Deriving Increased Limits Factors from Exposure Curves

Why an Exposure Curve Produces an ILF

An increased limits factor (ILF) compares expected losses under two policy limits. A severity exposure curve makes that comparison convenient: it expresses limited expected loss at each limit relative to a common reference amount. The factor is then the ratio of the curve value at the requested limit to its value at the base limit.

The ratio is meaningful only when both curve values use the same loss distribution, coverage terms, and normalization. In particular, a curve for ground-up losses cannot automatically be combined with a limited expected loss calculated after a deductible. Define the payment under the policy terms first; then use a curve that represents that same payment basis.

Definition: Let $Y_L$ be the insurer's payment for one loss under the specified coverage terms when the applicable limit is $L$, and let $K>0$ be a reference expected loss that does not vary with $L$. The limited expected loss is $\operatorname{LEL}(L)=E[Y_L]$. The normalized severity exposure curve is $C(L)=\operatorname{LEL}(L)/K$. For a base limit $L_b$ and requested limit $L_r$, the increased limits factor is $\operatorname{ILF}(L_r;L_b)=\operatorname{LEL}(L_r)/\operatorname{LEL}(L_b)=C(L_r)/C(L_b)$, provided $\operatorname{LEL}(L_b)>0$.

For ground-up loss $X\geq 0$ with no deductible, the usual limited payment is $Y_L=\min(X,L)$. A common normalization is $K=E[X]$, assuming $0<E[X]<\infty$. The curve is then $C(L)=E[\min(X,L)]/E[X]$. Other normalizations can be used, but the numerator and denominator of the ILF must use the same one so it cancels in the ratio.

Deriving the Ratio

The derivation follows directly from the definition of the exposure curve. At each limit, the curve value is the expected payment divided by the common reference amount. Dividing the requested-limit curve value by the base-limit curve value removes that reference amount.

\frac{C(L_r)}{C(L_b)} = \frac{\operatorname{LEL}(L_r)/K}{\operatorname{LEL}(L_b)/K} = \frac{\operatorname{LEL}(L_r)}{\operatorname{LEL}(L_b)}

Thus, if the expected loss at the base limit is known, the corresponding expected loss at the requested limit is the base expected loss multiplied by the ILF. This operation changes the expected loss component on the stated basis; it does not by itself adjust fixed expenses, commissions, profit provisions, or other premium components.

Worked Example: Deriving an ILF From a Discrete Severity Distribution

A hypothetical insurer models ground-up loss $X$, in thousands of dollars, as follows: $P(X=10)=0.50$, $P(X=30)=0.30$, and $P(X=80)=0.20$. Assume independent losses follow this same distribution for both contracts, with no deductible or other coverage modification. Find the ILF for a $50,000 limit relative to a $20,000 base limit. Interest and mortality assumptions are not applicable; the severity distribution and parameters are the three stated loss amounts and probabilities.

Method. Calculate the expected payment at each limit, then divide the requested-limit amount by the base-limit amount. The common mistake to avoid is dividing the curve value at the requested limit by 1 rather than by the curve value at the base limit.

First check the probabilities: $0.50+0.30+0.20=1.00$. The ground-up mean, in thousands, is

$$ E[X]=(10)(0.50)+(30)(0.30)+(80)(0.20)=5+9+16=30. $$

At the $20,000 limit, the insurer pays 10, 20, and 20 for the three possible losses. Therefore,

\operatorname{LEL}(20)=(10)(0.50)+(20)(0.30)+(20)(0.20)=5+6+4=15

thousand dollars. At the $50,000 limit, the payments are 10, 30, and 50, so

\operatorname{LEL}(50)=(10)(0.50)+(30)(0.30)+(50)(0.20)=5+9+10=24

thousand dollars. Using $K=E[X]=30$, the exposure-curve values are $C(20)=15/30=0.50$ and $C(50)=24/30=0.80$. The factor is

\operatorname{ILF}(50;20)=\frac{0.80}{0.50}=\frac{24}{15}=1.60.

The check is that $15(1.60)=24$ thousand dollars, exactly the requested-limit expected loss. The ILF is 1.60, not 0.80: 0.80 is the requested-limit exposure-curve value relative to ground-up mean loss, not relative to the base-limit expected loss.

Calculating From a Tabulated Curve

An exposure curve may be supplied directly rather than derived from individual severity probabilities. If the required limit is between listed limits, linear interpolation can estimate the curve value. This treats the curve as a straight line over that interval; it is an approximation, not a claim that the true curve is linear.

Worked Example: Interpolating Before Taking the Ratio

A hypothetical property portfolio has ground-up mean loss $E[X]=\$50,000$. Its exposure curve, normalized by that mean, has values $C(\$20,000)=0.54$, $C(\$40,000)=0.70$, and $C(\$80,000)=0.84$. Estimate the ILF for a $60,000 limit relative to a $20,000 base limit using straight-line interpolation between $40,000 and $80,000. The curve is assumed to represent the same ground-up loss basis at all limits; interest and mortality assumptions are not applicable.

Method. Interpolate the curve value at $60,000, then divide by the curve value at the base limit. The common mistake to avoid is dividing by the curve value at $40,000 merely because it is the lower endpoint of the interpolation interval.

The $60,000 limit is halfway between $40,000 and $80,000:

\frac{60{,}000-40{,}000}{80{,}000-40{,}000} = \frac{20{,}000}{40{,}000} =0.50.

The interpolated curve value is therefore

$$ C(60{,}000)=0.70+0.50(0.84-0.70)=0.70+0.07=0.77. $$

The estimated ILF relative to $20,000 is

\operatorname{ILF}(60{,}000;20{,}000)=\frac{0.77}{0.54}=1.425925\ldots\approx 1.426.

