PML & RE-INSURANCE’S INVOLVEMENT

Insurance is a business of Probability, spreading of losses when the probable event takes place. Spreading the losses takes place in the agreed business participation by way of Insurance & Re-insurance contracts.

Whenever, big Risks are underwritten, Insurance Company “Re-insures” portion of the Risks with the Re-insurance companies, who in turn participate in the risk spreading at agreed terms & conditions.

When Re-insurer accepts the Risk, it expects the Insurance Company, which cedes, to have some retention based on the capacity of that company, which in turn, indirectly have a control on the selection of risk transferred. It also expects the PML (Probable Maximum Loss) assessment (Estimation) to be based on their market knowledge & perception.

This leads to Re-insurer’s involvement in PML assessment/Estimation, whenever there is a fundamental & big difference by the insurer.

I wish to share, through this article, practical example of my assessment of PML of a big Refinery Risk in 1986 in India.

What is PML?

  • Probable means capable of being proved or
  • Which may reasonably be expected to happen
  • Wherein reasonable is meant in the sense of “Having Sound Judgement” or sensible or “as is judged appropriate or suitable to the circumstances or purpose”
  • Insurers are accepting risks on Probable Maximum Loss basis rather than on a Possible Maximum loss basis

My Practical experience & observations:

I had the privilege of assessing one of the big Refineries in south India in 1986, with a lot of urgency & differences of methods of assessing PML, at my level, corporate level, Regulatory level & Re-insurance level.

It was first estimated by Re-insurer on Sept. 9, 1985 for Material Damage & LOP for FIRE risk.

Suddenly I was forced to give PML within days. I went to the advisory committee local office & with the available information submitted my PML as 90% of Material damage & 100% of LOP.

Regulators started asking explanations with 4pager telexes.

Aghast with my calculations, they summoned all  Engineers of regulators & 2 Engineers from Re-insurer who had inspected the risk & had a marathon discussion within themselves without my presence!

Then they advised as under:

“Re: Fire Policies issued to ……Refineries Ltd. Dt. 25th March, 1986

Kindly refer to the PML report on the above risk assessing the PML around Rs. 341Crs. However, GIC have not accepted this PML Assessment and have referred the matter to Swiss Re Engineers who, based on the information made available to them at the time of their visit in September, 1985, have assessed the PML at Rs. 160 Crs. We enclose a copy of Re-insurer’s Report which we would request you to ask your engineers to go through & let us have your comments at the earliest. This may also be used as the basis for assessment of PML in similar cases.

In this connection, we have received a copy of the PML Assessment made by Mr. X, Engineer of the regulatory body on 7th March, 1986, where the PML has been assessed at Rs.1605 million. A copy of the report is also enclosed for your ready reference.”

My observations:

Points on PML of Refinery by SwissRe Engineer’s Report

At the outset, I would like to make it clear that the PML is a guess made on the assumptions based on certain technical data available &being an Electrical Engineer, I thought it fit to consult the chemical Engineer Mr. X of the Regulator on certain basics which I am not well conversant with (In Chemical Engineering) before my guessing the PML loss magnitude.

After going through the calculations sent by Re-insurer’s Engineer & Mr. X of Regulator, I feel the following

  1. As per the corporate letter dt. 25-08-86, if the basis of assessment of PML is to be followed as per Re-insurer’s report, a chemical Engineer will understand better than any other Engineer with different faculty of Engineering, since the formulae & the assumptions made need a lot of explanations to other faculty Engineers. Secondary explosions may create further explosions for which there are no guidelines or assumptions.
  2. I have been given to understand that PML assessments made in major loss events (Including the Flexborough) do not tally with the losses suffered. I have a feeling that the weightage of such experiences should find a place in our calculations
  3. If for the business operations, the Re-Insurance department wants to limit our PML for reinsurance underwriting considerations, it is better to include the appropriate factor of such nature in the form of guide lines.

Conclusion:

Business decisions & technical correctness are two different things & business decisions prevail.  Re-insurer’s intervention in big risks is welcome sign.

What way PML assessment affects Re-insurers?

No hard & fast rules can be laid down & no mathematical formulae can be prescribed to work out such estimates of a maximum loss.

Following key factors govern the method of thinking in the process of arriving at the maximum loss estimates.

  • Intimate knowledge of the hazards involved in each case
  • Experience gained by the knowledge of past losses
  • Logical approach & common sense
  • The estimates are essentially personal ones & a number of subjective factors are involved as it depends upon the estimator’s judgment of type of plants, processes, insured perils & knowledge and experience of comparable risks.
  • It, therefore, becomes imperative that these estimates are prepared by specialized people & are based on facts and figures relating to particular risk.
  • The estimates are exposed to numerous fluctuating factors
  • Rising costs on buildings / machineries such as Increasing concentration of values, Changes in process & technology etc. THEREFORE PML NEEDS TO BE CONTINUOUSLY REVIEWED with every treaty each time during renewal.

By: K. S. Burli, M.E.(Electrical). FIV, Risk Management Consultant (Insurance), Approved Valuer, Published in “The Insurance Times”, November, 2012

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.