Statistically Sound Machine Learning for

Algorithmic Trading of Financial Instruments

1.20 Edition David Aronson download

https://ebookname.com/product/statistically-sound-machine-
learning-for-algorithmic-trading-of-financial-
instruments-1-20-edition-david-aronson/

Get the full ebook with Bonus Features for a Better Reading Experience on ebookname.com
Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Machine Learning an algorithmic perspective 2ed.

Edition Marshland

https://ebookname.com/product/machine-learning-an-algorithmic-
perspective-2ed-edition-marshland/

Machine learning an algorithmic perspective 1st Edition

Stephen Marsland

https://ebookname.com/product/machine-learning-an-algorithmic-
perspective-1st-edition-stephen-marsland/

Market Risk Analysis Pricing Hedging and Trading

Financial Instruments v 3 Carol Alexander

https://ebookname.com/product/market-risk-analysis-pricing-
hedging-and-trading-financial-instruments-v-3-carol-alexander/

RF Components and Circuits 1st ed Edition Joe Carr

https://ebookname.com/product/rf-components-and-circuits-1st-ed-
edition-joe-carr/
Citizen Relationship Management A Study of CRM in
Government 1st Edition Alexander Schellong

https://ebookname.com/product/citizen-relationship-management-a-
study-of-crm-in-government-1st-edition-alexander-schellong/

The Business of Wine 1st Edition Per V. Jenster

https://ebookname.com/product/the-business-of-wine-1st-edition-
per-v-jenster/

Lying and Deception Theory and Practice 1St Edition

Edition Thomas L. Carson

https://ebookname.com/product/lying-and-deception-theory-and-
practice-1st-edition-edition-thomas-l-carson/

Walking a Line 1st Edition Tom Paulin

https://ebookname.com/product/walking-a-line-1st-edition-tom-
paulin/

Construction Law for Design Professionals Construction

Managers and Contractors 1st Edition Justin Sweet

https://ebookname.com/product/construction-law-for-design-
professionals-construction-managers-and-contractors-1st-edition-
justin-sweet/
Memory history forgetting Ric¶Ur

https://ebookname.com/product/memory-history-forgetting-
ric%c2%b6ur/
Great effort has been undertaken to ensure that the content of this book, as well
as the TSSB program, are correct. However, errors and omissions are
impossible to avoid in a work of this extent. Neither this book nor the TSSB
program are meant as providing professional advice. No guarantee is made that
these items are free of errors and omissions, and the reader assumes full
liability for any losses associated with use of these items.

About this book:

This is a tutorial, with instruction organized from easy to sophisticated. If the

reader just skims through the entire text, hoping to gain an idea of how to use the
TSSB program, the reader will be hopelessly dismayed by the vast complexity
of options. The correct approach is to begin with the first, very simple example
and implement it. Then progress to the next, and so forth. Each example builds
on experience gained from prior examples. In this way, the reader will
painlessly become familiar with the program.

T h e TSSB program may be downloaded (free) from the TSSB website:

http://tssbsoftware.com/

Copyright © 2013 David Aronson and TimothyMasters

All rights reserved

ISBN 978-1489507716
About David Aronson

David Aronson is a pioneer in machine learning and nonlinear trading system

development and signal boosting/filtering. He has worked in this field since
1979. His accomplishments include:

• Started Raden Research Group in 1982 where he oversaw the development

of PRISM (Pattern Recognition Information Synthesis Modeling).
• Chartered Market Technician certified byTheMarket Technicians
Association since 1992.
• Proprietary equities trader for Spear, Leeds and Kellogg 1997 – 2002.

• Was an Adjunct Professor of Finance, teaching a graduate level course in

technical analysis, data mining and predictive analytics to MBA and
financial engineering students from 2002 to 2011.

• Author of “Evidence Based Technical Analysis” published by John Wiley

& Sons 2006. This was the first popular book to deal with data mining
bias and the Monte Carlo Permutation Method for generating bias-free p-
values.
• Co-designer of TSSB (Trading System Synthesis and Boosting), a software
platform for the automated development of statistically sound predictive-
model-based trading systems.

• Developed a method for indicator purification and Pure VIX.

• Innovated the concept of signal boosting: using machine learning to
enhance the performance of existing strategies.

About Timothy Masters

Timothy Masters received a PhD in mathematical statistics with a specialization

in numerical computing. Since then he has continuously worked as a consultant
for government and industry.

• Early research involved automated feature detection in high-altitude

photographs while developing applications for flood and drought
prediction, detection of hidden missile silos, and identification of
threatening military vehicles.

• Worked with medical researchers in the development of computer

 algorithms for distinguishing between benign and malignant cells in needle
biopsies.

• Current focus is methods for evaluating automated financial market trading

systems.

• Authored five books on prediction, classification, and applications of

neural networks:
Practical Neural Network Recipes in C++ (Academic Press,
1993)
Signal and Image Processing with Neural Networks (Wiley,
1994)
Advanced Algorithms for Neural Networks (Wiley, 1995)
Neural, Novel, and Hybrid Algorithms for Time Series
Prediction (Wiley, 1995)
Assessing and Improving Prediction and Classification
(CreateSpace, 2013)

More information can be found on the author's website: TimothyMasters.info

 Table of Contents

Introduction
Two Approaches to Automated Trading
Predictive Modeling
Indicators and Targets
Converting Predictions to Trade Decisions
Testing the Trading System
Walkforward Testing
Cross Validation
Overlap Considerations
Performance Criteria
Model Performance Versus Financial Performance
Financial Relevance and Generalizability
Performance Statistics in TSSB
Desirable Program Features

A Simple Standalone Trading System

The Script File
The Audit Log
A Walkforward Fold
Out-of-Sample Results for This Fold
The Walkforward Summary

A Simple Filter System

The Trade File
The Script File
The Audit Log
Out-of-Sample Results for This Fold
The Walkforward Summary

Common Initial Commands

Market Price Histories and Variables
Quick Reference to Initial Commands
Detailed Descriptions
INTRADAY BY MINUTE
INTRADAY BY SECOND
MARKET DATE FORMAT YYMMDD
 MARKET DATE FORMAT M_D_YYYY
MARKET DATE FORMAT AUTOMATIC
REMOVE ZERO VOLUME
READ MARKET LIST
READ MARKET HISTORIES
MARKET SCAN
RETAIN YEARS
RETAIN MOD
CLEAN RAW DATA
INDEX
READ VARIABLE LIST
OUTLIER SCAN
DESCRIBE
CROSS MARKET AD
CROSS MARKET KL
CROSS MARKET IQ
STATIONARITY
A Final Example

Reading and Writing Databases

Quick Reference to Database Commands
Detailed Descriptions
RETAIN MARKET LIST
VARIABLE IS TEXT
WRITE DATABASE
READ DATABASE
READ UNORDERED DATABASE
APPEND DATABASE
IS PROFIT
A Saving/Restoring Example

Creating Variables
Overview and Basic Syntax
Index Markets and Derived Variables
An Example of IS INDEX and MINUS INDEX
Multiple Indices
Historical Adjustment to Improve Stationarity
Centering
Scaling
Normalization
 An Example of Centering, Scaling, and Normalization
Cross-Market Normalization
Pooled Variables
MEDIAN pooling
CLUMP60 Pooling
Mahalanobis Distance
Absorption Ratio
Trend Indicators
MA DIFFERENCE ShortLength LongLength Lag
LINEAR PER ATR HistLength ATRlength
QUADRATIC PER ATR HistLength ATRlength
CUBIC PER ATR HistLength ATRlength
RSI HistLength
STOCHASTIC K HistLength
STOCHASTIC D HistLength
PRICE MOMENTUM HistLength StdDevLength
ADX HistLength
MIN ADX HistLength MinLength
RESIDUAL MIN ADX HistLength MinLength
MAX ADX HistLength MaxLength
RESIDUAL MAX ADX HistLength MaxLength
DELTA ADX HistLength DeltaLength
ACCEL ADX HistLength DeltaLength
INTRADAY INTENSITY HistLength
DELTA INTRADAY INTENSITY HistLength DeltaLength
REACTIVITY HistLength
DELTA REACTIVITY HistLength DeltaDist
MIN REACTIVITY HistLength Dist
MAX REACTIVITY HistLength Dist
Trend-Like Indicators
CLOSE TO CLOSE
N DAY HIGH HistLength
N DAY LOW HistLength
Deviations from Trend
CLOSE MINUS MOVING AVERAGE HistLen ATRlen
LINEAR DEVIATION HistLength
QUADRATIC DEVIATION HistLength
CUBIC DEVIATION HistLength
DETRENDED RSI DetrendedLength DetrenderLength
Lookback
Volatility Indicators
 ABS PRICE CHANGE OSCILLATOR ShortLen Multiplier
PRICE VARIANCE RATIO HistLength Multiplier
MIN PRICE VARIANCE RATIO HistLen Mult Mlength
CHANGE VARIANCE RATIO HistLength Multiplier
MIN CHANGE VARIANCE RATIO HistLen Mult Mlen
ATR RATIO HistLength Multiplier
DELTA PRICE VARIANCE RATIO HistLength Multiplier
DELTA CHANGE VARIANCE RATIO HistLength
Multiplier
DELTA ATR RATIO HistLength Multiplier
BOLLINGER WIDTH HistLength
DELTA BOLLINGER WIDTH HistLength DeltaLength
N DAY NARROWER HistLength
N DAY WIDER HistLength
Indicators Involving Indices
INDEX CORRELATION HistLength
DELTA INDEX CORRELATION HistLength DeltaLength
DEVIATION FROM INDEX FIT HistLength
MovAvgLength
PURIFIED INDEX Norm HistLen Npred Nfam Nlooks
Look1
Basic Price Distribution Statistics
PRICE SKEWNESS HistLength Multiplier
CHANGE SKEWNESS HistLength Multiplier
PRICE KURTOSIS HistLength Multiplier
CHANGE KURTOSIS HistLength Multiplier
DELTA PRICE SKEWNESS HistLen Multiplier DeltaLen
DELTA CHANGE SKEWNESS HistLen Multiplier
DeltaLen
DELTA PRICE KURTOSIS HistLen Multiplier DeltaLen
DELTA CHANGE KURTOSIS HistLen Multiplier
DeltaLen
Indicators That Significantly Involve Volume
VOLUME MOMENTUM HistLength Multiplier
DELTA VOLUME MOMENTUM HistLen Multiplier
DeltaLen
VOLUME WEIGHTED MA OVER MA HistLength
DIFF VOLUME WEIGHTED MA OVER MA ShortDist
LongDist
PRICE VOLUME FIT HistLength
DIFF PRICE VOLUME FIT ShortDist LongDist
 DELTA PRICE VOLUME FIT HistLength DeltaDist
ON BALANCE VOLUME HistLength
DELTA ON BALANCE VOLUME HistLength DeltaDist
POSITIVE VOLUME INDICATOR HistLength
DELTA POSITIVE VOLUME INDICATOR HistLen
DeltaDist
NEGATIVE VOLUME INDICATOR HistLength
DELTA NEGATIVE VOLUME INDICATOR HistLen
DeltaDist
PRODUCT PRICE VOLUME HistLength
SUM PRICE VOLUME HistLength
DELTA PRODUCT PRICE VOLUME HistLen DeltaDist
DELTA SUM PRICE VOLUME HistLen DeltaDist
Entropy and Mutual Information Indicators
PRICE ENTROPY WordLength
VOLUME ENTROPY WordLength
PRICE MUTUAL INFORMATION WordLength
VOLUME MUTUAL INFORMATION WordLength
Indicators Based on Wavelets
REAL MORLET Period
REAL DIFF MORLET Period
REAL PRODUCT MORLET Period
IMAG MORLET Period
IMAG DIFF MORLET Period
IMAG PRODUCT MORLET Period
PHASE MORLET Period
DAUB MEAN HistLength Level
DAUB MIN HistLength Level
DAUB MAX HistLength Level
DAUB STD HistLength Level
DAUB ENERGY HistLength Level
DAUB NL ENERGY HistLength Level
DAUB CURVE HistLength Level
Follow-Through-Index (FTI) Indicators
Low-Pass Filtering and FTI Computation
Block Size and Channels
Essential Parameters for FTI calculation
Computing FTI
Automated Choice of Filter Period
Trends Within Trends
FTI Indicators Available in TSSB
 FTI LOWPASS BlockSize HalfLength Period
FTI MINOR LOWPASS BlockSize HalfLength LowPeriod
HighPeriod
FTI MAJOR LOWPASS BlockSize HalfLength LowPeriod
HighPeriod
FTI FTI BlockSize HalfLength Period
FTI LARGEST FTI BlockSize HalfLength LowPeriod
HighPeriod
FTI MINOR FTI BlockSize HalfLength LowPeriod
HighPeriod
FTI MAJOR FTI BlockSize HalfLength LowPeriod
HighPeriod
FTI LARGEST PERIOD BlockSize HalfLength LowPeriod
HighPeriod
FTI MINOR PERIOD BlockSize HalfLength LowPeriod
HighPeriod
FTI MAJOR PERIOD BlockSize HalfLength LowPeriod
HighPeriod
FTI CRAT BlockSize HalfLength LowPeriod HighPeriod
FTI MINOR BEST CRAT BlockSize HalfLength
LowPeriod HighPeriod
FTI MAJOR BEST CRAT BlockSize HalfLength
LowPeriod HighPeriod
FTI BOTH BEST CRAT BlockSize HalfLength LowPeriod
HighPeriod
Target Variables
NEXT DAY LOG RATIO
NEXT DAY ATR RETURN Distance
SUBSEQUENT DAY ATR RETURN Lead Distance
NEXT MONTH ATR RETURN Distance
HIT OR MISS Up Down Cutoff ATRdist
FUTURE SLOPE Ahead ATRdist
RSQ FUTURE SLOPE Ahead ATRdist

