Why the Emergence Pattern Matters
Expected claims methods need an assumption about how much of an ultimate loss amount has emerged by a given development age. That assumption is not the same for every loss measure. Payments may take time to be made, while case reserves can bring incurred amounts closer to ultimate sooner. Selecting an inappropriate pattern can therefore misstate the proportion treated as emerged and the share that remains unreported.
A development pattern summarizes that timing by development age. The actuary selects it from relevant historical experience, judgment, or both. The selection should match the measure being analyzed (paid or incurred), the type of claims, and the way the data are grouped and valued. It is a selection of assumptions about emergence, not a forecast of the dollar amount of ultimate losses.
For paid losses, \(C_k\) is cumulative payment through age \(k\). For incurred losses, it is paid loss plus case outstanding at that age. The ultimate amount and the age-specific amount must use a consistent loss basis, valuation, and coverage definition. Incurred amounts can change as case reserves are revised; the selected incurred pattern is an estimate of the relationship to ultimate, not a claim that incurred values always increase smoothly.
From Development Factors to Emergence
A common way to express a pattern is through selected age-to-age development factors. A factor from age \(j\) to the next age is the ratio of the cumulative amount at the next age to the amount at age \(j\). A factor above 1 indicates expected growth in the cumulative amount over that interval. A tail factor accounts for development beyond the last age shown in the selected sequence.
Here \(f_j\) is the selected factor from age \(j\) to the next age, \(m\) is the last age in the sequence before ultimate, and \(T\) is the selected factor from that age to ultimate. If no further development is assumed after the final observed age, \(T=1\). A tail assumption should be explicit: omitting a tail is a selection of no additional development, not a neutral calculation choice.
The resulting emergence proportions should ordinarily increase with age and remain at or below 1. This is a useful diagnostic, especially when factors have been selected separately or when incurred experience contains reserve reductions. An unusual pattern may be explainable, but it deserves investigation before use. The factor sequence, the cumulative factors, and the implied proportions should be reviewed together.
How to Select a Pattern
The data do not determine a selection mechanically. Historical factors are evidence, and the actuary must decide how much weight to place on each observation. Volume-weighted ratios give greater weight to larger cumulative amounts; a simple average gives equal weight to each observed ratio. Neither method is universally correct. A selected factor may also reflect changes in claim handling, coverage, inflation, or the composition of the claims being observed.
Use paid data to select a paid pattern and incurred data to select an incurred pattern. Check that the claims, coverage, valuation dates, and grouping are relevant to the subject being estimated.
Calculate historical age-to-age ratios from comparable triangle entries. Look for changes, outliers, sparse observations, and differences in volume or claim mix.
Choose factors using a defensible average or judgment informed by the data. State the tail assumption and why it is appropriate for the maturity and claim type.
Multiply the selected factors to obtain cumulative factors, take their reciprocals for emergence proportions, and check that the results are coherent across ages.
Worked Examples
Worked Example: Converting Selected Paid Factors into a Pattern
A hypothetical coverage has selected paid development factors of 1.50 from 12 to 24 months, 1.20 from 24 to 36 months, and 1.10 from 36 months to ultimate. Assume the same definition of paid loss applies at every age, and assume the selected factors are appropriate for the claims being analyzed. Find the cumulative emergence proportion and unreported share at 12, 24, and 36 months.
Method. Multiply the factors from each age through ultimate to obtain the cumulative factor. Its reciprocal is the emerged proportion; the complement is the unreported share.
At 12 months, the cumulative factor is \(1.50\times1.20\times1.10=1.98\). Thus emergence is \(1/1.98=0.5051\), and the unreported share is \(1-0.5051=0.4949\).
At 24 months, the cumulative factor is \(1.20\times1.10=1.32\). Emergence is \(1/1.32=0.7576\), and the unreported share is \(1-0.7576=0.2424\).
At 36 months, the cumulative factor is \(1.10\). Emergence is \(1/1.10=0.9091\), and the unreported share is \(1-0.9091=0.0909\).
Check. Emergence rises from 0.5051 to 0.7576 to 0.9091 as development age increases. At each age, emergence plus the unreported share equals 1. The common mistake avoided here is treating the cumulative factor itself as the percentage emerged; it is the reciprocal that gives that percentage.
Worked Example: Selecting Paid and Incurred Patterns Separately
For a hypothetical long-settling liability product, an actuary selects paid factors of 1.55 from 12 to 24 months, 1.25 from 24 to 36 months, and 1.08 from 36 months to ultimate. Separately, selected incurred factors are 1.30, 1.12, and 1.04 over the same intervals. Assume each sequence is appropriate for its own loss measure and the claim population is otherwise comparable. Find the 12- and 24-month emergence proportions for each measure.
Paid at 12 months. The cumulative factor is \(1.55\times1.25\times1.08=2.0925\). Emergence is \(1/2.0925=0.4779\); the unreported share is \(0.5221\).
Paid at 24 months. The cumulative factor is \(1.25\times1.08=1.35\). Emergence is \(1/1.35=0.7407\); the unreported share is \(0.2593\).
Incurred at 12 months. The cumulative factor is \(1.30\times1.12\times1.04=1.51424\). Emergence is \(1/1.51424=0.6604\); the unreported share is \(0.3396\).
