Real-Time Stock Market Analysis using Spark Streaming

Modeling high-frequency limit order book dynamics
with support vector machines
Alec N.Kercheval∗
Department of Mathematics
Florida State University
Tallahassee, FL 32306
Yuan Zhang†
Department of Mathematics
Florida State University
Tallahassee, FL 32306
October 24, 2013
Abstract
We propose a machine learning framework to capture the dynamics of high-
frequency limit order books in financial equity markets and automate real-time
prediction of metrics such as mid-price movement and price spread crossing. By
characterizing each entry in a limit order book with a vector of attributes such as
price and volume at different levels, the proposed framework builds a learning
model for each metric with the help of multi-class support vector machines
(SVMs). Experiments with real data establish that features selected by the
proposed framework are effective for short term price movement forecasts.
Key words: high-frequency limit order book; multi-class classifiers; support vector ma-
chines(SVMs); machine learning
Mathematics Subject Classification: 93E10, 93E14, 60G35, 60H07.
1 Introduction
Electronic trading systems have been widely adopted by many established ex-
changes including NYSE, NASDAQ, and the London Stock Exchange [9, 10, 30].
As a result, the traditional quote-driven markets have been increasingly replaced by
order-driven trading platforms. The rapid decrease in order processing time has lead
to high volume, high frequency trading as a growing fraction of total stock trading.
∗
kercheva@math.fsu.edu
†
yzhang@math.fsu.edu
1

For instance, among the trading transactions of US in 2012, high-frequency trading
accounted for 84% in stock trades and 51% in equity value [32]. Clearly, the char-
acteristics of order-driven trading systems change the dynamics of the markets and
demand new trading strategies that can capture short-term behavior of underlying
assets [5, 7, 16, 29].
Research on modeling limit order book dynamics can generally be grouped into
two main categories: statistical modeling and machine learning based methods. In
the former approach, statistical properties of the limit order book for the target fi-
nancial asset are developed and conditional quantities are then derived and mod-
eled [8, 10, 20, 33, 35]. On the other hand, machine learning methods represent data
from the limit order book in a systematic manner, then generalize the data so that
unseen data can be recognized and classified based on the generalization [22, 24].
Statistical methods usually impose strict and typically unvalidated mathematical as-
sumptions on models, parameters may be unobservable, and the intensive computa-
tion required by statistical methods makes them difficult to be deployed in real world
environments [1, 4, 17, 29]. On the other hand, machine learning methods, owing to
their data driven characteristics and light-weight computation, have been applied to
a variety of financial markets such as liquidity and volatility in bonds and index trad-
ing [18, 13, 25, 36]. In particular, support vector machines (SVMs – described below)
have been used in tracking the dynamics of foreign exchange markets [14]. However,
the application of machine learning techniques in financial markets, especially SVMs,
is still in its infancy.
In this paper, we employ “multi-class SVM” methods in a new way to capture
the dynamics in high-frequency limit order book data and forecast movements of the
mid-price and the direction of bid-ask spread crossings over short time intervals. Ex-
periments with real data from NASDAQ show that the multi-class SVM models built
in this paper not only predict various metrics with high accuracy, but also deliver
predictions efficiently. In addition, experiments with simple trading strategies using
these signals are observed to produce positive payoffs with controlled risk.
2

The paper is organized as follows: Section 2 describes the general approach
of machine learning via Support Vector Machines. Section 3 describes the limit or-
der book data we wish to model and the specific construction and validation of the
SVM model we employ. Section 4 presents our experiment results and analyzes their
performance, with Section 5 concluding.
2 Support Vector Machines
Machine learning techniques fall into two main categories unsupervised learning and
supervised learning. In the former, unlabeled training data are categorized by the
model into different groups that are not known a priori, and new data is assigned to
the closest group according to certain similarity measures. Popular examples of unsu-
pervised learning includes artificial neural networks [34], k-nearest neighbor methods
[3], and self-organizing maps [2]. In supervised learning, training data is labeled by
class in advance, and a model is used to assign new data to those classes. This in-
cludes the methods of logistic regression [37] and SVMs [11, 21].
The SVM method is the core of the model in [14] to predict the EUR/USD price
evolution, which has been the inspiration for our work here. Before we describe our
approach to SVM price forecasting, we need to describe the general SVM approach
and introduce the needed concepts and vocabulary.
2.1 The SVM framework
The SVM is a well-established technique in machine learning, which we briefly outline
in this section. For further details on the material of this section, see [6], [11], [14],
and [28].
The basic support vector machine is a kind of binary classifier. Imagine each
trial of some experiment produces a vector x ∈ Rn
of output data, the components of
which are the “attributes” of the data, and we wish to classify each trial output into
one of two classes (yes vs. no, good vs. bad, etc). The SVM is an algorithm which,
3

after some training, will automatically carry out the classification of new trial output
without user intervention.
The training data is a sequence of prior trials, x(1)
, . . . , x(m)
that have already
been classified. Without loss of generality, we suppose that classification means that
each x(i)
has been assigned a class value (or label) y(i)
∈ {−1, 1}, i = 1, . . . , m. After
training, the SVM will then assign a label y ∈ {−1, 1} to each new trial output x.
Conceptually, the method is as follows:
1. Choose a “feature mapping” φ : Rn
→ H, where H is some higher dimensional
Euclidean space (or an infinite dimensional Hilbert space – for simplicity we
suppose H is finite dimensional in this paper) called the “feature space”. For a
data instance x ∈ Rn
, the components of φ(x) are the “features” of the data to
be used to determine the correct label to be assigned.
2. Given the training data x(1)
, . . . , x(m)
, we consider the collection
F = {z(i)
= φ(x(i)
) : i = 1, . . . , m} ⊂ H,
and we expect that the feature mapping has been chosen so that the set of points
F(+1)
of F with label +1 are spatially separated in H from the set of points F(−1)
with label -1. “Spatially separated” means that there is a separating hyperplane
in H such that F(+1)
lies entirely on one side and F(−1)
on the other.
3. The SVM algorithm selects the optimal separating hyperplane L ⊂ H, in the
sense that distance from the hyperplane to F is maximized among all possible
separating hyperplanes.
4. New data x is then labeled according to which side of L the feature vector φ(x)
lies.
We now briefly describe, without full derivations, the SVM algorithm in terms
of the optimization problem mentioned above.
A hyperplane in H is described by an equation of the form wT
z + b = 0, where
w ∈ H, b ∈ R are parameters, z ∈ H is our independent variable and wT
denotes the
4

transpose of w. If this is a separating hyperplane for our training data set, we may
reverse the sign of (w, b) if needed so that wT
z(i)
+ b > 0 if y(i)
= +1 and otherwise
wT
z(i)
+ b < 0.
It is not difficult to show that the distance in H between the feature vector z(i)
and the plane is
γi = y(i) w
||w||
T
z(i)
+
b
||w||
.
We can then define the “geometric margin” of the plane (w, b) with respect to
the training set S = {(z(i)
, y(i)
), i = 1, . . . , m} to be the smallest of these individual
distances (or “margins”):
γ = min
i=1,...,m
γi.
The optimal separating hyperplane is then obtained as the solution to the fol-
lowing optimization problem that identifies the hyperplane maximizing the geometric
margin:
max
γ,w,b
γ (2.1)
s.t. y(i)
(wT
z(i)
+ b) ≥ γ||w||, i = 1, . . . , m. (2.2)
or equivalently, letting ˆγ = γ||w||,
max
ˆγ,w,b
ˆγ
||w||
(2.3)
s.t. y(i)
(wT
z(i)
+ b) ≥ ˆγ, i = 1, . . . , m. (2.4)
A better equivalent formulation of the problem is obtained when we take account
of the fact that the parameters (w, b) can be scaled by an arbitrary positive constant
without changing the hyperplane they determine, and hence without changing the
problem. We therefore select a scaling constraint so that ˆγ = 1 in the above problem,
which is then equivalent to
max
w,b
1
||w||
(2.5)
s.t. y(i)
(wT
z(i)
+ b) ≥ 1, i = 1, . . . , m. (2.6)
5

