From Risk Assessment to Ongoing Monitoring
A risk assessment is a view of exposure at a point in time. It helps an organization identify and prioritize risks, but it cannot show on its own whether conditions are changing or whether responses are working. Risk monitoring supplies that continuing feedback. Its purpose is not to produce more numbers; it is to notice decision-relevant changes early enough for someone with authority to assess and respond to them.
A monitoring cycle starts with a defined exposure and a decision the organization may need to make. It then observes a measure, compares that observation with an agreed reference point or threshold, investigates material signals, and routes concerns to the people responsible for decisions. After action, the organization checks whether exposure changed and whether the monitoring arrangement is still useful. The cycle repeats rather than ending when a report is issued.
The monitored exposure should connect to an organizational objective, process, or risk entry. For example, an insurer concerned about delayed claims handling needs to specify which claims process is in scope and why delay matters. Monitoring a measure outside that boundary, or reporting a measure without a decision it could inform, may create activity without useful risk oversight.
The Main Stages of the Cycle
The stages are related but not interchangeable. A threshold crossing is a signal to assess, not a complete diagnosis. Likewise, a discussion of possible causes is not a substitute for recording whether the agreed threshold was crossed.
Identify the exposure, the organizational boundary, and the decision the monitoring is intended to support. Keep the scope stable enough that observations can be compared over time.
Obtain the current observation for the agreed period. Check that it refers to the intended population and period before treating it as evidence of a change.
Compare the observation with the applicable threshold, prior observations, or other agreed expectations. Consider direction, size, persistence, and context; do not infer a cause from the number alone.
Route a material signal to the responsible person or forum. The recipient assesses significance, decides whether further investigation or action is warranted, and records the decision.
Assign actions and follow-up. Later observations help assess whether exposure changed and whether the measure, threshold, reporting route, or review frequency still serves its purpose.
A useful monitoring record preserves the link between observation and decision: what was observed, for which scope and period, how it compared with expectations, what explanation was considered, who was informed, and what happened next. The exact operational definition of a measure needs to be unambiguous, but the central point here is that the cycle must remain interpretable and actionable.
Comparing Changes Without Overstating Them
A simple comparison with a prior observation can help describe movement. If \(M_0\) is a prior value and \(M_1\) is the current value, the relative change is:
This formula expresses the difference as a share of the starting value. The numerator \(M_1-M_0\) is the absolute change; dividing by \(M_0\) scales that change to the prior level. For example, an increase of 10 from 20 is a 50% increase, whereas an increase of 10 from 100 is a 10% increase. The same absolute movement can therefore have different relative significance depending on the starting point.
Relative change is not a measure of severity by itself. It is undefined when \(M_0=0\), and it can look very large when the prior value is small. It should be considered alongside the actual values, the threshold, the exposure represented, and the context. A percentage also does not explain why a change occurred. Monitoring supports investigation; it does not replace it.
Worked Example: A Claims Backlog Signal
A hypothetical insurer monitors the number of open claims that have been awaiting a required next action for more than 14 days. The approved escalation threshold is 80 claims. The prior monthly observation was 72, and the current observation is 96. Determine the change, compare it with the threshold, and state the appropriate monitoring conclusion.
Method. First calculate the absolute and relative changes, then make the threshold comparison. The measure concerns the same defined claim population and monthly reporting point in both periods.
Absolute change. \(96-72=24\) claims. The backlog increased by 24 claims.
Relative change. \((96-72)/72=24/72=0.3333\ldots\), or approximately \(33.3\%\). The relative change is rounded to one decimal place.
Threshold comparison. The current value, 96, exceeds the threshold of 80 by \(96-80=16\) claims. The prior value, 72, was below the threshold.
Conclusion and next step. This is both an upward movement and a threshold breach. The responsible claims manager should receive the signal under the agreed escalation route and assess its operational significance. The comparison alone does not establish whether the increase reflects higher claim volume, slower processing, or another cause.
Check. The current value is 96, which is greater than 80, and the calculated increase of 24 is exactly \(33.3\%\) of the prior value 72 to one decimal place. A common mistake is to calculate \(24/96=25.0\%\) and call that the change. That divides by the current value; relative change uses the prior value as its base.
Interpret the Signal, Then Escalate
Interpretation asks what the observation means for the exposure and what decision may be required. A disciplined review separates at least three questions:
- Is the observation comparable? Confirm that the scope, period, and population match the intended comparison. An apparent change may result from a changed boundary or incomplete information.
- What does the comparison show? State whether the measure increased or decreased, by how much, and whether it crossed an approved threshold. Note relevant prior movement without assuming that one period establishes a trend.
- What decision follows? Determine who needs to know, whether further investigation is required, and who is accountable for deciding or completing any response.
Escalation should be proportionate and timely. A threshold is useful only if the organization has agreed what happens when it is approached or crossed, who receives the information, and how the concern is considered. A concern that is not yet above a threshold may still deserve attention if the movement is unusually rapid or the potential consequence is serious. Conversely, a breach should not be dismissed merely because its cause is not yet known.
The monitoring record should distinguish facts from hypotheses. “The count increased from 72 to 96 and exceeded the threshold of 80” is an observation. “A staffing shortage caused the increase” is a possible explanation that requires support. Keeping the distinction clear helps decision-makers avoid treating an untested explanation as a finding.
Worked Example: Comparing Rates When the Population Changes
A hypothetical insurer reviews overdue access recertifications for a claims system. In the previous review, 6 of 100 required recertifications were overdue. In the current review, 18 of 120 were overdue. The organization has set a review threshold of 10% overdue. Determine whether the rate crossed the threshold and explain why comparing counts alone is insufficient.
