Note: Descriptions are shown in the official language in which they were submitted.
CA 02910754 2015-10-29
SYSTEMS, METHODS AND DEVICES
FOR MODELLING OPERATIONAL RISK
FIELD
[0001] Aspects of the present disclosure relate to the field of operational
risk
management and particularly to systems, methods, devices and computer-readable
media
that model operational risk.
BACKGROUND
[0002]
Institutions such as banks have various lines of business which may be subject
to different classes of operational risks. The Advanced Measurement Approaches
(AMA)
set out by the Basel II Capital Accord is an operational risk assessment
framework which
identifies a series of problems for the financial industry to solve relative
to risk modelling.
[0003] Risks can
be associated with distributions, and modelling for such risk for a large
organization can require computational-heavy and time-consuming simulation
models.
Alternatives which can reduce computation, have shorter run times, or may be
more
credible would be beneficial.
SUMMARY
[0004] In
accordance with an aspect, a method for modelling operational risk is
provided. The method includes:
retrieving external loss data from at least one external
data source; retrieving internal loss data from at least one internal data
source; generating,
with at least one processor, mapped loss data by mapping the internal and
external loss
data by at least one of: source, unit and time period for unit-of-measure
processing; storing
the mapped loss data in at least one memory;
conducting, with the at least one
processor, model parameterization and exploratory data analysis on the mapped
loss data
to generate loss models based on the mapping; performing a simulation across
the loss
models to convolve frequency and severity components into an aggregate loss
distribution;
storing the aggregate loss distribution in the at least one memory; and
producing a
measure of operational risk based on the aggregate loss distribution.
[0005] In
accordance with another aspect, a device for modelling operational risk is
provided. The device includes: at least one memory; and at least one
processor. The at
least one processor is configured for: retrieving external loss data from at
least one
- 1 -
CA 02910754 2015-10-29
external data source; retrieving internal loss data from at least one internal
data source;
generating mapped loss data by mapping the internal and external loss data by
at least
one of: source, unit and time period for unit-of-measure processing; storing
the mapped
loss data in the at least one memory; conducting model parameterization and
exploratory
data analysis on the mapped loss data to generate loss models based on the
mapping;
performing a simulation across the loss models to convolve frequency and
severity
components into an aggregate loss distribution; storing the aggregate loss
distribution in
the at least one memory; and producing a measure of operational risk based on
the
aggregate loss distribution.
[0006] In accordance with another aspect, a non-transitory, computer-
readable medium
or media is provided. The medium or media has stored thereon instructions
which when
executed by at least one processor configure the at least one processor for:
retrieving
external loss data from at least one external data source; retrieving internal
loss data from
at least one internal data source; generating, with at least one processor,
mapped loss
data by mapping the internal and external loss data by at least one of:
source, unit and
time period for unit-of-measure processing; storing the mapped loss data in at
least one
memory; conducting, with the at least one processor, model parameterization
and
exploratory data analysis on the mapped loss data to generate loss models
based on the
mapping; performing a simulation across the loss models to convolve frequency
and
severity components into an aggregate loss distribution; storing the aggregate
loss
distribution in the at least one memory; and producing a measure of
operational risk based
on the aggregate loss distribution.
[0007] In various further aspects, the disclosure provides corresponding
systems and
devices, and logic structures such as machine-executable coded instruction
sets for
implementing such systems, devices, and methods.
[0008] In this respect, before explaining at least one embodiment in
detail, it is to be
understood that the disclosure is not limited in its application to the
details of construction
and to the arrangements of the components set forth in the following
description or
illustrated in the drawings. The disclosure is capable of other embodiments
and of being
practiced and carried out in various ways. Also, it is to be understood that
the phraseology
and terminology employed herein are for the purpose of description and should
not be
regarded as limiting.
- 2 -
CA 02910754 2015-10-29
[0009] Many further features and combinations thereof concerning embodiments
described herein will appear to those skilled in the art following a reading
of the instant
disclosure.
