Professional Documents
Culture Documents
‘Note: Page numbers followed by “f ” indicate figures and “t” indicate tables.’
577
578 Index
Autoregressive moving average Bottom up credit strategy, 14 Concave shape of market impact
(ARMA) (Continued ) Bottom-up approach, 99 function, 114e115
forecast intraday volumes profiles, Broker(s), 97, 405e408, 418, 543e544 Conjugate gradient step, 523
317e321, 319f broker-dealers, 24 Constant relative risk aversion (CRRA),
predicting remaining daily volume, models, 472 343e344
321e322 pretrade estimates, 554 Constraint description, 355e358
Autoregressive parameter, 315e316 trading desks, 16e17 objective function difficulty,
Average daily volumes (ADVs), 35, BroydeneFletchereGoldfarbeShanno 357
112, 234, 249, 254, 301e306, algorithm, 523 optimization objective function
305f, 309e310, 312, 431e432, Buy and hold investing, 503 simplification, 358
469, 484e486. See also Fore- Buy-initiated trade, 246 Continuous distribution functions
casting daily volumes (FDV) Buy-side firms, 520 chi-square distribution, 145
analysis, 304e305 Buy-side traders, 25 exponential distribution, 143e144
definitions, 303e304 log-normal distribution, 141
methodology, 303 C logistic distribution, 146e147
monthly volume forecasting model, Capital asset pricing model (CAPM), normal distribution, 137
304, 307t 246, 429e430 probability distribution function,
regression results, 306 Capital commitment. See Principal bid 132e135
Average return, 270 CAPM. See Capital asset pricing model standard normal distribution, 138
Average trade size, 35 (CAPM) student’s t-distribution, 139e140
Average trading cost component, 67 CART models, 228 triangular distribution, 148
Cash uniform distribution, 142
B deposit, 13 Convergence process, 524
Back-loaded profile. See Intraday dividend, 14 Conversions, 567
volume profile trading desk, 16 Convex shape of market impact
Back-testing, 470e471, 563 Cashflow trade, 13e14 function, 114
strategies, 487e490 cdf. See Cumulative density function Corporate action, 10
Balanced data sets, 96 (cdf) Corporate buy-back model, 565
Basket, 11e12 Chi-square distribution, 145 Corporations, 561
algorithms, 33, 349 Chi-square goodness of fit, 93e94 Correlation, 271
Benchmark index, 440e441 Classification analysis, 228e229, measures, 309
Benchmark price 529 Cost, 67, 115, 255
performance, 75e76 Classification based supervised learning, allocation method, 123e125
selection, 326e329 224 cost-adjusted frontier, 445e447
arrival, 326e327 Cluster analysis, 224e226, 529 costerisk optimization, 333
comparison, 329e345 cmf. See Cumulative mass function curves, 262, 469e470, 553, 564
future, 328e329 (cmf) index, 489e490
historical, 327e328 Co-location of servers, 52e53 measure, 66
Best execution, 58 CobbeDouglas production function, premium, 484
frontier, 448e449 200 Covariance, 271, 280e288
quest for, 451e456 Coefficient of variation, 42, 136 matrix, 287, 290, 298, 453e454
Beta, 273 Commission, 59 Cramer’s rule, 164, 172e173
investment allocation, 498 commission-based model, 45e46 Credit strategy, 14
Big data, 222 Commodities, 505 Cross sectional multi-factor model,
Binomial distribution, 149 Company news, 10 294e295
Black-box trading. See Algorithmic Company outlook, 10 Cross-sectional model, 296
trading Company specific risk. See Stock CRRA. See Constant relative risk
Block trading, 36, 432 specific risk aversion (CRRA)
desk. See Cash trading desk Competition, 504 Cumulative density function (cdf),
Bond(s), 14, 505 Completion, 355 130e131
index ETF, 506 Compliance managers, 545 Cumulative mass function (cmf),
Bootstrapping techniques, 211e212 Compliance officers, 561, 564 130
Index 579
Foreign exchange (FX), 501 rebate/liquidity trading, 52e54 scientific method, 235e245
