Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Algorithmic Trading Methods: Applications Using

Advanced Statistics, Optimization, and Machine
Learning Techniques 2nd Edition Robert Kissell

https://textbookfull.com/product/algorithmic-trading-
methods-applications-using-advanced-statistics-optimization-
and-machine-learning-techniques-2nd-edition-robert-kissell/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-loucas/

textboxfull.com

Machine Learning for Algorithmic Trading 2nd Edition

Stefan Jansen

https://textbookfull.com/product/machine-learning-for-algorithmic-
trading-2nd-edition-stefan-jansen/

textboxfull.com

Machine Learning for Algorithmic Trading 2nd Edition

Stefan Jansen

https://textbookfull.com/product/machine-learning-for-algorithmic-
trading-2nd-edition-stefan-jansen-2/

textboxfull.com

Machine Learning for Algorithmic Trading 2nd Edition

Stefan Jansen

https://textbookfull.com/product/machine-learning-for-algorithmic-
trading-2nd-edition-stefan-jansen-3/

textboxfull.com
Optimal Sports Math Statistics and Fantasy Robert Kissell
And Jim Poserina (Auth.)

https://textbookfull.com/product/optimal-sports-math-statistics-and-
fantasy-robert-kissell-and-jim-poserina-auth/

textboxfull.com

Machine Learning: A Bayesian and Optimization Perspective

2nd Edition Sergios Theodoridis

https://textbookfull.com/product/machine-learning-a-bayesian-and-
optimization-perspective-2nd-edition-sergios-theodoridis/

textboxfull.com

Python for Probability, Statistics, and Machine Learning

2nd Edition José Unpingco

https://textbookfull.com/product/python-for-probability-statistics-
and-machine-learning-2nd-edition-jose-unpingco/

textboxfull.com

Advanced Data Analytics Using Python: With Machine

Learning, Deep Learning and NLP Examples Mukhopadhyay

https://textbookfull.com/product/advanced-data-analytics-using-python-
with-machine-learning-deep-learning-and-nlp-examples-mukhopadhyay/

textboxfull.com

Algorithmic Aspects of Machine Learning 1st Edition Ankur

Moitra

https://textbookfull.com/product/algorithmic-aspects-of-machine-
learning-1st-edition-ankur-moitra/

textboxfull.com
Algorithmic Trading Methods
This page intentionally left blank
 Algorithmic Trading Methods
Applications using Advanced Statistics,
Optimization, and Machine Learning Techniques
Second Edition

Robert Kissell, Ph.D.

President
Kissell Research Group and Adjunct Faculty Member
Gabelli School of Business, Fordham University
Manhasset, NY, United States
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2021 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright
Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional practices,
or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein.
In using such information or methods they should be mindful of their own safety and the safety
of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
ISBN: 978-0-12-815630-8

For information on all Academic Press publications visit our website at

https://www.elsevier.com/books-and-journals

Publisher: Brian Romer

Editorial Project Manager: Gabrielle Vincent
Production Project Manager: Paul Prasad Chandramohan
Cover Designer: Matthew Limbert
Typeset by TNQ Technologies
Contents
Preface............................................................................................................................. xix
Acknowledgments........................................................................................................... xxi

CHAPTER 1 Introduction ................................................................................... 1

What is Electronic Trading? ......................................................................... 2
What is Algorithmic Trading? ...................................................................... 3
Trading Algorithm Classifications ................................................................ 3
Trading Algorithm Styles ............................................................................. 4
Investment Cycle .......................................................................................... 5
Investment Objective .................................................................................... 5
Information Content...................................................................................... 6
Investment Styles .......................................................................................... 7
Investment Strategies .................................................................................... 9
Research Data ............................................................................................. 15
Broker Trading Desks................................................................................. 16
Research Function....................................................................................... 17
Sales Function............................................................................................. 18
Implementation Types................................................................................. 19
Algorithmic Decision-Making Process ....................................................... 20
CHAPTER 2 Algorithmic Trading ....................................................................... 23
Advantages.................................................................................................. 25
Disadvantages ............................................................................................. 26
Growth in Algorithmic Trading.................................................................. 27
Market Participants ..................................................................................... 28
Classifications of Algorithms...................................................................... 31
Types of Algorithms ................................................................................... 31
Algorithmic Trading Trends ....................................................................... 35
Day of Week Effect .................................................................................... 37
Intraday Trading Profiles ............................................................................ 39
Trading Venue Classification...................................................................... 42
Displayed Market ................................................................................... 42
Dark Pool ............................................................................................... 42
Dark Pool Controversies ........................................................................ 43
Types of Orders .......................................................................................... 44
Revenue Pricing Models............................................................................. 45
Order Priority ......................................................................................... 46
Execution Options....................................................................................... 46

v
vi Contents

Algorithmic Trading Decisions................................................................... 47

Macro Level Strategies .......................................................................... 47
Micro Level Decisions ........................................................................... 48
Limit Order Models ............................................................................... 48
Smart Order Routers .............................................................................. 48
Algorithmic Analysis Tools........................................................................ 49
Pre-Trade Analysis................................................................................. 49
Intraday Analysis ................................................................................... 49
Post-Trade Analysis ............................................................................... 49
High Frequency Trading............................................................................. 50
Auto Market Making.............................................................................. 50
Quantitative Trading/Statistical Arbitrage.............................................. 52
Rebate/Liquidity Trading ....................................................................... 52
Direct Market Access.................................................................................. 54
CHAPTER 3 Transaction Costs .......................................................................... 57
What are Transaction Costs? ...................................................................... 57
What is Best Execution?............................................................................. 58
What is the Goal of Implementation?......................................................... 58
Unbundled Transaction Cost Components ................................................. 59
Commission............................................................................................ 59
Fees ........................................................................................................ 59
Taxes ...................................................................................................... 59
Rebates ................................................................................................... 60
Spreads ................................................................................................... 60
Delay Cost.............................................................................................. 60
Price Appreciation.................................................................................. 61
Market Impact ........................................................................................ 61
Timing Risk............................................................................................ 62
Opportunity Cost.................................................................................... 62
Transaction Cost Classification................................................................... 62
Transaction Cost Categorization................................................................. 65
Transaction Cost Analysis .......................................................................... 65
Measuring/Forecasting ........................................................................... 66
Cost vs. Profit and Loss ......................................................................... 67
Implementation Shortfall ............................................................................ 67
Complete Execution ............................................................................... 69
Opportunity Cost (Andre Perold)........................................................... 70
Expanded Implementation Shortfall (Wayne Wagner).......................... 71
 Contents vii

Implementation Shortfall Formulation........................................................ 74

Trading Cost/Arrival Cost...................................................................... 74
Evaluating Performance .............................................................................. 75
Trading Price Performance..................................................................... 75
Benchmark Price Performance............................................................... 75
VWAP Benchmark................................................................................. 76
Participation-Weighted Price Benchmark .............................................. 77
Relative Performance Measure .............................................................. 78
Pretrade Benchmark ............................................................................... 80
Index-Adjusted Performance Metric ...................................................... 80
Z-Score Evaluation Metric ..................................................................... 81
Market Cost-Adjusted Z-Score .............................................................. 83
Adaptation Tactic ................................................................................... 83
Comparing Algorithms ............................................................................... 84
Nonparametric Tests .............................................................................. 85
Paired Samples ....................................................................................... 86
Sign Test ................................................................................................ 86
Wilcoxon Signed Rank Test .................................................................. 87
Independent Samples .................................................................................. 89
ManneWhitney U Test.......................................................................... 89
Median Test ................................................................................................ 91
Distribution Analysis .................................................................................. 92
Chi-Square Goodness of Fit ....................................................................... 93
KolmogoroveSmirnov Goodness of Fit..................................................... 94
Experimental Design................................................................................... 95
Proper Statistical Tests ........................................................................... 95
Small Sample Size ................................................................................. 95
Data Ties ................................................................................................ 96
Proper Categorization............................................................................. 96
Balanced Data Sets................................................................................. 96
Final Note on Posttrade Analysis ............................................................... 97
CHAPTER 4 Market Impact Models ................................................................... 99
Introduction ................................................................................................. 99
Definition .................................................................................................... 99
Example 1: Temporary Market Impact................................................ 100
Example 2: Permanent Market Impact ................................................ 100
Graphical Illustrations of Market Impact............................................. 101
Illustration #1: Price Trajectory ........................................................... 101
Illustration #2: SupplyeDemand Equilibrium..................................... 102
viii Contents

Illustration #3: Temporary Impact Decay Function............................. 105

Example #3: Temporary Decay Formulation ...................................... 108
Illustration #4: Various Market Impact Price Trajectories .................. 108
Developing a Market Impact Model .................................................... 109
Essential Properties of a Market Impact Model .................................. 110
The Shape of the Market Impact Function .......................................... 112
Example: Convex Shape ...................................................................... 114
Example: Linear Shape ........................................................................ 114
Example: Concave Shape..................................................................... 114
Derivation of Models ................................................................................ 115
Almgren and Chriss Market Impact Model ......................................... 115
I-Star Market Impact Model ..................................................................... 120
Model Formulation ................................................................................... 120
I-Star: Instantaneous Impact Equation ................................................. 121
The Market Impact Equation ............................................................... 122
Derivation of the Model....................................................................... 122
Cost Allocation Method ....................................................................... 123
I* Formulation...................................................................................... 126
Comparison of Approaches.................................................................. 128
CHAPTER 5 Probability and Statistics .............................................................. 129
Introduction ............................................................................................... 129
Random Variables..................................................................................... 129
Probability Distributions ........................................................................... 130
Example: Discrete Probability Distribution Function.......................... 131
Example: Continuous Probability Distribution Function..................... 132
Descriptive Statistics ............................................................................ 135
Probability Distribution Functions............................................................ 136
Continuous Distribution Functions ........................................................... 137
Normal Distribution ............................................................................. 137
Standard Normal Distribution .............................................................. 138
Student’s t-Distribution ........................................................................ 139
Log-Normal Distribution...................................................................... 141
Uniform Distribution............................................................................ 142
Exponential Distribution ...................................................................... 143
Chi-Square Distribution ....................................................................... 145
Logistic Distribution ............................................................................ 146
Triangular Distribution......................................................................... 148
Discrete Distributions................................................................................ 149
Binomial Distribution........................................................................... 149
Poisson Distribution ............................................................................. 150
End Notes.................................................................................................. 150
 Contents ix

