OPERATIONAL RISK
RISK MANAGEMENT
Presented by:
GROUP 5
GROUP MEMBER
G.A.A AMANDA ATHALIA W NI KADEK ANISSA PRADNYADEWI
2208521068/07 2207521144/13
Definition of operational risk
Measurement of operational risk
TOPICS TO BE Calculating expected losses
Changes in operational
COVERED characteristics
Self-assessment to measure
operational risk
WHAT IS
OPERATIONAL RISK?
Operational risk refers to the potential loss that arises from
inadequate or failed internal processes, human factors, systems,
or external events.
For example: A small business faces a cash flow crisis
because of delayed payments from key customers, leading
to difficulties in meeting payroll and operational expenses.
TYPES OF OPERATIONAL RISK
Process Risk: These risks are related to the efficiency and effectiveness of
internal processes.
People Risk: This risks typically result from staff constraints,
incompetence, dishonesty, or a corporate culture that does not cultivate
risk awareness.
System Risk: This refers to risks stemming from the use of technology and
systems within an organization. Risk events might include bugs, system
failures.
External event Risk: These are risks arising from external factors beyond
the control of the organization.
MEASURING OPERATIONAL RISK
One technique for measuring operational risk is by using
the following two classifications:
1. Frequency or probability of risk occurrence
2. Severity of loss or impact of the risk
FIGURE 11.1
Page 197
FIGURE 11.2
Page 198
FIGURE 11.3
Page 200
CALCULATING
EXPECTED LOSSES
Suppose we want to calculate the expected loss if a certain
risk occurs. Using the framework of probability (frequency)
and severity, the expected loss is calculated as follows:
Expected Loss = Frequency (probability) × Severity
(magnitude of loss)
FOR EXAMPLE
What is the expected loss from workplace
accidents in the upcoming month? Using
the average values for frequency and loss,
the expected loss for the next month is:
Expected Loss = (Frequency) x (Severity)
= 5.25 x Rp2.4 million = Rp12.6 million
Page 201
ANALYTICAL APPROACH TO
CALCULATING EXPECTED LOSS
Suppose we know that the expected loss
level (average) is Rp10 million with a
standard deviation of Rp15 million. What
is the loss for a 95% confidence interval?
The loss value at the 5% limit can be
calculated as follows:
Loss Value = 10 million - 1.65 (10 million) =
-Rp6.5 million.
Here, 1.65 is the z-value corresponding to
a probability area of 5%. Thus, the
expected loss is Rp6.5 million.
SIMULATED APPROACH
Frequency X Severity
Expected Loss =
(how many losses occur to the company) X (how much loss occurs)
SIMULATED APPROACH
We can assume that after we did the evaluation there is a loss happening. In this case,
the frequency of losses can be explained by the Poisson Distribution method. The
average frequency of losses is 5 losses per month. The period evaluated is monthly (so
there are 5 losses every month on average).
In this method we also can evaluate the severity of losses, and conclude that the normal
distribution can explain the severity of past losses.
Suppose the average loss per loss event is Rp. 15,000,000 with
a standard deviation of Rp. 2,000,000. with 5 as the mean
In table 11.2 (page 203) column 2 represents
the Poisson probability distribution with an
expected value of 5. Column 3 presents the
cumulative probability. Column 4 presents
the numbers 0-99 to represent the
simulated numbers.
Here is an example of a one-sided
cumulative normal distribution, starting
from 0.5000 to 0.9990. For the other side,
we can start from 0.000 to 0.4999. The
total sequence of numbers will appear as
0.000 to 0.9990.
Table 11.3, Page 204
Based on this data, a simulation will be carried out. The steps are as follows:
01 02 03 04
Generate a Generate Multiply the Repeating steps
Random number random numbers frequency by the 1 to 3 several
for the for loss severity severity to times (e.g. 100
frequency of using the normal generate the times or 1,000
losses using the distribution total expected times)
Poisson loss in a given
distribution with period 3
an expected
value of 5
Suppose we generate 10 random numbers for steps 1 and 2
(simulation with 10 runs).
COLUMN 5 PRESENTS THE LOSS VALUE (SEVERITY) WHICH IS
CALCULATED AS FOLLOWS:
Z = (X - μ)/ 𝛔
If u = 15 million, standard deviation = 2 Million, then for z = 1.12 X is :
X = (1,12) x (2 million) + 15 Million = 17,24 Million
So that means that, the severity for that row is a loss of Rp. 17.24 million.
IF THE COLUMN 3 NUMBER IS BELOW 5000, THEN THE Z VALUE
IS CALCULATED AS
0.9990 - (random number/10000)
For example : for a number of 305,
z is (0.9990 - (305/10000)) = -1,86.
THE FOLLOWING TABLE PRESENTS THE CONTINUED DISTRIBUTION
OF LOSSES USING LOSS INTERVALS OF EVERY 10 MILLION.
THE FOLLOWING CHART SUMMARIZES THE RESULTS FROM THE
TABLE ABOVE IN GRAPHICAL FORM.
AT A GLANCE, IT APPEARS THAT THE DISTRIBUTION SHOWS A POSITIVE SKEWNESS. WE
CAN RUN SIMULATIONS UP TO 500 RUNS TO MAKE THE SIMULATION RESULTS SMOOTHER.
THE FOLLOWING CHART SHOWS THE TYPICAL RESULTS (LIKELY TO BE OBTAINED) IF WE
PERFORM A SIMULATION WITH 1,000 RUNS.
CHANGES IN OPERATIONAL RISK CHARACTERISTICS
Operational and other risks may change their characteristics over time. There are
several factors that affect the changing characteristics of a risk, including:
01 02 03 04 05
Globalization Automation Over-reliance on Outsourcing Cultural Change
Technology
ANY QUESTION???
THANK YOU
FOR LISTENING!
group 5