Operational Risk measurement

This is defined as “the risk of loss resulting from inadequate or failed internal processes,

people and systems or from external events. This includes legal risk, but excludes

strategic and reputation risk”.9 Such risks are likely to be significant in Islamic Banks

due to specific contractual features and the general legal environment. Specific aspects

that could raise operational risks in Islamic banks include the following:

(1) The cancellation risks in non binding murabahah and istisnah[’a contracts.

(2) Problems in internal control systems to detect and manage potential problems in

operational processes and back office functions.

(3) Technical risks of various sorts.

(4) The potential difficulties in enforcing Islamic Finance contracts in a broader legal

environment,

(5) The risk of non-compliance with Shariah requirements that may impact on

permissible income,

(6) The need to maintain and manage commodity inventories often in illiquid markets,

and

(7) The potential costs and risks in monitoring equity type contracts and the associated

Legal risks. In addition, increasing use structured finance transactions – specifically,

Securitization of loans originated by banks to manage risks on the asset side – could47

Expose banks to additional legal risks.

The three methods of measuring operational risks proposed in Basel II would need

considerable adaptations in Islamic Banks owing to the specificities noted earlier. The

use of gross income as the basic indicator for operational risk measurement could be

misleading in Islamic Banks, insofar as large volume of transactions in commodities, and

the use of structured finance raise operational exposures that will not be captured by

gross income. In contrast, the standardized Approach that allows for different business

lines would be better suited, but would still need adaptation to the needs of Islamic

Banks. In particular, agency services under mudarabah, the associated risks due to

potential misconduct and negligence, and operational risks in commodity inventory

management, all need to be explicitly considered for operational risk measurement.

3.3 Standardized models in measuring different types of risks :

3.3.1 GAP Analysis :

GAP analysis is an interest rate risk management tool based on the balance sheet. GAP

analysis focuses on the potential variability of net-interest income over specific time

intervals. In this method a maturity/re pricing schedule that distributes interest-sensitive

assets, liabilities, and off-balance sheet positions into time bands according to their

maturity (if fixed rate) or time remaining to their next re pricing (if floating rate) is

prepared. These schedules are then used to generate indicators of interest rate sensitivity

of both earnings and economic value to changing interest rates.

GAP models focus on managing net interest income over different time intervals. After

choosing the time intervals, assets and liabilities are grouped into these time buckets

according to maturity (for fixed rates) or first possible re pricing time (for flexible rates).

The assets and liabilities that can be re priced are called rate sensitive assets (RSAs) and

rate sensitive liabilities (RSLs) respectively, and GAP equals the difference between the

former and the latter.

Thus for a time interval, GAP is given by,

GAP = RSAs – RSLs

Note that GAP analysis is based on the assumption of re pricing of balance sheet items

calculated according to book value terms. The information on GAP gives the

management an idea about the effects on net-income due to changes in the interest rate.

For example, if the GAP is positive, then the rate sensitive assets exceed liabilities. The

implication is that an increase in future

interest rate would increase the net interest income as the change in interest income is

greater than the change in interest expenses. Similarly, a positive GAP and a decline in

the interest rate would reduce the net interest income. 48

3.3.2 Duration-GAP Analysis

Duration model is another measure of interest rate risk and managing net interest income

derived by taking into consideration all individual cash inflows and outflows. Duration is

value and time weighted measure of maturity of all cash flows and represents the average

time needed to recover the invested funds. The standard formula for calculation of

duration D is given by,

Where:

(CFt) is the value of cash flow at time t, which is the number of periods the cash flow

from the instrument is received, and

(I) is the instrument’s yield to maturity. The duration analysis compares the changes in

market value of the assets relative to its liabilities. Average duration gaps of assets and

liabilities are estimated by summing the duration of individual asset/liability multiplied

by its share in the total asset/liability.

A change in the interest rate affects the market value through the discounting factor

(1+i)-t.

Note that the discounted market value of an instrument with a longer duration will be

affected relatively more due to changes in the interest rate. Duration analysis, as such,

can be viewed as the elasticity of the market value of an instrument with respect to

interest rate.