Screening Variables
Chi-Square Tests
Options for the Chi-Square Test
Output of the Chi-Square Test
Running Chi-Square Tests from the Menu
Nonredundant Predictor Screening
 Options for Nonredundant Predictor Screening
Running Nonredundant Predictor Screening from the Menu
Examples of Nonredundant Predictor Screening

Models 1: Fundamentals
Overview and Basic Syntax
Mandatory Specifications Common to All Models
The INPUT list
The OUTPUT Specifier
Number of Inputs Chosen by Stepwise Selection
The Criterion to be Optimized in Indicator Selection
A Lower Limit on the Number or Fraction of Trades
Summary of Mandatory Specifications for All Models
Optional Specifications Common to All Models
Mitigating Outliers
Testing Multiple Stepwise Indicator Sets
Stepwise Indicator Selection With Cross Validation
When the Target Does Not Measure Profit
Multiple-Market Trades Based on Ranked Predictions
Restricting Models to Long or Short Trades
Prescreening For Specialist Models
Building a Committee with Exclusion Groups
Building a Committee with Resampling and Subsampling
Avoiding Overlap Bias
A Popularity Contest for Indicators
Bootstrap Statistical Significance Tests for Performance
Monte-Carlo Permutation Tests
An Example Using Most Model Specifications
Sequential Prediction

Models 2: The Models

Linear Regression
The MODEL CRITERION Specification for LINREG
Models
The Identity Model
Quadratic Regression
The General Regression Neural Network
The Multiple-Layer Feedforward Network
The Number of Neurons in the First Hidden Layer
The Number of Neurons in the Second Hidden Layer
 Functional Form of the Output Neuron
The Domain of the Neurons
A Basic MLFN Suitable for Most Applications
A Complex-Domain MLFN
The Basic Tree Model
A Forest of Trees
Boosted Trees
Operation String Models
Use of Constants in Operation Strings
Split Linear Models for Regime Regression
An Ordinary SPLIT LINEAR Model
The NOISE Version of the SPLIT LINEAR Model

Committees
Model Specifications Used by Committees
The AVERAGE Committee
The LINREG (Linear Regression) Committee
Constrained Linear Regression Committee
Models as Committees
Creating Component Models for Committees
Exclusion Groups
Explicit Specification of Different Indicators
Using Different Selection Criteria
Varying the Training Set by Subsampling
Varying the Training Set by Resampling

Oracles
Model Specifications Used by Oracles
Traditional Operation of the Oracle
Prescreen Operation of the Oracle
The HONOR PRESCREEN Option
The PRESCREEN ONLY Option
An Example of Prescreen Operation
More Complex Oracles

Testing Methods
Performance for the Entire Dataset
Walkforward Testing
Cross Validation by Time Period
Cross Validation using a Control Variable
 Cross Validation by Random Blocks
Preserving Predictions for Trade Simulation
Market States as Trade Triggers
An Example of Simple Triggering
Triggering Based on State Change
Triggering Versus Prescreening
Commands Common to All Four Examples
Example 1: Model Specialization via PRESCREEN
Example 2: Unguided Specialization
Example 3: Triggering on High Volatility
Example 4: Triggering on Low Volatility

Permutation Training
The Components of Performance
Permutation Training and Selection Bias
Multiple-Market Considerations

Transforms
Expression Transforms
Quantities That May Be Referenced
Vector Operations in Expression Transforms
Vector-to-Scalar Functions
An Example with the @SIGN_AGE Function
Logical Evaluation in Expression Transforms
An Example with Logical Expressions
A More Complex Example
Principal Component Transforms
Invoking the Principal Components Transform
Tables Printed
An Example
Linear and Quadratic Regression Transforms
A Regression Transform Example
The Nominal Mapping Transform
Inputs and the Target
Gates
Focusing on Extreme Targets
Declaring the Transform and its Options
A Nominal Mapping Example
The ARMA Transform
The PURIFY Transform
 Defining the Purified and Purifier Series
Specifying the Predictor Functions
Miscellaneous Specifications
Usage Considerations
A Simple Example

Complex Prediction Systems

Stacking Models and Committees

Graphics
Series Plot
Series + Market
Histogram
Thresholded Histogram
Density Map
Bivariate and Trivariate Plots
Trivariate Plots
Equity
Prediction Map
Indicator-Target Relationship
Isolating Predictability of Direction Versus Magnitude

Finding Independent Predictors

A FIND GROUPS Demonstration

Market Regression Classes

REGRESSION CLASS Demonstrations
The Hierarchical Method
The Sequential Method
The Leung Method

Developing a Stand-Alone System

Choosing Predictor Candidates and the Target
Choosing the Target
Quality Does Not Equal Quantity for Predictors
Predictor and Target Selection for this Study
Stationarity
The Problem of Outliers
Cross-Market Compatibility
 Data Snooping: Friend or Foe?
Checking Stability with Subsampling
How Long Does the Model Hold Up?
Finding Models for a Committee
The Trading System
The Final Test

Trade Simulation and Portfolios

Writing Equity Curves
Performance Measures
Portfolios (File-Based Version)
A Portfolio Example

Integrated Portfolios
A FIXED Portfolio Example
An OOS Portfolio Example
 Introduction
Many people who trade financial instruments would like to automate some or all
of their trading systems. Automation has several advantages over seat-of-the-
pants trading:

• Intelligently designed automated trading systems can and often do

outperform human-driven systems. An effective data-mining program can
discover subtle patterns in market behavior that most humans would not
have a chance of seeing.

• An automated system is absolutely repeatable, while a human-driven

system is subject to human whims. Consistency of decision-making is a
vital property of a system that can consistently show a profit.
Repeatability is also valuable because it allows examination of trades in
order to study operation and perhaps improve performance.
• Most properly designed automated trading systems are amenable to
rigorous statistical analysis that can assess performance measures such as
expected future performance and the probability that the system could have
come into existence due to good luck rather than true power.

• Unattended operation is possible.

Automated trading systems are usually used for one or both of two applications.
TSSB is a state-of-the-art program that is able to generate trading systems that
perform both applications:

• TSSB produces a complete, stand-alone trading system which makes all

trading decisions.

• TSSB produces a model which may be used to filter the trades of an

existing trading system in order to improve performance. It is often the
case that by intelligently selecting a subset of the trades ordered by an
existing system, and rejecting the other trades, we can improve the
risk/reward ratio. TSSB can also suggest position sizes according to the
likelihood of the trade’s success.

Two Approaches to Automated Trading

Whether the user’s goal is development of a stand-alone trading system or a
system to filter signals from an existing trading system, there are two common
approaches to its development and implementation: rules-based (IF/THEN
rules proposed by a human) and predictive modeling.

A rules-based trading system requires that the user specify the exact rules that
make trade decisions, although one or more parameters associated with these
rules may be optimized by the development software. Here is a simple example
of an algorithm-based trading system:

IF the short-term moving average of prices exceeds the long-term moving

average of prices, THEN hold a long position during the next bar.

The above algorithm explicitly states the rule that decides positions bar-by-bar,
although the exact definition of ‘short-term’ and ‘long-term’ is left open. The
developer might use software to find moving-average lookback distances that
maximize some measure of performance. Programs such as TradeStation®
include a proprietary language (EasyLanguage® in this case) by which the
developer can specify trading rules.

With the widespread availability of high-speed desktop computers, an

alternative approach to trading system development has become popular.
Predictive modeling employs mathematically sophisticated software to examine
indicators derived from historical data such as price, volume, and open interest,
with the goal of discovering repeatable patterns that have predictive power. A
predictive model relates these patterns to a forward-looking variable called a
target or dependent variable. This is the approach used by TSSB, and it has
several advantages over algorithm-based system development:

• Intelligent modeling software can discover patterns that are so complex or

buried under random noise that no human could ever see them.

• Once a predictive model system is developed, it is usually easy to tweak

its operation to adjust the risk/reward ratio to suit applications ranging
across a wide spectrum. It can obtain a desired trade off between
numerous signals with a lower probability of success and fewer signals
with a higher probability of success. This is accomplished by adjusting a
threshold that converts model predictions into discrete buy and sell
signals.

• Well designed software allows the developer to adjust the degree of

automation employed in the discovery of trading systems. Experienced
developers can maintain great control over the process and put their
knowledge to work creating systems having certain desired properties,
while inexperienced developers can take advantage of massive
automation, letting the software have majority control.
• In general, predictive modeling is more amenable to advanced statistical
analysis than algorithm-based system development. Sophisticated
statistical analysis algorithms can be incorporated into the model-
generating process more easily than they can be incorporated into systems
based on human-specified rules.
Predictive Modeling
The predictive modeling approach to trading system development relies on a
basic property of market price movement: all markets contain patterns that tend
to repeat throughout history, and hence can often be used to predict future
activity. For example, under some conditions a trend can be expected to
continue until the move is exhausted. Under other conditions, a sudden violent
move will, more often than not, be followed by a retracement toward the recent
mean price. A predictive model studies historical market data and attempts to
discover the patterns that repeat often enough to be profitable. Once such
patterns are discovered, the model will be on the lookout for their reoccurrence.
Based on historical observations, the model will then be able to predict whether
the market will soon rise, fall, or remain about the same. These predictions can
be translated into buy/sell decisions by applying thresholds to the model’s
predictions. We expand on notion this below.

Indicators and Targets

Predictive models do not normally work with raw market data. Rather, the
market prices and other series, such as volume, are usually transformed into two
classes of variables called indicators and targets. This is the data used by the
model during its training, testing, and ultimate realtime use. It is in the definition
of these variables that the developer exerts his or her own influence on the
trading system.

Indicators are variables that look strictly backwards in time. When trading in
real time, as of any given bar an indicator will be computable, assuming that we
are in possession of sufficient historical price data to satisfy the definition of the
indicator. For example, someone may define an indicator called trend as the
percent change of market price from the close of a bar five bars ago to the close
of this bar. As long as we know these two prices, we can compute this trend
indicator. The numerous indicators that TSSB can compute will be discussed in
detail here.

