EQC Cap Increase

EQC Cap Increase

By December 17, 2021Uncategorized

In October, we provided a high level summary of the proposed EQC changes and their qualitative impact on the industry. We now have the opportunity, courtesy of RMS catastrophe model outputs, to undertake a more quantitative analysis of the impact of the EQC cap change.

Our analysis is based on the results from the RMS catastrophe model of two hypothetical portfolios. One portfolio (High) has a higher than average exposure to Wellington and the other (Low) has a lower than average exposure to Wellington. Both portfolios have 2,500 risks, which while small is sufficient to give useful insights. RMS provided us with access to their catastrophe model outputs for these portfolios, with which we are able to understand the impact of the EQC cap change on the key metrics that drive reinsurance pricing.

Change in expected earthquake losses

The Average Annual Loss (AAL) is a quantitative measure of the expected earthquake losses per year, noting that the vast majority of losses will come in years with significant events. Table 1 below demonstrates the change in estimated AAL per risk from the RMS catastrophe model on a gross of reinsurance basis, but net of the EQC cap.

Table 1 – Estimated change in AAL per policy due to change in EQC cap (selected regions): High Portfolio

As expected the AAL reduction in the highest earthquake regions is significant. Low earthquake regions will see little change because the base earthquake risk is so low.

However, insurers are not pricing for earthquake claims directly but rather for the sum of their retained losses, internal capital load, and the reinsurance premium they pay to cover the majority of their earthquake losses. Reinsurers consider several factors in setting their premiums, two of which are the AAL and the volatility of the losses.

Change in volatility 

Figure 1 below demonstrates the change in the AAL and volatility (shown as the CoV = standard deviation/mean) for each layer of a hypothetical reinsurance programme purchased up to a 1 in 1000-year event as required by NZ solvency regulations.

Figure 1 – AAL and CoV by reinsurance layer: High Portfolio

As can be seen, the AAL is decreasing in each layer, and in fact the fourth layer has been dropped at the $300k cap as the cost of a 1 in 1000-year event has decreased below this layer. The volatility of each layer has, however, increased substantially. In effect, the increase in EQC cap truncates the loss distribution such that the losses that escalate to the reinsurance layers are only those losses that are larger and more volatile than at a lower cap.

Change in reinsurance costs

We have conducted a simplistic reinsurance pricing calculation by combining the AAL and the volatility using the ‘k factor’ approach[1]. Real world reinsurance pricing is more complicated but we believe the ‘k factor’ calculation is a reasonable simplification for the purposes of this article. Our calculations are summarised in Table 2 below.

Table 2 – Summary of pricing impacts per risk: High Portfolio

Interestingly, the proportional impact to the Low Portfolio is also very similar, despite having much lower exposure to Wellington, with the reinsurance premium expected to decrease by $63 per risk (60%).  For the High Portfolio, the reduction in reinsurance premium ($74 per risk vs $39 AAL) reflects a relatively high reinsurance loss ratio of 53% whereas the top layers of the earthquake reinsurance programmes can have implied loss ratios as low as 25-30%.  This is due to the reinsurance forgone having a relatively low volatility, increasing the volatility of the ongoing reinsurance programme.

Increase to the EQC levy

The levy will increase by a maximum of $180 per policy (plus GST) at 1 October 2022. Modelling from EQC and Treasury in support of the increase suggested that an increase of $138 would be required from the current ‘breakeven’ premium. That is, part of the increase we are seeing is ‘catch up’ to a breakeven rate because the levy has not been increased in some time.

For the High portfolio, the cost of insurance may decrease by approximately $81 per risk in the private market, offset by a $180 increase in the levy. If we allocated the reinsurance premium to regions based on AAL then Wellington, Gisborne, Hawke’s Bay, West Coast and Marlborough will, on average, have a reduction in reinsurance premium greater than the increase in the EQC levy. All other regions will have a lower reduction.

We highlight that a proper reinsurance allocation exercise should also consider aggregation of risk and would likely result in a greater decrease for Wellington as it typically drives the top of reinsurance programmes.

Summary

On both our hypothetical portfolios the theoretical reduction in reinsurance premium would be less than the increase in the EQC levy, therefore increasing the total insurance premium of these supposed insurance customers. And, in practical effect, this theoretical reduction in reinsurance premium would likely emerge slowly over time due to the reinsurance industry dynamics, as covered in our previous article.

The structure of the EQC scheme is such that the EQC will retain more of the (relatively) low volatility risk, leaving the more volatile risk with the private insurers. Reinsurers charge proportionally more for the high volatility risk which, as demonstrated above, means the reinsurance savings are modest. In essence we are relying on reinsurer capital for the most volatile risk, while the government balance sheet is underwriting the less volatile component of the risk.

According to this exercise, most New Zealanders will be paying more for home insurance from 1 October 2022. The key will be in how this is communicated to customers and, critically, government and industry being aligned in their messaging on the expected impacts of the changes.

If you would like to hear more insights, please speak to your Finity consultant or contact

Simon Young 

Ph: +64 (9) 306 7708

simon.young@finityconsulting.co.nz

 

[1] (AAL + (k * standard deviation))/(1 – (expenses + brokerage)).  Assuming a k-factor of 10% and expenses and brokerage of 12.5%