or, equivalently,
min
w,b
1
2
||w||2
(2.7)
s.t. y(i)
(wT
z(i)
+ b) ≥ 1, i = 1, . . . , m, (2.8)
which is now in the form of a convex quadratic objective with linear constraints.
While this is a tractable form of the problem, we still have a high dimensional
problem (the dimension of H) and a large number (m) of constraints. The SVM
approach improves this situation dramatically by considering the corresponding dual
optimization problem (see, e.g., [28]):
max
α
m
i=1
αi −
1
2
m
i,j=1
αiαjy(i)
y(j)
z(i)
, z(j)
(2.9)
s.t. αi ≥ 0, i = 1, . . . , m (2.10)
m
i=1
αiy(i)
= 0. (2.11)
Since the original (primal) problem 2.7 has a convex objective and convex constraints,
the so-called Slater condition says that this dual problem gives the same solution as
the primal problem if the constraints are strictly feasible, which will generically be
the case if the training data can be separated by any hyperplane. It can be shown
that the optimal hyperplane parameters are then given by
w =
m
i=1
αiy(i)
z(i)
and
b = −
1
2
( max
{i:y(i)=−1}
wT
z(i)
+ min
{i:y(i)=1}
wT
z(i)
).
This dual formulation has the convenient property that the optimal αi’s are
all equal to zero except for the feature vectors z(i)
whose distance to the optimal
hyperplane (w, b) is exactly equal to the geometric margin – these are called the
“support vectors”, giving the SVM approach its name. Once the optimal parameters
are found and z is a new feature vector that needs to be classiﬁed, we need only
6

compute the sign of
wT
z + b =
m
i=1
αiy(i)
z(i)
, z + b.
Since αi = 0 unless it corresponds to a support vector, many of the terms in this sum
are zero and we need only calculate inner products between z and the (typically small
number of) support vectors z(i)
of the optimal hyperplane.
In the remainder of this section, we describe how the SVM positively answers the
following two remaining questions: (a) the dimension of H might still be inconveniently
large; can the dimension of the computation be reduced?, and (b) can the algorithm
be modified to be less sensitive to spurious data in the training set?
2.2 Kernels
The feature mapping φ : Rn
→ H from the space of data attributes to the feature
space may need to have a high-dimensional range in order to allow the data to be
linearly separable.
As an example, suppose we take n = 3, choose c > 0, and define φ : R3
→ R13
by
φ(x1, x2, x3) = [x1x1, x1x2, x1x3, x2x1, . . . , x3x3,
√
2cx1,
√
2cx2,
√
2cx3, c]T
That is, the features are all monomials in the attributes of degree 2 or less. This
feature mapping will be effective if the data can be separated by quadratic functions
of the attributes. Here, c is a parameter controlling the relative weighting between
the first and second order terms. It is straightforward to verify for general n and for
x, ˆx ∈ Rn
, that
φ(x), φ(ˆx) =
n
i,j=1
(xixj)(ˆxi ˆxj) +
n
i=1
(
√
2cxi)(
√
2cˆxi) + c2
= (xT
ˆx + c)2
.
In other words, the inner product of feature vectors z = φ(x) can be computed in terms
of the much lower dimensional inner product xT
ˆx on Rn
via the “kernel function”
K(x, ˆx) = (xT
ˆx + c)2
.
7

Similarly, it can be verified that the kernel
K(x, ˆx) = (xT
ˆx + c)d
corresponds to a feature mapping that includes all monomials up to degree d. In this
case, the SVM classifier will be effective if the data can be separated by polynomial
hypersurfaces of degree d or less, which is a very general situation.
There is a general fact known as Mercer’s Theorem ([26]) stating that whenever
K : Rn
× Rn
→ R is symmetric positive semi-definite, there is a feature mapping φ
such that
z, ˆz ≡ φ(x), φ(ˆx) = K(x, ˆx).
The great advantage of this is that our classification problem now becomes
computing the sign of
wT
z + b =
m
i=1
αiy(i)
z(i)
, z + b (2.12)
=
m
i=1
αiy(i)
K(x(i)
, x) + b, (2.13)
which, for the kernels above, only requires computation of an n-dimensional inner
product.
2.3 Regularization
We may need to take account of a small number of spurious training data instances
that are labeled incorrectly, which could have a large unwanted effect on the optimal
separating hyperplane. To handle this, the standard method is to relax the constraint
and introduce a compensating term in the objective function to penalize data on the
wrong side of the hyperplane, leading to this reformulated problem:
min
w,b,ξ
1
2
||w||2
+ C
m
i=1
ξi (2.14)
s.t. y(i)
(wT
z(i)
+ b) ≥ 1 − ξi, i = 1, . . . , m (2.15)
ξi ≥ 0, i = 1, . . . , m. (2.16)
8

The constant C > 0 controls the relative weight between the competing goals
of making the total margin large, and ensuring as many examples as possible are on
the correct side of the hyperplane.
From this, the dual version can be formulated as
max
α
m
i=1
αi −
1
2
m
i,j=1
αiαjy(i)
y(j)
z(i)
, z(j)
(2.17)
s.t. 0 ≤ αi ≤ C, i = 1, . . . , m (2.18)
m
i=1
αiy(i)
= 0. (2.19)
In terms of the quadratic kernel K(x, ˆx) = (xT
ˆx + c)2
described above, we get
our final version of the optimization problem1
:
max
α
m
i=1
αi −
1
2
m
i,j=1
αiαjy(i)
y(j)
K(x(i)
, x(j)
) (2.20)
s.t. 0 ≤ αi ≤ C, i = 1, . . . , m (2.21)
m
i=1
αiy(i)
= 0. (2.22)
Using the solutions αi, and letting S = {s1, s2, . . . } be the set of support vectors
(corresponding to α > 0), the label for a new example x is assigned as +1 if
|S|
i=1
αiyiK(si, x) + b (2.23)
is positive, and −1 otherwise. The kernel functions used in this paper are the polyno-
mial kernels K(xi, xj) = (xi · xj + 1)d
with d = 2. (We use the constraint parameter
C = 0.25, though in experiments this choice does not affect results much.)
3 Model Architecture
In this section, we first describe the limit order book and key metrics to characterize
price evolution, then present the rationale behind our SVM design framework for price
1
This optimization problem is solved experimentally using a JAVA implementation of the Sequen-
tial Minimal Optimization algorithm [31].
9

movement forecasting. We also introduce various aspects of multi-class SVM – the
workhorse of the proposed framework, and delineate our tactics to validate models
built with SVMs and methods to evaluate their performance.
3.1 Limit Order Book Dynamics and Metrics
Records of high-frequency trading activity are organized into a database with two
major components: the message book and the order book. The former stores basic
information on each trading event, such as time of occurrence and transaction type,
while the latter keeps unexecuted limit orders for both bid and ask. A sample mes-
sage book and order book extracted from the NASDAQ stock AAPL are depicted in
Table 1. In the table, each row of the message book represents a trading event that
could be either a limit order submission, limit order cancellation, or market order
execution as shown in column “Event Type”. The arrival time of an event given in
column “Time”, measured from midnight, is in seconds and nanoseconds; the order
price is in US dollars, and the Volume is in number of shares. The “Direction” col-
umn indicates the type of limit order that is executed via an incoming market order
of the opposite type – e.g. an incoming market sell order executes against an existing
limit bid order, etc. On the other hand, each entry of the order book, also shown in
Table 1, groups ask and bid events on n diﬀerent price levels (we take n = 10) along
with their volume sizes. The best ask and best bid are listed ﬁrst, and the next best
second, etc.
Message book
Time(sec) Price($) Volume Event Type Direction
k − 1 34203.011926972 585.68 18 execution ask
k 34203.011926973 585.69 16 execution ask
... ... ... ... ... ...
k + 4 34203.011988208 585.74 18 cancellation ask
k + 5 34203.011990228 585.75 4 cancellation ask
... ... ... ... ... ...
k + 8 34203.012050158 585.70 66 execution bid
k + 9 34203.012287906 585.45 18 submission bid
k+10 34203.089491920 586.68 18 submission ask
10