DESCRIPTION OF THE FIGURES
[0010] In the figures, embodiments are illustrated by way of example. It is
to be
expressly understood that the description and figures are only for the purpose
of illustration
and as an aid to understanding.
[0011] Embodiments will now be described, by way of example only, with
reference to
the attached figures.
[0012] FIG. 1 is a block diagram illustrating aspects of an example system
according to
some embodiments.
[0013] FIG. 2 is a flowchart depicting aspects of an example method according
to some
embodiments.
[0014] FIG. 3 is a schematic diagram showing aspects of an example of
computing
device, exemplary of an embodiment.
DETAILED DESCRIPTION
[0015] Embodiments of methods, systems, and apparatus are described through
reference to the drawings.
[0016] The following discussion provides many example embodiments of the
inventive
subject matter. Although each embodiment represents a single combination of
inventive
elements, the inventive subject matter is considered to include all possible
combinations of
the disclosed elements. Thus if one embodiment comprises elements A, B, and C,
and a
second embodiment comprises elements B and D, then the inventive subject
matter is also
considered to include other remaining combinations of A, B, C, or D, even if
not explicitly
disclosed.
[0017] In some embodiments, systems, methods, devices, and computer-readable
media are provided generally in relation to operational risk management and
more
- 3 -
CA 02910754 2015-10-29
particularly, in relation to modelling and/or providing alerts or indications
associated with
the management of operational risk.
[0018] In some
embodiments, the framework described herein may be based on Loss
Distribution Approach (LDA) under Basel/OSFI (Office of the Superintendent of
Financial
Institutions) guidelines. Aspects of the present disclosure may relate to
testing the
goodness-of-fit (GoF) of a distributional family to operational loss data. In
some
embodiments, systems, methods, devices, and computer-readable media described
herein
may provide a framework which can manage a device's handling of operational
loss
models in a manner which may be computationally faster, more efficient and/or
more
accurate.
[0019] FIG. 1 is
a block diagram illustrating aspects of an example system 100 including
aspects suitable for use in modelling operational risk. As illustrated in the
embodiment
shown, the system 100 can include one or more systems or devices 102, 102A,
102B,
104, 106, 108, and network(s) 150. Some systems, devices and/or networks may
be
associated with one or more entities, for example financial institutions,
insurance
companies, etc. The systems and devices 102, 102A, 102B, 104, 106, 108 may
include
computing resources such as computers, servers, mainframes, mobile devices,
multiprocessor systems, shared or distributed computing resources, data
sources, data
storage devices, memory devices, and the like.
[0020] In some embodiments, a system 102 or group of devices 102A, 102B,
network(s)
150, and/or shared resources may be associated with an entity such as a bank.
Loss data
associated with the entity can be treated as internal loss data, and loss data
associated
with a different entity or an unassociated device/system can be treated as
external loss
data.
[0021] The systems and devices can, in some embodiments, include one or more
processors in a single device or split across any number of devices in the
system. These
processor(s) can be configured to perform one or more aspects of an
operational risk
management/modelling process.
[0022] The systems/devices may be connected over any suitable connection
including
over data bus(ses), local connection(s), communication network(s), and the
like.
- 4 -
CA 02910754 2015-10-29
[0023] In some embodiments, the system(s) or device(s) may include one or more
processor(s), memory/storage device(s), input/output interface(s), input
device(s),
resource(s) and/or any other devices or mechanisms suitable for or involved in
performing
aspects of the methods and functions described herein.