Targets are variables that look strictly forward in time. (In classical regression
modeling, the target is often referred to as the dependent variable.) Targets
reveal the future behavior of the market. We can compute targets for historical
data as long as we have a sufficient number of future bars to satisfy the
definition of the target. Obviously, though, when we are actually trading the
system we cannot know the targets unless we have a phenomenal crystal ball.
For example, we may define an indicator called day_return as the percent
market change from the open of the next day to the open of the day after the next.
If we have a historical record of prices, we can compute this target for every
bar except the last two in the dataset. Targets that TSSB can compute are
discussed here.

The fundamental idea behind predictive modeling is that indicators maycontain

information that can be used to predict targets. The task of a predictive model is
to find and exploit any such information. Consider the following hypothetical
values of two indicators that we choose to call trend and volatility, along with
a target variable that we will call day_return:

Suppose we provide several years of this data to a model and ask it to learn
how to predict day_return from trend and volatility. (This process is called
model training.) Then, we may at a later date calculate from recent prices that
trend=0.225 and volatility=1.244 as of that day. The trained model may then
make a prediction that day_return will be 0.152. (These are all made-up
numbers.) Based on this prediction that the market is about to rise substantially,
we may choose to take a long position.

Converting Predictions to Trade Decisions

Intuition tells us that we should put more faith in extreme predictions than in
more common predictions near the center of the model’s prediction range. If a
model predicts that the market will rise by 0.001 percent tomorrow, we would
not be nearly as inclined to take a long position as if the model predicts a 5.8
percent rise. This intuition is correct, because in general there is a large
correspondence between the magnitude of a prediction and the likelihood of
success of the associated trade. Predictions of large magnitude are more likely
to signal profitable market moves than predictions of small magnitude.

The standard method for making trade decisions based on predicted market
moves is to compare the prediction to a fixed threshold. If the prediction is
greater than or equal to an upper threshold (usually positive), take a long
position. If the prediction is less than or equal to a lower threshold (usually
negative), take a short position. The holding period for a position is implicit in
the definition of the target. This will be discussed in detail here.
It should be obvious that the threshold determines a tradeoff in the number of
trades versus the accuracy rate of the trades. If we set a threshold near zero, the
magnitude of the predictions will frequently exceed the threshold, and a position
will be taken often. Conversely, if we set a threshold that is far from zero,
predicted market moves will only rarely lie beyond the threshold, so trades will
be rare. We already noted that there is a large correspondence between the
magnitude of a prediction and the likelihood of a trade’s success. Thus, by
choosing an appropriate threshold, we can control whether we have a system
that trades often but with only mediocre accuracy, or a system that trades rarely
but with excellent accuracy.

TSSB automatically chooses optimal long and short thresholds by choosing them
so as to maximize the profit factor for long systems and short systems
separately. (See here for the definition of profit factor.) In order to prevent
degenerate situations in which there is only one trade or very few trades, the
user specifies a minimum number of trades that must be taken, either as an
absolute number or as a minimum fraction of bars. In addition, TSSB has an
option for using two thresholds on each side (long and short) so as to produce
two sets of signals, one set for ‘normal reliability’ trades, and a more
conservative set for ‘high reliability’ trades. Finally, in many applications,
TSSB prints tables that show performance figures that would be obtained with
varying thresholds.

Computation of thresholds and interpretation of trade results based on

predictions relative to these thresholds are advanced topics that will be
discussed in detail here. For now, the user needs to understand only the
following concepts:

• The user specifies indicator variables based on recent observed history

and target variables that portray future price movement.

• TSSB is given raw historical market data (prices and perhaps other data,
such as volume) and it generates an extensive database of indicator and
target variables. One or more models are trained to predict the target
given a set of indicators. In other words, the model learns to use the
predictive information contained in the indicators in order to predict the
future as exemplified by the target.

• Every time a prediction is made, the numerical value of this prediction is

compared to a long or upper threshold. If the prediction is greater than or
equal to the long threshold, a long position is taken. Similarly, the
prediction is compared to a short or lower threshold, which will nearly
always be less than the long threshold. If the prediction is less than or
equal to the short threshold, a short position is taken.
• The holding period for a position is inherent in the target variable. This
will be discussed in detail here.

• TSSB will report results for long and short systems separately, as well as
net results for the combined systems.

Testing the Trading System

TSSB provides the ability to perform many tests of a predictive model trading or
filtering system. The available testing methodologies will be discussed in detail
here. However, so that the user may understand the elementary trading/filtering
system development and evaluation presented in the next chapter, we now
discuss two general testing methodologies: cross validation and walkforward
testing. These are the primary standards in many prediction applications, and
both are available in TSSB in a variety of forms.

The principle underlying the vast majority of testing methodologies, including

those included in TSSB, is that the complete historical dataset available to the
developer is split into separate subsets. One subset, called the training set or
the development set, is used to train the predictive model. The other subset,
called the test set or the validation set, is used to evaluate performance of the
trained model. (Note that the distinction between the terms test set and
validation set is not consistent among experts, so the increasingly common
convention is to use them interchangeably. The same is true of training set and
development set.)

The key here is that no data that takes part in the training of the model is
permitted to take part in its performance evaluation. Under fairly general
conditions, this mutually exclusive separation guarantees that the performance
measured in the test set is an unbiased estimate of future performance. In other
words, although the observed performance will almost certainly not exactly
equal the performance that will be seen in the future, it does not have a
systematic bias toward optimistic or pessimistic values. Having an unbiased
estimate of future performance is one of the two main goals of a trading system
development and testing operation. The other goal is being able to perform a
statistical significance test to estimate the probability that the performance level
achieved could have been due to good luck. This advanced concept will be
discussed on Pages here and here.

In the earliest days of model building and testing, when high speed computers
were not readily available, splitting of the data into a training set and a test set
was done exactly once. The developer would typically train the model using
data through a date several years prior to the current date, and then test the
model on subsequent data, ending with the most recent data available. This is an
extremely inefficient use of the data. Modern development platforms should
make available cross validation, walkforward, or both. These techniques split
the available data into training sets and test sets many times, and pool the
performance statistics into a single unbiased estimate of the model-based
trading system’s true capability. This extensive reuse of the data for both
training and testing makes efficient use of precious and limited market history.
Walkforward Testing
Walkforward testing is straightforward, intuitive, and widely used. The
principle is that we train the model on a relatively long block of data that ends a
considerable time in the past. We test the trained model on a relatively short
section of data that immediately follows the training block. Then we shift the
training and testing blocks forward in time by an amount equal to the length of
the test block and repeat the prior steps. Walkforward testing ends when we
reach the end of the dataset. We compute the net performance figure by pooling
all of the test block trades. Here is a simple example of walkforward testing:

1) Train the model using data from 1990 through 2007. Test the model on 2008
data.
2) Train the model using data from 1991 through 2008. Test the model on 2009
data.
3) Train the model using data from 1992 through 2009. Test the model on 2010
data.

Pool all trades from the tests of 2008, 2009, and 2010. These trades are used to
compute an unbiased estimate of the performance of the model.

The primary advantage of walkforward testing is that it mimics real life. Most
developers of automated trading systems periodically retrain or otherwise
refine their model. Thus, the results of a walkforward test simulate the results
that would have been obtained if the system had been actually traded. This is a
compelling argument in favor of this testing methodology.

Another advantage of walkforward testing is that it correctly reflects the

response of the model to nonstationarity in the market. All markets evolve and
change their behavior over time, sometimes rotating through a number of
different regimes. Loosely speaking, this change in market dynamics, and hence
in relationships between indicator and target variables, is called
nonstationarity. The best predictive models have a significant degree of
robustness against such changes, and walkforward testing allows us to judge the
robustness of a model.

TSSB’s ability to use a variety of testing block lengths makes it easy to evaluate
the robustness of a model against nonstationarity. Suppose a model achieves
excellent walkforward results when the test block is very short. In other words,
the model is never asked to make predictions for data that is far past the date on
which its training block ended. Now suppose the walkforward performance
deteriorates if the test block is made longer. This indicates that the market is
rapidly changing in ways that the model is not capable of handling. Such a
model is risky and will require frequent retraining if it is to keep abreast of
current market conditions. On the other hand, if walkforward performance holds
up well as the length of the test block is increased, the model is robust against
nonstationarity. This is a valuable attribute of a predictive model. Look at
Figure 1 on the next page, which depicts the placement of the training and testing
blocks (periods) along the time axis.

Figure 1: Walkforward testing with short and long test periods

Figure 1 above shows two situations. The top section of the figure depicts
walkforward with very short test blocks. The bottom section depicts very long
test blocks. It can be useful to perform several walkforward tests of varying test
block lengths in order to evaluate the degree to which the prediction model is
robust against nonstationarity.

Walkforward testing has only one disadvantage relative to alternative testing

methods such as cross validation: it is relatively inefficient when it comes to
use of the available data. Only cases past the end of the first training block are
ever used for testing. If you are willing to believe that the indicators and targets
are reasonably stationary, this is a tragic waste of data. Cross validation,
discussed in the next section, addresses this weakness.

Cross Validation
Rather than segregating all test cases at the end of the historical data block, as is
done with walkforward testing, we can evenly distribute them throughout the
available history. This is called cross validation. For example, we may test as
follows:

1) Train using data from 2006 through 2008. Test the model on 2005 data.
2) Train using data from 2005 through 2008, excluding 2006. Test the model on
2006 data.
3) Train using data from 2005 through 2008, excluding 2007. Test the model on
2007 data.
4) Train using data from 2005 through 2008, excluding 2008. Test the model on
2008 data.

This idea of withholding interior ‘test’ blocks of data while training with the
surrounding data is illustrated in Figure 2 below. In cross validation, each step
is commonly called a fold.

Figure 2: Cross validation

The obvious advantage of cross validation over walkforward testing is that

every available case becomes a test case at some point. However, there are
several disadvantages to note. The most serious potential problem is that cross
validation is sensitive to nonstationarity. In a walkforward test, only relatively
recent cases serve as test subjects. But in cross validation, cases all the way
back to the beginning of the dataset contribute to test performance results. If the
behavior of the market in early days was so different than in later days that the
relationship between indicators and the target has seriously changed,
incorporating test results from those early days may not be advisable.

Another disadvantage is more philosophical than practical, but it is worthy of

note. Unlike a walkforward test, cross validation does not mimic the real-life
behavior of a trading system. In cross validation, except for the last fold, we are
using data from the future to train the model being tested. In real life this data
would not be known at the time that test cases are processed. Some skeptics
will raise their eyebrows at this, even though when done correctly it is
legitimate, providing nearly unbiased performance estimates. Finally, overlap
problems, discussed in the next section, are more troublesome in cross
validation than in walkforward tests.

Overlap Considerations
The discussions of cross validation and walkforward testing just presented
assume that each case is independent of other cases. In other words, the
assumption is that the values of variables for a case are not related to the values
of other cases in the dataset. Unfortunately, this is almost never the situation.
Cases that are near one another in time will tend to have similar values of
indicators and/or targets. This generally comes about in one or both of the
following ways:

• Many of the targets available in TSSB look further ahead than just the next
bar. For example, suppose our target is the market trend over the next ten
bars. This is the quantity we wish to predict in order to make trade
decisions. If this value is high on a particular day, indicating that the
market trends strongly upward over the subsequent ten days, then in all
likelihood this value will also be high the following day, and it was
probably high the prior day. Shifting ahead or back one day still leaves an
overlap of nine days in that ten-day target window. Such case-to-case
correlation in time series data is called serial correlation.

• In most trading systems, the indicators look back over a considerable time
block. For example, an indicator may be the market trend over the prior 50
days, or a measure of volatility over the prior 100 days. As a result,
indicators change very slowly over time. The values of indicators for a
particular day are almost identical to the values in nearby days, before and
after.

