There has been a lot of optimism recently regarding improvements in insurance pricing. Unfortunately, it is hard to know how much credence to put into this optimism as insurance company executives suffer from “overexuberance complexitis” continuously predicting pricing turns that usually die before they get started.
Thus, today’s topic is assessing how much optimism is warranted this time. I will review some history, provide some context, and even delve into some psychology. To provide the conclusion first: there is some reason for hope, but success is by no means assured and never comes pain free.
First, Some Background
Before diving into the weeds, let’s recap the situation on the ground. The cause for optimism is pricing is starting to rise across multiple lines of business for the first time since 2011.
The big difference vs. 2011 is that improvement then was driven by losses in workers compensation and low interest rates. The impetus for more rate came from the large primary carriers, notably Travelers. This time the pricing is being led by the wholesale (E&S) markets which is more typically a sign of stress. There is also some pressure in specialty lines, most notably D&O. I will discuss the dynamics behind these pressures a little later, but first a logic review…
Necessary vs. Sufficient
Why do so many insurance execs typically overestimate the likelihood of better pricing? Because they confuse a necessary condition with a sufficient condition. Yes, we’re going back to college math class!
A necessary condition is something that is required for an outcome. So, in order to keep breathing, we need oxygen. However, necessary conditions aren’t necessarily sufficient. You could have all the oxygen you need, but if you don’t also have access to water, you will die. So there is no one necessary condition to remain alive. Rather, there are several necessary conditions and only the combination of those necessary conditions are sufficient to continue life.
Insurance pricing is similar in that there is no one sufficient condition that guarantees better pricing. The common error is to assume a spike in losses will “turn the market”, particularly after say a large hurricane. However, increased losses, while necessary, are not a sufficient condition to produce better pricing.
The other necessary condition for a market turn is a withdrawal of capacity. This can come through a number of potential causes (none of which are necessary but any of which has the potential to be sufficient), including increased capital requirements, poor underwriting performance, large investment losses, and the catch-all “change in the perception of risk”.
What makes all this tricky is we can identify some necessary conditions, but no sufficient conditions. The only truly sufficient condition is fear! Fear is usually the product of a number of those potentially sufficient conditions coming together at the same time.
For example, 2009 seemingly had all the necessary components for a market turn and there were certainly companies predicting it. There were large investment losses, AIG teetering on the brink, several other large carriers whose balance sheets deserved multi notch downgrades, and a lot of uncertainty about the future. Everything was there but the fear! The excess market should have been afraid to write layers above AIG’s lead but for whatever reason it wasn’t and the moment passed.
Lessons from 2001
The reason the market turned after 9/11 was we had fear. There were plenty of necessary conditions – years of poor pricing, poor investment results, failed companies like Reliance, bad mergers, the explosion of asbestos claims, a spike in economically sensitive claims like D&O due to the recession, abuse of the med mal legal system – but the fear from 9/11 became the sufficient condition. Take another look at that list. My God there were a lot of bad things going on! And yet, they weren’t sufficient without 9/11.
One might look at that list and say “that seems kind of similar to today“. Well, maybe not exactly, but yes, it rhymes…years of poor pricing, loss cost pressure bubbling up, large torts like talc and Monsanto, low interest rates, bad mergers…it certainly rhymes.
For those readers with long memories, you will recall pricing actually began to move up in 2000 as companies were aware of the bad hand they were holding but unable to achieve much beyond small price increases. Without fear, the best option was not to make it worse. They tried jawboning the market to change the tide. Sort of sounds like what is happening at Lloyd’s today. Again, it rhymes.
Just like now, things started with E&S as that’s where the pressure built first as it has the most leverage (aside from reinsurance, but the reinsurers always seem to be the last to know when the tide goes out). It wasn’t enough though to really make a dent. It made things less worse, but the business written in 2000 & 2001 was still bad.
To really move pricing, someone would have to admit how bad their reserves were first…and nobody wanted to do that! It was a classic prisoner’s dilemma. In isolation, it seems crazy that 9/11 happening would make St.Paul admit it had a med mal problem or Travelers admit it had an asbestos problem or XL admit it made bad deals or AIG admit it screwed up pretty much its whole book, but that’s what happened. Nobody would confess until it became acceptable. They would all take their pain together. 9/11 broke the prisoner’s dilemma and made it OK to tell the truth about the sorry state of reserves.
The Compounding of Errors
I want to re-emphasize the point above about the pre-9/11 price improvement. Just because the price went up, it didn’t make the business good. It made it less bad and still worse than the business they wrote the prior year.
When companies say today they are getting rate in excess of trend, be cautious! We only need to go back a few years to recall companies saying rate was now above trend in commercial auto. And yet, every year since the reserves have developed adversely and the new picks are still above the original picks from five years ago (though not above the fully developed numbers from recent accident years = optimism still reigns).
If a company tells you rate is in excess of trend, you should assume one of two things: 1) They are underestimating their trend. 2) They are indeed getting rate over trend but the base loss ratio they are starting from is several points above the pick they are showing in their financials. At this stage of the cycle, most casualty lines are showing early signs of stress and those rarely get better with a little bit of price. Usually, they get materially worse first.
Introducing IANS Reserves
So let’s do some more logic for a moment. First, some assumptions:
1) Companies have an information edge over investors.
2) Companies typically only pursue price after bad news.