[0024] FIG. 2 depicts a flowchart illustrating aspects of an example method
200 for
modelling operational risk. At 205 and 210, processor(s) associated with an
entity retrieve
loss data. External loss data can include data regarding an actual loss
experienced by an
external entity, and internal loss data can include data regarding an actual
loss
experienced by the entity itself. Internal and/or external loss data can, in
some
embodiments, include business line and/or unit(s), loss type and/or
category(ies), loss
amount, exchange rate, loss occurrence date, loss detection date, associated
people/processes/systems/group, geographic and/or political region, consumer
price
indices, cause(s), risk class(es) and/or any other suitable data associated
with a loss or
near loss. In some embodiments, loss data regarding an internal loss may
include more
data and loss data regarding an external loss.
[0025] In some examples, the external loss data can be retrieved or otherwise
obtained
from device(s), system(s) associated with those other entities. In some
examples, the
external loss data can be aggregated at one or more central data sources such
as a
device/system 104 associated with the Operational Riskdata eXchange
Association
(ORX). In some examples, the external loss data can be retrieved continuously,
periodically and/or in real or near-real time. In some examples, the external
loss data can
be stored internally at a system/device associated with the entity and can be
updated
whenever new data is received. In some examples, retrieving external loss data
from an
external data source can include retrieved an internal copy of the external
loss data.
[0026] In some embodiments, internal loss data can be received, aggregated,
compiled
or otherwise obtained from one or more device(s)/system(s) associated with the
entity.
[0027] At 215, processor(s) generate mapped loss data from the internal and
external
loss data. In some embodiments, this can include generating a mapping based on
one or
more categories of loss data for unit-of-measure (UOM) processing. In some
embodiments, the mapping can be generated based on at least one of source,
unit, and
time period. In some examples, generating the mapping can include translating,
- 5 -
CA 02910754 2015-10-29
converting, rearranging or otherwise calibrating the loss data so that it is
suitable for unit-
of-measure processing. In some examples, this may allow for more rapid data
extraction
and/or unit-of-measure processing.
[002E3] In some embodiments, generation of the loss data mapping can be done
periodically, continuously, upon receipt of a request and/or when new loss
data is
retrieved.
[0029] At 220, processor(s) can store the mapped loss data in at least one
memory. In
some examples, the at least one memory can be a storage device at a computing
device
such as a hard drive, solid state memory, flash drive, optical drive, tape
drive, other non-
volatile memory(ies) and the like. In some examples, the at least one memory
can include
a cache or volatile memory such as RAM (random-access memory), and the like.
[0030] In some examples, the mapped loss data can be stored at a memory of a
computing, storage or other device via one or more network(s) 150. Such
memory(ies) can
include cloud storage, network attached storage, etc.
[0031] At 225, processor(s) conduct model parameterization and/or exploratory
data
analysis (EDA) on the mapped loss data to generate loss models. In some
examples, EDA
and model parameterization can include analyzing, estimating and/or generating
different
loss models based on the mapped loss data. In some embodiments, the loss
models can
include models for frequency, severity, correlations and/or potential losses.
The loss
models can, in some examples, be modelled for each UOM. In some examples, the
UOMs
can be based on time period sizes, source, unit, etc.
[0032] In some examples, the loss models can be based on different levels
of
granularity. For example, granular levels can include regional or political
boundaries,
people/group/unit risks, etc.
[0033] Processor(s) can generate frequency and severity models by estimating
probability distributions and selecting an appropriate frequency model and
severity
distribution for each UOM. Processor(s) can generate correlation models which
captures
inter and intra relationships of various UOMs. The correlation models may
includes testing
frequency autocorrelation, frequency-to-severity correlation, frequency to
frequency
correlations and/or inter-business line loss correlations across cells / UOMs.
- 6 -
CA 02910754 2015-10-29
[0034] In some examples, model parameterization and/or EDA can include
regression
analysis such as linear regression, upper tail analysis, fat tail regression,
frequency
regression, or other models.
[0035] In some examples, model parameterization and/or EDA can be based on one
or
more parameters. These parameters can be defined by the system/processor(s) or
through administrator input. In some examples, the parameters can include
clustering,
internal/external severity distribution weighting, severity distribution
selections, and the
like.
