Verisk recently released their updated estimates of industry cat risk. Amazingly, average annual loss (AAL) surpassed $150B, up from $100B just four years ago. Similarly, the 1 in 100 aggregate loss (PML) grew to $400B from $300B four years ago.
These are stunning rates of growth beyond what we have seen previously and suggest that the industry still is underestimating its cat exposure.
At the beginning of the survey in 2012, AAL was ~7.5% of US direct premiums. In 2024, it will be nearly 10%.
Note, the US is ~2/3 of industry AAL so I have estimated US AAL below and compared to industry US premiums.
| Year | AAL | US AAL (E) | US DPW | US AAL/DPW |
| 2012 | $59.3B | $39.5B | $523.9B | 7.5% |
| 2013 | $67.4B | $44.9B | $546.3B | 8.2% |
| 2014 | $72.6B | $48.4B | $570.8B | 8.5% |
| 2015 | $74.4B | $49.6B | $591.8B | 8.4% |
| 2016 | $80.0B | $53.3B | $613.4B | 8.7% |
| 2017 | $78.7B | $52.5B | $642.5B | 8.2% |
| 2018 | $85.7B | $57.1B | $678.3B | 8.4% |
| 2019 | $91.8B | $61.2B | $712.5B | 8.6% |
| 2020 | $99.6B | $66.4B | $729.0B | 9.1% |
| 2021 | $106.3B | $70.9B | $797.8B | 8.9% |
| 2022 | $123.3B | $82.2B | $876.1B | 9.4% |
| 2023 | $133.0B | $88.7B | $965.9B | 9.2% |
| 2024 | $151.1B | $100.7B | $1.05T (E) | 9.6% |
Given not all lines of business (e.g. workers comp, much of liability) are exposed to cat risk, this means the increase in cat load relative to property premium has grown even more dramatically.
See below, the trend is similar if we compare AAL to US industry capital.
| Year | AAL | US AAL (E) | US Surplus | US AAL/Surplus |
| 2012 | $59.3B | $39.5B | $595.2B | 6.6% |
| 2013 | $67.4B | $44.9B | $666.7B | 6.7% |
| 2014 | $72.6B | $48.4B | $688.7B | 7.0% |
| 2015 | $74.4B | $49.6B | $687.9B | 7.2% |
| 2016 | $80.0B | $53.3B | $712.5B | 7.5% |
| 2017 | $78.7B | $52.5B | $765.7B | 6.9% |
| 2018 | $85.7B | $57.1B | $757.1B | 7.5% |
| 2019 | $91.8B | $61.2B | $865.9B | 7.1% |
| 2020 | $99.6B | $66.4B | $929.2B | 7.1% |
| 2021 | $106.3B | $70.9B | $1.0531T | 6.7% |
| 2022 | $123.3B | $82.2B | $981.5B | 8.4% |
| 2023 | $133.0B | $88.7B | $1.042T | 8.5% |
| 2024 | $151.1B | $100.7B | $1.100T (E) | 9.2% |
Even more shocking is what happens when we estimate US PML relative to US surplus.
| Year | PML | US PML (E) | US Surplus | US PML/Surplus |
| 2012 | $205.9B | $171.6B | $595.2B | 28.8% |
| 2013 | $219.4B | $182.8B | $666.7B | 27.4% |
| 2014 | $231.5B | $192.9B | $688.7B | 28.0% |
| 2015 | $232.8B | $194.0B | $687.9B | 28.2% |
| 2016 | $252.9B | $210.8B | $712.5B | 29.6% |
| 2017 | $246.9B | $205.8B | $765.7B | 26.9% |
| 2018 | $270.9B | $225.8B | $757.1B | 29.8% |
| 2019 | $288.2B | $240.2B | $865.9B | 27.7% |
| 2020 | $301.1B | $250.9B | $929.2B | 27.0% |
| 2021 | $320.5B | $267.1B | $1.0531T | 25.4% |
| 2022 | $345.0B | $287.5B | $981.5B | 29.3% |
| 2023 | $372.0B | $310.0B | $1.042T | 29.7% |
| 2024 | $400.0B | $333.3B | $1.10T (E) | 30.3% |
The US 1 in 100 PML is 30% of US surplus. Wow!!! That is a level typically associated with cat reinsurers.
Sure, I get a lot of that PML doesn’t sit on US balance sheets due to reinsurance protection, but on a gross basis, the average US diversified insurer looks like a cat reinsurer!
At the end of the day, the primaries are paying for this one way or the other. Just because they’re paying it through ceded premium, doesn’t mean the cost isn’t there.
This calls into question whether US insurers are truly adequately capitalized to withstand a 1 in 100 event or are we heading towards another post Andrew reckoning where the industry learns it didn’t hold enough capital for cat risk?