3) Recall from earlier, nobody wants to admit bad news first so there is a bias to delay confession.
Ergo: If companies are pursuing price without an obvious need for it, the losses are probably worse than they have disclosed publicly.
There are certainly obvious ways to keep this information from investors like not taking up your assumptions on new cases even when there is evidence of emerging inflation. Maybe that is happening I don’t know, but I am not suggesting anything of that sort.
What I am referring to is something more subtle. An observant industry participant can be aware of certain adverse trends emerging in the system. If that person were say a chief actuary, he or she might see these worsening trends and realize they have an emerging reserve problem. However, if there are a lack of new cases reported, it might be hard to figure out how that trend could impact this person’s own company.
A prudent action might be to put up more IBNR as a precaution. But again, nobody wants to go first…prisoner’s dilemma. And we can see IBNR trends falling across the industry in casualty lines suggesting that indeed most chief actuaries are choosing to look away.
I call this tendency to look away from bad stuff we know is out there but hasn’t reached the threshold where we can confidently estimate them…Incurred And Never Spoken. That’s right, IANS!
These are shadow reserves. They don’t exist on a balance sheet, but managements know they are there and know they need to raise pricing before IANS emerge from the shadows and get reported as case. It is the specter of IANS reserves that is driving pricing behavior but just like Volde****, we must never speak of them out loud or we may have to admit that our loss picks are probably going to develop adversely.
Before anyone starts saying I am accusing companies of lying in their financials, no, I am not doing that. I am saying this is part of what happens under the self-graded exam approach. It is easy to punt on hard decisions because there is not enough clarity yet to make a call.
And for those who want to tell me that’s not how it works, well, how do you explain all the recent acquisitions where the buyer ends up adding a bunch of reserves to the target (ideally in the closing transactions so no EPS impact!) or buying an ADC for protection? Yet, peer companies of that target who remain independent continue to reserve as usual!
Clearly, there must be a quantum reserving effect where measuring reserves after they have been acquired changes them from the continuous state they were in when independent. My discovery of this effect might qualify me for a Nobel!
A Recent IANS Example: Jebi
Normally, I think of IANS referring to casualty or specialty lines. Certainly asbestos in 2000 was IANS. D&O is a great example where one knows post a bear market that losses will spike but you can not identify from where so you leave your pick alone and wait for the cases to emerge.
However, we just had an interesting example in property. Typhoon Jebi saw some early adverse loss development. There was plenty of speculation that the ultimate loss could be much worse than the original consensus. Yet, it was easy for an individual reinsurer to be in denial. “We haven’t been put on notice” or “it’s not going to reach our layer”. Plenty of plausible deniability. Plausible deniability is textbook IANS.
Reinsurers knew they were likely to see additional claims but they could kick the can due to a lack of evidence. Then, renewal came and still no new claims! Everything was great! Maybe there were no IANS?
Then, a funny thing happened. Once renewals came through with modest price increases, new claims emerged. Oops! So much for no IANS. Instead, we had Incurred and Never Spoken, But Now Reported = IANSBNR (sorry, had to do it)! The point though is IANS can only be deferred for so long. Eventually, you have to face up to reality. In Jebi, it didn’t take long. In commercial casualty, it might take another two or three years.
What Is There to Fear?
OK, let’s wrap up that aside and get back to the main event which is what will it take to really get a significant pricing turn? What is there to be afraid of other than creepy clowns or stalkers in hockey masks?
I don’t think any kind of natural disaster will do it. Even the Great Seattle Earthquake will be a property only event. It’s going to take something man made.
The most obvious candidate is a mega cyber clash event that exposes correlations across corporate America. Many cyber carriers are taking a lot of frequency risk (because they are afraid of being wrong on severity) and it’s not clear they have a good handle on their clash potential.
The other one I have is an attack on the electric grid, whether it be something like a massive solar flare or an electromagnetic terror attack. Anything that causes a lengthy disruption to the grid (lengthy meaning days not hours) exposes all kinds of BI risk not to mention the potential risk of lost financial data and other electronic information.
Are events like this likely? No. Does that suggest we are likely to maintain this status quo for some time with mild pricing and reserve charges looming in the distance? Probably. One might even suggest the real thing to fear is that nothing changes.
What Have We Learned?
1) The path the industry is currently on resembles the story of the late 90s…deteriorating underwriting standards, a spike in large losses, excessive merger premiums, and an attempt to reform by raising prices gradually before everything hits the fan.
2) There is no historical precedent for the industry self-correcting without pain. The closest analog is 2011 though that had different drivers so is not the most likely scenario in this environment.
3) The difference between reality and reported financials is greatest at the late stage of the cycle due to the unbooked IANS shadow reserves.
4) The prisoners dilemma behavior will likely continue barring a mega cat. This only serves to delay the ability to get the necessary pricing.
Conclusion: It is now officially late cycle. The good is that means we are closer to a positive inflection. The bad is it means earnings are overstated.
Is it possible the industry follows a 2011 like path and proactively continues to raise pricing while spreading out the reserve pain? Yes, it’s possible. Is it likely? No, it is not likely.
As an investor, the key decision to make is do you think the benefits of improved pricing (via better multiples) will come before the pain from lower earnings emerges due to the need to rebase picks? Or will the bad news come first and it is better to buy the stocks at their resulting lows in front of the resulting pricing response to the admission of trouble?