[0036] At 230, processor(s) perform simulations across one or more of the loss
models
to convolve frequency and severity components into an aggregate loss
distribution. In
some embodiments, the simulations can include value-at-risk (VaR)
simulation(s) of loss
aggregates across all UOMs. The VaR can, in some examples, be calculated
through the
aggregate loss distribution using a Student-t copula applied at the Business
Lines level. In
some examples, defined operational risk loss scenarios or parameters can
inform the
severity profiles used in the VaR simulation or may confirm that the severity
selections are
inline with expected views.
[0037] At 235, the aggregate loss distribution(s) can be stored in at least
one memory.
similar to the mapped loss data or otherwise. In some examples, the at least
one memory
can be a storage device at a computing device such as a hard drive, solid
state memory,
flash drive, optical drive, tape drive, other non-volatile memory(ies) and the
like. In some
examples, the at least one memory can include a cache or volatile memory such
as RAM
(random-access memory), and the like.
[0038] In some examples, the aggregate loss distribution(s) can be stored
at a memory
of a computing, storage or other device via one or more network(s) 150. Such
memory(ies)
can include cloud storage, network attached storage, etc.
[0039] At 240, processors produce a measure of operational risk based on the
aggregate loss distribution(s). In some examples, the loss distributions of
each cell can be
combined to determine the overall diversified regulatory capital.
- 7 -
CA 02910754 2015-10-29
[0040] In some embodiments, the VaR results at the 99.9 percentile from these
correlated loss simulations can then be allocated among the business lines to
meet Pillar 1
regulatory requirements.
[0041] In some embodiments, processor(s) can generate message or otherwise
communicate measure(s) of operational risk to one or more devices/systems
and/or output
devices.
[0042] In some embodiments, the measure of operational risk can be used to
determine
whether one or more units of the entity or the entity as a whole meets
regulatory
requirements such as capital reserve requirements.
[0043] In some
embodiments, if reserve requirements are not met, or if capital reserves
and within a defined threshold, processor(s) can be configured to generate an
alert. Such
alerts can include generating an email or other message(s), generating a
displayed
warning or message, generating a log entry, generating an audible alert, and
the like.
[0044] In some embodiments, processor(s) can be configured to generate reports
or
logs and/or to pause or otherwise allow an administrator to review various
stages of the
process such as the parameterization, correlation modelling, VaR simulation
and
aggregations and measure of operational risk. In some examples, the
processor(s) can
receive input(s) to modify one or more parameters at one or more stages. In
some
examples, business environment and internal control factors (BEICF) can be
used to
facilitate forward-looking measures and/or to provide credible
representation(s) of forward-
looking operational risk exposure. In some examples, inputs can be received to
adjust
parameters to reflect business and/or regulatory conditions.
[0045] In some embodiments, one or more of the aspects shown in FIG. 2, or
other
aspects can include selecting a severity distribution. The distribution
ideally will best reflect
the behavior of losses in the upper tail. The processor(s) can be configured
to select a
severity distribution by applying a goodness-of-fit (GOF) test.
[0046] In some examples, the processors can apply a GOF test belonging to the
family
of weighted Cramer-van Mises tests. With such a test, the processor(s) can
evaluate the
- 8 -
CA 02910754 2015-10-29
GOF of a cumulative distribution function F(e,a) compared to a given empirical
distribution
function Fn(*).
[0047] For simplicity, it may be assumed that 3 is a parametric family,
i.e., 3 =
f.F(0,8).6 E el, where 0 is an open set in Rd and F(., e)is the cumulative
distributive
function ("cdr). This can also be formulated as He :F(x) E (with a suitable
alternative).
For a random sample of size n, the test statistic is then
S)):
[0048] (g5 = 71 f_+: " = dF(x, e), (1)
-Tu4,o)i
[0049] Where n and F., (.1 are the sample size and empirical cdf,
respectively, and Q is
a test statistic for the GOF test.