trading cost, 506 Historical data and covariance, I-Star model, 99, 197, 380, 407e413,
Forward-looking view of portfolio 280e288 520
construction, 451 degrees of freedom, 283e288 broker cost estimates, 414t
Fractional linear regression model, 417 false relationships, 281e283 regression analysis using All Broker
Fractional polynomial model, 199 Historical look-back period, 310 MI Data, 415t
Fractional regression model, 152, 152f, Historical Moving Average (HMA), trading cost estimates, 409te410t
172e173, 199 274e275 Idiosyncratic risk. See Stock-specific
Front-loaded profile. See Intraday vol- Historical price benchmark, 327e328 risk
ume profile Historical trading cost, 489 Imbalance, 252e253
Funds, 547 Historical volatility, 269 size, 249
hedge, 28 HMA. See Historical Moving Average issues, 250
manager, 6e7 (HMA) Implementation, 5
mandates, 430e431 HMA-VIX adjustment model, 274, goal, 58e59
Futures, 497e498, 504e505 276e277, 390 types to investors, 19e20
price benchmark, 328e329 Hyperaggressive algorithm, 336 Implementation shortfall (IS), 67e74
views reflection, 431 Hypothesis testing, 205e206 algorithm, 33
FX. See Foreign exchange (FX) complete execution, 69
I expanded IS, 71e74
G I-Star market impact model, 120, formulation, 74e75
trading cost/arrival cost, 74e75
Game-playing algorithms, 223 234e235. See also Almgren and
Chriss market impact model opportunity cost, 70e71
GARCH. See Generalized auto-regres-
(AC market impact model) Implied volatility, 269
sive conditional heteroscedastic-
data set, 245e250 In-house market impact models,
ity (GARCH)
data definitions, 248 473e474
Gaussian distribution. See Continuous
estimating model parameters, 255e267 Independent samples, 89e91
probability distribution function
cost curves, 262 ManneWhitney U test, 89e91
Generalized auto-regressive conditional
sensitivity analysis, 257e261 Independent sampling approach, 85
heteroscedasticity (GARCH),
statistical analysis, 262e265 Index
274, 276, 390
stock-specific error analysis, arrival, 81
Generalized reduced gradient, 523
265e267 index-adjusted cost, 81
Global equity markets, 500e501
formulation, 121e122, 126e128 manager. See Passive manager
GPS, 223
model verification, 250e255 model, 292e293, 296
Group centroid, 225
ADV, 254 multi-index models, 293
Grouping data, 182, 257
analysis period, 252 single-index model, 292e293
annualized volatility, 254 reconstitution, 12
H cost, 255 research, 15
Heat-maps, 221e222 error analysis, 252 Indication of interests (IOIs), 52e53
Hedge funds, 28 first price, 254 Industry buzz words, 223
quant, 30 graphical illustration, 251 Industry event, 7
Hedging, 13 imbalance, 252e253 Inequality constraint matrix, 374
Heisenberg uncertainty principle number of data points, 252 Information
of finance, 100 POV rate, 255 content, 6e7
of trading, 490e491 regression analysis, 251 information-based rebalance, 12e13
Heteroskedasticity, 256 side of imbalance, 253 information-based trade, 6
HFT. See High frequency traders/trading size of imbalance, 254 information-less investing, 9
(HFT) stock universe, 252 leakage, 25
Hidden transaction costs, 65 time period, 252 set, 58
High frequency traders/trading (HFT), 4, turnover, 253 Initial parameter value, 524
24, 28e30, 50e54, 563 volume, 253 Input data, 176
AMM, 50e52 VWAP, 253 Instantaneous impact equation,
quant trading/statistical arbitrage, 52 z-score analysis, 251 121e122
Index 581
Market impact (MI), 61e63, 353, 439, convex shape, 114 Money
454, 513, 520, 547 linear shape, 114 management, 23e24
cost, 459e460, 470, 483 temporary market impact, 100 managers, 487e488
expression, 363, 380e382, 384e387, Marketable limit order, 44 Monte Carlo Techniques, 210e211,