Order book
Ask1
Bid1
Ask2
Bid2
Ask3
Bid3
...
Price Vol. Price Vol. Price Vol. Price Vol. Price Vol. Price Vol. ...
k − 1 585.69 16 585.44 167 585.71 118 585.40 50 585.72 2 585.38 22 ...
k 585.71 118 585.44 167 585.72 2 585.40 50 585.74 18 585.38 22 ...
... ... ... ... ... ... ... ... ... ... ... ... ... ...
k + 4 585.71 118 585.70 66 585.72 2 585.44 167 585.75 4 585.40 50 ...
k + 5 585.71 118 585.70 66 585.72 2 585.44 167 585.80 100 585.40 50 ...
... ... ... ... ... ... ... ... ... ... ... ... ... ...
k + 8 585.71 100 585.44 167 585.80 100 585.40 50 585.81 100 585.38 22 ...
k + 9 585.71 100 585.45 18 585.80 100 585.44 167 585.81 100 585.40 50 ...
k+10 585.68 18 585.45 18 585.71 100 585.44 167 585.80 100 585.40 50 ...
Table 1: sample: AAPL order book & message book
It is evident from Table 1 that a new entry in the message book typically causes
one fresh record to be added into the order book. For instance, the transaction event
at the k-th row of the message book of Table 1, execution of an ask order at the
price $585.69 with 16 shares, exactly cancels out the best ask price and its volume in
Row k − 1 of the order book, making the next best ask price, $585.71, become the
new best ask price as shown in Row k of the order book. It can also be observed
from the message book that multiple trading events could arrive within milliseconds
as demonstrated from Row k − 1 to k + 10, leading to drastic fluctuation of prices
and volumes in the order book. Although a variety of “metrics” have been designed
to capture the price fluctuation, in this paper we select as metrics (a) the occurrence
and direction of mid-price movement, and (b) the occurrence and direction of bid-ask
spread crossing, described below.
The mid-price is defined as the mean of the best ask price Pask
t and best bid price
Pbid
t at time t, Pmid
t = 1
2
(Pask
t +Pbid
t ). The three possible scenarios of mid-price move-
ments — upward, downward and stationary – are illustrated in Table 1. For example,
compared to Row k −1, the mid-price at Row k +4 increases to ($585.71+$585.70)/2
= $585.705 from ($585.69+$585.44)/2 = $585.565 due to upward movement of both
the best ask price and the best bid price. While the mid-price is stationary as time
advances from Row k+4 to k+5 owing to the motionless best ask/bid prices, it moves
downward to ($585.68 + $585.45)/2 = $585.565 when transactions further proceed to
11

Row k + 10.
Mid-price movement is a statistical indicator (though not a guarantee) of poten-
tial trading profits. In contrast, bid-ask spread crossing is a less-frequent occurrence
that however does assure a profit if correctly identified in advance. There are three
scenarios: 1) an upward price spread crossing occurs when the best bid price at t+∆t
exceeds the best ask price at time t (Pbid
t+∆t > Pask
t ); 2) a downward price spread
crossing happens when the best ask price at t + ∆t is less than the best bid price at
time t, (Pask
t+∆t < Pbid
t ); and 3) no price spread crossing takes place if Pask
t+∆t ≥ Pbid
t and
Pbid
t+∆t ≤ Pask
t . For example, an upward price spread crossing appears as indicated by
Row k − 1 and k + 4 of the order book in Table 1, — Pask
tk−1
< Pbid
tk+4
— the trader
makes a profit with a long position on the asset with best ask price $585.69 at time
k − 1 and then selling with a higher best bid price $585.70 at time k + 4.
These two metrics can work independently or together to provide guidance for
trading strategies depending on particular scenarios. Of course, the premise for the
arbitrage opportunities described above is that the directions of mid-price movement
and price spread crossing can be predicted accurately, which is the main task ad-
dressed in this paper.
3.2 Proposed Design Framework
As the trading day progresses, the limit order book contains a massive amount of
rapidly evolving data, with the possibility of important patterns forming and dissolv-
ing at a frequency too high for a human observer – hence the motivation for machine
learning.
To build a learning model for a given metric, such as mid-price movement, a
set of labeled samples, termed training data, should be prepared, in which each data
point is characterized with a vector of attributes, an SVM model is constructed, and
then subject to a validation procedure to verify soundness and robustness. To keep
up with the rapidly changed dynamics of the limit order book, the training data are
frequently updated so that models can be refreshed to learn and subsume any new
12

characteristics appearing in the data.
Generally speaking, separate learning models should be built for different metrics
depicting limit order book dynamics. To build a machine learning model for a specific
metric, the following four-phase process is employed.
• Features representation: the data in the order book and message book is
converted into a format suitable for machine learning methods to manipulate.
• Learning model construction: an SVM model is constructed.
• Learning model validation: the model is evaluated and validated using par-
ticular performance measurements.
• Unseen-data classification: the constructed learning model automates the
forecasting of the chosen metric in real time.
Figure 3.1: Architecture of framework for forecasting order book dynamics.
In this four-phase process, the first three phases are conducted offline while the
last phase works online to provide predictions for real time trading events. Details of
13

the main modules in Figure 3.1 will be provided in subsequent sections.
3.3 Feature Extraction and Training Data Preparation
The “feature representation” module is designed to convert the order book and mes-
sage book data into suitable inputs to the SVM. Each data point, which we take to be
the data recorded in a single row of the order book along with the corresponding row
in the message book, must be represented as a “feature vector” of specific attributes,
lying in a Euclidean space Rn
, where n is the total number of attributes selected.
(These feature vectors will then be mapped to the feature space by the feature map-
ping, according to the previous discussion of the SVM.) With well-formatted feature
vectors for data points in place, the module “feature representation” further randomly
samples some data points to construct the training data set for each metric used to
characterize the dynamics of the limit order book.
We denote the set of training data as T = {xi, yi}, (i = 1, ..., m), where
xi ∈ Rn
is the feature vector for the ith data point and yi ∈ Y = {l1, ..., lk}
is its true label identifying the category this data point belongs to. For exam-
ple, in the case of predicting the direction of mid-price movement, the label set is
Y = { upward, downward, stationary }.
To better profile the original data and fully represent all label categories of the
metric in question, the population in the training data set, although randomly sam-
pled, should follow the distribution of metric labels in the entire collection of data.
For example, assuming that the ratio of labels for metric mid-price (i.e., up, down,
and stationary) of stock AAPL shown in Tables 1 is 1:1:2 and a set of training data
with size 4000 is constructed, then ideally the numbers of samples in the training
data set with label upward, downward and stationary should be 1000, 1000, and
2000, respectively. It is typical that each metric has its own unique label set and
label distribution, making it necessary to build different training data sets for distinct
metrics. Meanwhile, the size of the training data set for different metrics may vary
as well depending on the characteristics of the data and the metrics. In this regard,
14

the size of a training data set for a metric is determined in the proposed framework
by a cross-validation process that monitors the relationship between size of training
data and the performance of the model built on the corresponding training data to
avoid overfitting. Moreover, the module frequently reconstructs the training data by
replace the oldest data points with the most up-to-date ones to ensure freshness of
the training data.
In Table 2 we identify a collection of proposed attributes that are divided into
three categories: basic, time-insensitive, and time-sensitive, all of which can be directly
computed from the data. Attributes in the basic set are the prices and volumes at
both ask and bid sides up to n = 10 different levels (that is, price levels in the order
book at a given moment), which can be directly fetched from the order book shown
in Table 1. Attributes in the time-insensitive set are easily computed from the basic
set at a single point in time. Of this, bid-ask spread and mid-price, price ranges, as
well as average price and volume at different price levels are calculated in feature sets
v2, v3, and v5, respectively; while v5 is designed to track the accumulated differences
of price and volume between ask and bid sides. By further taking the recent history
of current data into consideration, we devise the features in the time-sensitive set
presented in Table 2.
15