CHAPTER 6 Linear Regression Models ............................................................. 151

Introduction ............................................................................................... 151
Linear Regression Requirements ......................................................... 154
Regression Metrics............................................................................... 154
Linear Regression ..................................................................................... 156
True Linear Regression Model ............................................................ 156
Simple Linear Regression Model......................................................... 156
Solving the Simple Linear Regression Model ..................................... 157
Standard Error of the Regression Model ............................................. 159
R2 Goodness of Fit............................................................................... 159
T-test: Hypothesis Test......................................................................... 159
F-test: Hypothesis Test ........................................................................ 159
Example: Simple Linear Regression.................................................... 159
Multiple Linear Regression Model ...................................................... 161
Solving the Multiple Linear Regression Model................................... 162
Standard Error of the Regression Model ............................................. 164
R2 Goodness of Fit............................................................................... 164
T-test: Hypothesis Test......................................................................... 164
F-test: Hypothesis Test ........................................................................ 165
Example: Multiple Linear Regression ................................................. 165
Matrix Techniques .................................................................................... 167
Estimate Parameters ............................................................................. 167
Compute Standard Errors of b ............................................................. 168
R2 Statistic ............................................................................................ 169
F-Statistic ............................................................................................. 169
Log Regression Model.............................................................................. 169
Example: Log-Transformation ............................................................. 169
Example: Log-Linear Transformation.................................................. 171
Polynomial Regression Model.................................................................. 171
Fractional Regression Model .................................................................... 172
CHAPTER 7 Probability Models....................................................................... 175
Introduction ............................................................................................... 175
Developing a Probability Model............................................................... 175
Comparison of Linear Regression Model to Probability Model ......... 177
Power Function Model......................................................................... 177
Logit Model.......................................................................................... 178
Probit Model ........................................................................................ 179
Outcome Data....................................................................................... 180
Grouping Data...................................................................................... 182
x Contents

Solving Binary Output Models ............................................................ 182

Step 1: Specify Probability Function ................................................... 182
Step 2: Set Up a Likelihood Function Based on Actual Outcome
Results for all Observations. For Example, If We Have n
Observations, the Loss Function is Specified as Follows................... 183
Solving Probability Output Models .......................................................... 184
Examples ................................................................................................... 186
Example 7.1 Power Function............................................................... 186
Example 7.2 Logit Model .................................................................... 190
Comparison of Power Function to Logit Model ...................................... 191
Example 7.3 Logistic Regression ........................................................ 191
Conclusions............................................................................................... 195
CHAPTER 8 Nonlinear Regression Models......................................................... 197
Introduction ............................................................................................... 197
Regression Models.................................................................................... 198
Linear Regression Model ..................................................................... 198
Polynomial Regression Model ............................................................. 198
Fractional Regression Model ............................................................... 199
Log-linear Regression Model............................................................... 200
Logistic Regression Model .................................................................. 200
Nonlinear Model .................................................................................. 201
Nonlinear Formulation .............................................................................. 201
Solving Nonlinear Regression Model....................................................... 202
Estimating Parameters............................................................................... 202
Maximum Likelihood Estimation (MLE) ............................................ 202
Nonlinear Least Squares (Non-OLS)........................................................ 204
Step I: Define the Model...................................................................... 204
Step II: Define the Error Term............................................................. 204
Step III: Define a Loss FunctiondSum of Square Errors................... 205
Step IV: Minimize the Sum of Square Error ....................................... 205
Hypothesis Testing.................................................................................... 205
Evaluate Model Performance.................................................................... 206
Sampling Techniques................................................................................ 207
Random Sampling..................................................................................... 208
Sampling With Replacement .................................................................... 209
Sampling Without Replacement ............................................................... 210
Monte Carlo Simulation............................................................................ 210
Bootstrapping Techniques......................................................................... 211
Jackknife Sampling Techniques ............................................................... 212
Important Notes on Sampling in Nonlinear Regression Models......... 218
 Contents xi

CHAPTER 9 Machine Learning Techniques........................................................ 221

Introduction ............................................................................................... 221
Types of Machine Learning...................................................................... 224
Examples ................................................................................................... 225
Cluster Analysis ................................................................................... 225
Classification ............................................................................................. 228
Regression................................................................................................. 229
Neural Networks ....................................................................................... 231
CHAPTER 10 Estimating I-Star Market Impact Model Parameters .......................... 233
Introduction ............................................................................................... 233
I-Star Market Impact Model ..................................................................... 234
Scientific Method ...................................................................................... 235
Step 1: Ask a Question ........................................................................ 235
Step 2: Research the Problem .............................................................. 235
Step 3: Construct a Hypothesis............................................................ 235
Step 4: Test the Hypothesis ................................................................. 236
Step 6: Conclusions Communicate ...................................................... 236
Underlying Data Set............................................................................. 245
Data Definitions ................................................................................... 248
Imbalance/Order Size ........................................................................... 249
Average daily volume .......................................................................... 249
Actual market volume.......................................................................... 249
Stock volatility ..................................................................................... 249
POV Rate ............................................................................................. 249
Arrival Cost.......................................................................................... 250
Model Verification ............................................................................... 250
Model Verification #1: Graphical Illustration...................................... 251
Model Verification #2: Regression Analysis ....................................... 251
Model Verification #3: z-Score Analysis............................................. 251
Model Verification #4: Error Analysis ................................................ 252
Stock Universe ..................................................................................... 252
Analysis Period .................................................................................... 252
Time Period.......................................................................................... 252
Number of Data Points ........................................................................ 252
Imbalance ............................................................................................. 252
Side....................................................................................................... 253
Volume ................................................................................................. 253
Turnover ............................................................................................... 253
VWAP .................................................................................................. 253
xii Contents

First Price ............................................................................................. 254

Average Daily Volume ........................................................................ 254
Annualized Volatility ........................................................................... 254
Size ....................................................................................................... 254
POV Rate ............................................................................................. 255
Cost ...................................................................................................... 255
Estimating Model Parameters .............................................................. 255
Sensitivity Analysis.............................................................................. 257
Cost Curves .......................................................................................... 262
Statistical Analysis ............................................................................... 262
Stock-Specific Error Analysis .............................................................. 265
CHAPTER 11 Risk, Volatility, and Factor Models ................................................. 269
Introduction ............................................................................................... 269
Volatility Measures ................................................................................... 270
Log-Returns.......................................................................................... 270
Average Return .................................................................................... 270
Variance ............................................................................................... 270
Volatility............................................................................................... 271
Covariance............................................................................................ 271
Correlation............................................................................................ 271
Dispersion............................................................................................. 271
Value-at-Risk........................................................................................ 272
Implied Volatility ...................................................................................... 272
Beta ...................................................................................................... 273
Range.................................................................................................... 273
Forecasting Stock Volatility ..................................................................... 274
Volatility Models ................................................................................. 274
Determining Parameters via Maximum Likelihood Estimation .......... 277
Historical Data and Covariance ................................................................ 280
False Relationships............................................................................... 281
Degrees of Freedom ............................................................................. 283
Factor Models ........................................................................................... 288
Matrix Notation.................................................................................... 289
Factor Model in Matrix Notation......................................................... 290
Types of Factor Models ............................................................................ 292
Index Model ......................................................................................... 292
Macroeconomic Factor Models............................................................ 293
Statistical Factor Models...................................................................... 295
 Contents xiii

CHAPTER 12 Volume Forecasting Techniques ..................................................... 301

Introduction ............................................................................................... 301
Market Impact Model ............................................................................... 301
Average Daily Volume ............................................................................. 303
Methodology ........................................................................................ 303
Definitions ............................................................................................ 303
Monthly Volume Forecasting Model................................................... 304
Analysis................................................................................................ 304
Regression Results ............................................................................... 306
Observations Over the 19-Year Period: 2000e18 ................................... 306
Observations Over the Most Recent 3-Year Period: 2016e18 ................ 308
Volumes and Stock Price Correlation.................................................. 309
Forecasting Daily Volumes....................................................................... 309
Our Daily Volume Forecasting Analysis is as Follows....................... 310
Variable Notation ................................................................................. 311
ARMA Daily Forecasting Model......................................................... 311
Analysis Goal ....................................................................................... 311
Step 1. Determine Which is More Appropriate: ADV or MDV
and the Historical Look-Back Number of Days ................................. 312
Conclusion #1 ...................................................................................... 312
Step 2. Estimate the DayOfWeek(t) Parameter .................................... 314
Conclusion #2 ...................................................................................... 314
b
Step 3. Estimate the Autoregressive Parameter b................................ 315
Forecast Improvements ........................................................................ 316
Daily Volume Forecasting Model........................................................ 316
Conclusion #3 ...................................................................................... 316
Forecasting Intraday Volumes Profiles ................................................ 317
Forecasting Intraday Volume Profiles.................................................. 320
Predicting Remaining Daily Volume................................................... 321
CHAPTER 13 Algorithmic Decision-Making Framework ......................................... 323
Introduction ............................................................................................... 323
Equations................................................................................................... 324
Variables............................................................................................... 324
Important Equations ............................................................................. 325
Algorithmic Decision-Making Framework............................................... 325
Select Benchmark Price ....................................................................... 326
Comparison of Benchmark Prices ............................................................ 329
Specify Trading Goal ........................................................................... 330
Specify Adaptation Tactic.................................................................... 337
xiv Contents