As a check, the implied limited expected losses are $0.54(\$50,000)=\$27,000$ at the base limit and $0.77(\$50,000)=\$38,500$ at the requested limit. Their ratio is $\$38,500/\$27,000=1.425925\ldots$, matching the curve ratio. The factor is rounded to three decimal places.

Deductibles and the Payment Basis

The curve method still works when policy terms include a deductible, but the payment definition must reflect how the limit is stated. For a ground-up loss $X$, deductible $D$, and a limit $L$ applying to the insurer's payment above the deductible, the payment is $Y_L=\min((X-D)^+,L)$, where $(z)^+=\max(z,0)$. A curve built for that payment can be normalized by any common positive reference amount, such as the expected unlimited payment $E[(X-D)^+]$, if it is finite and positive.

This is different from a contract in which $L$ is the maximum total covered loss before subtracting the deductible. For that contract the payment would be $(\min(X,L)-D)^+$. These formulas generally produce different payments, so the analyst must read the policy wording rather than infer the limit convention from a number alone.

Worked Example: Applying the Ratio With a Per-Payment Limit Above a Deductible

A hypothetical liability policy has a $5,000 deductible and a limit applying to the insurer's payment above that deductible. Ground-up loss $X$, in thousands of dollars, has the discrete distribution $P(X=4)=0.25$, $P(X=12)=0.35$, $P(X=32)=0.25$, and $P(X=72)=0.15$. Find the ILF for a $30,000 payment limit relative to a $10,000 payment limit. Assume this same severity distribution and deductible apply at both limits. Interest and mortality assumptions are not applicable.

Method. Calculate $\min((X-5)^+,L)$ for each outcome and take the probability-weighted mean at each limit. The common mistake to avoid is subtracting the deductible from the stated payment limit a second time.

The probabilities sum to $0.25+0.35+0.25+0.15=1.00$. For $X=4$, the loss is below the deductible and the payment is zero at both limits. For $X=12$, the payment before applying the limit is $12-5=7$, so the payment is 7 at both limits. For $X=32$, it is 27, giving payments 10 and 27. For $X=72$, it is 67, giving payments 10 and 30.

Ground-up loss $X$ (thousands)	Probability	Payment at limit 10	Payment at limit 30
4	0.25	0	0
12	0.35	7	7
32	0.25	10	27
72	0.15	10	30

The limited expected losses, in thousands, are

\operatorname{LEL}(10)=(0)(0.25)+(7)(0.35)+(10)(0.25)+(10)(0.15) =0+2.45+2.50+1.50=6.45,

\operatorname{LEL}(30)=(0)(0.25)+(7)(0.35)+(27)(0.25)+(30)(0.15) =0+2.45+6.75+4.50=13.70.

The unlimited payment above the deductible has expected value $2.45+6.75+10.05=19.25$ thousand dollars. Using this as the common normalizer gives $C(10)=6.45/19.25$ and $C(30)=13.70/19.25$; the normalizer cancels:

\operatorname{ILF}(30;10)=\frac{13.70}{6.45}=2.124031\ldots\approx 2.124.

The check is $6.45(2.124031\ldots)=13.70$. A mistaken interpretation that reduces the stated payment limits to 5 and 25 would instead calculate payments under different contracts. It would give limited expected losses of $3.75$ and $12.45$, respectively, and a factor of $12.45/3.75=3.32$. Those are not the requested payment limits and therefore do not answer this problem.

Interpretation and Limits of Use

An ILF is a relative expected-loss measure, not a standalone premium and not necessarily a multiplier for every component of a rate. If a base-limit pure premium is \$120 and the applicable ILF is 1.25, the corresponding expected-loss amount is \$150, assuming the base premium is on the same loss basis. A fixed policy expense should not automatically be multiplied by 1.25; doing so would change the expense assumption rather than adjust expected loss for the increased limit.

The ratio also depends on the base limit. With the same exposure curve, the factor from one requested limit can differ when measured against a different base limit. State both limits, and confirm that both curve values refer to the same class of business, peril, deductible convention, development and trend basis, and payment definition. The ratio is not valid when the numerator and denominator come from incompatible bases or when the base expected loss is zero.

Key takeaway: To derive an ILF from an exposure curve, identify the payment and common normalization first, obtain the curve values at the requested and base limits, and divide requested by base. Check that multiplying the base limited expected loss by the factor recovers the requested limited expected loss.

Check Your Understanding

For each problem, use the specified curve or payment basis and give the ILF to three decimal places unless the result is exact.

A ground-up exposure curve has $C(15)=0.48$ and $C(45)=0.78$. Find the ILF for limit 45 relative to limit 15. (Answer: 1.625)
A hypothetical portfolio has $E[X]=\$60,000$, $C(\$25,000)=0.55$, and $C(\$50,000)=0.73$. Find each limited expected loss and the ILF for $50,000 relative to $25,000. (Answer: \$33,000; \$43,800; ILF $=1.327$)
A curve has $C(30)=0.60$ and $C(70)=0.82$. Using straight-line interpolation, estimate the ILF for limit 50 relative to limit 30. (Answer: 1.183)
For the same loss basis, the base limited expected loss is \$18,000, and the curve values are $C(L_b)=0.60$ and $C(L_r)=0.76$. Find the requested-limit limited expected loss and ILF. (Answer: \$22,800; ILF $=1.267$)
Ground-up loss $X$, in thousands, has $P(X=8)=0.20$, $P(X=20)=0.50$, and $P(X=60)=0.30$. A $10,000 deductible applies, and each stated limit is a maximum insurer payment above the deductible. Calculate the ILF for a $40,000 limit relative to a $15,000 limit. (Answer: 1.789)

What You'll Learn