These facts have several important implications. Because indicators change

only slowly, the model’s predictions also change slowly. Hence market
positions change slowly; if a prediction is above a threshold, it will tend to
remain above the threshold for multiple bars. Conversely, if a prediction is
below a threshold, it will tend to remain below that threshold for some time. If
the target is looking ahead more than one bar, which results in serial correlation
as discussed above, then the result of serial correlation in both positions and
targets is serial correlation in returns for the trading system. This immediately
invalidates most common statistical significance tests such as the t-test, ordinary
bootstrap, and Monte-Carlo permutation test. TSSB does include several
statistical significance tests that can lessen the impact of serial correlation. In
particular, the stationary bootstrap and tapered block bootstrap will be
discussed here. Unfortunately, both of these tests rely on assumptions that are
often shaky. We’ll return to this issue in more detail later when statistical tests
are discussed. For the moment, understand that targets that look ahead more than
one bar usually preclude tests of significance or force one to rely on tests having
questionable validity.

Lack of independence in indicators and targets has another implication, this one
potentially more serious than just invalidating significance tests. The legitimacy
of the test results themselves can be undermined by bias. Luckily, this problem
is easily solved with a TSSB option called OVERLAP. This will be discussed
here. For now we will simply explore the nature of the problem.

The problem occurs near the boundaries between training data and test data. The
simplest situation is for walkforward testing, because there is only one (moving)
boundary. Suppose the target involves market movement ten days into the future.
Consider the last case in the training block. Its target involves the first ten days
after the test block begins. This case, like all training set cases, plays a role in
the development of the predictive model. Now consider the case that
immediately follows it, the first case in the test block. As has already been
noted, its indicator values will be very similar to the indicator values of the
prior case. Thus, the model’s prediction will also be similar to that of the prior
case. Because the target looks ahead ten days and we have moved ahead only
one day, leaving a nine-day overlap, the target for this test case will be similar
to the target for the prior case. But the prior case, which is practically identical
to this test case, took part in the training of the model! So we have a strong
prejudice for the model to do a good job of predicting this case, whose
indicators and target are similar to the training case. The result is optimistic
bias, the worst sort. Our test results will exceed the results that would have
been obtained from an honest test.

This boundary effect manifests itself in an additional fashion in cross validation.

Of course, we still have the effect just described when we are near the end of
the early section of the training set and the start of the test set. This is the left
edge of the red regions in Figure 2. But we also have a boundary effect when we
are near the end of the test set and the start of the later part of the training set.
This is the right edge of each red region. As before, cases near each other but on
opposite sides of the training set / test set boundary have similar values for
indicators and the target, which results in optimistic bias in the performance
estimate.

The bottom line is that bias due to overlap at the boundary between training data
and test data is a serious problem for both cross validation and walkforward
testing. Fortunately, the user can invoke the OVERLAP option to alleviate this
problem, as will be discussed here.

Performance Criteria
Prior sections discussed the evaluation of performance using cross validation or
walkforward testing. Those sections dealt with the mechanics of partitioning the
data into training and test sets. Other considerations will be presented here.
Several terms should be defined first:

• A fold is a single partitioning of the available data into a training set, a test
set, and perhaps some cases from the dataset that are temporarily omitted
in order to handle overlap. In Figures 1 and 2, a fold would be one of the
horizontal strips consisting of the green training block and the red test
block.

• The in-sample (or IS) performance for a fold is the performance of the
model or trading system in the current training set. Because the training
process for the model optimized some aspect of its performance, the in-
sample performance will usually have an optimistic bias, often called the
training bias.

• The out-of-sample (or OOS) performance for a fold is the performance of

the model or trading system in the current test set. Because this data did
not take part in training the model, it is, for all practical purposes,
unbiased. In other words, on average it will reflect the true capability of
the model or trading system.

Model Performance Versus Financial Performance

Performance statistics for model-based trading systems fall into two categories.
One is the predictive performance of the model that determines trade decisions.
This may include statistics such as the mean squared error of the predictions, or
the model’s R-squared. The other is the financial performance of the trading
system, such as its profit factor or Sharpe ratio. Naturally, there is a degree of
correspondence between statistics in these two categories. If a model has
excellent performance, the trading system probably will as well. By the same
token, a poorly performing model will most likely produce a poorly performing
trading system.
Exploring the Variety of Random
Documents with Different Content
1713 : trau*lated to Clogher, 1717 ; vice-chancellor, Dublin
University, 1721 : gave money to erect printing-house, 1726;
founded exhibitions HI Trinity College, Dublin ; bequeathed aUo
books and valuable manuscript* ; his 'Tractates de VisiUtione
Inflrmorum' (1697) translated as 'The Curate's Manual,' 1M". [liv.
196] STERNE, LAURENCE (1713-1768), humorist and sentimentalist
; great-grandson of Richard Sterne [q. v.] ; born at Clonmel ; the sou
of a subaltern in the army ; after some years' wandering spent eight
years at school in Halifax ; left penniless ; sent to Cambridge by a
cousin (Richard) : sizar, Jesus College, 1733, exhibitioner, 1734 :
matriculated, 1736 ; intimate with John Hall-Stevenson [q. v.] ; M.A.,
1740 ; curate of Buckden ; became vicar of Sutton-iu-tbe-Forest,
1738, and prebendary of York, 1741, also obtaining a sinecure office
; married Miss Lamley, 1741 ; first used epithet ' sentimental ' in
letter, 1740 ; obtained, by wife's influence, Stillingtou, parish
adjoining Sut ton, 1743, being also chaplain to fourth Earl of Aboyue
; 1 in local private enclosure act*, 1766, 1766; dabbled In painting
and music ; joined in orgies of ' Demouiacks ' at Hall-Mcven-on •
house : showed affection for hU only child; troubled by his mother's
demands for money : said to have let her die of nturvation ;
published, 1750, assize sermon, inserted in 'Tristram Shandy';
satirised Yorkshire lawyer as ' Trim ' in sketch first published, 1769,
as' A Political Romance addressed to Esq. of York ' (rareX afterwards
reprinted as • History of a Warm Watch Coat' ; unfaithful to his wife,
who became insane, 1768 ; began 'Tristram Shandy,' 1769 ; carried
on flirtation with Mademoiselle Fourmantelle; vols. i. and ii. of his '
Tristram Shandy ' orbited at York by publisher of sermons, 1760 ;
taken by Croft to London, where he was well received in society :
second edition of bis ' Tristram ' and S-MIUHK of Mr. Yorick ' issued
by Dodsley, 1760, the latter being commended by Gray ; his '
Tristram Shandy * unpopular in Yorkshire on account of its
recognisable portraits of local characters, such as Dr. John Burton
(1710-1771) [q. v.]('S]op'); d« -mm need by Dr. , Richardson,
Horace Walpole, Uoldnmith, and o moral and literary grounds :
ineffectual attempt* made by Warburton to restrain his obscenity :
many pamphlet* issued against him, 1760-1 ; received perpetual
curacy of Cox wold, 1760 ; called bouse at Cbxwold (which now baa
tablet) Shandy Hall ; visited London to superintend publication of
vols. lit and Iv. of ' Trirtram,' 1761 : preached at Foundling Hospital,
London ; vols. v. ami vi. of • Tristram ' dedicated to Lord Spencer,
lasoed for him by Becket,
The text on this page is estimated to be only 23.79%
accurate

STERNE 1244 STEVENS December 1761 ; went abroad for

his health, January 1762 ; well received in I'ars >.-iloiis ; entertained
by Fox at St. (.iermain ; joined by wife and daughter ; lived at
Toulouse, and, till February 1764, in the south of France ; left wife
and daughter behind at Montauban, by their wish ; spent April-May
1764 in Paris, seeing much of Wilkes and preaching at English
ambassador's chapel ; published books vii. and viii. of ' Tristram,'
1765 : painted by Gainsborough at Bath, 1765; began seven
months' tour described in ' Sentimental Journey,' October 1765;
went by Paris, Lyons, and Savoy to Italy ; well received at Rome,
where he met Smollett ('Smelfungus'); in Naples, February 1766 ; in
Yorkshire again, June 1766 ; issued book ix. of 'Tristram Shandy,'
with dedication to Chatham, containing sentence supposed to have
suggested Burns's lines about the 'guinea stamp,' 1767; also
published vols. iii. and iv. 6f 'Sermons,' Voltaire being among the
subscribers : during visit to London, December 1766 to May 1767,
met, at house of Sir William James [q. v.], Mrs. Eliza Draper [q. v.] ;
kept journal addressed to her ('The Bnimine's Journal'; manuscript in
British MnseumX April to August 1767, after her departure for India ;
after two months' visit from wife and daughter at Cox wold,
arranged permanent separation : parted reluctantly with daughter;
his 'Sentimental Journey' issued, 1768 : died of pleurisy in Old Bond
Street lodgings, London ; buried in St. George's cemetery, Bays
water Road, London ; body said to have been ' resurrected ' ami sold
to Charleg Gollignon [q. v.], the skeleton being long preserved at
Cambridge; inscription on stone at St. George's erected by 'two
brother masons' inaccurate ; left no will, and died insolvent; wife
and daughter relieved through subscriptions collected by Hall-
Stevenson and Mrs. Draper, and by publication (1769) of three more
volumes of sermons; publication of his letters to Mr.?. Draper
threatened by widow In default of blackmail ; his letters published
by his daughter (Madame Medalle), 1775 ; publication of ' Letters of
Yorick to Eliza ' ( 1766-7) authorised by Mrs. Draper, 1775. The »
Letters from Eliza to Yorick,' 1775, and 'Letters supposed to have
been written by Yorick and Eliza,' 1779, are forgeries. Among
fraudulent imitations of his writings were John Carr's third volume of
'Tristram Shandy,' 1760, J. HallStevenson's continuation of the
'Sentimental Journey,' 17G9, and Richard Griffith's ' Posthumous
Works of a late celebrated Genius,' 1770(included in first collected
edition). His works contain many literary thefts, and the general
scheme of 'Tristram' closely resembles that of John Dunstan's 'A
Voyage round the World ... the rare adventures of Don Kainophilus'
(1720?). But his style is original, and his characters are of the first
class. First collective edition of 'Tristram Shandy' published, 1767,
last, 1779; 'Sermons of Mr. Yorick,' first reissued collectively, 1775,
last, 1787; 'Sentimental Journey' reissued with plates, 1792; first
collective edition of complete works (without letters) published, 1779
(Dublin): best early edition (with letters and Hogarth's plates)
published, 1780; Dr. J. P. Browne's (containing much newly
recovered correspondence), 1873. 'The 'Sentimental Journey ' has
been often translated. [iiv. 199] STERNE, RICHARD (1596?-1683),
archbishop of York ; scholar, Trinity College, Cambridge, 1614, M.A.,
SXJHfc^!? (°*ford, 1627); fellow, Beiiet (Corpus Chnsti) College,
1620; master of Jesus College, Cambridge, 1634 (deprived, 1644);
chaplain to Laud, c. 1633 ; rector of Yelverton and Harleton, 1634-
44; D.D., 1635arrested for royalism at Cambridge and imprisoned
16421645 ; allowed to attend Laud in Tower of London,' 1645 ;
bishop of Carlisle, 1660-4; said to have been a reviser of prayer-
book, 1662; archbishop of York, 1664-83; founded scholarships at
Jesus and Corpus Christi Colleges Cuii.Sri.lire ; assisted Brian Walton
[q. v.] with Polyglott! and published 'Summa Logics,' 1«85 : ' Whole
Duty of dan, ascribed to him among others, probably by Richard
Allestree [q. v.] [liv> 22i] STERNHOLD, THOMAS (c/. 1549), joint
versifier of tl .- RntaM with John Hopkins (d. 1570) fa. v.] ; according
to Wood, of Christ Church, Oxford groom of the robeg to Henry VIII,
1638, receiving legacy from him! probably M.P., Plymouth, 1645-7;
nineteen psalms by Lium-lf only (in metre of 'Chevy Chase')
contained in st Miition of Psalms (undated, dedicated to Eda VI);
geven added in second edition (1549), and three in edition of 1561 ;
forty in complete collections. [Iiv. 223] 8TEREY, NATHANIEL (d.
1698), dean of Booking fellow of Merton College, Oxford, 1649 ; B.A.
Cambridge. 1G4K : 15.D. Oxford, 1675 ; dean of Bocking, 1674-98.
STERRY, PETER (d. 1672), CromweM's^chaphiin ; brother of
Nathaniel Sterry [q. v.] ; fellow of Emmanuel College, Cambridge,
1636; M.A., 1637; member of Westminster Assembly; Cambridge
platonist ; preacher to council of state, 1649 ; after Restoration held
conventicle • satirised in ' Htidibras '; chief work, ' Discourse of the
Freedom of the Will ' (posthumous), 1675. [Iiv. 224] STEUART. [See
also STEWARD, STEWART, and STUART.] STEUART, 6m HENRY
SETON, first baronet (17591836), of Allanton ; agriculturist; claimed
for his family ancestry of the Stewarts of Lennox, Durnley, and
Castlemilk, 1799 ; served in the army, 1778-87 ; advocated canal
from Lancashire coalfields to Edinburgh, 1801- edited Sallust, 1806;
LL.D. (Edinburgh) and F.R.S.E.: his system of transplantation
adopted by Sir Walter Scott at Abbotsford ; his ' Planter's Guide,'
1828 (reissued, 1848) reviewed by Sir Walter Scott and 'Christopher
North'* created baronet, 1814. [iiv. 225] STEUART or STEWART, SIR
JAMES (1635-1715), lord-ad vooate ; barrister, 1661 ; outlawed as
contriver of Argyll's rebellion, 1685 ; pardoned by William of Orange;
as lord-advocate of Scotland, 1692-1709, introduced legal reforms.
£iiv. 227] STEUART, SIR JAMES, the elder (afterwards DKXHAM)
(1712-1780). [See DENHAM.] STEUART (formerly DENHAM), SIR
JAMES the younger (1744-1839). [See DENHAM.] STEUART-SETON,
REGINALD MACDONALD (1778-1838), sheriff of Stirlingshire and
friend of Scott • originally Macdonald of Staffa , son-in-law of Sir
Henry Seton Stctiart [q. v.] [iiv. 227] STEVENS, ALFRED (1818-
1875), artist ; son of a house-painter ; enabled to study art in Italy,
1833-42, by private liberality; employed by Thorwaldsen at Rome
1841-2; teacher in school of design, London, 1846-7employed by
Hoole & Co. of Sheffield ; gained first prize for designs in metal-
work at exhibition of 1851 ; executed Wellington monument at St.
Paul's Cathedral (unfinished)designed vases and lions at British
Museum. [Iiv. 227] STEVENS, FRANCIS (1781-1823), landscape-
painter ; exhibited at Royal Academy, 1804-5, 1819, and 1822;
member of Water-colour Society, 1809. [Hv. 229] STEVENS, GEORGE
ALEXANDER (1710-1784), author of ' A Lecture upon Heads ' ; wrote
' The History of Tom Fool' (1760), and various other pieces: his
lectures first given in the Haymarket, London, 1764, and afterwards
in the country and in America ; his ' Lecture on Heads ' published
spuriously, 1770 ; sold to Charles Lee Lewes [q. v.], 1774 (first
authentic edition, 1785); published 'Songs, Comic and SatyrLcal'
(cuts by Bewick), 1772; sole dramatic success, 'The Trip to
Portsmouth* (acted, 1773). [llv. 2291 STEVENS or STEPHENS, JOHN
(d. 1726), translator and antiquary ; said to have served in James
II's army iii Ireland ; described in his books as 'Captain '; translated
Bteda's 'Ecclesiastical History,1 and Portuguese, Spanish (including
Quevedo's 'Pablo de Segovia 'X and French works; published also
translation and abridgment of Dugdale's ' Monasticou,' 1718
(continuation, 1722). [liv. 231] STEVENS, RICHARD JOHN SAMUEL
(1757-1837), musician ; friend of Samuel Birch (1757-1841) [q. v.] ;
organist at Temple Church, London, 1786, at Charterhouse, London,
1796-1837 ; Gresham professor of music, 1801-37 ; edited collection
of sacred music, 1802 ; composed glees, sonatas, and songs. [Hv.
232] STEVENS, WILLIAM (1732-1807), biographer and editor of
William Jones of Nayland [q. v.], 1801; published posthumous
sermons of his cousin, bishop George Home [q. v.]; his 'Essay'
(1799) against relaxation of subscription to Thirty-nine Articles thrice
reissued ; joined William Jones in founding 'Society for the
Reformation of Principles'; 'Nobody's Friends' (club) founded in his
honour, c. 1800, originating from pseudonym used by him in
pamphlet ; treasurer, Queen Anne's tliv 233]
The text on this page is estimated to be only 20.29%
accurate