Projected Cost ...................................................................................... 338

Comparison Across Adaptation Tactics ................................................... 345
Modified Adaptation Tactics..................................................................... 346
How Often Should we Reoptimization Our Tactic?............................ 347
CHAPTER 14 Portfolio Algorithms and Trade Schedule Optimization ...................... 349
Introduction ............................................................................................... 349
Trader’s Dilemma ..................................................................................... 351
Variables............................................................................................... 351
Transaction Cost Equations ...................................................................... 352
Market Impact ...................................................................................... 353
Price Appreciation................................................................................ 353
Timing Risk.......................................................................................... 354
One-Sided Optimization Problem ........................................................ 354
Optimization Formulation......................................................................... 354
Constraint Description.......................................................................... 355
Portfolio Optimization Techniques ........................................................... 358
Quadratic Programming Approach ...................................................... 358
Trade Schedule Exponential ................................................................ 360
Residual Schedule Exponential............................................................ 361
Trading Rate Parameter........................................................................ 362
Comparison of Optimization Techniques ............................................ 364
Portfolio Adaptation Tactics ..................................................................... 368
Description of AIM and PIM for Portfolio Trading ............................ 370
How Often Should we Reoptimize? .................................................... 371
Appendix .............................................................................................. 372
CHAPTER 15 Advanced Algorithmic Modeling Techniques ..................................... 375
Introduction ............................................................................................... 375
Trading Cost Equations ............................................................................ 375
Model Inputs ........................................................................................ 376
Trading Strategy................................................................................... 376
Percentage of Volume .......................................................................... 377
Trading Rate......................................................................................... 377
Trade Schedule..................................................................................... 378
Comparison of POV Rate to Trade Rate ............................................. 378
Trading Time ............................................................................................ 378
Trading Risk Components ........................................................................ 379
Trading Cost ModelsdReformulated.................................................. 380
Market Impact Expression ................................................................... 380
Timing Risk Equation .......................................................................... 382
 Contents xv

Derivation of the 1/3 Factor................................................................. 384

Timing Risk For a Basket of Stock ..................................................... 387
Comparison of Market Impact Estimates............................................. 388
Forecasting Covariance ........................................................................ 390
Efficient Trading Frontier .................................................................... 391
Single Stock Trade Cost Objective Function....................................... 393
Portfolio Trade Cost Objective Function ............................................. 394
Managing Portfolio Risk...................................................................... 395
Residual Risk Curve ............................................................................ 395
Minimum Trading Risk Quantity......................................................... 397
Maximum Trading Opportunity........................................................... 398
When to Use These Criteria? ............................................................... 400
Program-Block Decomposition............................................................ 400
CHAPTER 16 Decoding and Reverse Engineering Broker Models with Machine
Learning Techniques .................................................................... 405
Introduction ............................................................................................... 405
Pre-Trade of Pre-Trades............................................................................ 406
I-Star Model Approach ........................................................................ 407
Neural Network Model Approach ....................................................... 413
Portfolio Optimization .............................................................................. 416
What Should the Portfolio Manager Do? ............................................ 419
Deriving Portfolio Optimization Market Impact Models .................... 419
Example: Share Quantity Regression Model....................................... 420
Example: Trade Value Regression Model ........................................... 425
CHAPTER 17 Portfolio Construction with Transaction Cost Analysis ........................ 429
Introduction ............................................................................................... 429
Portfolio Optimization and Constraints .................................................... 430
Transaction Costs in Portfolio Optimization ............................................ 434
Portfolio Management Process ................................................................. 437
Example: Efficient Trading Frontier With and Without Short
Positions .............................................................................................. 437
Example: Maximizing Investor Utility ................................................ 438
Trading Decision Process ......................................................................... 439
What is the Appropriate Optimal Strategy to Use? ............................. 440
Unifying the Investment and Trading Theories........................................ 441
Which Execution Strategy Should the Trader Use? ............................ 443
Cost-Adjusted Frontier.............................................................................. 445
Determining the Appropriate Level of Risk Aversion ............................. 447
Best Execution Frontier ............................................................................ 448
xvi Contents

Portfolio Construction with Transaction Costs......................................... 449

Quest for Best Execution Frontier ....................................................... 451
Example .................................................................................................... 456
Important Findings ............................................................................... 464
Conclusion ................................................................................................ 466
CHAPTER 18 Quantitative Analysis with TCA ...................................................... 469
Introduction ............................................................................................... 469
Quantitative Overlays........................................................................... 469
Market Impact Factor Scores ............................................................... 469
Cost Curves .......................................................................................... 469
Alpha Capture ...................................................................................... 470
Investment Capacity............................................................................. 470
Portfolio Optimization.......................................................................... 470
Backtesting ........................................................................................... 470
Liquidation Cost................................................................................... 471
Sensitivity Analysis.............................................................................. 471
Are The Existing Models Useful Enough For Portfolio
Construction?....................................................................................... 472
Current State of Vendor Market Impact Models ................................. 473
Pretrade of Pretrades ................................................................................. 475
Applications ......................................................................................... 476
Example #1 .......................................................................................... 476
Example #2 .......................................................................................... 477
Example #3 .......................................................................................... 477
Example #4 .......................................................................................... 477
How Expensive is it to Trade? ................................................................. 478
Acquisition and Liquidation Costs....................................................... 481
Portfolio ManagementdScreening Techniques .................................. 484
Backtesting Strategies ............................................................................... 487
Market Impact Simulation ........................................................................ 490
Simulation Scenario ............................................................................. 491
Multi-Asset Class Investing ...................................................................... 495
Investing in Beta Exposure and Other Factors .................................... 495
Example #5 .......................................................................................... 495
Multi-Asset Trading Costs........................................................................ 498
Global Equity Markets ......................................................................... 500
Multi-Asset Classes.............................................................................. 501
Market Impact Factor Scores .................................................................... 508
Current State of Market Impact Factor Scores .................................... 510
 Contents xvii

Market Impact Factor Score Analysis ...................................................... 510

Alpha Capture Program ............................................................................ 512
Example #6 .......................................................................................... 513
Example #7 .......................................................................................... 514
Alpha Capture Curves.......................................................................... 516
CHAPTER 19 Machine Learning and Trade Schedule Optimization.......................... 519
Introduction ............................................................................................... 519
Multiperiod Trade Schedule Optimization Problem................................. 520
Setting up the Problem......................................................................... 520
Trader’s Dilemma Objective Function................................................. 521
Nonlinear Optimization Convergence ...................................................... 523
Newton’s Method................................................................................. 525
Example #1 .......................................................................................... 526
Example #2 .......................................................................................... 527
Machine Learning ..................................................................................... 528
Neural Networks .................................................................................. 530
Neural Network Errors......................................................................... 531
Machine Learning Training Experiment................................................... 531
Step I: Generating Simulated Trade Baskets ....................................... 532
Step II: Compile Stock and Basket Data Statistics.............................. 532
Step III: Solve the Multiperiod Trade Schedule Optimization
Problem ............................................................................................... 534
Step IV: Train the NNET..................................................................... 535
Step V. Calculate the Initial Parameter Values for the NNET ............ 536
Principal Component Analysis............................................................. 536
Stepwise Regression Analysis.............................................................. 536
Neural Network Structure .................................................................... 539
Neural Network Error .......................................................................... 539
Performance Results ................................................................................. 540
Conclusions............................................................................................... 542
CHAPTER 20 TCA Analysis Using MATLAB, Excel, and Python ................................ 543
Introduction ............................................................................................... 543
Transaction Cost Analysis Functions ....................................................... 545
Transaction Cost Model............................................................................ 547
MATLAB Functions................................................................................. 549
Excel and Python Functions ..................................................................... 549
TCA Report Examples.............................................................................. 550
Conclusion ................................................................................................ 558
xviii Contents

CHAPTER 21 Transaction Cost Analysis (TCA) Library ........................................... 559

Introduction ............................................................................................... 559
TCA Library......................................................................................... 560
Transaction Cost Analysis Using the TCA Library ................................. 561
List of TCA Functions ......................................................................... 565

Bibliography.................................................................................................................... 569
Index ............................................................................................................................... 577
Preface
Any intelligent fool can make things bigger and more complex. It takes a touch of
genius — and a lot of courage to move in the opposite direction.
Albert Einstein

“Algorithmic Trading Methods: Applications using Advanced Statistics, Optimization, and

Machine Learning Techniques,” Second Edition is a sequel to “The Science of Algorithmic
Trading & Portfolio Management.” This book greatly expands the concepts, foundations,
methodology, and models from the first edition, and it provides new insight into Algorithmic
Trading and Transaction Cost Analysis (TCA) using advanced mathematical techniques,
statistics, optimization, machine learning, neural networks, and predictive analytics.

Algorithmic Trading Methods provides traders, portfolio managers, analysts, students,

practitioners, and financial executives with an overview of the electronic trading environment,
and insight into how algorithms can be utilized to improve execution quality, fund
performance, and portfolio construction.

We provide a discussion of the current state of the market and advanced modeling techniques
for trading algorithms, stock selection and portfolio construction.

This reference book will provide readers with:

Insight into the new electronic trading environment.

Overview of transaction cost analysis (TCA) and discussion of proper metrics for cost mea-
surement and performance evaluation.

Description of the different types of trading algorithms: VWAP/TWAP, Arrival Price,
Implementation Shortfall, Liquidity Seeking, Dark Pols, Dynamic Pricing, Opportunistic,
and Portfolio Trading Algorithms.

Proven market impact modeling and forecasting techniques.

Trading costs across various asset classes: equities, futures, fixed income, foreign exchange,
and commodities.

Advanced forecasting techniques to estimate daily liquidity, monthly volumes, and ADV.

An algorithmic decision-making framework to ensure consistency between investment and
trading objectives.

An understanding of how machine learning techniques can be applied to algorithmic trading
and portfolio management.