Basic Set Description(i = level index, n = 10)
v1 = {Pask
i , V ask
i , Pbid
i , V bid
i }n
i=1, price and volume (n levels)
Time-insensitive Set Description(i = level index)
v2 = {(Pask
i − Pbid
i ), (Pask
i + Pbid
i )/2}n
i=1, bid-ask spreads and mid-prices
v3 = {Pask
n − Pask
1 , Pbid
1 − Pbid
n , |Pask
i+1 − Pask
i |, |Pbid
i+1 − Pbid
i |}n
i=1, price differences
v4 = { 1
n
n
i=1 Pask
i , 1
n
n
i=1 Pbid
i , 1
n
n
i=1 V ask
i , 1
n
n
i=1 V bid
i }, mean prices and volumes
v5 = {
n
i=1(Pask
i − Pbid
i ),
n
i=1(V ask
i − V bid
i )}, accumulated differences
Time-sensitive Set Description(i = level index)
v6 = {dPask
i /dt, dPbid
i /dt, dV ask
i /dt, dV bid
i /dt}n
i=1, price and volume derivatives
v7 = {λla
∆t, λlb
∆t, λma
∆t , λmb
∆t , λca
∆t, λcb
∆t } average intensity of each type
v8 = {1{λla
∆t>λla
∆T }, 1{λlb
∆t>λlb
∆T }, 1{λma
∆t >λma
∆T }, 1{λmb
∆t >λmb
∆T }}, relative intensity indicators
v9 = {dλma
/dt, dλlb
/dt, dλmb
/dt, dλla
/dt}, accelerations(market/limit)
Table 2: Feature vector sets
In feature set v6, average time derivatives of price and volume are computed
over the most recent 1 second. The average intensity – defined as the recent short-
term average arrival rate of a certain trading type – is calculated in feature set v7
for limit ask/bid orders (denoted as λla
∆t and λlb
∆t), market ask/bid orders (denoted
as λma
∆t and λmb
∆t ), as well as cancellation ask/bid orders (denoted as λca
∆t and λcb
∆t).
We choose ∆t = 1 second. Features in v8 focus on the discrepancy between short
term and long term intensities for different trading type denoted as λtype
∆t and λtype
∆T ,
respectively, where type can be limit ask/bid orders as well as market ask/bid or-
ders (denoted as la, lb, ma, and mb). The indicator function 1{λtype
∆t >λtype
∆T } determines
whether the trading type in question has intensified in the most recent past. In our
experiments, we select ∆t = 10 seconds, and ∆T = 900 seconds. The acceleration of
a trading type presented as the derivative of its intensity is captured in feature set
v9, computed as an average rate of change over the previous 1 second. For instance,
if both dλma
/dt and dλlb
/dt have positive sign, then market ask orders and limit bid
orders are arriving faster, which could drive up the bid-ask spread and may lead to
an upward spread crossing.
16

Armed with the attributes described in Table 2, the feature vector xi for the
i-th data point in the training data set can be formed by simply selecting one or
more feature sets and then concatenating them together. However, we need to resist
the temptation to simply include as many attributes as possible: some features may
depict similar characteristics of the underlying data, making them redundant; some
features may not capture the intrinsic hidden patterns in the data and therefore only
introduce noise into model training; and the computation time required to carry out
the SVM model is sensitive to the size of the feature vectors. Therefore, we select a
smaller set of attributes in the feature space, which we can call an “economized fea-
ture set”, selected according to their contributions to the performance of the resulting
model as measured in terms of information gain [12],[15].
Entropy is a measure of uncertainty or unpredictability of a system, which is
defined as H(Y ) = − y∈Y p(y) log2(p(y)) for a given variable Y with various values
(y ∈ Y ), where p(y) represents the probability that a data point in the sample (e.g. a
training data set T) has value y ∈ Y . If all data points are equally likely, p(y) is the
total proportion |T(y)|/|T| of the data in the training set having label y, where T(y)
is the subset of data in T having label y.
If X is an attribute with values x ∈ X, then the conditional probability p(y|x)
is defined to be the proportion |T(x, y)|/|T(x)| of data with label y among all data
with attribute x. In this case the conditional entropy of Y , after observing feature X,
is defined as
H(Y |X) = −
x∈X
p(x)
y∈Y
p(y|x) log2(p(y|x)).
The information gain of Y contributed by X is then defined as
IG(X) = H(Y ) − H(Y |X),
and is a measure of the reduction of uncertainty about Y resulting from knowledge
of attribute X. We can then order all the attributes according to their information
gain, and select the attributes with the largest IG such that the performance of the
model built with these attributes is comparable to that obtained from the full original
17

feature set. (See Appendix.)
3.4 Learning Model Construction and Validation
In the proposed framework, we use Support Vector Machine techniques as described
in Section 2.
When the label set Y = {l1, ..., lk} has more than two members (k > 2), the clas-
sification problem is no longer binary. Instead, a multi-class categorization problem
should be solved. In this paper, we reduce the multi-class learning problem into a set
of binary classification tasks and build a binary classifier independently for each label
lk with a one-against-all method [19, 23]. In the one-against-all training method,
k binary SVM models are constructed, one for each class in Y ; to this end, when
training the ith SVM (i = 1, ..., k), examples in the ith class are treated as positive
while samples in other classes as negative. After k binary SVMs are built, an unseen
data point is assigned to the class with the largest value (equation (2.23)) generated
by these k binary models.
Before a newly built learning model for a metric is actually applied to predict
out-of-sample data, it is subject to evaluation and validation. We use the concepts
of precision (P), recall (R), and Fβ as measures of quality of the prediction process,
defined as follows. Consider a set T training data samples and we are trying to assign
the label y of a particular class. Say Y is the subset of T of examples that have the
true label y, and Z is the subset of examples that our SVM assigns to label y. Then
1. precision P is #(Y ∩ Z)/#(Z)
2. recall R is #(Y ∩ Z)/#(Y )
3. Fβ = (1 + β2
)PR/(β2
P + R), the weighted harmonic mean of P and R. When
β = 1, meaning that both precision and recall are equally important, we obtain
the F1-measure F1 = 2PR/(P + R).
18

We use an n-fold cross validation process that works as follows: the training
data set T is first divided into n equally-sized subsets {Ti, i = 1, ..., n}, each of which
has the same distribution as T with respect to the metric in question, and then the
following steps are repeated n times guided by iteration index i = 1, ..., n:
• Training: Set Ti is set aside as validation (test) set while the remaining n − 1
sets are combined to form a new training set Ti = ∪j=iTj, which is used to build
learning model Li.
• Predicting: Model Li is employed to label samples in validation set Ti and a
sample is predicted correctly if its assigned label by Li agrees with its true label.
• Measuring: Performance measures such as accuracy, precision, recall and Fβ-
measure are computed for model Li.
In this n-fold cross validation process, the average of the P, R, F1 measures from
n iterations are used as the overall measurements for the learning model.
4 Experiments and Results
This section summarizes our experimental results. Model efficiency as measured by
training and prediction time are demonstrated in 4.1. The model performance for
different feature sets and the analysis of feature selection are shown in 4.2. Illustrated
by the tables of corresponding p-values, the model’s performance is further justified
by the results of paired t-tests in 4.3. Finally, results of a test of a simple trading
strategy built on the spread crossing model is given in 4.4. With a full trading day
length, real data of 5 stocks from NASDAQ are used in the experiments. The pre-
diction time horizon, which specifies how long ahead of current time for the model
to project forecasting, is denoted by the notation ∆t and measured as the number of
events. All experiments are performed using Linux, with a 2.9GHz Intel Core i7 CPU
and 8GB 1600MHz memory.
19