STEVENS 1-Jl.-, STEVENSON STEVENS, WILLIAM UMJSHAW

(1760-1800). I> 17-jj I; M.A., i77:«. D.H., i; 177H IHOO: rhaplain to
, v.l: inciiiiiiM-ntof omtAlns renrints from periodicals, and an
nnprinted contains repriuto from periodicals, fragment of family
history.
The text on this page is estimated to be only 24.46%
accurate

STEVENSON 1246 STEWART )N, SETH WILLIAM (1781-

1853), antiquary ; son of William Stevenson (1741-1821) [q. v.] ;
proprietor and editor of ' Norfolk Chronicle ' ; mayor of -;•_': K.S.A..
1827; his « Dictionary of Roman Coins' (completed by Frederic
William Madden) published, 1889. [liv. 254] STEVENSON, THOMAS
(1818-1887), engineer and meteorologist ; son of Robert Stevenson
[q. v.] ; educated at Edinburgh ; joint-engineer to board of northern
lighthouses, 1853-85 ; continued experiments of Alan Stevenson [Q.
T.] in lighthouse illumination ; invented and perfected 'azirautbal
condensing system'; president, Royal Society of Edinburgh, 1885;
president, Royal Scottish Society of Arts, 1859-60 ; M.I.O.E., 1864 ;
honorary secretary, Scottish Meteorological Society, 1871 ; designed
( 1864) Stevenson screen for thermometers, and made other original
contributions to meteorology ; wrote on lighthouse illumination and
harbour construction, [liv. 255] STEVENSON, WILLIAM (1719?
-1783), physician; M.D. Edinburgh ; served in army ; practised at
Coleraine, Wells (Somerset), Bath, uud Newark ; Jacobite ; opposed
bleedings and prevalent use of drugs ; published medical [liv. 256]
STEVENSON, WILLIAM (1741-1821), proprietor of 'Norfolk Chronicle'
and publisher; edited Campbell's ' Lives of British Admirals.' [liv. 254]
STEVENSON, WILLIAM (1772-1829), keeper of treasury records ;
nonconformist minister, classical tutor, Manchester Academy ; farmer
; editor of ' Scots Magazine,' c. 1797; keeper of the records in the
treasury, 1806-29 ; published agricultural and other works, compiled
' Annual Register,' and contributed to ' Edinburgh Review ' and other
periodicals. [liv. 256] STEVENSON, W. B. (fl. 1803-1825), author of '
A Historical and Descriptive Narrative of twenty years' residence in
South America' (1825); imprisoned by Spanish at Concepcion,
Callao, and Lima ; joined revolutionists at Quito; governor of the
Esmeraldas, 1810; secretary to Lord Oochraue, 1818. [liv. 257]
STEVENSON, WILLIAM FLEMING (1832-1886), Irish presbyterian
divine ; M.A. Glasgow, 1851 ; hon. D.D. Edinburgh, 1881; studied
also in Germany; town missionary in Belfast, 1857 ; minister of
Rathgar, Dublin, 1860-86; convener of Irish General Assembly's
foreign missions, 1873; made missionary tour round the world, 1877
; professor of evangelistic theology, New College, Edinburgh, 1879-
80 ; moderator of general assembly, 1881 ; chaplain to viceroy of
Ireland, 1886 ; published ' Praying and Working,' 1862. [liv. 258]
STEWARD. [See also STEUART, STEWART, and STCART.] STEWARD
or STEWART, RICHARD (1593 ?-1651), j dean designate of St. Paul's
Cathedral and Westminster : ' of Westminster and Magdalen Hall,
Oxford ; M.A., 1615, ! D.C.L., 1624 ; fellow of All Souls College,
Oxford, 1613 ; canon of Salisbury ; clerk of the closet to Charles 1,
1633 ; dean of Chichester, 1635 ; provost of Eton, 1639-43 ; dean of
the Chapel Royal, 1643 ; dean designate of St. Paul's, j London,
1641, of Westminster, 1645 ; deprived by parliament ; defended
episcopacy at Uxbridge conference^ 1645 ; ecclesiastical adviser to
Charles II ; published theological works ; died in Paris. [liv 258]
STEWARD, ROBERT THE (1316-1390). ROBERT IL] [See STEWARD,
8TYWARD, or WELLS, ROBERT (d. 1557), first dean of Ely ; M.A.
Cambridge, 1520 ; prior of Ely, 1522; maintained validity of Henry
VIII's marriage with Catherine of Arragon, 1529, but afterwards
carried out royal policy; surrendered Ely monastery, 1559, receiving
pensions; dean of Ely, 1541-57; continued ' Historia Eliensis ' from
1486 to 1554. [liv. 260] STEWARD, SIR SIMEON (d. 1629 ?), poet ;
related to , Robert Steward [q. v.]; of Trinity Hall, Cambridge;
knighted, 1603; M.P., Shaftesbury, 1614, Aldeburgh, 1627; his 'Fairy
King' (1635) reprinted, 1656, and later I by Bliss, 1813, and Mr. A. E.
Wait*, 1888 (' Elfin Music '). STEWARD, THOMAS (16697-1753),
preLbyterian divine; minister at Debenham, 1689-1706, Cook Street,
Dublin. 1706-24, Bury St. Edmunds, 1724-53 ; hon. D.D. Aberdeen,
1733 ; correspondent of Francis Hutcheson (1694-1746) [q. v.] and
Philip Doddridge [q. v.] [liv. 261] STEWARDSON, THOMAS (1781-
1859), portrnitpainter ; exhibited at Royal Academy, 1804-29 ;
among his sitters Canning, Lord Liverpool, and Grote, [liv. 2621
STEWART. [See also STEUART, STEWARD, and STUART.] STEWART,
ALEXANDER, EARL OP BUCHAX and LOUD OK HAI>I-:\«ICH (1343 ?
-1405 ?), 'the Wolf of Badenoch' ; son of Robert II of Scotland by
Elizabeth Mure : granted Badenoch, 1371 ; king's lieutenant north of
j Forth, 1372; became Earl of Buchan and acquired !;.«>by
marriage, c. 1382 ; censured for deserting wife by bishops of Moray
and Ross, 1389 ; in revenge burnt Fonvand Elgin, 1390 ; his tomb in
church of Dunkeld. [liv. 262] STEWART, ALEXANDER. EARL OF MAR
(1375 ?1435), natural son of Alexander Stewart, the ' Wolf of
Badenoch ' [q. v.] ; having previously instigated the murder of her
husband, besieged at Kildrummy Isabel (d. 1408), countess of Mar,
and heiress of the Douglas estates, 1404 ; compelled her to make
him his heir and to marry him ; ambassador to England, 1406-7 ;
helped to restore John of Burgundy at Liege, 1408 ; defeated Lord of
the Isles at Harlaw, 1411 ; conservator of truce with England and
warden of the marches, 1424; though adherent of Albany not
disgraced by James I. [liv. 263] STEWART, ALEXANDER, DUKE OF
ALBANY and EARL OF MARCH (1454 ?-1485), second son of James
II of Scotland [q. v.] ; earl of March and lord of Auuandale, 1455 :
duke of Albany before 1458 ; captured by English ship on way to
Guelderland, but released, 1454; high admiral of Scotland and
warden of the marches; lieutenant of Scotland, 1472 ; fortified
Dunbar against royal forces and was imprisoned at Edinburgh, 1479
; escaped to Paris; well received by Louis XI; married (second wife)
Anne de la Tour d'Auvergne, 1480 ; concluded with Edward IV treaty
of Fotheringay, 1482, agreeing to rule Scotland as his vassal ;
returned to Scotland with English army, but agreed to be faithful to
James III if restored to his estates, 1482 ; made truce with English,
surrendering Berwick; created earl of Mar and Gairloch, and
reconciled to James III ; attempted to seize James Ill's person, 1482
; made new treaty with Edward IV, 1483, agreeing to help in
conquest of France ; after death of Edward IV indicted and
outlawed, 1483 ; fled to England ; with Douglas made raid on
Lochmaben ; accidentally killed in France. [liv. 264] STEWART,
ALEXANDER (H93 ?- 1513), archbishop of St. Andrews ; natural son
of James IV ; appointed archbishop in boyhood; visited the
Netherlands and France, c. 1506 ; studied under Erasmus at Padua,
1508 ; praised in Erasmus's 'Adagia'; lord chancellor of Scotland, c,
1510 ; co-founder of college of St. Leonard's, St. Andrews, 1512 ;
killed at Floddeu. . [liv. 267] STEWART, ALEXANDER, fifth EARL OF
MORAY (d. 1701), Scottish statesman; justice- general, 1674; a lord
of the treasury, 1678 ; extraordinary lord of session. 1680; secretary
of state, 1680-8; active in repressing covenanters ; high
commissioner to Scottish parliament, 1686 ; K.T., 1687 ; deprived at
revolution. [liv. 268] STEWART, ALEXANDER, fifth LORD BLANTYRE
(d. 1704), raised regiment for William III, and received pension;
seceder from convention, 1702; took seat in Scottish parliament,
1703. [liv. 268] STEWART, ALEXANDER (d. 1795), of Inveruahyle,
Jacobite; 'out' with Stewarts of Appiu, 1716 and 1745 ; distinguished
at Prestonpans, 1745 ; wounded at Culloden, 1746; pardoned under
act of indemnity; introduced Sir Walter Scott to the highlands, 1787.
[Hv. 269] STEWART, ALEXANDER PATRICK (1813-1883), physician;
M.D.Glasgow, 1838; physician, Middlesex Hospital, 1855-66 ;
F.R.O.P., 1855 ; active member of British Medical Association; his
paper differentiating typhus from typhoid fever (Parisian Medical
Society, 1840) reprinted, 1884. [liv. 269] STEWART, ANDREW, first
BARON AVANDALK or AVONDALE (d. 1488), chancellor of Scotland ;
grandson of Murdac Stewart, second dukeof Alb.any [q. v.] ;
knighted, and probably educated in England ; member of general
The text on this page is estimated to be only 21.76%
accurate