A best execution process to ensure funds are positioned to achieve their maximum level of
performance.

A TCA library that allows investors to perform transaction cost analysis and develop algo-
rithmic trading models on their own desktop.

xix
xx Preface

A methodology to decode broker models and develop customized market impact models
based on the investment objective of the fund.
Readers will subsequently be prepared to:

Develop real-time trading algorithms customized to specific institutional needs.

Design systems to manage algorithmic risk and dark pool uncertainty.

Evaluate market impact models and assess performance across algorithms, traders, and brokers.

Implement electronic trading systems.

Incorporate transaction cost directly into the stock selection process and portfolio optimizers.
For the first time, portfolio managers are not forgotten and will be provided with proven
techniques to better construct portfolios through:

Stock Selection

Portfolio Optimization

Asset Allocation

MI Factor Scores

Multi-Asset Investing

Factor Exposure Investing
The book is categorized in three parts. Part I focuses on the current electronic market
environment where we discuss trading algorithms, market microstructure research, and
transaction cost analysis. Part II focuses on the necessary mathematical models that are used
to construct, calibrate, and test market impact models, as well as to develop single stock and
portfolio trading algorithms. The section further discusses volatility and factor models, as
well as advanced algorithmic forecasting techniques. This includes probability and statistics,
linear regression, probability models, non-linear regression, optimization, machine learning
and neural networks. Part III focuses on portfolio management techniques and TCA, and
shows how market impact can be incorporated into the investment decisions stock selection
and portfolio construction to improve portfolio performance. We introduce readers to an
advanced portfolio optimization process that incorporates market impact and transaction
costs directly into the portfolio optimization. We provide insight in how MI Factor Scores
can be used to improve stock selection, as well as a technique that can be used by portfolio
managers to decipher broker dealer black box models.

The book concludes with an overview of the KRG TCA library. This chapter providers readers
with insight into how the models and methodologies presented in the book can be packaged and
utilized within numerous software packages and programming languages. These include:
MATLAB, Excel Add-Ins, Python, Java, C/Cþþ, .NET, and standalone applications as
.EXE and .COM application files.

And like the Albert Einstein quote above asks, Algorithmic Trading Methods dares to be
different and exhibits the courage to move in new direction. This book presents the
simplicity behind the algorithmic trading curtain, and shows that the algorithmic trading
landscape is not nearly as complex as Einstein’s intelligent industry fools would have us
believe. This book is a must read for all financial investors and students.
Acknowledgments
There have been numerous people over the years who have made significant contributions
to the field and to the material introduced and presented throughout the text. Without their
insights, comments, suggestions, and criticisms, the final version of this book and these
models would not have been possible.
Roberto Malamut, Ph.D., from Cornell University, was instrumental in the develop-
ment of the numerous methodologies, frameworks, and models introduced in this book.
His keen mathematical insight and financial market knowledge helped advance many of
the theories presented throughout the text. He provided the foundation for multiperiod
trade schedule optimization and is by far one of the leading experts in algorithmic trading,
finance, statistics, and optimization. Roberto is a coauthor of the CFA Level III reading
“Trade Strategy and Execution,” CFA Institute 2019, and he and I have coauthored more
leading-edge algorithmic trading papers in peer-reviewed journals than either one of us can
probably remember.
Morton Glantz, my coauthor from Optimal Trading Strategies, provided invaluable
guidance and direction, and helped turn many of our original ideas into formulations that
have since been put into practice by traders and portfolio managers, and have now become
mainstream in the industry.
The All-Universe Algorithmic Trading Team: Roberto Malamut (Captain), Andrew
Xia, Hernan Otero, Deepak Nautiyal, Don Sun, Kevin Li, Peter Tannenbaum, Connie Li,
Nina Zhang, Grace Chung, Jungsun (Sunny) Bae, Arun Rajasekhar, Mustaq Ali, Mike
Blake, Alexis Kirke, Agustin Leon, and Pierre Miasnikof. Thank you for all your con-
tributions and ideas, which have now become ingrained into the algorithmic trading
landscape. Your contribution to algorithmic trading is second to none and has shaped the
industry.
Wayne Wagner provided valuable direction and support over the years. His early
research has since evolved into its own science and discipline known as transaction cost
analysis (TCA). His early vision and research have helped pave the way for making our
financial markets more efficient and investor portfolios more profitable. Robert Almgren
and Neil Chriss provided the groundbreaking work on the efficient trading frontier, and
introduced the appropriate mathematical trading concepts to the trading side of the in-
dustry. Their seminal paper on optimal liquidation strategies is the reason that trading
desks have embraced mathematical models and algorithmic trading.
Victoria Averbukh Kulikov, Director of Cornell Financial Engineering Manhattan,
allowed me to lecture at Cornell on algorithmic trading (Fall 2009 & Fall 2010) and test
many of my theories and ideas in a class setting. I have a great deal of gratitude to her and
to all the students for correcting my many mistakes before they could become part of this
book. They provided more answers to me than I am sure I provided to them during the
semester. Steve Raymar and Yan An from Fordham University for encouraging me to
continue teaching algorithmic trading and to push and encourage students.
Connie Li, Quantitative Analyst and Algorithmic Trading Expert, M.S. in Financial
Engineering from Cornell University, provided invaluable comments and suggestions

xxi
xxii Acknowledgments

throughout the writing of the book. And most importantly, Connie corrected the errors in
my math, the grammar in my writing, and helped simplify the many concepts discussed
throughout the book. Connie Li is also a coauthor of the CFA Level III reading “Trade
Strategy and Execution,” CFA Institute 2019.
Nina Zhang, M.S. in Quantitative Finance and Statistics from Fordham University,
provided insight and suggestions that led to the development of the TCA functions and
TCA Libraries for MATLAB and Excel. Nina is a coauthor of the paper “Transaction Cost
Analysis with Excel and MATLAB” (JOT, Winter 2017).
Grace Chung, M.S. in Mathematical Finance from Rutgers University, provided insight
and suggestions to help incorporate TCA into the portfolio optimization process. Grace is a
coauthor of the paper “An Application of Transaction Cost in the Portfolio Optimization
Process” (JOT, Spring 2016).
Jungsun “Sunny” Bae, M.S. in Information Systems from Baruch College, is a leading
researcher and practitioner in machine learning and natural language processing. Sunny
was instrumental in helping to develop machine learning applications for their use in
multiperiod trade schedule optimization. She is coauthor of the paper “Machine Learning
for Algorithmic Trading and Trade Schedule Optimization” (JOT Fall 2018).
Scott Bentley, Ph.D., was my publisher for Algorithmic Trading Methods, Second
Edition. He provided invaluable guidance, suggestions, comments, and encouragement
throughout both projects. He is a major reason for the success of both books.
Scott Wilson, Ph.D., provided invaluable insight and direction for modeling trading
costs across the various asset classes, and was influential in helping to structure the
concepts behind the factor exposure allocation scheme.
John Carillo, Jon Anderson, Sebastian Ceria, Curt Engler, Marc Gresack, Kingsley
Jones, Scott Wilson, Eldar Nigmatullin, Bojan Petrovich, Mike Rodgers, Deborah Bere-
bichez, Jim Poserina, Tom Kane, Dan Keegan, and Diana Muzan for providing valuable
insight, suggestions, and comments during some of the early drafts of this manuscript. This
has ultimately led to a better text. The team at Institutional Investor (now IPR), Brian
Bruce, Allison Adams, Debra Trask, and Melinda Estelle, for ongoing support and the
encouragement to push forward and publish with new ideas and concepts.
A special thanks to Richard Rudden, Stephen Marron, John Little, Cheryl Beach, Russ
Feingold, Kevin Harper, William Hederman, John Wile, and Kyle Rudden from my first
job out of college at R.J. Rudden Associates (now part of Black and Veatch) for teaching
the true benefits of thinking outside the box, and showing that many times a nontraditional
approach could often prove to be the most insightful.
Additionally, Trista Rose, Hans Lie, Richard Duan, Alisher Khussainov, Thomas
Yang, Joesph Gahtan, Fabienne Wilmes, Erik Sulzbach, Charlie Behette, Min Moon,
Kapil Dhingra, Harry Rana, Michael Lee, John Mackie, Nigel Lucas, Steve Paridis,
Thomas Reif, Steve Malin, Marco Dion, Michael Coyle, Anna-Marie Monette, Mal Selver,
Ryan Crane, Matt Laird, Charlotte Reid, Ignor Kantor, Aleksandra Radakovic, Deng
Zhang, Shu Lin, Ken Weston, Andrew Freyre-Sanders, Mike Schultz, Lisa Sarris, Joe
Gresia, Mike Keigher, Thomas Rucinski, Alan Rubenfeld, John Palazzo, Jens Soerensen,
 Acknowledgments xxiii

