Tuesday, December 2, 2014

Doing the sums for Crisafulli

The numbers in the video currently promoting cyclone preparedness come from Risk Frontiers a research unit at Macquarie University. I never did get around to another insurance post following the recent distressing experience of listening to Warren Entsch for an hour on the topic. Relevant to this disaster scenario, Wazza expressed that Cyclone Yasi shouldn't impact on Cairns premiums because there wasn't much damage in Cairns. Homer Simpson for Leichhardt  o_O

Among other risk delusions are that the insurers are ripping us off by exaggerating cyclone risk relative to event such as storms in SEQ and bushfires. This is a common misconception frequently also heard on talkback radio. It also popped up this week from Local Government Minister Crisafulli down in Townsville: NQ insurance rip off must stop
“I’m not asking insurers to charge people in other parts of the state any more but I am asking them to stop unfairly targeting the north,” he said. “This storm proves there is no greater risk in North Queensland, which has been the reason used by some insurers to jack up premiums.
“Whether it is bushfires in Vic, hail storms in Sydney or Brisbane or floods in central Queensland, we all live with a level of risk. “North Queenslanders are happy to pay our fair share for coverage but we just don’t want to be the whipping boys because of an unfair ­perception that there is a greater risk in our part of the country.”
Crisafulli is just plain wrong and is also directly contradicting the findings of the Australian Government Actuary. It also reminded me of a blog post discovered a while ago written by John Devaney at local FNQ insurance broker Joe Vella. This was I think written a couple of year ago and John has done a very simple back-of-the-drink-coaster analysis. It doesn't pretend to be anything sophisticated with some loose assumptions but is certainly superior to the public-bar-windbag anecdotal derp of Entsch and Crisafulli. This is an extract:

Written by John Devaney.

There are some parts of Warren Entsch's Opinion piece [Time to find answers on North's insurance crisis Cairns Post Oct 20] that correctly challenges the Insurance Council of Australia [ICA] to be more frank with consumers. Clearly there is an issue with the massive hikes in certain classes of property insurance.

That said I have a fundamental problem with the Entsch article. He focusses on what he sees as the inequitable price of insurance compared to other regions of Australia. It's good politics that score emotive points and much of that is understandable but it's inaccurate none the less. And it distracts from how we must face and attempt solve the problem.
Yes it's true that house and other property insurance in Brisbane, Sydney, Melbourne or even Darwin might be cheaper than Far North Queensland but Darwin has had one Cyclone in fifty years. Our region has had several major Cyclones in thirty years, two of those in the last six years.
The brutal reality is the mathematical imperatives that are in play. The following is pretty crude but it illustrates my point. The cost of claims is sourced from the ICA website.
The population figures are courtesy of Mr Google.
Table 1: Cyclone Yasi Cost per North QLD "Dwelling"
  • Total Cyclone Yasi Insurance Claims Cost $1,412,239,000
  • Population of Cairns 150,920 + Townsville 190,000 + Mackay 121,000 = 461,920
  • Divide Population by say, 4 persons representing 1 Family Unit = 115,480 "Dwellings"
  • Spread of Claims Cost across each Dwelling or Family Unit: $12,229
Now let's look at the Brisbane Floods
Table 2: SE QLD Flood Event Cost per South East QLD "Dwelling"
  • Total SE QLD Flood Event Claims Cost $2,387,624,000
  • Population Brisbane in a 200 KM radius from the Brisbane Post Office: 3,005,000
  • Divide Population by say, 4 persons representing 1 Family Unit = 751,250 "Dwellings"
  • Spread of claims Cost across each Dwelling or Family Unit: $3,178
This is the harsh, stark reality. The Flood event was almost double the cost of Yasi yet spread across the respective population bases Yasi cost Far North "dwellings" 4 times that of the Flood Event's cost to an equivalent "dwelling." Sure, the above figures include commercial losses in North QLD but they include commercial losses in South East QLD as well. The numbers would still crunch the same way.

Warren Entsch suggested we compare ourselves with Darwin. Fair enough.
Table 3: Cyclone Tracy amortised over 38 years (there have been no other major weather events)
  • Cyclone Tracy Claims Cost $ 200,000,000
  • Population of Darwin 120,000 divided by our 4 family average = 30,000 "Dwellings"
  • Spread of Claims across each Dwelling or Family Unit = $6,666
  • Amortised over 38 years = $175
Compare that to our region
Table 4: FNQ Cyclones & Weather Events amortised over 38 years
  • Winifred ($40 M) + Aivu ($26M) + Rona ($4M) + Steve ($11M) + Tessi ($15M) Larry $540 + NQ Deluge 2009
  • ($19M) + Yasi ($1,412,239,000) = $2,067,239,000
  • Spread of Claims cost across each dwelling or Family Unit = $17,901
  • Amortised over 38 years = $471
We have two fundamental problems we have to solve. We don't have the spread of numbers to sustain ever increasing losses from more frequent weather events.

So what happens now?

In my almost forty years in the industry there has only been three occasions that I can recall where the insurance market corrected itself, sometimes known as a 'hard market.' They were the 1990's recession, the 2001 post HIH crash and September 11 tragedy and in 2012 the floods, and Cyclone Yasi. In between those 3 corrections the market has been relatively defined as 'soft' in short, too under-priced.

Some insurance industry commentators suggest that the market will start to soften in the next two years. This time, I'm not so sure. There is far more accurate technology than ever before. In my view we have entered the age of computerised actuarial discipline – certainly when it comes to commoditised insurance products like house and motor vehicle policies. Insurers used to think – "let's dominate the market by cutting the price." The other classic rationale was "Jeez those chaps in North QLD are caning our profits. Thank goodness the Brisbane, Sydney, Melbourne markets can prop up the figures." That thinking is redundant. Now the market mantra is "Each region must be self-sustaining."

I think the Brisbane storm is currently up to about $200 million but it will go nowhere close to a big cyclone hitting a regional city relative to the written insurance premiums in each area based on historical experience and wont shift the analysis much. I have plenty of issues with the recent insurance situation in Tropical Queensland but sticking our heads in the sand and deluding ourselves on the relative risks avoids the core issues and understanding. It is simply seeking a subsidy from the rest of Australia.

Crisafulli is in fact asking other people to pay more even when he says he isn't. It can't be otherwise on any analysis of the numbers. Perhaps Crisafulli should instead take on the issue of state insurance stamp duty in cabinet which as posted here previously is a fiscal transfer from his constituents to Brisbane.

Update: Allianz has been selected as preferred buyer of TIO: https://theconversation.com/will-the-territory-insurance-office-sale-push-up-premiums-33741

No comments:

Post a Comment