Method. Calculate the overdue rate separately for each review, then compare the current rate with the 10% threshold. Do not compare the counts as if the populations were the same size.
Previous rate. \(6/100=0.06=6\%\).
Current rate. \(18/120=0.15=15\%\).
Threshold comparison. The current rate exceeds the 10% threshold by \(15\%-10\%=5\) percentage points. Relative to the threshold, the excess is \(5\%/10\%=50\%\), but the direct threshold statement is that the rate is five percentage points above it.
Change over time. The rate rose from 6% to 15%, a difference of 9 percentage points. The overdue count tripled, from 6 to 18, but the number of required recertifications also increased from 100 to 120. The rate accounts for that denominator change.
Conclusion and next step. The current observation breaches the stated rate threshold and is materially higher than the previous rate. It should be escalated as agreed. The responsible owner should check whether the population and reporting scope are comparable and investigate the reasons for overdue reviews before proposing a response.
Check. \(18/120=0.15\), so the current rate is 15%, not 18%. A common mistake is to compare the count of 18 directly with a threshold expressed as a percentage; the units do not match.
Close the Loop Through Action and Review
Monitoring is incomplete if escalation produces no decision or if action is assigned but never revisited. A decision may be to investigate, change a control, allocate resources, accept the exposure within authority, or take no immediate action with a stated reason. The cycle should make that choice visible and provide a way to check its consequences.
Review has two purposes. First, it asks whether the exposure changed after the response. Second, it asks whether the monitoring arrangement remains fit for purpose. A measure may cease to be informative if the process changes, the underlying exposure moves elsewhere, or decision-makers no longer use the information. A threshold or reporting route may also need reconsideration under appropriate governance. Such review does not mean changing a threshold simply to make an unfavorable observation disappear.
Worked Example: Following a Catastrophe-Claims Signal Through the Cycle
A hypothetical property insurer monitors catastrophe claims awaiting first customer contact more than two business days after notice. Its approved escalation threshold is 30 claims. At the prior review, 18 claims were overdue. During the current review, 22 new claims entered this overdue category and 6 claims left it after contact was completed. Determine the current count, the threshold status, and a suitable follow-up cycle.
Method. Reconcile the movement in the count, compare the resulting observation with the threshold, then specify escalation, action, and a later review.
Reconcile the count. Start with 18 overdue claims, add 22 that entered the category, and subtract 6 that left it: \(18+22-6=34\) claims.
Compare with the threshold. The current count is 34, which exceeds the threshold of 30 by \(34-30=4\) claims. The prior count of 18 was below the threshold.
Escalate and decide. The breach should be sent to the designated claims decision-maker. That person should confirm that the count covers the intended claims and period, assess whether the overdue cases have a common operational cause, and decide whether temporary capacity or another response is warranted.
Act and follow up. Suppose the decision-maker assigns additional contact staff and requests daily counts for the next two weeks. The later counts are 29, then 21. These observations are below the threshold, and the sequence is consistent with improvement, but it does not by itself prove that the staffing change caused the decline. The owner should confirm that the same scope and counting rule were used and assess whether the response should continue.
Review the monitoring. At the end of the two weeks, the decision-maker should review whether the measure identified a useful service risk, whether the escalation was timely, and whether daily reporting remains necessary or can return to the normal schedule. The review should retain the reason for its decision.
Check. The net change is \(22-6=16\); adding that to 18 gives 34. A common mistake is to add the 22 new cases but forget to subtract the 6 cases that exited the overdue category, producing an incorrect count of 40.
Common Monitoring Traps
Several errors can weaken an otherwise well-designed cycle:
- Treating a number as an explanation. A change identifies a signal; it does not prove its cause. Validate and investigate before presenting a hypothesis as fact.
- Ignoring the denominator or scope. Counts may rise because the population grew. Rates can also mislead if the numerator and denominator refer to different populations or periods.
- Using the threshold as the whole analysis. A below-threshold observation can still warrant attention, while a breach requires a governed response rather than an automatic conclusion about cause or severity.
- Escalating without ownership. A report that has no named recipient, decision responsibility, or follow-up can become a record of concern without a management response.
- Failing to review after action. Without follow-up, the organization cannot determine whether exposure changed or whether the monitoring process needs adjustment.
Check Your Understanding
For each situation, identify the monitoring conclusion or calculation requested and consider what follow-up the signal supports.
- A hypothetical insurer recorded 45 late premium reconciliations last month and 60 this month. The threshold is 55. Calculate the absolute and relative change using last month as the base, then state whether the current observation breaches the threshold. (Answer: Increase of 15, relative increase of 33.3%, and a breach by 5 reconciliations.)
- A service unit reports 9 unresolved complaints from 150 cases; its threshold is 5%. Calculate the rate and state the threshold status. (Answer: \(9/150=6\%\), which is 1 percentage point above the threshold.)
- A hypothetical insurer has 12 claims awaiting review at the start of a week. During the week, 17 enter that category and 8 leave it. The threshold is 20. Calculate the ending count and describe its status. (Answer: \(12+17-8=21\), a breach by 1 claim.)
- A monitoring count falls from 8 to 4, while the threshold is 10. A manager says the risk has been eliminated. Give one reason that conclusion is not established by these observations. (Answer: The count is lower and below threshold, but the observations alone do not establish that exposure is eliminated; scope, persistence, and context still need review.)
- A hypothetical insurer observes 24 overdue vendor reviews among 160 reviews, compared with 15 overdue among 100 previously. The threshold is 12%. Calculate both rates and determine whether the current rate breaches the threshold. (Answer: Previous rate \(15/100=15\%\); current rate \(24/160=15\%\). The current rate is unchanged and exceeds the threshold by 3 percentage points.)