Adam Denny, Diane Neligan, Rahul Grover, Rana Chammaa, Stefan Balderach, Chris
Sinclaire, James Rubinstein, Frank Bigelow, Rob Chechilo, Carl DeFelice, Kurt Burger,
Brian McGinn, Dan Wilson, Kieran Kilkenny, Kendal Beer, Edna Addo, Israel Moljo,
Peter Krase, Emil Terazi, Emerson Wu, Trevor McDonough, Simon (I still do not know
his last name), Jim Heaney, Emilee Deutchman, Seth Weingram, and Jared Anderson.
My previous algorithmic trading students who provided tremendous insight into
algorithmic trading models through numerous questions:
Fall 2009 (Cornell): Sharath Alampur, Qi An, Charles Bocchino, Sai Men Chua,
Antoine Comeau, Ashish Dole, Michael Doros, Ali Hassani, Steven Hung, Di Li, Sandy
Mtandwa, Babaseyi Olaleye, Arjun Rao, Pranav Sawjiany, Sharat Shekar, Henry Zhang,
Xiaoliang Zhu.
Fall 2010 (Cornell): Alisher Khussainov, Shubham Chopra, Jeff Yang, Woojoon Choi,
Ruke Ufomata, Connie Li, Poojitha Rao, Zhijiang Yang, Seung Bae Lee, Ke Zhang, Ming
Sheng.
Fall 2015 (Fordham): Lu An, Chad Brown, Tyler Carter, Isabel Du Zhou, Tianzuo
Feng, Ying Gao, Zhen Huang, Xichen Jin, Aditya Khaparde, Hanchao Li, Shuang Lin, Yi
Liu, Xiaomin Lu, Jinghan Ma, Jinshu Ma, Fupeng Peng, Boyang Qin, Zilun Shen, Fen-
gyang Shi, Alton Tang, Jiahui Wang, Xiaoyi Wang, Jieqiu Xie, Jiaqi Yang, Anya Zaitsev,
Ning Zhang, Yufei Zhang, Yi Zheng, Ruoyang Zhu, Yuhong Zhu, Yunzheng Zhu.
Spring 2016 (Fordham): Amit Agarwal, Yash Bhargava, Richard Brewster, Yizhuoran
Cao, Liangao Chen, Nan Chen, Zhaoyi Ding, Ruiqun Fan, Rui Ge, Tianyuan He, Yue Jia,
Anqi Li, Yang Li, Hsin-Han Lin, Dongming Liu, Kuan-Yin Liu, Bingwan Liu, Yunpeng
Luo, Shihui Qian, Yisheng Qian, Wenlu Qiao, Shao Qiu, Vadim Serebrinskiy, Nitin
Sharma, Yuxin Shi, Hongyi Shu, Ethan Soskel, Shuyi Sui, Zhifang Sun, Ming Wang, Wen
Xiong, Chen Xueqing, Kaicheng Yang, Siqi Yi, Shuwei Zhan, Huidong Zhang, Diyuan
Zhang, Xianrao Zhu, Ying Zhu.
Summer 2016 (Fordham): Eric Adams, Mohammad Alhuneidi, Sergei Banashko,
Zheng Duan, Alexander Flaster, Yuting Guo, Junchi Han, Christian Hellmann, Yushan
Hou, Yangxiu Huang, Ziyun Huang, Hanchen Jin, Xi Jing, Yuxiao Luo, Edward
Mccarthy, Ryan McNally, Francesk Nilaj, Xiaokang Sun, Yinxue Sun, Guoliang Wang,
Melanie Williams, Zihao Yan, Yitong Zheng.
Fall 2017 (Fordham): Maha Almansouri, Seongjun Bae, Jinshuo Bai, Subhasis Bhadra,
Shiwen Chen, Taihao Chen, Yutong Chen, Yichen Fan, James Ferraioli, Patrick Fuery,
Ziqing Gao, Bingxin Gu, Yang Hong, Tingting Huang, Keyihong Ji, Owen Joyce, Jiayuan
Liu, Xun Liu, Xin Lu, Rui Ma, Mengyang Miao, Xueting Pan, Xiao Tan, Yaokan Tang,
Mengting Tian, Hongyuan Wang, Ning Wang, Yimei Wang, Jiajin Wu, Hansen Xing,
Zheng Zheng, Yuan Zhou.
Spring 2017 (Fordham): Jianda Bi, Sean Burke, Beilun Chen, Yilun Chen, Tamar
Chikovani, Niclas Dombrowski, Chong Feng, Fangfei Gao, Masoud Ghayoumi, Man
Avinash Gill, Jiangxue Han, Yuze Ji, Shuxin Li, Lianhao Lin, Chang Liu, Xinyi Liu, Yi
Luo, Tianjie Ma, Nicholas Mata, John Mitchell, Mathieu Nicolas, Boyuan Su, Haoyue
Sun, Yifan Tang, Tuo Wang, Weixuan Wang, Minqi Wang, Qijin Wu, Junhao Wu, Yifan
xxiv Acknowledgments

Wu, Wei Wu, Xin Xiong, Mao Yang, Kai Yang, Kirill Zakharov, Tongqing Zhang,
Haojian Zheng, Jiri Beles, Brian Block, Chaitanya Motla, Rongxin Wang, Ye Zhang.
Fall 2018 (Fordham): Jennifer Dunn, Nadir Bajwa, Riley Coyne, Robert Genneken III,
Connor Griffin, Liesel Judas, Lubaba Khan, Steve Kotoros, Ryan Mertz, James O’Hara,
Lokesh Sivasriaumphai, Roberto Stevens.
Spring 2019 (Fordham): Charles Blackington, Rong Deng, Lei Feng, Luoyi Fu,
Tianlun Gao, Yutong Guo, Tong Han, Yuxuan He, Danny Hemchand, Alana Higa, Boyu
Hu, Lan Huang, Brandon Jaskulek, Zhiyan Jiang, Athanasios Katsiouris, Rachel Keough,
Tanya Krishnakumar, Anqi Li, Chenxi Li, Yiteng Li, Kuiren Liao, Xin Liu, Xuwen Lyu,
Yaakov Musheyev, Ziwen Pan, Luman Sun, Chengxin Wang, Jingzhi Wang, Yutong
Xing, Yuxin Xu, Yazhuo Xue, Deyi Yu, Xiang Ning Zang, Yixiao Zhang, Zhizhe Zhao,
Lujun Zhen, Zepeng Zhong, Ruiyang Zou, Yingting Zou.
In addition to all those who have made significant contribution to the field of algo-
rithmic trading, unfortunately, there are also those who have impeded the progress of
algorithmic trading and quantitative finance. These individuals have been mostly moti-
vated by greed and their never-ending quest for self-promotion. The nice thing is that
many are no longer in the financial industry, but there are still a few too many who remain.
A list of those individuals is available upon request.

Best regards,
Robert Kissell, Ph.D.
 Chapter
1
Introduction
To say that electronic algorithmic trading has disrupted the financial envi-
ronment is truly an understatement. Algorithmic trading has transformed
the financial marketsdfrom the antiquated days of manual labor, human
interaction, pushing, yelling, shoving, paper confirmations, and the occa-
sional fist-fightdinto a system with electronic audit trails and trading facil-
itated using computers, complex mathematical formulas, machine learning,
and artificial intelligence.
Nowadays, the trading floors of these antiquated exchanges more resemble
a university library than they do a global center of trade and commerce.
Many of the glamourous trading floors of years ago, such as the floor of
the New York Stock Exchange, have been relegated to just another stop
on a historical walking tour of downtown New York City.
Trading floors are no longer an active center of trading commerce. Trading
floors are relatively quiet and are no longer littered with filled will paper or-
ders and confirmations. Today, all trading occurs electronically in data cen-
ters with computers rather than people matching orders.
In 2019, electronic trading comprised approximately 99.9% of all equity
volume and algorithmic trading comprised approximately 92% of all equity
volume.1 The remaining 8% of the orders that are not executed via an algo-
rithm are still transacted electronically. But in these situations, brokers still
route orders via a computer trading system to different exchanges, venues,
and/or dark pool to be transacted in accordance with specified pricing rules
define by these brokers.
Fig. 1.1 illustrates the evolution of electronic and algorithmic trading over
the period 2000e19. Electronic trading in the early years was dominated
by firms such as Instinet, Island, and ITG/Posit, and occurred mostly in
NASDAQ/OTC stocks. Electronic trading grew from 15% in 2000 to
99.9% in 2019. The only trading that does not occur electronically today

1
Source: Kissell Research Group, www.KissellResearch.com.
Algorithmic Trading Methods, Second Edition. https://doi.org/10.1016/B978-0-12-815630-8.00001-6
Copyright © 2021 Elsevier Inc. All rights reserved. 1
2 CHAPTER 1 Introduction

Electronic and Algorithmic Trading

120.0%

100.0%

80.0%

60.0%

40.0%

20.0%

0.0%

Electronic Trading Algorithmic Trading

n FIGURE 1.1 Electronic and Algorithmic Trading. Source: Kissell Research Group.

(<0.1%) occurs through special situation sales and transactions. Otherwise,

all trading in US markets occurs electronically.
Over the same period, algorithmic trading exploded grew from 1% of total
market volume in 2000 to 93% in 2019. The biggest increase in algorithmic
trading occurred first in August 2007 due to the quant meltdown and then
again in 2008e09 due to the financial crisis. It was during these periods of
high volatility, trading difficulty, and rapid price changes that institutions
realized the benefits of algorithmic trading. During these times, investors
were faced with rapidly adverse price movement and decreased transactable
liquidity. Time delay encountered in when disseminating orders to brokers
for immediate execution were often met with information leakage, less
favorable transaction prices, and lower profits margins. To avoid these hos-
tile trading conditions, investors turned to the more advanced trading sys-
tems and the usage of algorithms so that they could better control their
orders, keep their trading intentions hidden, and achieve more favorable
transaction prices.
The old-fashioned trading system environment we once know where we call
our broker, market-maker, or specialist over the phone are long gone.
Welcome to the new financial environment.

WHAT IS ELECTRONIC TRADING?

Electronic Trading is the process of transacting orders over a computer sys-
tem or network rather than via a phone call or fax sent to your broker where
you need to state your order, trading intentions, and any special instruction.
Electronic trading could be as simple as entering a buy order into a retail
trading system using a computer terminal, or more recently, via a mobile
app. Electronic trading could also be more advanced and complex such
 Trading Algorithm Classifications 3

as situations where investors route orders to different trading venues for

execution at specified prices, within price spreads, or for execution at
different times of the day. In all these cases, electronic trading encompasses
any order that generates via a computer connection.
Electronic trading should not be anything new. In fact, much of our daily lives
have become mobile and we are connected to the internet almost twenty-four
seven. Think of all the purchases we make on the internet including every-
thing from movie, sports, theater, and entertainment tickets; clothing, travel,
hotel, and airfare, automobile, and at times even home purchases.

WHAT IS ALGORITHMIC TRADING?