4.1 Model Performance on Processing Time
Using training data equipped with all the features (basic, time-insensitive, time-
sensitive) from Table 2, the training and prediction times for the two metrics are
outlined in Table 3. Several random sets of 1500 training data points are constructed
by replicating the distribution of each metric in the original data. The numbers illus-
trated are the average time that models use for identifying each unseen example. Data
in “Training time” and “Testing time” columns are given in seconds and milliseconds
respectively. Table 3 indicates that, for the same ticker, the time used to train a
model as well as the time for predicting new data vary by metric. This difference may
be due to the number of support vectors of the models trained, which directly relates
to computation workload. For instance, the mid-price movement model of INTC has
171 support vectors and spread crossing has 291, which results in the training and
prediction times for mid-price (3.884s and 0.0250ms respectively), less than that of
spread crossing (5.020s and 0.0366ms).
Training time(s) Prediction time(ms)
Ticker mid-price spread crossing mid-price spread crossing
MSFT 1.116 2.960 0.0250 0.0180
INTC 3.884 5.020 0.0250 0.0366
AMZN 5.199 5.920 0.0683 0.0210
AAPL 4.246 5.390 0.0567 0.0311
GOOG 3.255 3.640 0.0783 0.0237
Mean 3.540 4.586 0.0507 0.0260
Table 3: Models’ average processing time with different tickers: all features are used for
training, training set size = 1500, ∆t = 5 (events)
Similarly, the time used for training and prediction varies for different tickers.
For example, there were 211 support vectors in the spread crossing model of MSFT,
so the 291 support vectors make the training and prediction time for INTC (5.020s
and 0.0366ms) slightly longer than for MSFT (2.960s and 0.0180ms). In the last row
“Mean”, the data indicate the efficiency of the models in an overall view.
The computation time for prediction is sufficiently fast for certain simple ap-
20

plications. For example, since the prediction time of AAPL, 0.0311ms, is less than
0.0612ms, which is the time diﬀerence between the upward spread crossing events
from the Row k −1 to Row k +4 in Table 1, the model could in principle perform fast
enough to inﬂuence corresponding trading decisions. A commercial implementation
would likely need to be faster to fully capture such spread crossing events — however,
our experiments with unoptimized code and equipment are only intended to illustrate
the concept.
4.2 Feature Selection
For each metric and ticker, we rank our original set of attributes by information gain.
A subset of attributes called the “economical set” is determined by sequentially includ-
ing one attribute at a time, beginning with the attribute with the highest information
gain, until the F1 measure of performance reaches 95% or more of the F1 measure of
the original set of attributes.
Original Set Economical Set
Ticker Label P(%) R(%) F1(%) P(%) R(%) F1(%)
AAPL
U (↑) 77.8 74.7 76.2 75.1 74.2 74.6
D (↓) 80.4 83.3 81.8 80.5 82.3 81.4
S (–) 99.1 98.8 99.0 99.3 98.6 99.0
GOOG
U (↑) 85.4 86.8 86.1 85.5 82.6 84.0
D (↓) 83.0 79.3 81.2 79.5 80.5 80.0
S (–) 98.6 99.5 99.0 98.6 99.5 99.0
Table 4: Performance measurement before and after feature selection: the original set is the
training set before feature selection (82 features), and the economical set is the training set
after feature selection (53 attributes for GOOG, 60 attributes for AAPL). Metric = spread
crossing; Training set size = 2000. ∆t = 5 (events).
Table 4 shows the spread crossing model training performance of the original
and economical feature sets for AAPL and GOOG. Among 82 original attributes,
53 and 60 attributes are selected for GOOG and AAPL, respectively. To illustrate
the distribution of attributes chosen as having highest information gain, a graphical
21

description is shown in Figure 4.1 for AAPL, with the top row for the mid-price metric
and the bottom row for spread-crossing.
More detailed information about the attributes selected via information gain is
given in the Appendix. In the remainder of this paper, we always limit ourselves to
the “economical” subset of attributes, a proper subset of the full original set of 82
attributes that varies by ticker and metric.
Figure 4.1: Economical training data set feature distribution selected by information gain.
The upper pie charts (grey and red) are the distribution of mid-price prediction features,
and the lower pie charts (blue and green) is the distribution of spread-crossing prediction
features. Basic = basic feature set; T-insen = time-insensitive feature set; T-sen = time-
sensitive feature set; Ask side = features only related to ask side; Bid side = features only
related to bid side; Price = features only related to price; Volume = features only related
to volume.Ticker = AAPL.
To compare the performance with different feature set configurations, we exam-
ine the SVM performance for the mid-price metric for AAPL in Table 5. The first
column outlines the precision (P), recall (R) and F1-measure (F1). The columns ∆(i,j)
denotes the difference Fi −Fj of entries in the prior columns, while the column “Avg.”
is the average of the last three columns.
22

AAPL: Mid-price
Label F0 F1 F2 F3 ∆(1,2) ∆(0,1) ∆(0,2) ∆(0,3) Avg.
P(%)
U (↑) 83.6 89.5 85.3 84.7 -4.2 5.9 1.7 1.1 2.9
D (↓) 75.2 88.0 84.9 84.3 -3.1 12.8 9.7 9.1 10.5
S (–) 79.8 90.0 80.1 81.9 -9.9 10.2 0.3 2.1 4.2
R(%)
U (↑) 76.5 89.5 77.4 76.5 -12.1 13.0 0.9 0.0 4.6
D (↓) 80.5 88.0 82.9 83.2 -5.1 7.5 2.4 2.7 4.2
S (–) 81.0 90.0 83.8 82.4 -6.2 9.0 2.8 1.4 4.4
F1(%)
U (↑) 79.9 89.5 81.4 80.6 -8.1 9.6 1.5 0.7 3.9
D (↓) 77.8 88.0 83.9 83.4 -4.1 10.2 6.1 5.6 7.3
S (–) 80.4 90.0 82.0 82.2 -8.0 9.6 1.6 1.8 4.3
Table 5: AAPL model mid-price prediction measurements with diﬀerent feature
sets, F0 = {Basic features}, F1 = {Basic and time-insensitive features}, F2 =
{Basic and time-sensitive features}, and F3 = {All features}. ∆(i,j) = value of column
Fj − value of column Fi. U = upward; D = downward; S = stationary. Training set size =
1500; ∆t = 5 (events)
From Table 5 we see that training data with the extended feature sets generally
outperform the basic feature set. However, more is not necessarily better: the perfor-
mance of feature set F3 is generally lower than those of either F1 or F2. For AAPL,
F1 tends to perform better than F2 or F3, but this depends on the stock.
4.3 Model Performance with Tests of Signiﬁcance
The model’s task is to forecast which of three possible classes the state of the order
book will be in at a time ∆t in the future. The possibilities are: the mid-price will
be higher, lower, or the same (mid-price metric), or there will be an upward spread
crossing, a downward spread crossing, or no spread crossing (spread crossing metric).
We measure the look-ahead interval ∆t in number of events, with possible values 5,
10, 15, or 20 events ahead of the current time. At each new event, the model will
make a new forecast.
Figures 4.2 and 4.3 report the prediction accuracy (percentage of correct pre-
dictions) for feature set F3 for a sample of 1200 data points. Generally, accuracy
declines with ∆t. However, we can see that the spread crossing accuracy for INTC
23

is higher for ∆t = 10 or 15 than for ∆t = 5, suggesting that the modeler will need
to investigate the natural time scales of each ticker’s price movements to choose the
optimal prediction horizon.
5 10 15 20
0
20%
40%
60%
80%
100%
t (events)
CorrectlyPredictedPercentage
MSFT INTC AMZN AAPL GOOG
Figure 4.2: Correct prediction percentage for different prediction horizons. Training data
set size = 1200, metric = mid-price movement; ∆t = 5, 10, 15, 20 (events)
5 10 15 20
0
20%
40%
60%
80%
100%
t (events)
CorrectlyPredictedPercentage
MSFT INTC AMZN AAPL GOOG
Figure 4.3: Correct prediction percentage for different prediction horizons. Training data
set size = 1200, metric = spread crossing; ∆t = 5, 10, 15, 20 (events)
To investigate the question of whether our positive results are merely due to
luck we use the paired t-test to examine the statistical significance of our performance
measurements. The benchmark for our t-test is the ZeroR classifier. As the sim-
plest classification method, the ZeroR classifier predicts all the data points to be the
24