388t Markowitz ETF, 436 449e450
for basket of stock, 382 MATLAB functions, 549 Month over month (MOM), 303e305
comparison, 388e395, 389f Matrix notation, 289e292 Moving median daily volume (MDV),
ETF, 391e392 Matrix techniques, 167e169 309e310, 312
forecasting covariance, 390e391 compute standard errors of b, 168 MSE. See Mean sum of square errors
I-Star, 380 estimate parameters, 167 (MSE)
portfolio trade cost objective func- F-statistic, 169 MSR. See Mean square regression (MSR)
tion, 394e395 log regression model, 169e171 Multi-asset class investing, 495e498
for single stock order, 380e382 log-linear transformation, 171 allocation schedule, 500t
single stock trade cost objective R2 statistic, 169 in beta exposure and other factors, 495
function, 393 Maximum likelihood estimation tech- equities, 495e496
factor scores, 469, 508e510 niques (MLE techniques), 180, ETF, 496e497
analysis, 510e512 183, 190, 198, 202e204, 275 example, 495e498
current state, 510 determining parameters via, 277e280 futures, 497e498
formula, 434 likelihood function, 203 Multi-asset trading costs, 498e506,
parameters, 353t, 418 Maximum trading opportunity, 398e399 507t. See also Trading costs
portfolio optimization model, 425 MDV. See Moving median daily bonds, 505
price trajectories, 108e109 volume (MDV) commodities, 505
simulation, 490e494 Mean, 135, 221e222 currency, 506
scenario, 491e494 mean-variance optimization, 350, 441 equities, 504
summary, 550e551 trading cost component, 67 equity market trading cost analysis,
Market impact models, 99e115, of probability model, 182 502t
301e303 Mean square regression (MSR), 155 ETF, 504
derivation of models, 115e120 Mean sum of square errors (MSE), 155 futures, 504e505
AC market impact model, 115e120 Median, 135 global equity markets, 500e501
I-Star market impact model, 120 test, 91e92 multi-asset classes, 501e506
developing, 109e110 MI. See Market impact (MI) observations, 504
essential properties, 110e111 Micro strategy, 20 room for improvement, 506
graphical illustrations, 101 Microadaptation tactics, 54 Multi-index model, 292e293
market impact price trajectories, Microlevel decisions, 48 Multiperiod
108e109 Microorder placement strategies, 50 optimization problem, 449e450
price trajectory, 101e102 Mind-maps, 221e222 trade schedule optimization problem,
supplyedemand equilibrium, Minimum trading risk quantity, 397 520e523
102e105, 103f Mispricing, 10e11 setting up problem, 520e521
temporary decay formulation, 108 liquidity, 503 trader’s dilemma objective function,
temporary impact decay function, MLE techniques. See Maximum likeli- 521e523
105e107 hood estimation techniques trade schedule optimization solution
model formulation, 120e128 (MLE techniques) times, 519
comparison of approaches, 128 Mode, 135 Multiple linear regression models,
cost allocation method, 123e125 Model 161e164. See also Simple linear
derivation of model, 122e123 driven, 12 regression models
I* formulation, 126e128 inputs, 376 calculate model performance statistics,
I-Star model, 121e122, 126e128 model-driven rebalance, 12e13 164
market impact equation, 122 performance measurement, 202, estimate model parameters, 162e164
permanent market impact, 100 278e279 example, 165e166
shape of market impact function, MOM. See Month over month (MOM) output, 166te167t
112e113 Momentum. See Price appreciation test for statistical significance of
concave shape, 114e115 Momentum cost, 339 factors, 164
Index 583
Portfolio construction (Continued ) transaction costs in, 434e436 in algorithmic trading, 129
with transaction cost analysis, 429, Portfolio risk management, 395e403 distributions, 130e136
449e456 maximum trading opportunity, continuous, 132e135