Algorithmic trading in its simplest form is the computerized execution of a
financial instrument following a prespecified set of trading rules and instruc-
tions. Investors, instead of sending an order to a broker for execution or
routing an order to an exchange, simply enter the order into the algorithm
for execution. Algorithms then slice larger orders into smaller pieces for
execution over the day, and at various trading venues, to achieve the best
market prices and reduce overall trading costs.
The primary goal of algorithmic trading is to ensure that the implementation
of the investment decision is be consistent with the investment objective of
the fund and to manage the overall transaction costs of the order and
achieve the favorable prices.

TRADING ALGORITHM CLASSIFICATIONS

Trading algorithms are classified into three categories: execution algo-
rithms, profit seeking algorithms, and high frequency trading algorithms.
n Execution Algorithms: An execution algorithm is tasked to transact the
investment decision made by the investor or portfolio manager. The man-
ager determines what to buy or sell based on their investment style and
fund investment objective, and then enters the order into the algorithm.
The algorithm will then execute the order and implement the decision
following a set of rules specified by the portfolio manager.
n Profit Seeking Algorithms: A profit seeking algorithm is an algorithm
that will both determine what to buy and/or sell in the market and will
execute those decisions without interaction by the portfolio manager.
For example, these algorithms will use real-time price information and
market data such as prices, volume, volatility, and price spreads to deter-
mine what to buy or sell, and will then implement a trade when the con-
ditions are favorable to the investor. Profit seeking algorithms seek to
4 CHAPTER 1 Introduction

earn a profit based on a quantitative model, market mispricing, or a stat-

arb strategy based on pairs, index funds, or ETFs.
n High Frequency Trading: High frequency trading (HFT) is a type of
profit-seeking algorithm that seeks to earn a short-term trading profit.
The holding period for an HFT trader will often only last for a few seconds
or less and HFT trading will tend not hold any overnight position as it ex-
poses the fund to much increment risk. HFT algorithms are characterized by
very high turnover rates and they seek to profit by exploiting market mis-
pricing and liquidity conditions across different exchange, venues, and
dark pools. HFT algorithms are also notorious for trying to uncover the
buying and/or selling intentions of long-term investor through processing
market data, prices, and quotes, and then using this information to their
advantage to achieve a profit. All HFT algorithms are profit seeking algo-
rithms, but not all profit seeking algorithms are HFT.

TRADING ALGORITHM STYLES

There are many different types of trading algorithms in the market, each
with a unique name that often does not adequately describe how the algo-
rithm will transact in the marketplace. To help managers differentiate algo-
rithms, they are often classified as aggressive, working order, or passive.
Managers need to determine the algorithm that will transact in a manner
consistent with the investment object of the fund. These are:
n Aggressive: Algorithms that will trade aggressively in the market
with the goal of transacting shares at a specified price or better. These
algorithms have often been described as liquidity seeking algorithms
and/or liquidity sweeping algorithms. These algorithms will likely trade
aggressively in the market and take liquidity across multiple venues
when there is volume at the specified price or better. They tend to trade
with more market order than limit orders.
n Working Order: Algorithms that trade in the market following prescribed
rules based on the needs of investors. These algorithms will often seek to
balance the tradeoff between trading cost and market risk, as well as seek
to maximize the specified investment objectives. These algorithms will
trade with an appropriate balance and mix of limit and market orders.
n Passive: Algorithms that trade in a very passive manner and using mostly
limit orders. These algorithms will also seek to trade greater quantities of
shares in dark pools to minimize information leakage and to ensure that
the execution of the order does not provide the market with signals per-
taining to the trading intentions of the fund.
Another Random Scribd Document
with Unrelated Content
The Project Gutenberg eBook of The mill house
mystery
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: The mill house mystery

Author: Florence Warden

Release date: March 21, 2024 [eBook #73223]

Most recently updated: May 1, 2024

Language: English

Original publication: London: Jarrold & Sons, 1911

Credits: an anonymous Project Gutenberg volunteer

*** START OF THE PROJECT GUTENBERG EBOOK THE MILL

HOUSE MYSTERY ***
 THE MILL HOUSE
MYSTERY
BY
FLORENCE WARDEN
AUTHOR OF “THE INN BY THE SHORE,” “THE HOUSE ON
THE MARSH,” ETC.

LONDON
JARROLD & SONS, WARWICK LANE, E.C.
 CONTENTS.
I. AN ACCIDENT
II. RHODA PEMBURY’S DISCOVERY
III. TEN YEARS AFTER
IV. RHODA RETURNS TO MILL-HOUSE
V. LADY SARAH’S RECOGNITION
VI. JACK ROTHERFIELD
VII. THE SCARRED HAND
VIII. THE MISSING SNUFF-BOXES
IX. RHODA’S WATCHFULNESS
X. THE STOLEN “ROMNEY”
XI. THE PICTURE RECOVERED
XII. LADY SARAH’S DUPLICITY
XIII. SIR ROBERT SEEKS ADVICE
XIV. JACK ROTHERFIELD’S EFFRONTERY
XV. SELF-ACCUSATIONS
XVI. A FRUSTRATED ELOPEMENT
XVII. SIR ROBERT’S PLANS
XVIII. THE COMPANION’S ORDEAL
XIX. OTHERWISE THAN INTENDED
XX. SIR ROBERT’S SECRET
 THE
MILL HOUSE MYSTERY