majority class of the training data set. Maintaining the correctness of the major-
ity class, the performance of ZeroR therefore outperforms a random classifier, which
predicts data points in each class with equal probability. Hence, although there is
no prediction power in ZeroR, it is useful for determining a baseline performance for
other classification methods.
The paired t-test is performed between the multi-class SVM model and the
ZeroR using repeated 10-fold cross-validation, as explained below. The null hypothesis
is that the P, C and F1 measures of our SVM model have the same mean as those
of the ZeroR classifier. The paired t-test tells us the probability of obtaining the
observed average measurements conditional on the null hypothesis.
For our 10-fold cross-validation, we select 2000 data points from our total sample
and divide this into 10 segments of n2 = 200 data points. For each of 10 trials, we take
one of the 200 point segments as the test data, and the remaining n1 = 1800 points as
training data. We compute the performance (P, C, F1) on the test data. Doing this
for each 200 point segment in turn completes a 10-fold cross-validation experiment.
We then repeat this experiment J = 10 times, each with a new independent sample
of 2000 points drawn from the original total sample (about 400,000 rows of the order
book, in our case). For each performance measure X ∈ (P, C, F1), we compute, for
the ith test, i = 1, . . . , 100, di = X(SV M(i))−X(ZeroR(i)). The mean and standard
deviation of {d1, . . . , d100} are µd and sd, and our null hypothesis is H0 : µd = 0.
The t-test we use is the “corrected resampled t-test” from [27] to account for
the fact that the cross-validation trials are not completely independent. We use
t =
µd
sd
1
J
+ n2
n1
.
To reject the null hypothesis at confidence level α, we compute this t-statistic and use
p-values for the t(k) distribution, with k = 99 degrees of freedom.
Results are reported in Tables 6 and 7, which shows the p-values well below
typical significance levels α = 0.05 or 0.01. We deduce that the P, C, and F1 measures
are all statistically significantly better than the performance of our baseline classifier.
25

ZeroR Multi-class SVMs
Ticker P(%) C(%) F1(%) P(%) C(%) F1(%)
AAPL 36.22 59.14 44.21 91.03 (1E-08) 90.25 (8E-07) 90.01 (7E-08)
GOOG 34.11 56.43 43.01 92.00 (5E-09) 91.33 (1E-07) 91.51 (4E-08)
Table 6: Results of paired t-test of spread crossing prediction, ∆t = 5 (events), training
data size = 2000. The measurements are weighted precision (P%), percentage of correct
prediction (C%) and weighted F1-measure (F1%). Cross-validation n = 10, Trials = 10; The
values in parentheses are the p-values obtained from the paired t-test in scientific notation.
Values less than 0.01 represent significance to the 99% confidence level.
ZeroR Multi-class SVMs
Ticker P(%) C(%) F1(%) P(%) C(%) F1(%)
MSFT 36.00 52.73 41.50 74.34 (1E-07) 74.65 (1E-04) 72.00 (1E-06)
INTC 41.00 51.63 43.00 75.00 (7E-07) 73.54 (1E-04) 71.20 (2E-06)
Table 7: Results of paired t-test of mid-price prediction, ∆t = 5 (events), training data size
= 4000. Notation as in Table 6.
4.4 Testing a Simple Trading Strategy
As a final test of the effectiveness of the SVM model, we test it with a simple-minded
trading strategy against our data. The strategy below is not meant to represent a
realistic trading implementation, but rather a simple sanity test to illustrate whether
the model produces useful information in the context of sample market data.
For a window of time ∆T = 300, 600, 900, 1800 or 3600 seconds, we run the
spread-crossing model for a prediction horizon of ∆t = 5 events, using four hours of
AAPL order book data on June 21, 2012. At each event, if the model predicts no
spread crossing, we do nothing. If it predicts an upward spread crossing at horizon
∆t, we submit a market buy order for a nominal $100 at best ask and after ∆t = 5
events, no matter what happens, sell it back for the current best bid. Similarly for a
prediction of downward spread crossing. We assume no transaction costs and collect
all our profit and loss at the end of the trading window. The model is retrained using
the most recent 2000 events once per hour. Table 8 shows the results averaged over
30 overlapping windows of each size across the four hour testing period.
26

AAPL
Duration (s) Mean (Tick) Std. Dev. (Tick) Avg. # of Trades
300 1.5390 1.4352 13.0
600 1.7781 1.1113 22.0
900 1.9781 1.5113 31.0
1800 2.8658 1.8491 68.0
3600 3.5862 1.1083 126.0
Table 8: Trading test by using various length of sliding windows based on the prediction of
spread crossing, ∆t = 5 (events). In each trading window, the simulated trader goes long
$100.00 on upward spread crossing and short back 5 events later; goes short with $100.00
if downward spread crossing occurs; no moves if stationary prediction occurs. After the
window duration ends, the accumulated profits/loss is computed. In total, 30 windows are
included in the test. The mean and standard deviation of the profit of different window
lengths are shown in the table. 1 tick = $0.01; Ticker = AAPL.
Figure 4.4 illustrates this further for the case of a window of 1800 seconds for
AAPL. Most events are correctly labelled as stationary, and indicated by the purple
curve on the graph. Every so often there is an up or down spread crossing event;
these are labeled according to whether the model correctly predicted them. In the
picture is also the size of the bid-ask spread for reference, with scale on the right
axis. Profit and loss for our simple trading strategy is reported in Figure 4.5. In
this trial, we can see a realized profit of about 3 basis points over this 30 minute
window. In the places where the profit-loss curve drops down, there are events that
are mis-predicted by the model. For example, the profit-loss curve drops around the
time 600s in Figure 4.5, which is due to an incorrect upward prediction when there
actually follows a down event. For similar reasons, upward movements in the profit
curve indicate where correct predictions are made.
27

583
583.5
584
584.5
585
585.5
586
586.5
587
587.5
588
588.5
589
589.5
590
Price($)
0 200 400 600 800 1000 1200 1400 1600 1800
0
0.5
1
1.5
2
2.5
3
Bid−AskSpread($)
Time Elapsed (s)
correct stationary label (mid−price)
correct up label
correct down label
down labelled as up
stationary labelled as down
up labelled as stationary
down labelled as stationary
bid−ask spread
Figure 4.4: Example: prediction with labelled signals and bid-ask spread evolution. Left
y-axis is the stock price in US dollar; the right y-axis is the bid-ask spread in US dollar.
Time duration = 1800 seconds. The purple box contains a time period and the prediction,
which will be zoomed in the Figure 4.6. ∆t = 5 (events); Ticker = AAPL.
583
583.5
584
584.5
585
585.5
586
586.5
587
587.5
588
588.5
589
589.5
590
Price($)
0 200 400 600 800 1000 1200 1400 1600 1800
−2
−1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
x 10
−4
Profit($)
Time Elapsed (s)
correct stationary label (mid−price)
correct up label
correct down label
down labelled as up
stationary labelled as down
up labelled as stationary
down labelled as stationary
profit−loss curve
0 200 400 600 800 1000 1200 1400 1600 1800
−2
−1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
x 10
−4
Profit($)
Time Elapsed (s)
0 200 400 600 800 1000 1200 1400 1600 1800
−2
−1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
x 10
−4
Profit($)
Time Elapsed (s)
zero profit line
Figure 4.5: Example: prediction with labelled signals and proﬁt-loss curve. Left y-axis
is the stock price in US dollar; the right y-axis is the bid-ask spread in US dollar. Time
duration = 1800 seconds. ∆t = 5 (events); Ticker = AAPL.
28