STEWART 1247 STEWART council at Stirling, 1440; in

household of Jama II; .'.aucellor of Scotland, Orkney and Sbetl.i;
Dciiinark : !>. -'.•_•,•.! Allunv's castle at Dunbar, 1479 ; deserted
James III at Lander. 1 >S2 ; lu-lped to effect -:' ; Hmbawador to
France, 1484. [HT. 170] STEWART, ANDREW, second BARON O<
HILTRRK (./f. 1548-1598), a lord of the congregation, ISM bond for
expulsion of French from Scotland, and subBcribed book of
discipline. 1560: act-ompanial Holyrood, isr.3 : joined Moray's rising
against Darnley'i marriage, 1568, and the conspiracy againat
wounded by Henries at Langside, 1668; n. ton's privy council, 1578 ;
attempted mediation between Huntly and Moray (second earlX 1692;
ufti-r Hmitly1* treachery tried to attack him in the highland*, i.ut liild
to retire to England, 1594; panloned by James VI, 1595, on agreeing
to abandon Bothwell ; lieutenant on the . 15V)H. [liT. >71]
STEWART, ANDREW (rf . 1671 X Irish divine ; minister of
Donaghadee, c. 1615 ; grouted government salary, 1664 ; helped to
draw up • Act of fiangor,' 1664 ; ejected, 1661 ; imprisoned on
suspicion of complicity in Blood's plot, 1663; part of his 'Short
Account' of the Irish church printed by Dr. W. D. Killeu, 1866. [liv.
373] STEWART, ANTHONY (1773-1846), miniaturepainter ; executed
earliest miniatures of Queen Victoria ; exhibited at Royal Academy.
[liv. 273] STEWART or STUART, LADY ARABELLA (15751615). [See
ARABELLA.] STEWART or STUART, ARCHIBALD .IAMF.S EDWARD
(1748-1827). [See DOUGLAS, AHOIIBAU) JAMBS EDWARD, first
BARON DOUGLAS ox DOUGLAS.] STEWART, BALFOUR (1828-1887),
physicist and meteorologist: assistant to Professor Forbes at
Edinburgh, 1856; director of Kew Observatory, 1859-71; F.R^S.,
1862: secretary to government meteorological committee, 1867-9;
professor of natural philosophy, Owens College, Manchester, 1870-
87; a founder of Society for Psychical Research (president, 1885-7) ;
president of Physical Society and of Manchester Literary and
Philosophical society, 1887 ; Rumford medallist, 1868, for researches
on radiant heat, which helped to lay foundation of spectrum analysis
: demonstrated applicability of law of radiation to polarised rays of
light, 1860 ; suggested variations in a primary electric current in the
sun as cause of aurorte, magnetic storms, and earth currents 1860;
investigated sun-spots: made calculations as to fellow of AU Souls
College, Oxford, 1796: D.D.. 1816; [llv.177] STEWART (afterwards
YAH1), CHAKLl. LIAM, third MARQOH or • - • T ..:;:...--. ,,- • : . ...-. .
-....'. .. .' . -i.l.r. 1 .-....-. I. r i V,.r. «taff In Netherlands, 1794-5 ; with
Aurtrmus on Rhine ..':••;.-..;••, .,.'.'... ..:..; lieutenant-colonel, 5th
dragoon*, 1797, Iftth l '.»» : nnder-seorcurj In Irelan-l M.I'..
Thomi^u.wn. 17HH 1HU» : M.I', co. Deny, 1800-14; commanded
with sequel, 1878. Other works include text-books on physics and
work on sun-spots (with Warren de la Rue ' [q. v.] and Benjamin
Loewy), 1865-8. [liv. 273] STEWART, BERNARD or BERADLT, third
SKIONEUR D'AuBiONY (1447 ?-1508). [See STUART.] STEWART,
LORD BERNARD, titular EARL OF LICHFIELD (1623 7-1645). [See
STUART.] STEWART, CHARLES, sixth DUKK OF LBNNOX and third
DUKE OF RICHMOND (1640-1672). [See STUART.] STEWART,
CHARLES (1775 - 1812), lieutenantcolonel ; lieutenant, 71st
Highlanders, 1791; wounded at Seringapatam, 1792 ; lieutenant-
colonel, 50th foot, 1805 : commanded first battalion in Walcberen,
1809, and in the Peninsula, 1811-12 ; died at Corla. [liv. 276]
STEWART, CHARLES (1764-1837), orientalist; served In Bengal army,
1781-1808; assistant-professor of Persian, Fort William College,
Calcutta, 1800-6; mnf^ii i of Arabic, Persian, and Hindustani at
Haileybury, 1807-27 ; edited and trauslated oriental works. [liv. 276]
STEWART or STUART, CHARLES EDWARD (1720-1788), the Young
Pretender. [See CHARLES BDWARD Louis PHILIP CASIMIR.]
STEWART, CHARLES JAMES (1775-1837), bishop of Quebec; brother
of Sir William Stewart (1774-1827) [q v.] ; B.A. Corpus Christ!
College, Oxford, 1795 ; under-secretory for war, 1807-9 ; brigade
under Sir John Moore, 1806 ; adjutant-general to Wellington, 1809-
12; major-general, 1 at Talavera and Poentes d'Onoro; groom of the
bedchamber, 1812; K.B., 1813; British minister to Prussia, 1813 ;
signed treaty with Prussia and Rtuaia at Dresden ; severely wounded
at Kulm. 1813 : Induced BenudoUe to take active part fur allies ;
created privy councillor, and Baron Stewart, 1814 ; ambassador at
Vienna during congress, 1814 : envoy at Troppau, 1820, Laybach,
1821, Verona, 1822 ; lord of the bedchamber, 1890-7- succeeded
Castlereagh as marquis, 1822 ; took surname of Vane on second
marriage, 1819 ; appointed to St. Petersburg embassy, 1835, but
withdrew owing to parliamentary 'opposition; travelled in eastern
and southern Europe; fought duel with GrattunV son, 1839; general,
1887; 1852 ; pall-bearer at Wellington's funeral : published
narrative* of Peninsular campaigns, 1828, of campaigns of 1813-14.
1830, travels, and 'Memoir and Correspondence of Viscount
Castlereagh,' 1848-63. [liv. 278] STEWART, DAVID. DUKR OF
ROTHKSAY (137X71402), son of Robert III of Scotland ; Earl of
Carrick on father's accession ; pacified northern Scotland, e. 1396;
lieutenant of the kingdom and Duke of Rotheaay, 1399: married
Elizabeth Douglas, 1399; defeated March and bis English allies at
OockburnspHth : seised by Albany (his uncle) and Imprisoned at
Falkland, 1402: *. have been starved to death. [liv. 281 ] STEWART,
DAVID (1772-1829), major-general : of Garth ; joined 42nd
hu'hlaiiden, 1787 ; fought In Flanders, 1794, West Indies, 1795-7;
captured on way to Minorca, 1798: severely wounded at Alexandria,
1801 ; wounded at Maida, 1806 ; lieutenant-colonel. West India
rangers, 1808 ; at capture of Guadeloupe, 1810 ; C.B., 1815 :
majorgeneral, 1825 ; published 'Sketches ... of the Highlanders of
Scotland ; with Details of the Military Service of Highland
Regiments,' 1822 ; died governor of St. Lucia. [liv. 182] STEWART,
SIR DONALD MARTIN, first baronet (1824-1900), field-marshal ;
entered Bast India Company's service as ensign, 9th Bengal native
Infantry, 1840 : major (Bengal staff corps), 1866 : general, 1881 ;
fieldmarshal, 1894 ; served with distinction In Indian mutiny, 1857-8
; brigadier-general in Abyssinian expedition, 18671868; C.B.. 1868;
reorganised convict settlement in Andaman Islands, 1869 ;
commander of Lahore division, 1876 ; commanded Quetta army in
Afghan war, 1878-80; K.C.B., 1879; made celebrated march to Kabul,
1880: G.O.B. and created baronet, 1880: commander-in-cblef In
India, 1880-6 : member of council of India, 1886, till death ;
G.C.S.I., 1885 ; governor of Chelsea Hospital, 1895. fSuppl. HI. 367]
STEWART, DUGALD (1753-1828;, philosopher: son of Matthew
Stewart (1717-1785) [q. v.] : educated at Edinburgh High School,
and at Edinburgh and Glasgow universities : friend of Thomas Heid
(1710-1796) fq. v.] and (Sir) Archibald Alison ; associated with bis
ather in Edinburgh mathematical professorship, 1775 ; lectured for
Adam Ferguson [q. v.], 1778-9 : Edinburgh professor of moral
philosophy, 1785 ; gave up lectur 1809, but retained chair till after
death (1820) of coadjutor, Thomas Brown (1778-1820) [q. ?.] ; met
ft at latrine, 1786; visited France in summers of 1788 and 1789 ;
gave offence by sympathy with revolution : supported Sir John Leslie
(1 766-1832 )[q. vl, 1806; received sinecure from wblgs. 1806 ;
accompanied Lords Lenderdale and Henry Petty to Paris, 18O6 ;
monument erected to him upon Calton Hill. Edinburgh ; Palmerston,
RusteU, ud Luudoww (Locd Bsnrj ;••••, . umoog hit . ipib.*»d Scott
among friends ; while approximating to empirical
The text on this page is estimated to be only 24.57%
accurate