Duration gap (DGAP) reflects the differences in the timing of asset and liability cash

flows and given by,

DGAP = DA - u DL

Where DA is the average duration of the assets, DL is the average duration of liabilities,

and u is the liabilities/assets ratio. Note that a relatively larger u implies higher leverage.

A positive DGAP implies the duration of assets is greater than that of liabilities. When

interest rate increases by comparable amounts, the market value of assets decrease more

than that of liabilities resulting in the decrease in the market value of equities and

expected net-interest income. Similarly, a d cline in the interest rate decreases the market

value of the equity with a positive DGAP. Banks can use DGAP analysis to immunize

portfolios against interest rate risk by keeping DGAP close to zero.49

3.3.3 Credit Value at Risk (VaR)

Used for market risk, Displaced Commercial risk.

Value at Risk (VaR) is one of the newer risk management tools. The VaR indicates how

much a firm can lose or make with a certain probability in a given time horizon. VaR

summarizes financial risk inherent in portfolios into a simple number. Though VaR is

used to measure market risk in general, it incorporates many other risks like foreign

currency, commodities, and equities. VaR has many variations and can be estimated in

different ways. We outline the underlying concept of VaR and the method of estimating it

below.

Assume that an amount A0 is invested at a rate of return of r, so that after a year the value

of portfolio is A= A0 (1+r). The expected rate of return from the portfolio is μ with

standard deviation σ. VAR answers the question of how much can the portfolio lose in a

certain time period t (e.g., month). To compute this, we construct the probability

distribution of the returns r. We then choose a confidence level c (say 95) percent. VaR

tells us what is the loss (A*) that will

not be exceeded c percent of the cases in the given period t. In other words, we want to

find the loss that has a probability of 1-c percent of occurrence in the time period t. Note

that there is a rate of return r* corresponding to A*. Depending on the basis of

comparison, VaR can be estimated in the absolute and relative sense. Absolute VaR is the

loss relative to zero and relative VaR is the loss compared to the mean μ.

A simpler parametric method can be used to estimate VaR by converting the general

distribution into a standard normal distribution. This method is not only easier to use but

also gives more accurate results in some cases. To use the parametric method to estimate

VaR, the general distribution of the rates of return are converted into a normal

distribution in the following way

-α= (-׀r*׀-μ)/σ

Note that α represents the standard normal distribution equivalent loss corresponding to

confidence level of 1-c of the general distribution (i.e., r*). Thus, in a normal distribution,

α would be 1.65 (or 2.33) for a confidence level c=95 (or c= 99 percent). Expressing time

period T in years (so that one month would be 1/12), the absolute and relative VaRs using

the parametric method are then given as

VaR zero = A0 (α σ(T)0.5 – μ T ) And VaR mean= A0 ασ(T)0.5 50

respectively. Say, for a monthly series the VaR (zero) is estimated to be ‘y’ at 95 percent

confidence level. This means that under normal market conditions, the most the portfolio

can lose over a month is an amount of y with a probability of 95 percent.

Figure 3 : basic concept of VAR

Example on VaR :

Assume an investment portfolio marked to the market is valued at SR 100 million has

expected rate of return of 5 percent and standard deviation of 12 percent. We are

interested to estimate VaR for holding period of one month at 99 percent confidence

interval. Using the symbols in the text, this information can be written as follows:

A0=100 million, μ = 5 percent, σ = 12 percent, c=99, α=2.33, and T=1/12.

Note that 99 percent confidence interval yields α=2.33 in a normal distribution. Given the

above we can estimate the two variants of VaR as:

VaR mean = A0 (ασ(T)0.5 − μT )

= 100 × 2.33 × 0.12 × (1/12)0.5 = 8.07

and,

VaR zero = A0 ασ(T)0.5

=100[2.33 × 0.12 × (1/12)0.5 – 0.05× (1/12)] = 8.07-0.42= 7.65

The result in the relative sense (i.e. relative to mean) implies that under normal conditions

there is a 99 percent chance that the loss of the portfolio will not exceed SR 8.07 million

over a month. In the absolute sense (i.e. relative to zero) this amount is SR 7.65 million.