CHAPTER I.
AN ACCIDENT

The July sun was pouring floods of blinding, glaring light upon the
town of Dourville, which, lying in a great chasm between two high
white lines of cliff, and straggling under the foot of them to east and
west, bears witness, in its massive castle, and in its old relics of
stone buildings among the commonplace iron frames and plate-glass
windows of the new, to the notable part it has taken in England’s
history.
The long straight road that goes northwards up through the town
and out of the town, rising, at first by slow degrees, and latterly by a
steep ascent, to a point from which one can look down upon town
and sea, soon leaves small shops for queer old-fashioned rows of
houses; and these in their turn give place to roomy old residences of
greater pretension.
At the back of one of these, a sombre, plain building, roomy rather
than dignified, there stretches a splendid expanse of garden and
pleasaunce, where a stream runs among meadows and lawns in a
direct line towards the sea.
This stream once supplied the power that worked a great paper-
mill, which was the foundation of the prosperity of the Hadlow family.
But three generations back, the reigning Hadlow, more enterprising
than his predecessors, had speculated outside his little world, had
prospered, and finally blossomed into the great philanthropist, whose
magnificent endowment of certain royal charities had earned him a
baronetcy.
Rich as the family had grown, the Hadlows clung to the old nest
with a pertinacity which had in it something of dignity; and only the
condition in which the grounds were kept, nothing in the appearance
of the house itself, would have betrayed that now, under the third
baronet, the place was the property of a man of great wealth.
The trees grew thickly within the high dark wall that shut the
grounds in from the road. And under their shade Sir Robert Hadlow,
in a light linen suit and shady planter’s hat, could saunter at his ease
in the heat of the day. A man of middle height, slight and almost
boyish in figure, with a close-trimmed dark beard and large, mild,
grey eyes, Sir Robert Hadlow, at thirty years of age, looked rather
older by reason of the quiet gravity of his manners and the leisurely
dignity of his movements.
A man of leisure, he had devoted himself early and
enthusiastically to the study of the antiquities of the neighbourhood
in which he was born; and something of the far-away look of the
student softened and mellowed the expression of his eyes, and gave
a certain measured dignity to his gait.
Stopping from time to time to peep between the branches of the
lilac-bushes at the stream as it sparkled in the bright sunlight
beyond, he was sauntering towards the house, when a succession of
piercing screams, followed by the shouts of men, reached his ears
from the road outside.
“Stop her!” “Look out!” “She’ll be killed!”
These, among others, were the cries which came to Sir Robert’s
ears as he hurriedly made his way to one of the wooden doors in the
high wall, and inserting into the lock his own private key, let himself
through into the public street.
Looking up the road, to the left, he saw the figure of a woman, in a
light dress, coming swiftly down the hill on a bicycle, of which it was
evident that she had lost control. A glance to the right showed him a
traction engine coming slowly up the hill with a couple of waggons
trailing behind it, and the confused cries of the bystanders called his
attention to the fact that it was a collision between this and the
bicycle which they all feared.
 Stepping forward into the road, and watching the light machine
vigilantly as it came quickly down upon him, Sir Robert prepared for
his rather risky attempt to save the woman from her danger. As the
bicycle reached him he turned to run with it down the hill, at the
same time seizing the handlebar with so much dexterity that he
neither stopped the machine nor threw off its rider.
The woman was muttering incoherent thanks in a faint voice, and
Sir Robert became suddenly conscious that there was a fresh
danger to be averted.
“Keep your head. Steady! Hold tight! You’re all right,” he cried as
he still ran with the bicycle, upon which he was now acting as a
brake.
But his words fell on ears that scarcely heard; and before he could
bring the machine absolutely to a standstill, when he was within
three or four yards of the traction-engine, which had been stopped,
the rider fell to the ground with a moan.
There was a crowd round the group already, and there were
shrieking women and curious men streaming towards the spot,
where Sir Robert, with an air of authority, was giving directions to
such of the more intelligent among the crowd as seemed likely to be
of use in the emergency. Thus, he sent one man for a doctor and
another for his own servants, while he himself knelt down by the
roadside, and raised the unconscious victim of the accident.
She had struck her head against the kerb-stone, and one side of it
was cut and bleeding.
“Poor child! She isn’t dead. She’ll be all right presently,” said Sir
Robert, answering the alarmed comments of the women who
pressed round him. “I’m going to have her taken into my house,
where the doctor will see her.”
The accident had occurred within twenty yards of the entrance to
Sir Robert’s house, and five minutes later the baronet and his butler
were carrying the unconscious girl under the little portico and up the
staircase into a pleasant room at the back of the house, overlooking
the grounds and the flowing stream.
A couple of children, a boy and a girl, the orphaned nephew and
niece of Sir Robert and permanent members of his household,
watched the arrival from the upper staircase with eager interest.
 “Look at the blood, Minnie!” observed the boy, in an awestruck
whisper. “And look at her eyes—all shut!” he added with thrilling
interest.
The girl, younger and more tender-hearted, began to cry.
“She’s dead!” sobbed she. “Oh, George, don’t yook at her. She’s
been killded.”
“No, she hasn’t,” said he sturdily. “Uncle Robert won’t let her die.”
Their hissing whispers had by this time attracted attention, and
Bessie, the old family nurse in whose charge they were, beckoned to
them from below with an austere frown.
“If you don’t both go back into the nursery this minute——”
There was no need to say more: in an instant the scampering of
small feet, followed by the banging of an upper door, showed that the
young people, who were known in the household as The Terrors,
were for the moment quelled.
In the meantime the victim of the accident had been laid upon a
bed in a darkened room, and Bessie and her master were looking at
her with sympathetic interest.
“Why, the poor dear’s but a lass, sir,” said the sympathetic Bessie,
as she loosened the girl’s clothes and peered keenly into the pale
face.
“Yes, not more than eighteen or nineteen, I should think,” said Sir
Robert. “She had a narrow escape. Search her clothes, Bessie, for
some indication as to her name and address. Her people will be
alarmed about her, whoever she is, and whoever they may be.”
“Yes, sir. I’ll have a hunt as soon as the doctor’s here.”
She had not to wait long. And by the time the doctor had come,
examined the patient, and reported that the victim was suffering from
concussion of the brain and must be kept quiet, that she had
sustained an injury to the right wrist and severe bruises, the old
nurse had made a search of the girl’s pockets, and had discovered
an opened letter in one of them directed to “Miss Rhoda Pembury” at
an address in Deal.
This was enough for Sir Robert, who telegraphed at once to the
address, to the name of Pembury, to the effect that Miss Rhoda had
met with a slight accident, but that she was safe and going on well.
 Within a couple of hours the girl’s father and mother had arrived at
the Mill-house, and proved to be a London physician and his wife,
who were staying at Deal with their family, of whom Rhoda was the
eldest.
They were deeply grateful to Sir Robert, who insisted that they
should leave their daughter where she was until she was fully
recovered, a suggestion which poor Mrs. Pembury, the harassed
mother of half a dozen children, gratefully accepted, it being
arranged that Dr. Pembury should cycle over every day to see how
his daughter was getting on.
Within a few days Rhoda, very pale still, and with deep dark lines
under her large, plaintive, blue eyes, was sitting at the window of the
room that had been assigned to her, permitted for the first time to
leave her bed.
She was a tall, thin slip of a girl, not yet fully developed, but
languid and almost sickly of appearance by reason of the rapidity of
her recent growth. At seventeen she was five feet seven inches in
height, with a lean, fair-skinned face, a mass of pale golden hair that
looked as if a silver veil had been thrown over it, and a look of
listlessness that told of weakly health.
She confessed to Bessie that she had only had her hair “up” within
the last month, and that, in her present enfeebled condition, she
preferred to leave it loose again, tied with a black ribbon, to the
fatigue of doing it herself, or even of having it done for her.
So that she looked like a child as she sat at the open window, with
her white dressing-gown on, and her head thrown back against the
pillows provided for her.
In the garden below were two figures, upon whom her attention
was fixed with interest so deep that Bessie watched her in furtive
surprise, wondering at the look of vivid excitement which was making
the blue eyes glow and the white skin flush.
The old nurse looked out, and saw that the objects of the girl’s
interest were Sir Robert Hadlow, sauntering in the grounds in his
linen coat and broad-brimmed hat, and his handsome young ward,
Jack Rotherfield, a tall, well-made man of two and twenty, whose
dark-skinned, beardless face and curly black hair and dark eyes had
earned him the reputation of the handsomest man in Kent.
 The expression upon the girl’s face, as she gazed out at the two
men, was so unmistakably one of admiration of the most vivid kind,
that Nurse Bessie smiled indulgently.
“A good-looking fellow, isn’t he?” she said with a nod in the
direction of the two figures.
To her surprise, the girl turned towards her with a look of ecstasy
in her thin face.
“Good-looking!” she echoed in an awestruck tone. “Oh, don’t call
him that! He’s so much more than that! It seems to me,” she added,
in a low voice, as again her eyes wandered in the direction of the two
gentlemen, “that I’ve never seen, no, and never even imagined, any
face either so handsome or—so—noble. It’s because he’s so good,
so much better and greater than other men that he is so handsome.”
The old nurse sat amazed and perplexed by this enthusiasm,
which exceeded so far even her own warmth of admiration. She did
not dare to smile, although the girl’s tone was so outrageously,
childishly vehement as to throw her into considerable astonishment.
“Well, Mr. Rotherfield is generally thought to be nice-looking,” she
said, “but I don’t know as he’s all you take him for.”
The girl’s fair face, out of which the glow of colour brought by her
enthusiasm had already faded a little, looked at her with a frown of
slight perplexity.
“Mr. Rotherfield? Who is that?” she asked.
Old Bessie stared.
“Why, the gentleman you’ve admired so much, the young
gentleman that’s walking with Sir Robert. That’s his ward, Mr.
Rotherfield.”
A deeper flush than had yet appeared in Rhoda’s face now spread
quickly over it, and she lowered her eyelids quickly.
“I didn’t notice him,” she said. “I was speaking about Sir Robert.”
The old nurse uttered a low cry of surprise.
Then a smile, indulgent, amused, appeared on her face.
“Well, it’s the first time I’ve ever heard him called so nice-looking,”
she said. “He’s very well, of course, and he’s got a good face, and a
nice face, but I’ve never heard tell he was considered handsome.”
The girl looked up again, the most innocent surprise in every
feature.
 “Not handsome!” she said under her breath. “Why, it seems to me
I never saw any face so—so beautiful! He’s like a picture, not like
any man I ever saw before. To look at him makes me feel humiliated
at the thought that I should have been the means of causing him to
hurt himself, and yet it makes me proud too to think that he should
have done what he did for me!”
Beginning timidly, the girl grew more and more enthusiastic as she
went on, till she ended with fire in her blue eyes, and sat with her lips
parted in a sort of ecstasy, gazing out of the window at the figure of
the wholly unconscious gentleman who was now sauntering back
towards the house.
Sir Robert, who had hurt his arm in his efforts to stop the runaway
bicycle, carried it in a sling, and Rhoda’s eyes softened and filled
with tears as she noted the fact.
The old nurse’s face began to grow prim.
“You mustn’t let Lady Sarah hear you speaking so admiring of her
intended, or she’ll be jealous,” said she.
A sudden shadow passed over the girl’s face.
“Lady Sarah! Who is she?” she asked quickly, in a stifled voice.
Bessie peered at her rather anxiously.
“Dear, dear, miss, you mustn’t get so excited about it, or I shall feel
I didn’t ought to have told you so much,” she said.
A faint, mechanical smile appeared on Rhoda’s face.
“Nonsense,” she said. “Of course I’m not excited, only interested.
Who is Lady Sarah?”
The nurse hesitated a moment, but seeing that a red spot was
beginning to burn in each of the invalid’s cheeks, she decided that it
would be better to tell her what she wanted and have done with it.
“Lady Sarah,” she said, gravely and deliberately, hoping that the
style and title of the persons she was about to mention would duly
impress her hearer, “is the youngest daughter of the Marquis of
Eridge, and she is engaged to be married to Sir Robert Hadlow, who
is madly in love with her.”
A look of dismay, so ingenuous, so complete as to be touching,
appeared on Rhoda’s face. Then she glanced quickly at the nurse,
reddened deeply, and subduing her feelings, whatever those might
be, answered in a matter-of-fact tone, in words which surprised
Bessie.
“The Marquis of Eridge! Oh, yes, I know. He was made bankrupt
two years ago, and he has four of the most beautiful daughters
possible.”
Bessie was taken aback by the completeness of the girl’s
information.
“I’m sure I don’t know anything about the Marquis’s affairs,” she
said, somewhat stiffly. “But a Marquis is a Marquis, and Lady Sarah
is a most beautiful young lady. And Sir Robert is crazy about her,
and to look at her it’s no wonder. But you’ll see her for yourself, I
dare say, before you go away. She lives up in the Vale, at the Priory,
and she and Lady Eridge are here most days when Sir Robert
doesn’t go to the Priory.”
Rhoda bent her head without speaking. And the nurse, though she
reproached herself for the feeling and said to herself that it was
‘rubbish,’ felt a momentary wonder whether it would not have been
better for Sir Robert, with his studious habits and his grave
demeanour, to have loved an earnest, simple little girl like the blue-
eyed, fair-haired Rhoda with the devotion in her eyes, rather than the
brilliant and slightly disturbing creature whom he had chosen for his
wife.
 CHAPTER II.
RHODA PEMBURY’S DISCOVERY

