|
ALLOCATION OF SURPLUS.
One early attempt to make a pricing decision that explicitly reflects the
value of surplus was the New Jersey Automobile Remand Decision in 1972. The
Supreme Court asked then Commissioner of Insurance Robert Clifford to
determine a fair profit provision for automobile insurance. Clifford
determined both the fair return to the insurers' equity and the means of calculating that equity. After extensive hearings, Clifford found that only
underwriting profits and the investment income earned from investing
"policyholder supplied funds" were to be considered in ratemaking. A total
return from these sources of 6% was appropriate on the "needed surplus", but
only 1% was appropriate for "surplus surpus."
Clifford held that for automobile insurance, surplus was needed only up to
one-half of written premiums, the excess was "surplus surplus." Using this
reasoning, and a one-to-one industry average premium-to-surplus ratio he
allocated both "needed" and "surplus" surplus in proportion to written
premiums. Averaging the 6% and 1% returns, he found that a 3.5% return from
insurance operations was appropriate then (3). In the years since the
decision, Clifford's arithmetic has not been revised to reflect changes in
eitherthe industry's capital structure or profit levels outside the
industry, The 3.5% operating profit margin has been used as a fixed
target (4).
This appears to have been the understanding of Clifford's decision from the
first. Dineen, quoting ISO, "summed up Commissioner Clifford's decision"
as:
"The effect of Commissioner Clifford's Determination is to establish a
3.5% after Federal Income Tax provision for underwriting profit from
which investment income on policyholder-supplied funds, variable by
line, must be deducted and that result then increased to a pre-Federal
Income Tax basis for inclusion in the ratemaking formula."
Even though Clifford's allocation method has not attracted much attention,
it is still worth examining. Clifford considered the surplus of the
industry, but the technique is applicable to a single insurer.
 |
(1) |
The surplus allocated to a section of the insurer's book (S,) is found by
multiplying that section's net written premium (WP,) by the ratio of the
insurer's total surplus to total net written premiums.
This is shown for some groups of lines of insurance for 1985 industry totals
in Table I.
Table I -- Surplus Allocated on Annual Written Premium
 |
| Line |
1985 Written Premiums |
Allocated
December 31, 1985 Surplus |
 |
| Auto Liability |
$37,576,765 |
$18,388,661 |
| Auto Physical Damage |
25,519,959 |
12,488,512 |
| Homeowners |
14,473,884 |
7,082,977 |
| Other Property |
12,196,058 |
5,968,294 |
| Workers' Compensation |
19,263,729 |
9,426,947 |
| Medical Professional |
3,218,076 |
1,574,806 |
| Other Liability |
16,048,871 |
7,853,716 |
| Miscellaneous |
26,008,331 |
12,727,503 |
 |
| Total |
$154,305,673 |
$75,511,417 |
 |
| Note: |
Figures are in thousands of dollars. Unless otherwise noted,
all industry totals are taken from the 1986 edition of
Best's Aggregates and Averages. |
One practical problem with this allocation is that changing the relative
rate levels of the different sections changes the allocation. The amount of
surplus allocated to a section decreases when the rates are decreased. This
is counter-intuitive.
An alternative that avoids this problem is to allocate surplus on accident
year incurred losses and loss adjustment expenses. However, this method has
the problem of accurately estimating incurred losses during the accident
year. If that estimate could reliably be made, surplus would hardly be
needed.
Both allocation methods have several other practical problems. Neither
method actually considers how much surplus is needed to support a section of
the insurer's book. For example, a line of insurance that is no longer
written may take many years to run off. No surplus would be allocated to a
line running off because it has no written premiums and no accident year
losses. However the danger of adverse runoff still limits the insurer's
capacity and should be reflected in any allocation of surplus. Several
major insurers were acutely aware of this in the 1980's as they experienced
significant adverse development on Medical Malpractice reserves but had
stopped writing this line in the seventies (5). Clearly, surplus funded the
development, whether allocated to the line or not.
The opposite problem can occur in a rapidly growing section of the book.
Too much surplus would be allocated there. Since these methods do not
distinguish between different sections of the book to reflect special
circumstances, such as the type of reinsurance, growth patterns, and reserve
margins, they are extremely limited.
Either allocation formula relies on an annual flow (either premiums or
losses) to allocate year-end surplus. This choice of a one-year history of
the flow of funds is arbitrary. The choice also causes a peculiar factor in
the allocation formula. A section's surplus is found in equation (1) by
multiplying the section's annual written premium by the quantity (S/WP). S
is valued in dollars, and WP is valued in dollars-per-year, so (S/UP) must
be valued in years. This is a peculiar result to at least one actuary (6).
Calculation based on these allocations may not even need to consider
surplus. Consider a pricing methodology that uses a return-on-net-worth
calculation as a target.
 |
(2) |
________________________________________________________________
(3) R. E. Dineen, "An Early Look at the Decision in the New Jersey Remand
Case" NAIC Pioceedings, 1974, Volume II.
(4) See, for example, recent N.J. private passenger rate filings.
(5) Source: A. M. Best Standard Computer Tapes, Schedule P.
(6) Charles Niles' remarks at the 1984 Casualty Loss Reserve Seminar
(CLRS proceedings, p. 901.) addressed the premium-to-surplus ratio
as a measure of leverage, but they should apply here as well.
|