appropriate level of risk aversion, 398e399 descriptive statistics, 135e136
447e448 minimum trading risk quantity, 397 discrete, 131e132
best execution frontier, 448e449 program-block decomposition, functions, 136
cost-adjusted frontier, 445e447 400e403 Probability density function (pdf),
example, 456e466 residual risk curve, 395e396 130e131
findings, 464e466 values, 400 Probability mass function (pmf), 130
investment and trading theories, Posttrade analysis, 49e50, 66, 97, 455, Probability models, 175, 186
441e445 547, 555e558, 562, 565 comparison of linear regression model
portfolio management process, POV. See Percentage of volume (POV) to, 177, 178f
437e439 Power function, 176, 191f, developing, 175e184
in portfolio optimization and con- 194f grouping data, 182
straints, 430e436 grouped limit order data, 193t likelihood function based on actual
portfolio optimization and stock limit order data, 187te189t outcome results, 183e184
data, 457t logistic regression summary results, logit model, 178e179
quest for best execution frontier, 194t outcome data, 180e182
451e456 model, 177, 412 mean, 182
trading decision process, 439e441 Predictive analytics, 223e224 model formulation, 182
Portfolio management, 437e439, Predictive modeling. See Predictive variance, 182
484e487, 567 analytics power function model, 177
ETF with and without short positions, Pretrade probit model, 179e180
437e438, 438f analysis, 49, 65, 545, 562, 565 comparison of, 181f, 191e194
maximizing investor utility, 438e439, benchmark, 80 solving binary output models, 182
439f cost calculations, 473 solving probability output models,
Portfolio managers, 5e7, 9e10, 12e13, impact models, 473 184e185
57, 419, 421e423, 431, market impact parameters, 562 examples, 186e190
473e474, 488, 495, 510, 544, of pretrade approach, 247, 406e416, power function, 191
546, 560e561, 564 475e478, 513 specifying probability function,
decision, 439 applications, 476 182e183
and traders, 406, 433 examples, 476e478 Profit and loss (PnL), 67
Portfolio optimization, 416e429, I-star model approach, 407e413 algorithms, 24
449e450, 470 NNET model approach, 413e416, Profit seeking algorithms, 3e4
comparison of optimization techniques, 416f Program, 11e12
364e368, 366t Price program-block decomposition,
optimization solution time, 368t appreciation, 61, 353, 512, 548 400e403, 402f
solving portfolio objective problem, discovery, 27 sales trader, 18e19
365e368 evolution, 512 trading desk, 16e17
and constraints, 430e433 price-based algorithms, 34e35 Project Information Model tactics (PIM
deriving portfolio optimization market price-size priority, 46 tactics), 368e371, 370f
impact models, 419e420 price-time models, 46 PWP. See Participation-weighted price
portfolio managers, 419 return, 512 (PWP)
QP approach, 358e360 trajectory, 101e102 Python functions, 549
residual schedule exponential, trend, 512
361e362 volatility, 379 Q
share quantity regression model, Principal bid, 16e17, 19e20, 46e47 Quadratic programming (QP)
420e424, 422te423t Principal component analysis (PCA), approach, 358e360
trade schedule exponential, 360e361 295, 536. See also Transaction formulation, 350
trade value regression model, 425e428, cost analysis (TCA) subproblem, 523
426te427t Principal trade. See Principal bid trade schedule optimization technique,
trading rate parameter, 362e364 Probability 372e373
Index 585
Quant trading/statistical arbitrage, 50, 52 Registered investment advisors (RIAs), RMSE. See Root mean square error
Quantitative analysis with TCA. 28e29 (RMSE)
See also Transaction cost anal- Regression, 229e231 RMZSE. See Root mean Z-Score
ysis (TCA) analysis, 151, 251, 529 squared error (RMZSE)