1013.25294782 1013.25294783 1013.25294784 1013.25294785 1013.25294786 1013.25294787
586.45
586.5
586.55
586.6
586.65
586.7
Best Ask
Best Bid
586.67
586.69
586.50
586.59
Upward
Prediction
Figure 4.6: Zoom in: upward crossing prediction with the best ask and best bid prices
in the purple box of Figure 4.4. The y-axis is the stock price in US dollars; the x-axis is
the event time stamps (to the nanosecond); the green arrow indicates an upward prediction
occurs at that moment. ∆t = 5 (events); Ticker = AAPL.
Figure 4.6 shows a close-up view of the region of the purple box in Figure 4.4.
We see that at time t = 1013.2594782s, the model gives an upward prediction denoted
by a green arrow, so a long position is taken at the best ask price $586.59. At 5 events
later, t = 1013.2594787s, the prediction turns out to be correct, and the asset is sold
at the best bid price $586.67 to earn a profit.
5 Conclusions
The machine learning modeling method described in this paper provides a framework
to automate the prediction process for limit order book dynamics in real time. Treat-
ing prediction of various metrics as a supervised learning problem, our method first
reduces the problem at hand into a set of binary classification tasks and then builds
multi-class models by using SVMs for each binary task.
To improve the efficiency in training models and labeling unseen samples, fea-
tures are selected according to information gain so that only those attributes sig-
29

nificantly contributing to the performance of the resulting model are retained. The
validity and robustness of models are verified with a n-fold cross-validation process
that evaluates model performance in terms of precision, recall and F1-measure.
Our experiments with real-world data show that the feature sets designed in the
proposed framework are effective. For instance, the models built for mid-price move-
ment by using only the basic features can attain on average 79.6%, 82.6%, and 80.3%
for precision, recall and F1-measure, respectively; the performance can be further im-
proved by about 6.3% and 3.2% when time-insensitive and time-sensitive features are
included.
Experiments presented in this paper also demonstrate the efficiency of the frame-
work, as the time spent on classifying an unseen data point is less than a tenth of
a millisecond using unoptimized code on and a single CPU implementation. Our
comparison study with a baseline ZeroR predictor indicates that our framework sig-
nificantly outperforms the latter, and the improvement in performance is statistically
significant. Moreover, simple trading strategies based on the model’s forecasts can
achieve profitable returns tested against historical data, with low risk.
A related multi-class SVM approach has been used successfully in [14] to fore-
cast high frequency foreign exchange prices, but ours is the first work we know to
demonstrate that this approach is effective also for equity limit order book forecast-
ing.
We conclude with a few comments comparing our approach with [14]. A multi-
kernel SVM modeling method in [14] forecasts the EUR/USD price evolution and tags
a given transactional event with respect to spread crossing direction based on three
SVM models. Each of these three models is trained with a weighted sum of multi-
ple kernels and recognizes one of the three labels: up, down or stationary. There, a
prediction for a given data sample is not made unless the three independent SVMs
unanimously agree on the classification. As a result, the prediction coverage – pro-
portion of labeled data – of the model could be low, which may represent a loss of
trading information.
30

In contrast, the models in the present paper label every input data sample by
assigning the label corresponding to the hyperplane with the largest positive distance,
which means there is complete prediction coverage. In addition, extensive compara-
tive experiments show that the multi-class SVM modeling techniques proposed in this
paper reduce the training cost compared to the three multi-kernel SVMs in [14], but
still attain robust performance in terms of precision and recall.
Acknowledgments: The authors would like to thank the Chair of Economet-
rics at Humboldt-Universität zu Berlin, Germany, for providing data used in this
paper.
6 Appendix
As described in section 4.2, the attributes used in this study are those in the “eco-
nomical set” selected according to those having the largest information gain, a subset
of the original set of attributes described in Table 2. For reference, we report more
detail here on the specific features selected by this method.
In Tables 9 and 10 we show the top ten features selected for GOOG for each of
the two metrics, along with the Information Gain (IG) values.
In Table 11 we define an enumeration of the original set of attributes, so that we
may specify in Table 12 the complete set of economical features used our experiments
for AAPL and GOOG.
31

Spread crossing
Rank IG Type Description
1 0.3711 Basic volume at best ask
2 0.3186 Basic volume at best bid
3 0.2512 T-insen mid-price at best ask and bid
4 0.2490 T-insen mid-price at level 4
5 0.2098 T-insen bid-ask spread at best ask and bid
7 0.1957 T-insen bid-ask spread at level 4
9 0.1675 T-sen acceleration of limit bid order arrival
10 0.1639 T-insen bid-ask spread at level 8
... ... ... ...
Table 9: Examples: features selected with top 10 information gain for spread crossing, along
with the feature type and descriptions. Basic = basic feature; T-insen = time-insensitive
feature; T-sen = time-sensitive feature; ∆t = 5 (events), Ticker = GOOG.
Mid-price
Rank IG Type Description
3 0.2032 T-insen average price on ask side
4 0.1907 Basic ask price at level 3
5 0.1871 T-insen bid-ask spread at best ask and bid
6 0.1593 Basic ask price at level 2
7 0.1568 T-sen average intensity of bid cancellation
8 0.1566 T-sen average intensity of market bid order
9 0.1565 T-sen average intensity of limit bid order
10 0.1559 T-insen average volume at bid side
... ... ... ...
Table 10: Examples: features selected with top 10 information gain for mid-price movement,
along with the feature type and descriptions. ∆t = 5 (events), Ticker = GOOG.
32

Basic Set Description(i = level index)
v1 = {ak
1 }4n
k=1 = {Pask
i , V ask
i , Pbid
i , V bid
i }n
i=1, price and volume(n levels)
Time-insensitive Set Description(i = level index)
v2 = {ak
2 }2n
k=1 = {(Pask
i − Pbid
i ), (Pask
i + Pbid
i )/2}n
i=1, bid-ask spreads and mid-prices
v3 = {ak
3 }2n
k=1 = {|Pask
i+1 − Pask
i |, |Pbid
i+1 − Pbid
i |}n
i=1, price differences
v4 = {ak
4 }4
k=1 = { 1
n
n
i=1 Pask
i , 1
n
n
i=1 Pbid
i , 1
n
n
i=1 V ask
i , 1
n
n
i=1 V bid
i }, mean prices and volumes
v5 = {ak
5 }2
k=1 = { n
i=1(Pask
i − Pbid
i ), n
i=1(V ask
i − V bid
i )}, accumulated differences
Time-sensitive Set Description(i = level index)
v6 = {ak
6 }4n
k=1 = {dPask
i /dt, dPbid
i /dt, dV ask
i /dt, dV bid
i /dt}n
i=1, price and volume derivatives
v7 = {ak
7 }6
k=1 = {λla
∆t, λlb
∆t, λma
∆t , λmb
∆t , λca
∆t, λcb
∆t } average intensity of each type
v8 = {ak
8 }4
k=1 = {1{λla
∆t>λla
∆T
}, 1{λlb
∆t>λlb
∆T
}, 1{λma
∆t >λma
∆T
}, 1{λmb
∆t >λmb
∆T
}}, relative intensity indicators
v9 = {ak
9 }4
k=1 = {dλma/dt, dλlb/dt, dλmb/dt, dλla/dt}, accelerations(market/limit)
Remarks:
∗ ak
1 (k = 4i − j) denotes the ask price, volume; bid price, volume at level i for j = 3, 2, 1, 0, respectively.
∗ ak
2 (k = 2i − j) denotes the bid-ask spread and mid-price at level i for j = 1, 0, respectively.
∗ ak
3 (k = 2i − j) denotes the price differences for ask and bid at level i for j = 1, 0, respectively. (Pn+1 = P1)
∗ ak
6 (k = 4i − j) denotes derivative of ask price, volume; bid price, volume at level i for j = 3, 2, 1, 0, respectively.
Table 11: Feature vector sets enumerated.
Ticker List of Attributes ak
l in Economical Set (Mid-price)
AAPL
a4
1, a2
1, a3
2, a15
2 , a4
2, a2
2, a8
2, a32
1 , a13
2 , a1
7, a3
9, a14
1 , a12
1 , a34
1 , a28
1 , a4
9, a10
1 , a40
1 , a8
1, a24
1
a18
1 , a5
1, a13
1 , a9
1, a25
1 , a33
1 , a2
4, a17
1 , a21
1 , a1
1, a20
2 , a2
3, a2
8, a2
6, a4
6, a2
9, a2
5, a7
2, a14
2 , a39
1
GOOG
a20
2 , a8
2, a1
4, a10
1 , a1
2, a5
1, a6
7, a4
7, a2
7, a4
4, a2
1, a13
2 , a15
2 , a4
9, a28
1 , a27
1 , a30
1 , a25
1 , a1
5, a29
1
a31
1 , a7
2, a3
7, a3
6, a1
6, a1
8, a24
1 , a4
8, a1
7, a5
7, a1
3, a19
2 , a20
1 , a40
1 , a8
1, a6
1, a2
4, a16
1
Ticker List of Attributes ak
l in Economical Set (Spread Crossing)
AAPL
a20
2 , a2
1, a3
4, a1
4, a1
6, a1
2, a3
6, a2
2, a1
9, a4
9, a1
7, a29
1 , a9
1, a13
1 , a25
1 , a37
1 , a5
1, a1
1, a2
6, a2
9, a14
2
a17
1 , a33
1 , a7
2, a20
2 , a6
7, a21
1 , a2
5, a2
3, a31
1 , a35
1 , a39
1 , a4
7, a2
7, a1
8, a2
8, a4
6, a3
9, a27
1 , a11
1 , a19
1
a23
1 , a15
1 , a3
1, a7
1, a4
8, a11
2 , a8
2, a20
1 , a10
2 , a15
2 , a9
2, a6
1, a3
2, a13
2 , a4
1, a34
1 , a4
2, a5
2, a30
1
GOOG
a2
1, a4
1, a2
2, a8
2, a1
2, a4
2, a7
2, a14
2 , a2
9, a15
2 , a19
2 , a9
2, a1
3, a9
1, a17
2 , a5
1, a6
7, a4
7, a2
7, a2
3, a2
6
a1
1, a13
1 , a1
2, a19
1 , a2
5, a27
1 , a15
1 , a1
5, a2
8, a3
1, a1
9, a39
1 , a3
9, a7
1, a4
6, a3
2, a15
2 , a13
2 , a6
1, a8
2
a16
2 , a22
1 , a1
4, a35
1 , a23
1 , a11
2 , a10
2 , a11
1 , a17
1 , a3
4, a29
1 , a21
1 , a20
2 , a4
9, a1
6
Table 12: Complete list of the selected attributes for the economical sets. ak
l denotes the
k-th attribute from the vector vl in Table 11, l = 1, 2, ..., 9.
33