Incurred at 24 months. The cumulative factor is \(1.12\times1.04=1.1648\). Emergence is \(1/1.1648=0.8585\); the unreported share is \(0.1415\).
Check. For each measure, the 24-month emergence proportion exceeds the 12-month proportion. Incurred emergence is higher than paid emergence at both ages under these selections, consistent with the assumed factors. That comparison is not a universal rule: it follows from these selected patterns. The mistake avoided is applying paid factors to incurred amounts merely because both are shown in the same triangle.
Worked Example: Volume-Weighted Selection and an Explicit Tail
A hypothetical claims triangle gives these paid cumulative amounts for observed 12-to-24-month development: 100 to 140, 300 to 480, and 200 to 300. For 24-to-36-month development, two observations are 140 to 168 and 480 to 576. The actuary selects a factor of 1.05 from 36 months to ultimate based on relevant tail analysis. Assume these observations are comparable and the volume-weighted ratio is selected for each observed interval. Calculate the selected pattern and emergence proportions.
12 to 24 months. The total at 12 months is \(100+300+200=600\). The total at 24 months is \(140+480+300=920\). The volume-weighted factor is \(920/600=1.5333\), selected as 1.53 to two decimal places for this example.
24 to 36 months. The total at 24 months is \(140+480=620\), and the total at 36 months is \(168+576=744\). The selected factor is \(744/620=1.20\). The tail factor from 36 months to ultimate is 1.05, as given.
At 12 months, the cumulative factor is \(1.53\times1.20\times1.05=1.9278\). Emergence is \(1/1.9278=0.5187\), and the unreported share is \(0.4813\).
At 24 months, the cumulative factor is \(1.20\times1.05=1.26\). Emergence is \(1/1.26=0.7937\), and the unreported share is \(0.2063\). At 36 months, the cumulative factor is 1.05, so emergence is \(1/1.05=0.9524\) and the unreported share is \(0.0476\).
Check. The tail keeps 36-month emergence below 1, as expected when additional development is selected. All reported proportions are rounded to four decimal places. The common mistake avoided is dropping the tail from the 12- and 24-month cumulative factors, which would overstate emergence at those ages.
Common Selection Traps
A well-calculated pattern can still be inappropriate if the data or assumptions do not match the intended use. In particular, the factor direction and the loss basis should be checked before calculating emergence.
- Reversing factor direction. If \(F_k\) is ultimate divided by the amount at age \(k\), then emergence is \(1/F_k\), not \(F_k\). For example, if \(F_k=1.50\), the wrong calculation \(e_k=1.50\) implies 150% emerged. The correction is \(e_k=1/1.50=0.6667\), with an unreported share of \(0.3333\).
- Mixing paid and incurred patterns. Paid and incurred amounts develop differently. Select and label each pattern for the loss measure to which it applies.
- Ignoring the tail. A tail factor greater than 1 increases the cumulative factor and therefore lowers emergence at earlier ages. Include it wherever the selected sequence calls for development beyond the observed ages.
- Overweighting thin or atypical experience. A ratio from a small or unusual group may be volatile. Compare the underlying volumes, claim mix, and operational conditions before allowing it to drive the selection.
- Assuming observed incurred amounts only rise. Reserve revisions can reduce incurred amounts. Investigate resulting ratios and judge whether they are credible for selecting a pattern rather than imposing a smooth shape without support.
The selected pattern matters because it supplies the age-specific emergence assumption used in expected-loss methods. A higher emergence proportion means a smaller complementary unreported share; a lower emergence proportion means a larger one. The pattern does not by itself determine an ultimate loss estimate. It provides the development-timing assumption, and its suitability depends on relevant experience and a consistent basis. This treatment of emergence-pattern selection is aligned with the CAS Exam 5 (Basic Ratemaking and Estimated Claim Liabilities) syllabus.
Check Your Understanding
For each question, state the emergence proportion and, where requested, the unreported share. Round proportions to four decimal places.
- A hypothetical paid pattern has a selected cumulative factor of 2.00 at 12 months. What are the 12-month emergence proportion and unreported share? (Answer: emergence 0.5000; unreported share 0.5000)
- Selected factors from 24 months to ultimate are 1.25 from 24 to 36 months and 1.10 from 36 months to ultimate. Find 24-month emergence and the unreported share. (Answer: emergence 0.7273; unreported share 0.2727)
- An incurred pattern has selected factors of 1.40 from 12 to 24 months and 1.10 from 24 months to ultimate. Find 12-month emergence and unreported share. (Answer: emergence 0.6494; unreported share 0.3506)
- A paid pattern has a selected factor of 1.20 from 24 to 36 months and a tail factor of 1.05 from 36 months to ultimate. Find the 24-month cumulative factor, emergence, and unreported share. (Answer: cumulative factor 1.26; emergence 0.7937; unreported share 0.2063)
- At 12 months, selected paid factors are 1.60 from 12 to 24 months and 1.25 from 24 months to ultimate. The corresponding incurred factors are 1.30 and 1.10. Find the emergence proportion for each measure and the difference, incurred minus paid. (Answer: paid 0.5000; incurred 0.6993; difference 0.1993)