Just because we have better models now, doesn’t mean we can’t make the same mistakes in new ways! Housing investors had much better tools to assess risk in the 2000s than the 1990s but they made much bigger mistakes.
History vs. Models
Why might the industry be underestimating cat risk? My guess is that, while it claims to rely on models, much of the industry is still benchmarking to history.
What do I mean? Take a look at this chart. It’s annual insured cat losses since 1970 in today’s dollars.
What do you see? I see there have only been three years above the $150B AAL!
This likely is partially due to exposure growing faster than inflation so that the chart isn’t truly comparable to the past.
However, you certainly don’t see anything approaching a 1 in 100. This data is over 50 years of history. Why have there been no events approaching the 1 in 100?
Remember, a 1 in 100 can happen in year 1. The odds are 42% the 1 in 100 should have already happened in this 53 year sample.
Of course, sometimes it takes over 100 years for it to happen (37% of the time it takes longer than 100 years).
But if the industry is using history as a guide, it is severely underestimating the likelihood of that $400B year.
Increasing At An Increasing Rate
Let’s look at the AAL data again. This time we’ll look at rates of change. You’ll notice there has been an acceleration of trend so far this decade.
| Year | AAL yoy | PML yoy | DPW yoy | Surplus yoy |
| 2013 | 13.7% | 6.6% | 4.3% | 12.0% |
| 2014 | 7.7% | 5.5% | 4.5% | 3.3% |
| 2015 | 2.5% | 0.6% | 3.7% | -0.1% |
| 2016 | 7.5% | 8.6% | 3.6% | 3.6% |
| 2017 | -1.6% | -2.4% | 4.7% | 7.5% |
| 2018 | 8.9% | 9.7% | 5.6% | -1.1% |
| 2019 | 7.1% | 6.4% | 5.0% | 14.4% |
| 2020 | 8.5% | 4.5% | 2.3% | 7.3% |
| 2021 | 6.7% | 6.4% | 9.4% | 13.3% |
| 2022 | 16.0% | 7.6% | 9.8% | -6.8% |
| 2023 | 7.9% | 7.8% | 10.2% | 6.2% |
| 2024 | 13.6% | 7.5% | 8.7% (E) | 5.5% (E) |
If I split the sample down the middle, you’ll see through 2018, AAL was increasing 6%/yr. Since then, it has increased at 10%.
| Year | AAL yoy | PML yoy | DPW yoy | Surplus yoy |
| 2012-2018 | 6.3% | 4.7% | 4.4% | 4.1% |
| 2018-2024 | 9.9% | 6.7% | 7.6% | 6.4% |
That’s faster than the 7% growth in premium and 6% growth in capital. So something has clearly changed. While part of the growth can be attributed to general inflation, that wouldn’t have impacted ’18-’20.
We can debate causes, but both the Swiss Re Sigma report and Verisk attribute ~1% annual growth to climate with the majority to exposure growth (partly more building in risky areas, partly inflation).
| Source of US loss inflation | Verisk | Swiss Re |
| Exposure ex-price | 2.0% | 2.3% |
| Inflation | 5.4% | 3.4% |
| Climate | 1.0% | 1.0% |
| Other | 0.7% (E) | 1.3% |
| Total | 9.2% (E) | 8% |
Hopefully cat exposures don’t keep expanding like growth in the universe because we have a big enough problem as is. Presumably, double digit annual growth isn’t sustainable over the long term.
But I bet 12 years ago, nobody would have bet on a 8% CAGR going forward. There is nothing on the horizon to suggest this problem is going to get better.
So what do we have to look forward to?
- further increases in annual cat loads and PMLs relative to capital
- $200+B cat loss years not being out of the norm
- continued demand for more reinsurance capital to offload risk
What’s A Carrier To Do?
For starters, raise your cat loads in your pricing, especially in admitted lines.
My guess is many companies need to admit that their reinsurance retentions need to be higher, and they should use the savings from buying low to buy higher up (and for multiple events).
A simple stress test would be to model a $200B year and see how well your reinsurance holds up to that.
Buying capital protection is more important than buying earnings protection. It is better to accept earnings volatility than have a capital stress in a bad year.
Remember, the 1 in 100 AAL is an aggregate, not occurrence, number. The really bad years tend to have multiple large events so buying protection for second and third events (with lower retentions) is just as important as buying all the way up for one event.
I would also explore synthetic capital options. One of the biggest mistakes companies make is, while they may pass their capital stress test today, they often won’t pass it the day after they’ve lost a large chunk of capital.
For example, buy some far out on the curve ILW beyond your gross exposures to account for potential model error. This can act like cheap capital in a stress event (if your actual loss is > modeled loss, this extra protection ends up replicating new capital).
But the main thing is to look in the mirror and be honest about your exposure while you can still do something about it.