C06 OPERATIONAL RISK

DEFINING OPERATIONAL RISK #

The Basel definition of operational risk is “the risk of direct and indirect loss resulting from inadequate or failed internal processes, people, and systems or from external events.” 

METHODS FOR CALCULATING OPERATIONAL RISK REQUIREMENTS:

The three methods for calculating operational risk capital requirements are:

(1) the basic indicator approach,

(2) the standardized approach, and

(3) the advanced measurement approach (AMA).

The basic indicator approach and the standardized approach determine capital requirements as a multiple of gross income at either the business line or institution level. The advanced measurement approach (AMA) offers institutions the possibility to lower capital requirements in exchange for investing in risk assessment and management technologies.

OPERATIONAL RISK CATEGORIES #

The Basel Committee on Banking Supervision disaggregates operational risk into seven types. A majority of the operational risk losses result from clients, products, and business practices.

1. Clients, products, and business practices. Failure (either intentional or unintentional) to perform obligations for clients.
2. Internal fraud. Disobeying the law, regulations, and/or company policy, or misuse of company property.
3. External fraud. Actions by a third party that disobey the law or misuse property.
4. Damage to physical assets. Damage occurring from events, such as natural disasters.
5. Execution, delivery, and process management. Failure to correctly process transactions and the inability to uphold relations with counterparties.
6. Business disruption and system failures. Examples include computer failures, both hardware- and software-related, or utility outages.
7. Employment practices and workplace safety. Actions that do not follow laws related to employment or health and safety.

LOSS FREQUENCY & LOSS SEVERITY #

• Loss frequency is defined as the number of losses over a specific time period (typically one year), and loss severity is defined as the value of financial loss suffered (i.e., the size of the loss).
• Loss frequency is most often modeled with a Poisson distribution. Loss severity is often modeled with a lognormal distribution.
• Loss frequency and loss severity are combined in an effort to simulate an expected loss distribution (known as convolution). The best technique to accomplish this simulation is to use a Monte Carlo simulation process.

DATA LIMITATIONS #

Banks should use internal data when estimating the frequency of losses and utilize both internal and external data when estimating the severity of losses. Regarding external data, banks can use sharing agreements with other banks (which includes scale-adjusted data) and public data.

FORWARD LOOKING APPROACH #

1. .Causal relationships are a convenient method of identifying potential operational risks. Relationships are analysed to check for a correlation between firm actions & operational risk losses.
2. A frequently used tools in operational risk identification and measurement is the risk and control self assessment (RCSA) program. The basic approach of an RCSA is to survey those managers directly responsible for the operations of the various business lines.
3. The identification of appropriate key risk indicators (KRIs) may also be very helpful when attempting to identify operational risks. In order to be valuable as risk indicators, the factors must (1) have a predictive relationship to losses and (2) be accessible and measurable in a timely fashion. The idea of utilizing KRIs is to provide the firm with a system that warns of possible losses before they happen.

SCORE CARD DATA #

One method for allocating  risk capital to each business unit

is the scorecard approach. This approach involves surveying each manager regarding the key features of each type of risk. Questions are formulated, and answers are assigned scores in an effort to quantify responses. The total score for each business unit represents the total amount of risk. Scores are compared across business units and validated by comparison with historical losses.

THE POWER LAW #

The power law is useful in extreme value theory (EVT) when we evaluate the nature of the tails of a given distribution. The use of this law is appropriate since operational risk losses are likely to occur in the tails. The law states that for a range of variables:

P(V> X) = K x X^–α

where: V =loss variable, X =large value of V, K and a = constants

DATA LIMITATIONS #

Managers have the option to insure against the occurrence of operational risks. Two issues facing insurance companies and risk managers are moral hazard and adverse selection.

• A moral hazard occurs when an insurance policy causes an insured company to act differently with the presence of insurance protection.
• Adverse selection occurs when an insurance company cannot decipher between good and bad insurance risks. Since the insurance company offers the same polices to all firms, it will attract more bad risks since those firms with poor internal controls are more likely to desire insurance.