Guest post: Buffett’s climate certainty

From Kate Mackenzie, former Alphavillain and current climate-finance think-tanker

______________

Warren Buffett’s annual letter last week badly lets down any reader hoping to understand the implications of climate change for the general insurance and reinsurance sector.

If Buffett had said climate change impacts are not a problem for ‘his’ insurance companies, because his managers are managing the risks thusly, that would be fine. It’d also be a fascinating read, if it went into some detail — unlikely though, because that would reveal competitive information.

Unfortunately he chose to apply it to all of the insurance sector:

When you are thinking only as a shareholder of a major insurer, climate change should not be on your list of worries.

In fact, climate change is a legitimate subject of inquiry for anyone who invests in or supervises an insurance company or a reinsurer. It certainly can materially raise risks to profits and even solvency if the insurer isn’t adequately assessing their climate risk (and arguably, even if they are, it raises the risk profile overall).

Before we explain why, here are the paragraphs in which Buffett reflects specifically on climate change and Berkshire Hathaway’s insurance businesses:

Think back to 1951 when I first became enthused about GEICO. The company’s average loss-per-policy was then about $30 annually. Imagine your reaction if I had predicted then that in 2015 the loss costs would increase to about $1,000 per policy. Wouldn’t such skyrocketing losses prove disastrous, you might ask? Well, no.

Over the years, inflation has caused a huge increase in the cost of repairing both the cars and the humans involved in accidents. But these increased costs have been promptly matched by increased premiums. So, paradoxically, the upward march in loss costs has made insurance companies far more valuable. If costs had remained unchanged, Berkshire would now own an auto insurer doing $600 million of business annually rather than one doing $23 billion.

Up to now, climate change has not produced more frequent nor more costly hurricanes nor other weather- related events covered by insurance. As a consequence, US super-cat rates have fallen steadily in recent years, which is why we have backed away from that business. If super-cats become costlier and more frequent, the likely ‘ though far from certain ‘ effect on Berkshire’s insurance business would be to make it larger and more profitable.

As a citizen, you may understandably find climate change keeping you up nights. As a homeowner in a low-lying area, you may wish to consider moving. But when you are thinking only as a shareholder of a major insurer, climate change should not be on your list of worries.

Let’s start with what he got right.

Many people, when thinking about “climate change” and “insurance” without knowing much about either topic, will jump to one of two conclusions:

Either: insurers are gonna be smashed by climate change!

Or (less often): insurers… they really must *get* climate change!

Buffett is right to point out that insurers and reinsurers have some protections against climate change. Twelve months’ cover is the norm although some specialised policies can cover a longer term. If the risk rises, you raise your premiums. Et, voila! Your business has grown. But, but… maybe your risk levels have grown by more? The 12-month exposures and the ability to raise premiums are only potential safeguards; they don’t work if they’re not used. Buffet tells us nothing about how his insurance companies have been quantifying the impact of climate change on their losses.

A key metric for insurers covering catastrophes is “annual recurrence interval”. What are the costliest losses you are likely to be covering, with a probability of 99.5 per cent or higher? This is also known as the “one-in-200-year” loss. Tail risks are huge, particularly in catastrophe underwriting. As an insurer or re-insurer, how you estimate risks in the 99.5th percentile and beyond matters a lot, because it determines your capital levels and your premium pricing – your overall resilience, in other words. And guess what increases the probability of extremes, and also increases uncertainty and volatility in tail risk? Climate change!

Look what it can mean for capital:

For skewed distributions of losses, which apply to catastrophe losses, the increase in risk based capital requirements are much greater than the increase in mean average losses. If Mr Buffet’s insurers are only pricing increases in the average annual loss then there’s a risk they are not allowing for enough return on capital for his investors at best, or greatly increasing the probability of insolvency at worst. In either case his investors deserve more information.

Mr Buffet’s response appears to be to walk away from insuring risks:

Up until now, climate change has not product more frequent nor more costly hurricanes nor other weather-related events covered by insurance. As a consequence, US super-cat rates have fallen steadily in recent years, which is why we have backed away from that business.

That’s great in the short run, but is it a viable strategy for the long run? Is it a viable strategy for the industry as a whole? A shrinking top line can never be good for investors.

Now, let’s skip up the letter to this part on page 11:

Indeed, Berkshire is far more conservative in avoiding risk than most large insurers. For example, if the insurance industry should experience a $250 billion loss from some mega-catastrophe ( a loss about triple anything it has ever experienced ) Berkshire as a whole would likely record a significant profit for the year because of its many streams of earnings. We would also remain awash in cash and be looking for large opportunities to write business in an insurance market that might well be in disarray. Meanwhile, other major insurers and reinsurers would be swimming in red ink, if not facing insolvency.

So now it is glaringly clear that Buffett thinks other reinsurers (and alternative suppliers of reinsurance capital, such as insurance-linked bond investors) are pricing the risk too low. It’s hard to compete if you’re more expensive (premium prices are the main decider for people choosing an insurer — along with cute animal advertisements, obviously). Berk Hath doesn’t want to price risk that cheaply though, so they cut their business. They’re happy with their risk profile, but everyone else is continuing to underwrite catastrophe risk too cheaply ,which Buffett warns could lead to carnage.

This is all fine, but then why, a couple of pages down, he says “investors in big insurers” shouldn’t worry about climate change? Capital markets are a factor in driving down prices, but Buffett himself says the relatively benign claims levels have been a factor.

