I and many others have proposed that TWIA reduce the maximum policy limits on residential properties from the current $1.8 million to some substantially lower figure as a way of reducing the likelihood of post-event bonding and insolvency. Such a reform would also have the effect of reducing subsidization of individuals who have more expensive homes than the average Texan paying for the subsidy. As shown in the recently released Alvarez & Marsal report It would also bring TWIA more in line with other coastal windstorm programs such as Alabama’s $500,000 or Florida’s $1.000,000 dwelling limit.
But, a legitimate question is how much value would this reform really create? The case for the reform is somewhat stronger if it would reduce TWIA exposure by, say, 20% than if it would do so by just 1%. One statistic advanced by Representative Craig Eiland (D. Galveston) is to note that only 1% of residences insured by TWIA are valued at over $1 million. This statistic needs to be augmented, however, by an appreciation that the more valuable the residence, the more it contributes towards the risks of TWIA. All residences should not count alike.
Data provided by TWIA permits a first stab at a better answer. I’ve presented it in the chart below. It shows that reducing limits to $1 million for residences will have only a 1% effect on TWIA’s total insured value. Somewhat disappointing. If we’re serious about cutting TWIA exposure, we have to dig deeper. Going to $500,000 gets about a 5% reduction in total residential insured value. A reduction to $250,000 (the federal flood limit) creates the big gain, reducing TWIA’s residential TIV by 29%.
A few more points.
1. Just because the reduction is smaller than ideal does not mean we should not do it. Every little bit helps. And, as I have said ad nauseum, the economic and moral case for subsidizing expensive beach homes seems rather small. It’s all the more so where we condition the continued reduction in maximum policy limits, as I have proposed, on a finding that excess insurance is available.
2. For actuarial math nerds only! The figures I’ve put up, although the best I can do with the data I have, are still not ultimately what one would want. What really needs to be done by organizations such as AIR, RMS and others that have better access to the underlying storm data, is to determine the effect of right-censoring the loss distribution on homes on the overall distribution of losses faced by TWIA after a correlative premium reduction is taken into account. One can then use the survival function of the transformed aggregate insured loss distribution to recompute the probability of TWIA needing to resort to various classes of securities and its risk of insolvency. My guess is that the relationship is somewhat sublinear because losses to the left of the censor point are more common than losses to the right. So, even though I am showing a healthy 29% reduction in TIV from a $250,000 cap, that will likely not reduce of TWIA insolvency by 29%. If you asked me for a wild guess, I’d guess 15%. That’s still good. Every little bit helps.
3. The data here is one reason I really like coinsurance as a way of limiting TWIA exposure. (See suggestion # 2 of my 10 point proposal here). It has a more direct effect than right-censoring individual loss distributions, makes it less necessary for people to purchase multiple insurance policies, and may result in greater mitigation.
4. In the spirit of transparency, I’m posting the spreadsheet that underlies this analysis here, along with a Mathematica notebook used to conduct the analysis.