The Derivatives Business: Focus on Financial Risk and Control Systems

1
The Derivatives BusinessFocus on Financial Risk
and Control Systems
?
MILAN, 30 JUNE 2004
2
Key Facts

• COMPANY PROFILE (December 2003)
• Conceived in 1998, formally spun off in 2000
• 569 employees, average age 35
• SP Rating AA-
• Total Assets EUR 47b
• Gross Revenues EUR 770m
• Net income EUR 340m
• Average daily VaR EUR 4.4m
• Cost/Income 24
• ROE 61
• Net revenues/employee EUR 598k
• SOME TRADING BOOK STATISTICS

3
Executive Summary

• Over 80 of UBMs gross revenues are generated
  from Financial Products sales trading (FP)
• FP includes institutional and corporate
  derivatives as well as fixed income and equity
  trading
• Over 85 of FPs gross revenues are generated
  from derivatives
• In total approximately 75 of UBMs gross
  revenues are generated from derivatives
• The purpose of this presentation is to illustrate
  the risk management framework behind this
  activity
• Four main risk classes are analysed
• Market risk
• Model risk
• Counterparty risk
• Operational risk

4
Financial Products - sales and trading

• From the beginning of operations, market risk
  management has been a distinctive element in UBM
• Derivative products are traded on all asset
  classes
• Ability to address a broad segment of corporate
  and institutional customers with innovative
  financial products
• Core skills in pricing, hedging and trading
• Products are unbundled into elementary risk
  components by proprietary pricing models and
  hedged through wholesale markets
• Limited back-to-back trading
• Industrial approach large volumes, high
  throughput, efficient time-to-market
• UBM transforms derivatives risks into market,
  model, counterparty and operational risks

5
The Risk Management Stages
DERIVATIVE PRODUCT
OPERATIONAL RISK
MARKET RISK
MODEL RISK
COUNTERPARTY RISK
CTPY1
UBM
CLIENT1
CTPY2
CLIENT2
CTPY3
CLIENTn
CTPYm
SALES
TRADING
RISK MANAGEMENT
6
Market Risk daily Value-at-Risk

• Value-at-Risk (VaR)
• UBMs portfolio affected by more than 40,000 risk
  factors
• VaR calculated daily through a proprietary VaR
  engine via historical simulation
• VaR parameters 99 double-tail confidence level,
  1-day holding period, daily update of time series
• VaR model has been validated by Italian
  regulators for capital requirement purposes
• UBM daily VaR limit is 7m, against a one-year
  average of 4.4m (max 6.6m)
• Daily back-testing against clean PL series

7
and daily Stress and Crash Tests

• Stress Tests
• A total of 8 stress scenarios are evaluated and
  the worst outcome is selected
• Stress test (as of 14 May 2004) 14.8m
• Crash Tests
• A total of 8 crash scenarios are evaluated and
  the worst outcome is selected
• The last scenario is calculated as a combination
  of the the worst event for each asset class
• Credit spread scenarios have been obtain using
  telecom industry and South America as benchmarks
• Crash test (as of 14 may 2004) 13.7m

STRESS
CRASH
8
Model Risk

• Team of dedicated professionals developing and
  implementing pricing models for exotic
  derivatives
• Implementation and maintenance of a large,
  high-throughput risk management system
• Over 200 proprietary pricing models deployed
• Dedicated Model Testing team based in London
• Front-office booking system with over 200,000
  trades
• Proprietary technology for capital guaranteed
  products with more than 500,000 web-based
  transactions executed, with real time tracking
  and stress tests
• Ubiquitous computing Enterprise-wide Parallel
  Processing (over 500 CPUs)
• Skew/smile proprietary pricing models based on
  stochastic volatility

The structure
Some examples
Market Surface
TCP
TCP, FTP
9
Counterparty Risk

• Measuring Counterparty Risk
• Derivative portfolio size gt 1,000b
• Market risks are hedged through OTC trades with
  other market counterparties
• Traditionally counterparty risk has been
  monitored using fixed coefficients
• Need to quantify the effective cost of
  substitution
• The cost of substitution is the value of each
  trade increased by the potential variation that
  may occur
• Estimate the cost of substitution as the
  marked-to-market valuation plus a simple add-on
• Methods advantages
• Quantifies more accurately the effective
  counterparty risk
• Uses standard product control techniques
• Can be used with risk mitigation tools such as
  close-out netting and collateral agreements
• When using netting, portfolio diversification is
  captured
• Dynamic methodology in line with market
  variations
• More in line with market practice and central
  bank models
• Can bring in market risk techniques (substitution
  of the simple add-on with VaR)

10
Counterparty risk a practical example

• Monitoring and analysis of historical data
• All counterparty information is stored in a
  database
• Marked-to-market valuations are available at
  individual deal level
• Simple add-on is compared with realized 10-day
  marked-to-market valuation moves
• There is evidence that current method
  overestimates effective exposure
• Currently implementing 10-day VaR in place of
  simple additive add-on
• This will factor in portfolio diversification and
  will bring a uniform methodology to exposure
  valuation

Simple Add-on vs 10-day delta MTM
1,400
1,200
1,000
800
DELTA MTM
LIMIT
ADD_ON
EXPOSURE
Value (/m)
MTM
600
COLLATERAL
400
200
-
-200
13-Feb-03
06-Mar-03
27-Mar-03
17-Apr-03
09-May-03
30-May-03
Date
11
Operational Risk

• Assessing and controlling operational risk
• Operational risk is defined as the risk of losses
  resulting from inadequate or failed internal
  processes, people and systems or from external
  events. This definition includes legal risk, but
  excludes strategic and reputational risk (Basel,
  September 2001)
• Risk control is achieved through process mapping,
  risk assessment and business process
  reengineering
• The Operational risk control project is an
  on-going activity that advances by gradual
  improvements
• Main steps
• Definition of roles and responsibility
• Implementation of techniques of process analysis
  and representation
• Application development to support historical
  loss data collection
• Study of risk assessment methodology to influence
  business process and control reengineering
• Banking activity is so heterogeneous and
  specialised that only process owners are able to
  effectively control, process and assess risk
• Internal workgroups have been constituted with
  line managers and Internal Audit
• These workgroups have permanent responsibility on
  process design, operational risk assessment and
  reengineering solution identification
• Processes are targeted and mapped out to workflow
  charts
• Workflow charts illustrate the role of every
  actor in the process both internal and external,
  as well as systems contribution
• Identification of critical process activities and
  improvements (e.g. by introducing new control
  points)
• Development of in-house database to support a
  structured collection of historical data
  describing operational accidents, related losses
  and recovery
• Implementation of reporting tools both internal
  and regulatory