[0050] In theoretical statistics, such tests have been known in generality
as Q.: for
0 g <2 based on a series of works of Deheuvels and Martynov. Chernobai
suggested
using Q,2i in the context of Operational Risk.
[0051] To test whether data arises from a distribution family, an estimate
of 6 can be
applied to F; however this may not detect significant departures from a null
hypothesis.
Accordingly, in some embodiments, processor(s) are configured to obtain an
estimate of 6
and determine a test statistic by applying it to F. The processor(s) can
obtain samples of F
(with the estimate of 0 applied), and applying a new estimate of 0 in a Monte
Carlo
approach. From this, the processor(s) can approximate p-values for the
asymptotic
distribution of equation (1). In some examples, this may be a computationally
heavy
process.
[0052] In some embodiments, the processor(s) can approximate a covariance
matrix of
the integral operator in (1) corresponding to the asymptotic distribution by
applying
jackknife estimations and influence functions. In some embodiments, the
processor(s) find
the matrix's eigenvalues and can approximate the p-value by applying
saddlepoint
approximation.
- 9 -
CA 02910754 2015-10-29
[0053] The embodiments of the devices, systems, media and methods described
herein
may be implemented in a combination of both hardware and software. These
embodiments may be implemented on programmable computers, each computer
including
at least one processor, a data storage system (including volatile memory or
non-volatile
memory or other data storage elements or a combination thereof), and at least
one
communication interface.
[0054] Program
code is applied to input data to perform the functions described herein
and to generate output information. The output information is applied to one
or more output
devices. In some embodiments, the communication interface may be a network
communication interface. In embodiments in which elements may be combined, the
communication interface may be a software communication interface, such as
those for
inter-process communication. In still other embodiments, there may be a
combination of
communication interfaces implemented as hardware, software, and combination
thereof.
[0055] Throughout the foregoing discussion, numerous references will be made
regarding servers, services, interfaces, portals, platforms, or other systems
formed from
computing devices. It should be appreciated that the use of such terms is
deemed to
represent one or more computing devices having at least one processor
configured to
execute software instructions stored on a computer readable tangible, non-
transitory
medium or media. For example, a server can include one or more computers
operating as
a web server, database server, or other type of computer server in a manner to
fulfill
described roles, responsibilities, or functions.
[0056] The technical solution of embodiments may be in the form of a software
product.
The software product may be stored in a non-volatile or non-transitory storage
medium,
which can be a compact disk read-only memory (CD-ROM), a USB flash disk, or a
removable hard disk. The software product includes a number of instructions
that enable a
computer device (personal computer, server, or network device) to execute the
methods
provided by the embodiments.
[0057] The embodiments described herein may be implemented by physical
computer
hardware, including computing devices, servers, receivers, transmitters,
processors,
memory, displays, and networks. Some embodiments described herein provide
useful
physical machines and particularly configured computer hardware arrangements.
Some
- 10-
CA 02910754 2015-10-29
embodiments described herein are directed to electronic machines and methods
implemented by electronic machines adapted for processing and transforming
electromagnetic signals which represent various types of information. Some
embodiments
described herein pervasively and integrally relate to machines, and their
uses; and the
embodiments described herein have no meaning or practical applicability
outside their use
with computer hardware, machines, and various hardware components.
Substituting the
physical hardware particularly configured to implement various acts for non-
physical
hardware, using mental steps for example, may substantially affect the way the
embodiments work. Such computer hardware limitations may be clearly essential
elements
of some embodiments described herein, and they cannot be omitted or
substituted for
mental means without having a material effect on the operation and structure
of the
embodiments described herein. In some embodiments, the computer hardware is
essential
to implement the various embodiments described herein and is not merely used
to perform
steps expeditiously and in an efficient manner. For example, it may be
impractical and
unreasonable to consider human implementation for computationally complex
encryption
and/or decryption, or maintaining and/or accessing vast databases of data with
myriad
interrelationships.