STEWART 1243 STEWART school disavowal it*

developments and retained ' intttitioni*m ' : uii.il.K- to study Kant.
His works, collected by Sir William Hamilton, 1854-60 (11 vols.),
include ' Klrrii-nt- of tin- Philosophy of the Human Mind ' (vols. i.,
1792, ii., 1814, Hi., 1827), ' Outlines of Moral Philosophy,' 1793
'Lectures on Political Economy' (delivered, 1801) ), and • Account of
Life and Writings of Thomas Reid,' 1802. [liv. 282] STEWART, ESME",
sixth SKH;M:I i'. D'AUBIGXY and fin* IH-KK OK LKXXOX (1542 ?
-1583). [See STUAUT.] STEWART or STUART, FRANCIS, fifth EARL
OP BOTH WELL (d. 1624). [See HEPBUKX.] STEWART, FRANCES
TERESA, DUCHESS OP RICHMOND AND LENNOX (1647-1702). [See
STUART.] STEWART, HELEN D'ARCY CRANSTOUN (17651838). [See
CUANSTOUX, HKLKX D'Aut Y.] STEWART, HENRY, first BARON
METHVEN ( 1495 ?1551 ?) : supported Margaret, Queen-dowager of
Scotland, in proclaiming James V, 1624 ; became James V's treasurer
and chancellor, and Queen Margaret's favourite; married her after
her divorce from Angus, 1528; imprisoned by Angus ; created Baron
Methven and master of the artillery by James V, 1528. [liv. 286]
STEWART or STUART, HENRY, LORD DARNLEY (1545-1567), son of
Matthew Stewart, fourth or twelfth earl of Lennox [q. v.] and Lady
Margaret Douglas [q. v.] ; skilful penman and lutenist and expert at
physical exercises, but mentally and morally weak ; sent by his
mother to France with a view to marriage with Mary Stuart, 1560;
with his mother confined in London by Qneeu Elizabeth, 1561, but
released and received into favour, 1563; allowed to go to Scotland at
Mary Stuart's request, 1565 ; created Duke of Albany and married to
Mary Stuart, 1565, the marriage being without real affection on
either side ; opposed by Moray ; was refused the crown matrimonial
and ousted from political influence by David Riccio [q. v.] ; his
jealousy of Riccio used by the nobles to make him join their
conspiracy, 1566 ; promised to establish protestantism in return for
crown, matrimonial and right of succession ; after death of Riccio
betrayed nobles to Mary Stuart and helped her to escape to Dunbar,
1566 ; temporary reconciliation with Mary Stuart dissolved by
discovery of boa plot: refused to attend baptism of James VI : only
prevented by illness from leaving Scotland ; induced by Mary Stuart
to leave Glasgow for Edinburgh ; murdered at Kirk o' Field. [liv. 287]
STEWART, HENRY, DUKE OP GLOUCESTER (16391660). [See
HENRY.] STEWART or STUART, HENRY BENEDICT MARIA CLEMENT
(1725-1807), styled by Jacobites Henry IX. [See HENRY.] STEWART,
SIR HERBERT (1843-1885), majorgeneral; ensign 1863; aide-de-
camp to commander of Allahabad division, 1868-70; deputy-
assistant quartermaster-general, Bengal, 1872-3, for conduct in
cholera outbreak, 1870 ; entered staff college and Inner Temple,
1877; brigade-major of cavalry during Zulu war, 1879; military
secretary to Wolseley and brevet lieutenantcolonel, 1880 ; as chief
staff officer to Sir George Pomeroy Oolley [q. v.] captured at Majuba,
1881 ; assistant adjutant-general of cavalry in Egypt, 1882 ; secured
Cairo after Tel-el-Kebir ; C.B., brevet-colonel, and aide-decamp to
Queen Victoria ; K.C.B. for services in Suakim campaign, 1884;
assistant adjutant and quartermasterKciK-ral, south-eastern district,
1884 ; in Lord Wolseley's (Jonlon relief expedition, 1884-5,
commanded desert column to proceed to Metemmeh; repelled Arab
attack at Abu Klea, but three days later was mortally wounded :
promoted major-general before death ; monuments at St. Paul's,
London, and Winchester. [liv. 291] STEWART, SIR HOUSTON (1791-
1875), admiral of the fleet ; entered navy, 1805 ; in Walcheren
expedition, 1809; Keith's signal-lieutenant in the Channel, 1812-13 :
flag-captain on Jamaica station, 1817-18 ; at reduction of Acre,
1840; C.B., 1840: controller- general of the coastguard, 1846-50 ; a
lord of the admiralty, 1850-2 ; rearadmiral, 1851 ; commanded at
reduction of Kinburn, Black Sea, 1855 ; commander-in-chief on
North American aUUou, 1856-60, at Devonport, 18GO-3; G.C.B.,
1865; admiral of the fleet, 187*. [liv. 293] STEWART, JAMES (d.
1309), high steward of Scotland, 1283-1309; a guardian of Scotland
under Queen Margaret, 1286 ; partisan of Bruce : made treaty with
France, 1295, but cann- to terms with Edward 1, l'J97: *ent to
negotiate with Wallace before battle of Stirling, 1297; joined him
afterwards; envoy to France, 1302; again submitted fc> English,
1305, but recognised Brace, 1309. [liv. 2U4] STEWART, JAMES,
DUKE OP Ross (1476 ?-1504), son of James III of Scotland : created
duke, 1488 ; archbishop of St. Andrews, 1498: seen by Ariosto at
Rome and eulogised in ' Orlando Furioso ' ; chancellor of Scotland,
1502. [liv. 295] STEWART, JAMES, EARL OP MORAY (1499 ?-1544),
natural son of James IV : granted earldom of Moray, 1501 ; accused
(1517) Lord Home of slaying James IV after Flodden ; a guardian of
James V and lieutenantgeneral of French forces in Scotland, 1523 ;
suppressed insurrection of the isles, 1531 ; warden of east and
middle marches, 1532-6 ; commissioner for marriage of James V and
Marie de Bourbon, 1535-6 ; specially hostile to England ; partisan of
Beaton ; member of council of state, 1543 ; served against Hertford,
1544. [liv. 295] STEWART, LORD JAMES, EARL OP MAR, and
afterwards of MORAY (1531 V - 1570), natural eon of James V ; half-
brother of Mary Queen of Scots ; granted Tantallon, 1536, priory of
St. Andrews, 1538; studied at St. Andrews, 1541-4; accompanied
Mary Stuart to France, 1548 ; repelled English raid on Fife, 1549 ;
visited France, 1550, 1552 ; legitimated, 1551 ; attended Kuox's
teaching at laird of Dun's, 1555; signed invitation to him to return,
1557 : fell ill after witnessing marriage of Mary Stuart to dauphin,
1557 ; at first supported regency of Mary of Guise ; on discovery of
her bad faith headed lords of the congregation, 1559 ; with Argyll
drove French from Perth and attacked Scone ; disavowed designs on
the crown ; procured suspension of the regent and negotiated with
Queen Elizabeth for help against the French ; concluded treaty at
Berwick providing for expulsion of French, but safeguarding of Mary
Stuart's sovereign rights, 1560; reoccnpied Edinburgh by English
help, and concluded treaty confirming that of Berwick, 1560 ;
deputed, 1561, by Scottish estates to sound Mary Stuart after death
of Francis II, 1560; strongly dissuaded her from attempting to
Romanise Scotland ; won confidence of Queen Elizabeth by
disclosing his conduct, but deprecated English attempts to prevent
or intercept Mary Stuart's return to Scotland ; opposed proposal to
debar Mary Stuart in Scotland from the mass ; privy councillor, 1561
; privately granted earldom of Moray, but assumed title Earl of Mar,
1562 ; virtually home secretary ; by expedition against Liddesdale
prevented Bothwell's establishment in southern Scotland ; formally
created Earl of Moray after personal expedition of Mary Stuart
against Huutly (holder of title), 1562 ; on Queen Elizabeth's refusal
to recognise Mary as her successor, supported projected Spanish
alliance of Mary, 1563 ; opposed the Darnley match ; thenceforth
estranged from Mary ; attempted capture of Mary and Darnley
before marriage, 1565; backed by Knox and Lennox's enemies, but
not by bulk of protestauts ; deceived by Queen Elizabeth ; publicly
disavowed and insulted by Queen Elizabeth, but granted asylum in
England and privately received, 1565 : after failure to procure
Riccio's intercession joined plot against him ; supported charge of
infidelity against Mary and promised to acknowledge Daruley ;
returned after Riccio's assassination, 1566 ; affected to .favour Both
well's pretensions and was nominally reconciled to Mary ; tacitly
sanctioned murder of Darnley, though probably not cognisant of
Bothwell's plan ; left Scotland for France immediately afterwards,
1567 ; returned on Mary Stuart's abdication at Lochleven : accepted
the regency at her personal request, 1567, making great show of
reluctance ; promoted declaration of her privity to the Darnley
murder ; secured punishment of subordinates, but took no steps
against principal in the crime ; took rapid measures to defeat Mary
after escape from Locbleven : secured his position as regent by
formal sanction of Queen Elizabeth, though pretending approval of
the Norfolk marriage scheme; voted against divorce from Bothwell,
but after discovery of Norfolk's intrigues excused his apparent
approval of them and made revelations ; caused Maitlaud of
Lethlngton to be accused of Darnley's murder, and imprisoned,
1569; captured and imprisoned Northumberland; pro 
The text on this page is estimated to be only 21.99%
accurate

STEWABT 1249 8TEWART pos«l to Queen Elizabeth

imprisonment of Mary Stuart , or Johanna [< in .-.-otliin.l, 15C9;
assassinated at Linlithgow by James ' Hamilton (.//. ir,G6-1880)
[q.v.]oC BothweUhaugh ; buned :it ' . i iil.-X Edinburgh ; called ' the
good regent.' riiv. »7] STEWART or STUART, JAMKS. *voi,d KMU. -F
MOUAY (-/. i.iai-), 'the bonny earl'; assumed title In ru-i.t of his wife,
daughter of the regent Moray, 1680 : commissioner under acts
against Spanish Armada, 1688, and the Jesuits, 1690; assisted laird
of Grant against Huntly, 1690 : induced to come south on promise of
pardon ; warrant granted to Huntly against him by James [q. v.],
queen-dowager of Bcouand, by the black knight of Lome : created
Karl of Atboll, r. 1467 : ••-.:... .:• . ••-. • , • • • '.< • ,:;.,: . . ,- •.....:,
•-.-,. . , ., ... ,,( -.•:,,:•. 148S; on ri Jam. UTi (ODera! , 1488; ....
pffaoosd !.. &m • iv.« .- ifftsj smidS hvook [Uv. SU] STEWART. JOHN,
third (or eleventh) RAUL or Lxxxox (d. 1826), succeeded, lilt ; Joined
Arran's party and ariaed Dumbarton, 1618 ; helped to blockade
Htirtlng. 1616 : at first opponent, afterwards partisan of Albany ;
AMm'spwtT and "--• - -— - « ^-=L VI, who was said to be jealoui of
his favonr with hl« queen : .-lain by Huntly's men at DonlbrisUe, his
motlier'* house having been burned ; his corpse long nnburied ;
popular Indignation expressed In traditional ballad. [llv. 307]
STEWART, JAMBS, of Bothwrilmuir, EARL or AURAN (d. 1896),
second son of Andrew Stewart, second baron Ochlltree [q. v.] ;
served states of Holland against Spanish ; gentleman of king's
chamber, 1880 ; on Lennox's : cold Wool by Kir lv. SU] behalf
accused Morton of Darnley's murder, 1880 ; privy councillor, 1681 ;
recognised as bead of the Hamilton* and granted earldom of Arran,
being cousin of James Hamilton, third earl [q. v.], now insane ;
forbidden the court for Insolence to Lennox, but temporarily
reconciled with him; caused raid of Ruthven (1881) by 'intrusion ' of
Robert Montpomerie (
The text on this page is estimated to be only 24.01%
accurate