The day after this conversation with Bessie, Rhoda was allowed
downstairs for the first time. Sir Robert was kindness itself to her,
though he was rather puzzled by the extreme reserve and timidity of
the girl whose life he had saved, never guessing, in his masculine
obtuseness, at the sentimental cause of her rather perplexing
demeanour.
Jack Rotherfield, who was staying with his guardian, was delighted
to welcome a new and pretty guest, and at once proceeded to exert
himself to amuse and interest the convalescent, so that old Bessie
used to smile demurely when she came into the room where the two
young people would be sitting together, Rhoda gentle and rather
listless, Jack energetically trying to rouse her from the somewhat
abstracted state in which she still remained.
Rhoda laughed at the idea of falling in love with Jack, a possibility
which Bessie plainly foresaw and made no scruple about
mentioning.
“He’s very nice,” she said, “and, I suppose, very good-looking. But
I don’t like him as much as I ought to do, considering how kind he is.
He always seems to me to be saying things to me which he must
have said before.”
Bessie looked surprised.
“Lor, miss, that’s not a bad guess, I’m afraid, if it is a guess,” she
admitted. “Mr. Jack is so nice-looking, and so merry and bright, that
the young ladies do make a fuss of him. Even Lady Sarah,” she
added in a rather lower tone.
A deep flush at once overspread Rhoda’s face.
“I should hardly think,” she said quite tartly, “that a girl who had the
good fortune to be liked by Sir Robert would care much for Mr.
Rotherfield.”
Bessie looked askance at her, but said nothing more on the
subject, until she presently remarked in a rather dry tone that Lady
Sarah was coming that afternoon, with her mother and one of her
married sisters, to play tennis and to have tea in the grounds.
Rhoda was excited by the news. She was exceedingly anxious to
see the woman with whom Sir Robert was in love, and Bessie noted
the trembling of her hands and the feverish light in her eyes as she
dressed to go downstairs.
Lady Sarah Speldhurst proved to be a very fascinating and lovely
little person. Not nearly so tall as Rhoda herself, nor with the
advantage of so good a figure as the younger girl, she was, at three
and twenty, mistress of all the arts by which a pretty young woman
makes the best of herself. Dark-eyed, with a brilliant complexion, and
with masses of wavy dark brown hair, she dressed in light colours for
choice, and was wearing, on this occasion, a tight-fitting lace dress
of creamy tint over a slip of lemon colour, and a big black hat with
black and white ostrich feathers.
“Sir Robert don’t know what a lady’s dress bill means—yet,”
remarked Bessie shrewdly, when she looked out and saw Lady
Sarah in the garden.
“What an odd dress to play tennis in!” was Rhoda’s matter-of-fact
comment.
Bessie smiled.
“She don’t play tennis much herself. Her ladyship likes the sitting
about with a racquet in her hand, and the cakes and the ices, better
than running in the sun and getting her face red,” she said.
Rhoda frowned a little. Pretty as Lady Sarah was, the younger girl
felt that a better, a more sincere and noble-natured person than Lady
Sarah appeared to be would have been a better match for the
generous and good Sir Robert who was her own idol. She went
downstairs slowly, resented the quick and almost supercilious
manner in which Lady Sarah appeared to sum her up at a glance
while shaking hands, and decided angrily that Sir Robert was
throwing himself away.
The baronet himself, however, was evidently by no means of the
same way of thinking. There was adoration in his mild grey eyes as
he watched the brilliant little brunette, there was tenderness in the
tone of his voice as he spoke to her, and it was abundantly clear that
his infatuation was complete.
 Jack Rotherfield, meanwhile, was less attentive to Rhoda than he
had been before the appearance of Lady Sarah. Rhoda did not mind
this, but she remarked it, and, sitting silent for the most part, she
noticed a good deal more, as the afternoon wore on, that might have
escaped the notice of a less observant or more talkative person.
For one thing she saw, and felt ashamed of seeing, that something
like a secret understanding existed between Jack Rotherfield and
Lady Sarah; their eyes would meet with a sudden look of sympathy
or mutual amusement from time to time, as, for instance, when Sir
Robert declared that nothing would induce him to replace the old
furniture and fittings of the house for more modern ones.
Rhoda felt ashamed of herself for thinking that it looked as if Lady
Sarah had already discussed that very subject with Sir Robert’s
ward, and in a manner not very sympathetic with the views of her
future husband.
Indeed it was clear to the most careless eyes that there was a
great gulf between the tastes of the Marquis’s lovely daughter, with
her French toilette and her brilliant if scarcely sincere manners, and
steady-going, quiet Sir Robert Hadlow with his grave demeanour and
quiet habits.
Rhoda found herself wondering what sort of a household theirs
would be, and which of the two would finally get the upper hand, as it
was plain that, in such an ill-assorted couple, one or other must
eventually do.
It seemed natural to suppose it would be the little, wilful, spoilt
beauty, as it was easy to see she was not in love with Sir Robert,
who, by keeping her head, would become the arbiter of the
household destinies. The baronet seemed, indeed, to be like wax in
her hands; and he was far too much in love to see that the sweet
looks and pretty smiles, the little words of tenderness, and the
gestures of caressing cajolery, were dictated by anything less than
love equal to his own.
The rest of the party soon went into the grounds, and Rhoda, who
was not yet allowed to exert herself much, was left alone in the
house. She sat near the window, watching the pretty figures of the
ladies in their light dresses as they flitted over the tennis-lawn, like
gay butterflies against the background of soft greenery, when she
heard a stealthy footstep behind her, and looking round, saw the
Terrors, George and Minnie Mallory, crouching close to her chair.
“When did you come in?” asked Rhoda quickly. “I didn’t hear you.”
The two children chuckled.
“Nobody never does hear us,” said Minnie, who was a long-
legged, short-frocked imp of six years of age. “We don’t never let ’em
hear us,” she added thoughtfully.
“But that’s not right. It’s like eavesdropping,” said Rhoda solemnly.
George nodded gravely.
“It doesn’t matter for us, ’cos we’re only children,” said he with a
shrewd air. “And we often hear things that we like to hear. We heard
Lady Sarah talking to Jack the other day, and saying how hard it was
for her to have to marry a rich man, ’cos rich men are always what
you don’t like.”
Rhoda uttered a sort of gasp. Then she recovered herself, and
scolded the boy.
“It’s very naughty to listen,” she said. “And very ungentlemanly too.
What would your uncle, who’s always so good and kind to you, say if
he thought his niece and nephew were not behaving like a lady and
gentleman?”
George was not abashed.
“I’ll behave like a gentleman when I grow up,” he said reflectively.
“I don’t see the good of beginning too soon. It’s nicer to do as we like
and hear what we want to.”
The comical gravity with which he spoke suddenly made Rhoda
want to laugh, so she was silent for a moment, and the children took
advantage of this to steal away out of the room, no doubt to follow
their favourite dubious occupations elsewhere.
But Rhoda did not heed them. She was filled with a terrible
thought. Her hero, the man she worshipped as the ideal of all that
was noble and worthy, was being deceived, grossly deceived, by the
woman he passionately loved. She had no doubt at all that the words
reported by the mischievous boy had really been uttered by Lady
Sarah, in confidential talk to Jack Rotherfield, between whom and
herself it was plain that an active flirtation was still going on.
Her heart was torn by the thought that her hero, instead of being
loved as he deserved to be loved, was being married for his money
alone by the woman he worshipped. If only he could learn the truth
before it was too late!
But how?
She could not tell him what she knew or guessed, and even if she
could, he would not believe her.
What could she do?
Staying day by day under the same roof with Sir Robert, she had
fallen more and more completely under the influence of his great
kindness and gentleness, of the nature that was ever self-sacrificing,
ever considerate for others, yet with a certain manliness and
firmness that made Rhoda wonder what he would be like if he should
ever find out that those he loved and trusted had deceived him.
She was still torn with her fears on his account when the baronet
came in, racquet in hand, and sitting down beside her, asked her
kindly how she felt.
The girl, pale and trembling, looked into his gentle, kindly face,
and the words that came to her lips refused to come further.
He smiled at her, and patted her hand.
“You’ve been overtiring yourself. I shan’t let you come downstairs
to-morrow,” he said.
Rhoda struggled to regain her self-command and answered
steadily:
“I must come down to-morrow, Sir Robert, for I must go back
home.”
“You are tired of us? That you are in such a hurry to get away?”
She shook her head.
“You are the best and the kindest people I’ve ever met,” she said
tremulously. “But I want to go back.”
He looked at her keenly.
“You have something on your mind,” said he.
Rhoda rose suddenly to her feet. Looking down upon him with
eyes that blazed, she said hoarsely:
“Yes. I want to warn you. Find out, Sir Robert, whether you are
loved as you deserve to be loved. That’s all.”
The baronet rose, frowning and displeased. She saw that he
looked upon her words as an impertinence, and she was cut to the
heart.
 Faltering, she stammered out an incoherent apology. Sir Robert
looked at her coldly.
“There is nothing to apologise for,” he said gravely. “I’m sure you
mean well. I was taken by surprise, that’s all.”
Rhoda felt that the room was spinning round her. She knew his
danger, and she saw that she was helpless to save him. There was
only one thing to be done; she must go away. She could not stay
another day now that she had offended him, nor could she watch the
progress of the harm she could not prevent.
On the pretext of fatigue, she staggered upstairs, assisted by Sir
Robert, as far as the foot of the staircase, where she gently refused
further help.
Rhoda had never seen the kindly Sir Robert angry before, and the
effect his displeasure had upon her was overwhelming. She,
however, was not to be the only person to offend him that day, for
Bessie, who came in with a little tray with the wing of a chicken for
the convalescent, brought with her the news that Sir Robert was
gravely displeased with his old servant, Langton, to whom he had
given notice to leave him.
“I don’t know the rights of it,” went on Bessie, “and I don’t want to
gossip. But it’s thought Langton told Sir Robert something he didn’t
want to hear, and didn’t believe, and this is the consequences!”
Rhoda listened in distressed silence. Had the faithful servant
dared to tell his master something that he had seen? Something that
concerned Lady Sarah and Jack Rotherfield? She would not
condescend to gossip with Bessie about it, but when she was alone,
she left her repast almost untasted, and, attracted by a soft murmur
of voices that came like a distant whisper through the open window,
crossed the floor and looked out, and saw, between the branches of
the trees, two figures sauntering along the avenue that ran inside the
outer wall of the grounds.
She had no difficulty in recognising them, and when, before they
had gone many steps, they stopped and the man put his arm round
the girl and kissed her, Rhoda knew that it was Jack Rotherfield
whom she had seen kissing the betrothed wife of his guardian.
Rhoda could bear no more; turning from the window, giddy and
almost sick with grief and horror, she resolved to leave the house
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com