References
[1] B. Biais, D. Martimort, and J.-C. Rochet. Competing mechanisms in a common value
environment. Ecnonmetrica, 68(4):799–837, July 2000.
[2] A. Blazejewski and R. Coggins. Application of self-organizing maps to clustering of
high-frequency financial data. In M. Purvis, editor, Australasian Workshop on Data
Mining and Web Intelligence (DMWI2004), volume 32, pages 85–90, Dunedin, New
Zealand, 2004. ACS.
[3] A. Blazejewski and R. Coggins. A local non-parametric model for trade sign inference.
Physica A, 348:481–495, March 2005.
[4] J.-P. Bouchaud, M. Mezard, and M. Potters. Statistical properties of stock order books:
Empirical results and models statistical properties of stock order books: Empirical
results and models. Quantitative Finance, 2(4):251–256, 2002.
[5] A. Bovier, J. Cerny, and O. Hryniv. The opinion game: Stock price evolution from mi-
croscopic market modeling. International Journal of Theoretical and Applied Finance,
9(1):91–111, February 2006.
[6] C. J. Burges. A tutorial on support vector machines for pattern recognition. Data
Mining and Knowledge Discovery, 2(2):121–167, June 1998.
[7] A. Cartea, S. Jaimungal, and J. Ricci. Buy low sell high: A high frequency trading
perspective, November 2011.
[8] R. Cont. Statistical modeling of high-frequency financial data. Signal Processing Mag-
azine, IEEE, 28(5):16–25, August 2011.
[9] R. Cont and A. de Larrard. Order book dynamics in liquid markets: Limit theorems
and diffusion approximations, Feburary 2012.
[10] R. Cont, S. Stoikov, and R. Talreja. A stochastic model for order book dynamics.
Operations Research, 58(3):549–563, May 2010.
[11] C. Cortes and V. Vapnik. Support-vector networks. Machine Learning, 20(3):273–297,
September 1995.
[12] T. M. Cover and J. A. Thomas. Elements of Information Theory. John Wiley & Sons,
Inc., 1991.
[13] K. Crammer and Y. Singer. On the algorithmic implementation of multiclass kernel-
based vector machines. Journal of Machine Learning Research, 2:265–292, March 2001.
[14] T. Fletcher and J. Shaww-Taylor. Multiple kernel learning with fisher kernels for high
frequency currency prediction. Computational Economics, 42(2):217–240, August 2013.
[15] H. Föllmer. On entropy and information gain in random fields. Zeitschrift für
Wahrscheinlichkeitstheorie und Verwandte Gebiete, 26(3):207–217, September 1973.
34

[16] T. Foucault. Order flow composition and trading costs in a dynamic limit order market.
Journal of Financial Markets, 2(2):99–134, May 1999.
[17] T. Foucault, O. Kadan, and E. Kandel. Limit order book as a market for liquidity.
Review of Financial Studies, 18(4):1171–1217, August 2005.
[18] J. Hasbrouck. Empirical Market Microstructure: The Institutions, Economics, and
Econometrics of Securities Trading. Oxford University Press, USA, January 2007.
[19] T. Hastie and R. Tibshirani. Classification by pairwise coupling. Annals of Statistics,
26(2):451–471, 1998.
[20] H. He and A. N. Kercheval. A generalized birth-death stochastic model for high-
frequency order book dynamics. Quantitative Finance, 12(4):547–557, April 2012.
[21] T. Joachims. Making Large-scale Support Vector Machine Learning Practical. MIT
Press, Cambridge, MA, USA, 1999.
[22] E. Jondeau, A. Perilla, and G. Rockinger. Optimal Liquidation Strategies in Illiquid
Markets, volume 553. Springer Berlin Heidelberg, 2005.
[23] S. Knerr, L. Personnaz, and G. Dreyfus. Single-layer learning revisited: A stepwise pro-
cedure for building and training a neural network. Neurocomputing, 68:41–50, January
1990.
[24] J. T. Linnainmaa and I. Rosu. Weather and time series determinants of liquidity in a
limit order market, 2008.
[25] A. W. Lo, A. C. MacKinlay, and J. Zhang. Econometric models of limit-order execu-
tions. Working Paper 6257, National Bureau of Economic Research, November 1997.
[26] J. Mercer. Functions of positive and negative type and their connection with the
theory of integral equations. Philosophical Transactions of the Royal Society of London,
209:415–446, 1909.
[27] C. Nadeau and Y. Bengio. Inference of the generalization error. Machine Learning,
52(3):239–291, September 2003.
[28] A. Ng. Cs229 lecture notes, part v: Support vector machines. Technical report, Stanford
University, 2012.
[29] C. Parlour. Price dynamics in limit order markets. Review of Financial Studies,
11(4):789–816, 1998.
[30] C. Parlour and D. J. Seppi. Limit Order Market: A Survey. Elsevier North-Holland,
2008.
[31] J. C. Platt. Fast training of support vector machines using sequential minimal opti-
mization. In B. Schölkopf, C. J. C. Burges, and A. J. Smola, editors, Advances in kernel
methods, volume 3, pages 185–208. MIT Press, Cambridge, MA, USA, September 1999.
35

[32] N. Popper. Times topic: High-frequency trading. Technical report, The New York
Times, 2012.
[33] I. Rosu. A dynamic model of the limit order book. Review of Financial Studies,
22:4601–4641, June 2009.
[34] J. Shabolt and J. Taylor. Neural Networks and the Financial Markets: Predicting,
Combining and Portfolio Optimisation. Springer UK London, 2002.
[35] H. H. S. Shek. Modeling high frequency market order dynamics using self-excited point
process, 2010.
[36] P. Tino, N. Nikolaev, and X. Yao. Volatility forecasting with sparse bayesian kernel
models. In 4th International Conference on Computational Intelligence in Economics
and Finance, pages 1052–1058, 2005.
[37] B. Zheng, E. Moulines, and F. Abergel. Price jump prediction in a limit order book.
Journal of Mathematical Finance, 3:242–255, April 2013.
36

Real-Time Stock Market Analysis using Spark Streaming

More Related Content

What's hot

Similar to Real-Time Stock Market Analysis using Spark Streaming

More from Sigmoid

Recently uploaded

Real-Time Stock Market Analysis using Spark Streaming