alpha capture program, 470, 512e518 based supervised learning, 225 Robo” trading. See Algorithmic trading
backtesting, 470e471 fractional, 199 Robo-investing, 223
strategies, 487e490 metrics, 154e156 Robotics, 223
cost curves, 469e470 linear, 198 Roll cost, 504e505
existing models for portfolio construc- log-linear, 200 Root mean square error (RMSE),
tion, 472e475 logistic, 200e201 278e279
investment capacity, 470 models, 177, 198e201 Root mean Z-Score squared error
liquidation cost, 471 nonlinear, 201 (RMZSE), 278e279
market impact polynomial, 198e199 RPM. See Relative performance mea-
factor scores, 469, 508e510 Regression sum of squares (SSR), 155 sure (RPM)
simulation, 490e494 Reinforcement learning, 223, 225
multi-asset class investing, 495e498 Relative performance measure (RPM), S
multi-asset trading costs, 498e506 75, 78e80 Sales function, 17e18
portfolio optimization, 470 Remaining daily volume (RDV), Sampling techniques, 207e208
pretrade of pretrades, 475e478 321e322 random, 208e209
quantitative overlays, 469 Research, 564e565 with replacement, 209
sensitivity analysis, 471 data of investment analysis, 15e16 without replacement, 210
trading costs, 478e487 function, 17e18 Scientific method, 233,
Quantitative analysts, 18, 561 Research sales. See Equity sales 235e245
Quantitative asset managers (Asset Residual risk curve, 395e396, 396f, asking question, 235
Mgmt-Quant), 30 399f communicating experimental results,
Quantitative fund, 8 Residual schedule, 355 236e245
Quantitative managers, 8, 12, 488 exponential, 361e362, 522 solution technique, 237e245
Quantitative portfolio managers, 16e17, Residual size bounds, 356 constructing hypothesis, 235e236
437 Residual trade schedule, 519 researching problem, 235
Quantitative process, 429e430 Retail investors, 28 testing hypothesis, 236
Quantitative research, 15 Returns, 274, 452e453 Screening techniques, 484e487
Quest for best execution frontier, Revenue pricing models, 45e46 Second wave of portfolio optimization,
451e456 Reverse engineering broker models, 434
return, 452e453 405e406 Self-financing, 356
risk, 453e456 portfolio optimization, 416e428 Sell-initiated trade, 246
Quicker trading, 124 pre-trade of pre-trades, 406e416 Sell-side firms, 520
RIAs. See Registered investment advi- Semisupervised learning, 225,
R sors (RIAs) 528e529
R2 Risk, 453e456 Sensitivity analysis, 257e261, 349, 471,
goodness of fit, 159, 164 aversion, 365, 455 545, 552f
statistic, 169, 239 appropriate level of, 447e448 Sequential quadratic programming, 523
R2000 index, 407e408 parameter, 33e34 Shadow liquidity, 503
Random forests, 228 factor, 292 Share quantity regression model,
Random sampling techniques, 208e209 hedging, 503 420e424, 422te423t,
Random variables, 129e130 idiosyncratic risk component, 292 469e470
Range, 273 management, 13, 356e357, 431 Sharpe ratio of trade, 447e448
RDV. See Remaining daily volume trade, 13 Short trading horizons, 50
(RDV) market, 292 Short-term alpha, 10
Realized volatility, 269 reduction, 13 Shrinking portfolio, 355
Rebates, 60 stock specific, 292 Side indicator function, 116
rebate/liquidity trading, 50, 52e54 systematic risk component, 292 Side of imbalance, 253
Redemption, 13 Risk bid. See Principal bid Sign test, 86e87
586 Index
Simple linear regression models, Steepest descent, 523 TR. See Timing risk (TR)
156e158. See also Multiple Stepwise regression analysis, 536e538 Trade, 6
linear regression models Stock cost distribution graph, 552e553
example, 159e161 stock-specific list, 11e13
model parameters estimation, 157e158 error analysis, 265e267 rate, 234, 519, 534
model performance statistics risk, 111, 292 bounds, 356
evaluation, 158 universe, 252 schedule, 32e33, 234, 356, 378
output, 161t volatility, 249, 482 exponential, 360e361, 522
