A critical but often overlooked aspect of a competent derivatives trading
operation are the computer systems used to manage the risk, account for the
positions on a mark-to-market basis, track derivatives-related events (such as
expiries and rollovers) and measure Value-at-Risk. Once you have read some of
the articles in the Derivatives section of the Financial Pipeline, you will
realize how quickly a portfolio of transactions can become complicated. In the
section on hedging swaps, we discussed some of these complications including
the problems associated with mismatched short term cash flows and maturity
bucket grouping. Options produce their own problems because of the convexity of
these products. Taking snapshots of delta, gamma and vega at an instantaneous
specification of prices is insufficient (although necessary) for competent
financial risk management.
One must also have an appreciation of how these risks change with the
progress of time and the evolution of prices. In the first edition of Risk
Professional magazine published by the Global Association of Risk Professionals
(see http://www.garp.com), Geoff Kates
establishes a framework for evaluating a financial risk management system and
he uses this framework to assess some of the more common off-the-shelf products
on the market. This article will discuss those criteria and it will explain the
importance.
Integration
At the beginning of the global derivatives market's development, almost
every bank pursued derivatives in a stand-alone asset-class-by-asset-class
fashion. That is to say, one group managed interest rate derivatives, another
group managed equity derivatives and a third group managed foreign exchange
derivatives. For many banks, this is still the case.
However, an increasing number of financial institutions are turning to a
more integrated approach, stripping the derivatives desks from each asset
class' cash group and combining them into a more efficient cross-marketing
machine. Once you understand interest rate derivatives, it is straightforward
to understand equity derivatives or foreign exchange derivatives. Conversely,
it is not necessarily the case that a manager who has spent his entire career
overseeing spot foreign exchange salespeople will be able to understand the way
in which a derivatives book works. It is not something you're likely to learn
from a book or a classroom. You have to have experience.
To buy or to build
The next question the bank's senior management must
ask itself is whether or not the bank should buy an off-the-shelf system or
build one using its own internal IT resources.
Buying a system is convenient, particularly if it is one that is in
widespread use. Popular systems have been tested and have had all of the kinks
worked out. The more popular the system, the less likely that it is vulnerable
to internal control irregularities. That is, the more popular the system, the
less likely it is possible for individuals to manipulate the bank's official
records for fraudulent purposes. Systems are typically very expensive, with
charges for both a site license and individual annual user permits. Many of the
companies that sell these systems make it easy for the user to customize
reports, batch files, pricing modules, etc.
However, many financial institutions are reticent to relinquish the
responsibility for risk management computer systems to a third party. The
managers of these institutions would prefer to have their own internal risk
management personnel design the system that is then implemented by the bank's
IT staff. Not only is this more expensive than buying an off-the-shelf system
in terms of up-front dollar cost and delays in implementation but the system is
vulnerable to the expertise of a handful of individuals. Let's say you are the
head of trading at ABC Bank and you commission your risk management department,
all of three people (Larry, Curly and Moe) to design and implement your
interest rate risk management system. If Larry, Curly and Moe leave to go as a
team to DEF Bank, you will have lost all of your core knowledge base and you
will have to start from scratch. There is also the possibility that Larry,
Curly or Moe designed secret entrances into the system for themselves so that
they could manipulate tickets and positions and profit and loss statements.
Speed
In order to be effective, risk management information must be at least as
fast as the markets to which it refers. On the face of it, for most people
using applications designed for home use, this is not problematic. However, for
financial institutions with portfolios consisting of thousands of different
instruments, some of which use very complicated formulae, and arrays of
parameters to revalue, this is a serious database design problem.
Interface
One of the key aspects of a well-designed system is its flexibility. A good
risk management system will have a user interface that is customizable. Many of
them are beginning to use the Internet as their interface platform. The
interface is also the mechanism in which reports are designed. For an example
of the kinds of reports dealers and risk managers require, see our earlier
article entitled "How Do Options
Traders Look At Their Portfolios".
Asset Class Coverage
Further to our discussion of the integration of asset classes in the
management of derivatives sales and trading operations at leading financial
institutions, a good risk management system will provide the senior management
with the ability to immediately access information on all of the derivatives
activities in which the financial institution is engaged, across all asset
classes.
It is not uncommon for banks to have systems in place that enable their
management to take a snapshot of the firm's financial price risk with the
simple click of a button at any point during the trading day, in real-time.
Covering all of the asset classes also allows for greater overall
risk-taking because it allows for the portfolio effects of diversification of
risk across the different asset classes.
Pricing Model Flexibility
Model risk refers to the problems
associated with discrepancies between the theoretical pricing of a financial
instrument and the way in which it actually trades in the market. The
difference in price, for a given set of input parameters, is a result of the
assumptions that are necessary for solution of the mathematical model of the
price of the financial product in question.
For some financial products, particularly the more exotic or novel ones, the
choice of pricing model is a controversial one. A good risk management system
will allow management to pick and choose the pricing model it prefers for a
particular instrument and it will also allow management to compare the model
risk in different market environments associated with individual pricing
models.
Ability to Link to Other Systems
The derivatives risk management system is only one of a handful of systems
with which the dealer at a financial institution must be familiar. Other
systems include ticketing systems for cash instruments, accounting systems,
credit risk management systems and, possibly, spreadsheets tracking customer
portfolios.
A dealer's life is made much easier when the primary system he uses on a
daily basis, the risk management system, can communicate its information to the
other relevant systems automatically. Otherwise, the dealer (or more likely his
assistant) will have to input multiple tickets for a single transaction. This
is not just a question of personal effort. It is an operational risk issue, as
well. Every time the dealer inputs a ticket, there is room for an error. Too
many errors and the bank begins to lose customers as well as money.
The key point here is that technological sophistication leads to better
management.
These are just some of the criteria that a financial institution risk
manager may choose to apply to the selection of a financial risk management
computer system. In the next wave of development, risk management platforms
will be entirely web-based. Already, Goldman Sachs and other leading American
investment banks are offering web-based risk management systems, including
pricing models, to their clients. Helping their clients understand the
financial price risk they face makes it easier for the client to understand the
efficacy of financial products (products which they hopefully transact with
Goldman Sachs).
Article by Chand Sooran, Principal Victory
Risk Management Consulting, Inc.
|