11. In addition, generate the simil
Page 316 and 317:
Portfolio TheoryUnderstanding portf
Page 318 and 319:
Chapter 9Here, is the portfolio var
Page 320 and 321:
Chapter 9var=w1**2*var1 +w2**2*var2
Page 322 and 323:
Chapter 9The graph is shown here:To
Page 324 and 325:
Chapter 9Optimization terminated su
Page 326 and 327:
Chapter 9Here, U is the utility fun
Page 328 and 329:
Chapter 9print("meanAnnual, varAnnj
Page 330 and 331:
[ 303 ]Chapter 9'SCD' 'SEF' 'SI' 'S
Page 332 and 333:
output.columns=[ticker]return outpu
Page 334 and 335:
The key command used is ret.T*ret.
Page 336 and 337:
Chapter 9def sharpe(R,w):var = port
Page 338 and 339:
2. Code for defining two functions:
Page 340 and 341:
Chapter 9In one of the previous pro
Page 342 and 343:
Chapter 9The output is shown here:A
Page 344 and 345:
Chapter 9ret = sp.array(mean_return
Page 346 and 347:
Chapter 9x = getData(ticker,begdate
Page 348 and 349:
Chapter 9Here is the output:Since t
Page 350 and 351:
Chapter 97. Click CSV on the right-
Page 352 and 353:
Chapter 90 10006 19570301 198407181
Page 354 and 355:
Chapter 911. For several stocks suc
Page 356 and 357:
Chapter 910. Update yanMonthly.pkl,
Page 358:
Chapter 916. Estimating the Sharpe
Page 361 and 362:
Options and Futures• Using the bi
Page 363 and 364:
Options and FuturesTaking the natur
Page 365 and 366:
Options and Futuress0=1.25rHome=0.0
Page 367 and 368:
Options and FuturescontractFuturesS
Page 369 and 370:
Options and FuturesTo create a grap
Page 371 and 372:
Options and FuturesA graphical repr
Page 373 and 374:
Options and FuturesEuropean versus
Page 375 and 376:
Options and FuturesGenerating our o
Page 377 and 378:
Options and FuturesThe first line o
Page 379 and 380:
Options and FuturesThe related grap
Page 381 and 382:
Options and Futuresx3=60; c3=5y1=(a
Page 383 and 384:
Options and FuturesThe correspondin
Page 385 and 386:
Options and FuturesNote that in the
Page 387 and 388:
Options and FuturesOn the other han
Page 389 and 390:
Options and FuturesThe following co
Page 391 and 392:
Options and FuturesThe correspondin
Page 393 and 394:
Options and Futuresplt.figtext(0.75
Page 395 and 396:
Options and FuturesHere, u is the u
Page 397 and 398:
Options and Futures# at level 2su=r
Page 399 and 400:
Options and Futuresd = 1.0 / ua = e
Page 401 and 402:
Options and FuturesFor the currency
Page 403 and 404:
Options and FuturesSimilarly, we co
Page 405 and 406:
Options and FuturesvolHigh=1.0cLow=
Page 407 and 408:
Options and Futuresmin_value=1000fo
Page 409 and 410:
Options and FuturesAppendix A - dat
Page 411 and 412:
Options and FuturesNote that -1 mea
Page 413 and 414:
Options and Futures16. The current
Page 416 and 417:
Value at RiskIn finance, implicitly
Page 418 and 419:
Chapter 11d1=stats.norm.pdf(0,0.1,0
Page 420 and 421:
Chapter 11s=sp.arange(-10,z,delta)f
Page 422 and 423:
Chapter 11ret=f(x)plt.plot(x,ret)x2
Page 424 and 425:
Chapter 11z=norm.ppf(1-confidence_l
Page 426 and 427:
Chapter 11#print(retNdays.head())po
Page 428 and 429:
Chapter 11p =getData(ticker, begdat
Page 430 and 431:
Chapter 11n=5000000#ret = random.no
Page 432 and 433:
Chapter 11VaR based on sorted histo
Page 434 and 435:
leftTail=int(n*(1-confidence_level)
Page 436 and 437:
std=np.std(ret)#n_simulation=5000sp
Page 438 and 439:
for i in np.arange(n):stock=tickers
Page 440 and 441:
Chapter 11Again, we expect to see 2
Page 442 and 443:
Chapter 11#sum=0.0for i in np.arang
Page 444 and 445:
Chapter 11Estimation of an n-day Va
Page 446 and 447:
Chapter 1110. When we estimate a Va
Page 448 and 449:
Monte Carlo SimulationMonte Carlo S
Page 450 and 451:
Chapter 12This program is equivalen
Page 452 and 453:
Chapter 12Here, f(x) is the density
Page 454 and 455:
Chapter 12import matplotlib.pyplot
Page 456 and 457:
Chapter 12Using simulation to estim
Page 458 and 459:
Chapter 12Generating random numbers
Page 460 and 461:
Chapter 12#for i in range(len(nonSt
Page 462 and 463:
Chapter 12The constraint specified
Page 464 and 465:
Chapter 12Simulation of stock price
Page 466 and 467:
Chapter 12Graphical presentation of
Page 468 and 469:
Chapter 12y2=rho*x1+sp.sqrt(1-rho**
Page 470 and 471:
dt=T/n_stepscall=sp.zeros([n_simula
Page 472 and 473:
Chapter 12Based on the preceding re
Page 474 and 475:
Chapter 12We have the following equ
Page 476 and 477:
print("max NPV of project=",round(m
Page 478 and 479:
Chapter 12One alternative is a so-c
Page 480 and 481:
Chapter 12totalCost=salaryCost + ad
Page 482 and 483:
[ 455 ]Chapter 12In other words, we
Page 484 and 485:
Chapter 12Constructing an efficient
Page 486 and 487:
Chapter 12The graph is shown here:I
Page 488 and 489:
Chapter 12The following program ref
Page 490 and 491:
Chapter 12On the other hand, if we
Page 492 and 493:
4. To convert random numbers from a
Page 494 and 495:
Credit Risk AnalysisThe objective o
Page 496 and 497:
Chapter 136 A3 A- A- 1 17 Baa1 BBB+
Page 498 and 499:
Chapter 13Source: Moody's (2007).No
Page 500 and 501:
Chapter 13A secured debt is a debt
Page 502 and 503:
If P is the default probability we
Page 504 and 505:
14 B2/B 455.800 481.600 505.200 531
Page 506 and 507:
Chapter 13Using the KMV model to es
Page 508 and 509:
Chapter 13The following Python prog
Page 510 and 511:
Chapter 13The related graph is show
Page 512 and 513:
Chapter 13output.append(round(R,5))
Page 514 and 515:
Chapter 13CDS P CDS P CDS P CDS P C
Page 516 and 517:
8. What are uses of the term struct
Page 518 and 519:
Exotic OptionsIn Chapter 10, Option
Page 520 and 521:
Chapter 14x=40. # exercise priceT=6
Page 522 and 523:
Chapter 14Here, call (T) is the cal
Page 524 and 525:
Chapter 14v[i][j]=max(v1,v2)return
Page 526 and 527:
Chapter 14n = mu.shape[0]Sigma_det
Page 528 and 529:
Chapter 14First, we should study th
Page 530 and 531:
Chapter 14sigma1=0.15sigma2=0.20pri
Page 532 and 533:
Chapter 14The final price would be
Page 534 and 535:
Chapter 14total+=sTprice_average=to
Page 536 and 537:
Chapter 14#if sT<barrier:in_=True#p
Page 538 and 539:
Chapter 14c=p4f.bs_call(s,x,T,r,sig
Page 540 and 541:
Chapter 14The Python code for this
Page 542 and 543:
Chapter 14If we use futures to hedg
Page 544 and 545:
Chapter 14Here, T is the maturity d
Page 546:
Chapter 14SummaryThe options we've
Page 549 and 550:
Volatility, Implied Volatility, ARC
Page 551 and 552:
Page 553 and 554:
Page 555 and 556:
Page 557 and 558:
Page 559 and 560:
Page 561 and 562:
Page 563 and 564:
Page 565 and 566:
Page 567 and 568:
Page 569 and 570:
Page 571 and 572:
Page 573 and 574:
Page 575 and 576:
Page 578 and 579:
IndexSymbols2-Step Approach 1522-st
Page 580 and 581:
merging 228-234merging, based on da
Page 582 and 583:
Monte Carlo Simulationabout 408, 42
Page 584 and 585:
Python SimPy module 449QQuasi Monte
show all

Python for Finance

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?