[0058] For simplicity only one computing device 500 is shown but system may
include
more computing devices 500 operable by users to access remote network
resources and
exchange data. The computing devices 500 may be the same or different types of
devices.
The computing device 500 at least one processor, a data storage device
(including volatile
memory or non-volatile memory or other data storage elements or a combination
thereof),
and at least one communication interface. The computing device components may
be
connected in various ways including directly coupled, indirectly coupled via a
network, and
distributed over a wide geographic area and connected via a network (which may
be
referred to as "cloud computing").
[0059] For
example, and without limitation, the computing device may be a server,
network appliance, set-top box, embedded device, computer expansion module,
personal
computer, laptop, personal data assistant, cellular telephone, smartphone
device, UMPC
tablets, video display terminal, gaming console, electronic reading device,
and wireless
hypermedia device or any other computing device capable of being configured to
carry out
the methods described herein.
-11-
[0060] FIG. 3 is a schematic diagram of computing device 500, exemplary of an
embodiment. As depicted, computing device 4 includes at least one processor
502,
memory 505, at least one I/O interface 506, and at least one network interface
508.
[0061] Each processor 502 may be, for example, any type of general-purpose
microprocessor or microcontroller, a digital signal processing (DSP)
processor, an
integrated circuit, a field programmable gate array (FPGA), a reconfigurable
processor, a
programmable read-only memory (PROM), or any combination thereof.
[0062] Memory 505 may include a suitable combination of any type of computer
memory that is located either internally or externally such as, for example,
random-access
memory (RAM), read-only memory (ROM), compact disc read-only memory (CDROM),
electro-optical memory, magneto-optical memory, erasable programmable read-
only
memory (EPROM), and electrically-erasable programmable read-only memory
(EEPROM),
Ferroelectric RAM (FRAM) or the like.
[0063] Each I/O interface 506 enables computing device 500 to interconnect
with one or
more input devices, such as a keyboard, mouse, camera, touch screen and a
microphone,
or with one or more output devices such as a display screen and a speaker.
[0064] Each network interface 508 enables computing device 500 to communicate
with
other components, to exchange data with other components, to access and
connect to
network resources, to serve applications, and perform other computing
applications by
connecting to a network (or multiple networks) capable of carrying data
including the
Internet, Ethernet, plain old telephone service (POTS) line, public switch
telephone
network (PSTN), integrated services digital network (ISDN), digital subscriber
line (DSL),
coaxial cable, fiber optics, satellite, mobile, wireless (e.g. Wi-Fi, WiMAX),
SS7 signaling
network, fixed line, local area network, wide area network, and others,
including any
combination of these.
[0065] Computing device 500 is operable to register and authenticate users
(using a
login, unique identifier, and password for example) prior to providing access
to
applications, a local network, network resources, other networks and network
security
devices. Computing devices 4 may serve one user or multiple users.
CAN_DMS: \144373489\1 - 12 -
Date Recue/Date Received 2022-03-01
CA 02910754 2015-10-29
[0066] Although the embodiments have been described in detail, it should be
understood that various changes, substitutions and alterations can be made
herein without
departing from the scope as defined by the appended claims.
[0067] Moreover,
the scope of the present application is not intended to be limited to the
particular embodiments of the process, machine, manufacture, composition of
matter,
means, methods and steps described in the specification. As one of ordinary
skill in the art
will readily appreciate from the disclosure, processes, machines, manufacture,
compositions of matter, means, methods, or steps, presently existing or later
to be
developed, that perform substantially the same function or achieve
substantially the same
result as the corresponding embodiments described herein may be utilized.
Accordingly,
the appended claims are intended to include within their scope such processes,
machines,
manufacture, compositions of matter, means, methods, or steps.
[0068] As can be understood, the examples described above and illustrated are
intended to be exemplary only.
- 13-