STEWART 1250 STEWART rot* against Balmerino at his

trial, but afterwards obtained his pardon, 1634: treasurer of
Scotland, 1636-41 : assisted Charles I to introduce the liturgy, but
advised cautious policy and moderation towards its opponents :
conveyed arms and ammunition to Dulkeith, but had to surrender to
covenanters, 1639 ; joined Charles I at York : king's commissioner to
Edinburgh assembly, 1639 ; assented to abolition of episcopacy and
signature of the covenant ; distrusted by both parties ; dismissed
and condemned to death, 1641 ; sentence remitted at Charles I's
instance; heavily fined, 1644; perhaps betrayed Montrow's plans to
Leslie, 1645 : readmitted to parliament, 1648- raised cavalry for the
'engagement'; taken at Preston, 1648; prisoner in England till 1654.
[liv. 326] STEWART, JOHN, called JOHN ROY (1700-1762), Jacobite ;
lieutenant in Scots greys ; resigned commission ; Jacobite agent;
fought in French army at Fonteuoy, 1745 ; commanded 'Edinburgh
regiment,' 1745-6 : favourite with Prince Charles Edward; escaped
with him to France ; died there. He was a noted Gaelic poet. [liv.
328] STEWART, JOHN (1749-1822), ' Walking Stewart ' ; refractory
at Harrow and Charterhouse School ; went to Iinlia, 1763 ; resigned
East India writership, 1765 ; general under Hyder All ; escaped
wounded : prime minister of nabob of Arcot ; travelled in Persia,
Ethiopia, and Abyssinia ; came to Europe through Arabian desert,
walking through France and Spain towards England, 1783 ; walked
from Calais to Vienna, 1784 ; in North America, 1791 ; met
Wordsworth, 1792, De Quincey, 1798-9 ; announced lectures in
London, 1803 ; his money claims on Arcot settled by East India
Company, c. 1813 ; published discursive philosophical work?,
including ' Travels to discover the Source of Moral Motion,' c. 1789.
[liv. 328] STEWART or STTJART, LOUISA, COUNTESS OF ALBANY
(1753-1824), wife of the Young Pretender. [See ALBANY.] STEWART,
LUDOVIOK, second DUKE OP LENNOX and DUKE OF RICHMOND
(1574-1624). [See STUART.] STEWART or STUART, MARIA
CLEMENTINA (1702-1735), wife of the Old Pretender ; daughter of
Prince James Sobieski, eldest son of the king of Poland ; married
James Francis Edward Stuart, the Old Pretender [q. v.], 1719 ; left
her husband and retired to a nunnery, 1724. [xxix. 202] STEWART
or STUART, MARY (1542-1587), queen of Soots. [See MARY.]
STEWART, MATTHEW, second (or tenth) EARL OF LKNNOX (d. 1513),
succeeded to Lennox and had sheriffdotn of Dumbarton ; slain at
Flodden, commanding Scots right wing. [liv. 314] STEWART,
MATTHEW, fourth (twelfth) EARL OF Li:v\ox ( 1516-1571), regent of
Scotland ; keeper of Dumbarton Castle, 1531 ; commanded Scots
men-at-arms in Provence, 1536 : naturalised in France, 1537 ;
induced by French party to return to Scotland, 1543 ; put forward as
rival to Arran as next heir after the Princess Mary ; seized Mary of
Guise, the queen dowager and Princess Mary at Edinburgh, 1543 ;
brought them to Stirling ; disappointed of marriage with Mary of
Guise; opened negotiations with Henry VIII for hand of Lady
Margaret Douglas [q. v.] ; joined English party, but came to
temporary agreement with Arran, 1544 ; went to London and signed
treaty agreeing to surrender Dumbarton and Bute and support
English overlordship in exchange for hand of Lady Margaret and the
governorship of Scot* Uind, 1544; received English estates;
naturalised and married; as English lieutenant for southern Scotland
made attempt against west coast, 1544-5 ; assisted Hertford's
invasion, 1545; outlawed in Scotland, 1545 ; imprisoned by Queen
Elizabeth for design to return to Scotlaud, 1562, his object being to
promote marriage of his son (Daruley) with Mary Queen of Scots ;
allowed to go, 1564 ; restored to title and lands, 1564 ; lieutenant
over western Scotland, 1565 ; privy to plot against Riccio ; warned
Mary of Darnley's wish to leave Scotland ; formally accused
Bothwell, 1567 ; prevented from appearing against him : provisional
regent after Mary's surrender ; accused Mary at Westminster
conference, 1568; lieutenant-general «nd regent of Scotland, 1570 ;
fought against Huntly and (be Hamlltons : held parliament at
Edinburgh ; surprised at Stirling by Kirkcaldy of Grange ; rescued by
Mar, but •tabbed by Captain Oalder. [liv. 331] STEWART, MATTHEW
(1717-1785), geometrician; studied at Glasgow and at Edinburgh
under Colin Maclaurin [q. v.] ; gained reputation by his ' General
Theorems,' 1746 ; minister of Roseneath, 1745-7 ; professor of
mathematics at Edinburgh University, 1747-85, the duties being
performed by his sou. Dugnld Stewart [q. v.], after 1772; F.R.S.,
1764; chief work, 'Tracts, Physical and Mathematical,' 1761, applying
geometrical demonstration to astronomy. [liv. 336] STEWART,
MURDAC or MURDOCH, second DUKE ! OF ALBANY (d. 1425),
govemor of Scotland ; known as ! Earl of Fife till death of father,
Robert Stewart, first I duke of Albany [q. V.], 1420 ; justiciary north
of Forth, 1392 ; captured at Homildon, 1402 ; prisoner in England 1
till exchanged for Sir Henry Percy, second enrl of Northi uinberland
[q. v.], 1415 ; suspected of delaying James I's liberation ;
incompetent and corrupt governor, 1420-4 ; j arrested and executed.
£liv. 337] STEWART, PATRICK, second EARL OF ORKNEY (d. 1614),
son of Robert Stewart, first earl of Orkney [q. v.] ; I granted charter
of earldom of Orkney and lordship of Zetland, 1600 ; exercised
almost independent sovereignty ; charged with tyranny and cruelty ;
tried and sentenced to imprisonment and loss of justiciarship, 1611 ;
released, but again imprisoned ; refused all terms ; executed for
instigating rebellion of his son. [liv. 337] STEWART, PATRICK (1832-
1866), major, royal (17th Bengal) engineers; lieutenant, Bengal
engineers, 1854 ; brevet major, 1858 ; attached to headquarters
staff during relief of Lucknow, 1867, and at siege and capture, 1858
; accompanied Lord Canning to Allahabad, 1858 ; served on
commission to inquire into cause of the great mortality from cholera,
1861-2; director-general of government Indo-European telegraph at
Bombay, 1863 ; laid cable from Gwadar to Fao ; C.B., 1864. [Suppl.
iii, 358] STEWART, ROBERT, first DUKE OF ALBANY (1340 11420),
regent of Scotland ; third son of Robert Stewart, earl of Strathearn
(afterwards Robert II) [q. v.] ; hostage in England, 1360; Baron of
Menteith on marriage, 1361; Earl of Fife and Menteith, 1371 ;
hereditary governor of Stirling, 1373 ; placed in line of succession ;
chamberlain of Scotland, 1382-1407 ; took part in raids into
Northumberland and Cumberland, 1385-6 ; led invasion of 1388 ;
provisional guardian of Scotland, 1389-99 ; created Duke of Albany,
1398 ; member of Rothesay's council, 1399 ; reinstated as governor
after Rothesay's arrest and death, 1402 ; his forces defeated by
English, 1402 ; supposed to have acted with Hotspur, 1403; regent
of Scotland on capture of Prince James and death of Robert III [q.
v.], 1406 ; his supposed connivance at imprisonment of James I in
England not substantiated ; crushed rebellion of Donald Macdonald,
second lord of the Isles [q. v.], who claimed Albany's earldom of
Ross, 1411 ; caused erection of Inverness Castle ; demanded release
of James 1, 1416 ; protected Thomas Warde, the pretended Richard
II of England ; made ' foul raid ' against Roxburgh, 1417 ; granted
charter, 1420 ; buried at Dunfermline. [liv. 339] STEWART or
STUART, ROBERT, SBIOKBUR D'AUBIONY (1470?-1543), brother of
Matthew Stewart, second or tenth earl of Lennox [q. v.] ; served
with Scots under Seigneur d' A ubigny ; as marshal of France
defeated Colonna at Villa Franca, 1615 ; fought at Marignano ;
captured at Pavia. [liv. 314] STEWART, LORD ROBERT, afterwards
EARL OF ORKNEY (d. 1592), natural son of James V : abbot of
Holyrood, 1539 ; privy councillor, 1652 ; acted with lords of
congregation ; prominent during French attack on Edinburgh, 1559;
pensioned by his half-sister, Mary Queen of Scots, 1666 ; reported to
have warned Darnley of the plot against him ; exchanged
temporalities of Holyrood for those of Orkney, 1669 ; a chief
conspirator against Morton, 1580 ; created Earl of Orkney, 1581. [liv.
842] STEWART, SIR ROBERT (d. 1670 ?), Irish royalist ;
accompanied James VI to England, 1603 ; In Swedish service, 1611-
17; granted Irish estates, 1617; raised troops in Scotland for
Sigismund III of Poland, 1623 ; at battle of Leipzig and taking of
Wurzburg, 1629 ; enlisted Irish troops for Sweden, 1637 ; governor
of Culmore Castle, 1638, of Londonderry, 1643-4 ; M.P.,
Londonderry, 1639; received royal commission against Irish rebels,
1641 ; defeated Sir Phelim O'Neill [q. v.] at Glenmaquin,
The text on this page is estimated to be only 21.92%
accurate

STEWART 1251 STEWART 1642; surprised Owen Roe

O'Neill [q. v.], U543 ; reluctantly took the covenant ; refitted to obey
parliamentary governor of m rry ; secured and M-nt to London, 1648
; escaped and joined Olanricarde in Ireland, 1649 ; after royalist
def« .r (•appointed to Derry and Culmore at Restoration. [liv 3431
STEWART, ROBERT, first MAK^I i> .,* : IIKRKY ( 173'.* -ihii j : Irish
M.P., oo. Down, 1769-83 ; promint nt delegate to second Duuganuon
convention, 1783 : Irish privy councillor ; created (Irish) Baron
LondonSTEWART. RODERT, second MARQUIS OF LOXDOVDEHRY,
better known as Vweoi'XT OASTLKIUUOH (17691822), statesman :
son of Robert Stewart, flrst marquis of Londonderry [q. v.] ; studied
at St. John's College, Cambridge, 1788, and abroad : M.P., oo. Down
(Irish parUamentX 1790; M.P., Tregony, 1704-6, Oxford, 1796-7;
supported enfranchisement of Irish Roman catholics ; at flrst voted
with opposition ; as keeper of Irish privy seal, 1797-8, was acting
chief seoretary to Ttceroy ; forestalled United Irish rebellion by
arresting leaden : procured English troops to replace Irish militia ;
chief secretary for Ireland, 1799-1801, being specially selected by
Pitt ; impressed on ministry necessity for immediate union ; solely
responsible for its passing through Irish parliament, 1800. though
reluctantly employing corruption ; threatened resignation on refusal
of ministry to recommend all peerages promised; pressed for
introduction of Irish Catholic Emancipation Bill ; resigned on George
Ill's refusal, 1801 : after union represented ob. Down, 1800-6 and
1812-20, Boroughbridge, 181)6, Plyrnptou-Earl, 18061812, Oxford,
1821-2 ; had unofficial charge of Addington's Irish measures, 1801 ;
prepared plans for Irish tithe commutation, and recommended state
payment of Roman catholic priests ; at Pitt's request became
president of (East India) board of control under Addington, 1802 ;
supported Lord Welle* ley against court of directors and the cabinet,
and conducted negotiations with Persia; appointed by Pitt to war and
colonial office in addition to board of control, 1805 ; responsible for
abortive Elbe expedition, 1805; attacked Grenville's foreign policy,
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!

ebookname.com