test for statistical significance of Stock-specific risk, 292 optimization process, 349, 447e448
factors, 159 Structured data, 222e223 size bounds, 356
Simple mean. See Arithmetic mean Student’s t-distribution, 139e140 strategy, 519
Simply trading PnL. See Trading price Suboptimal Markowitzian portfolio, 448 trajectory, 32e33
performance Sum of squared X, 155 value regression model, 425e428,
Single factor regression model, 246 Sum of squared XY, 156 426te427t
Single optimal trading strategy, 445 Sum of squared Y (SSY), 156 volume, 503
Single stock, 16 Supervised learning, 224e225, Trade cost analysis (TCA), 405e406
algorithms, 31 528e529 TradeMetrics
market impact Supplyedemand equilibrium, 102e105, broker pretrade estimates, 555f
analysis, 551 103f single stock
simulation, 494, 494f Systematic risk component, 292 cost curves, 554f
trade cost objective function, 393 market impact analysis, 551f
Single-index model, 292e293 T trade algorithm, 550e551, 550f
Size of imbalance, 254 t-distribution. See Student’s Traders, 25, 57, 544, 560
Skewness, 136, 175, 221e222 t-distribution dilemma, 63, 351e352
Slicing strategy, 349 T-test, 159, 164e165 objective function, 521e523
Small cap stocks, 38e39, 240, 308 Tag-clouds, 221e222 variables, 351e352
Smart order router/routing (SOR), Taker-maker pricing model, 45 Trading
20e21, 48e49, 54, 564 Taxes, 59 algorithm
Solution technique, 237e245 TCA. See Trade cost analysis (TCA); classifications, 3e4
SOR. See Smart order router/routing Transaction cost analysis (TCA) styles, 4
(SOR) Temporary decay formulation, 108 characteristics, 479t
SP500 index, 292, 407e408 Temporary impact decay function, decisions, 432
Specified tolerance value, 523e524 105e107, 107t optimal strategy, 440e441
Speculation, 503 Temporary market impact, 99e100 process, 439e441
Spreads, 60, 111, 241 cost, 123 desks, 16
SSE. See Error sum of squares (SSE) Third wave of portfolio optimization, floor function, 18e19
SSR. See Regression sum of squares (SSR) 435e436 goals, 330e337
Standard deviation, 135, 221e222 Third-party balance tradeoff between cost and
Standard error of regression model, consultants, 97 risk, 333
159, 164 TCA systems, 544 minimizing cost, 330
Standard normal distribution, 138 TIC data. See Treasury international minimizing cost with risk constraint,
Starting solution, 524 capital data (TIC data) 331
Static trade schedule algorithm, 86 Time period, 252 minimizing risk with cost constraint,
Static trading parameters, 84e85 Time weighted average price algorithm 333
Statistical/statistics (TWAP algorithm), 32 price improvement, 334
in algorithmic trading, 129 Timing risk (TR), 62, 354, 379, 392, liquidity, 503
analysis, 262e265 439, 442e443, 454, 520, 548 price performance, 75
error analysis, 262e265 for basket of stock, 387e388 rate, 377, 522
factor models, 295e299 equation, 382e388 comparison of POV rate to, 378
factor selection, 297e299 expression, 363e364 variable, 377
models, 52 Top down credit strategy approach, 14 rate parameter, 362e364
tests, 95 Total sum of squares, 155 market impact expression, 363
Index 587
Volume, 253 Volume-weighted average price Wilcoxon signed rank test, 87e89
forecasting techniques, 301 (VWAP), 58, 246, 253, 550e551 WilcoxoneManneWhitney ranks test, 95
ADV, 303e306 algorithm, 32 Working order of algorithm, 4, 31
market impact model, 301e303 benchmark, 76e77
observations over 19-year period, frontier, 447 Y
306e308 strategy, 389, 432e433, 440, Yield curve strategy, 14
observations over 3-year period, 445, 477, 486
308e309
W Z
and stock price correlation, 309
Wagner’s IS. See Expanded IS Z-score
variance, 379
“What-if” analysis, 473e474 analysis, 251
evaluation metric, 81e83