You may well be wondering how we can say that climate change is a factor when Buffett also says “up until now, climate change has not impacted on claims…”

Don’t be misled. He’s not saying that costly climate impacts aren’t already being felt anywhere, or that it’s not raising costs for insurers anywhere in the world. In fact, there is extremely high confidence that climate impacts are being felt already. As of quite recently the role of climate change can be fairly confidently identified in individual events — as the UK floods demonstrate. Buffett is talking about insured, US, catastrophe losses. Hurricanes are particularly difficult to link to climate change just now, compared to heatwaves, droughts, and changes in rainfall. Buffett is also presumably talking about US commercially-insured losses which exclude a lot of flood damage, as properties with high flood risk have long been able to get cheap insurance from the federal government.

In fact, worldwide catastrophe losses have been unusually high over the past decade or so.

Those numbers alone doesn’t implicate climate change; lots of other factors determine insurance losses from catastrophe and extreme weather. One huge factor is the value and the location of what is insured. More people living in more vulnerable living (coastal living, hello) and owning ever-more valuable stuff are a very big part of it. Inflation is also a factor.

It’s possible to “normalise” for all those factors, and you still are left with an increase over the past decade; some of which could be down to natural weather patterns like El Nino and La Nino. Or it could just be a really unlucky, bad, tragic, but statistically plausible decade?

Let’s assume at least some of those normalised catastrophe loss increases aren’t from natural weather patterns. As climate change is affecting us already, it’s probable that some of it this increase is climate-related.

In 2014 a couple of S&P analysts conducted a small experiment: what would it mean if the past decade’s biggest loss for each reinsurer was in fact a typical 1-in-10 year loss, going ahead?

Due to the complexity of climate systems, there is significant uncertainty about the impact of climate change and how that impact will develop over time. Nevertheless, we consider that the possibility that climate change is already affecting the frequency and severity of extreme events cannot be ruled out, even if it cannot be quantified exactly.

To understand the effect climate change could have on reinsurers’ financial strength, we have analyzed a simple scenario based on the assumption that the catastrophe experience over the last 10 years indicates the current probabilistic distribution of extreme events–in other words, that it represents a typical decade.

In otherwords, they tested out what it would mean if the past decade is the “new normal”.

They added a lot of caveats, but the limitations of the experiment don’t negate its power:

We recognize that 10 years is a short period from which to build a robust probabilistic view and that using longer time periods (50-100 years) leads to more robustly calibrated stochastic models. However, it is possible that the risk of extreme events may be underestimated if most of the data comes from a period when climate change had less of an effect on the underlying weather risk.

Reinsurers, they found, aren’t adequately taking that possibility into account. They’re still capitalised like it’s 2004; as if Hurricane Katrina, Superstorm Sandy, and the Thai floods were aberrations, the like of which we shouldn’t expect to see in the next 12 months or 10 years. But if the next decade IS like the past one:

On average, our scenario’s one-in-10-year loss is around 50% higher than the one-in-10-year loss reinsurers are modeling. (Most reinsurers are in the range 0% to 80%). We then extrapolated from the one-in-10-year loss to find the one-in-250-year loss under the scenario, because we input this level into our capital model to assess a reinsurer’s capital needs for catastrophe risk. We estimate that on a gross basis, reinsurers may be understating both the one-in-10 and the one-in-250-year loss by the same magnitude (around 50%).

So Buffett is correct in saying that anyone can raise their premiums in response to a change in events. But the S&P paper suggests that at least some of them haven’t reflected this in their capital levels.

If an insurer has incorrectly estimated the scale of their one-in-200- year loss, they will be undercapitalised and they quite possibly will be under-pricing risk. Which – funnily enough – Buffett agrees is happening now in catastrophe risk!

How does an insurance company investor know whether or not the company has incorporated robust and up-to-date climate science to make sure their pricing and capitalising correctly from one year to the next?

That’s difficult. These companies are not prone to disclosing detailed information about their climate modelling and assumptions. Hence the shareholder resolution.

It’s not just about catastrophe!

Buffett was commenting in response to a shareholder resolution proposal which mentioned different types of climate risk; yet his letter only addressed one type of risk – catastrophe losses, via extreme weather and natural disaster. He should surely be aware that climate change doesn’t only mean increased catastrophe loss risks for insurers. Several types of risks were spelled out very clearly by the UK’s Prudential Regulatory Authority back in September – and mentioned in the supporting text of the resolution:

— physical risk (catastrophes would be the most obvious manifestation, but other losses might include, say, health insurance, or massive disruptions in the globalised supply chain).

— transition risk; for example from coal miners’ share prices crashing or technological change destroying the value of some polluting assets (this is the exposure from investments; insurers are highly-capitalised and so they rely to some extent on investment income from this capital).

— lastly, there’s liability risk: company directors and officers are sued for failing to act on risks related to climate change which they should’ve reasonably been aware of. That’s where it gets very interesting, and where most people (including apparently Buffett) have not yet done a lot of thinking. This is an emerging risk, but as the PRA points out, asbestos-related liabilities could illustrate the problem:

Although the health dangers of asbestos were already well known, it was not until 1985 that asbestos bans were first introduced in the UK and that insurers imposed comprehensive asbestos exclusions on US liability risks. This meant that for a long period until then, insurers of employers’ liability in the UK and product liability in the US had covered asbestos risks. This, coupled with the very long latency period for mesothelioma, drove the insured losses from small beginnings to net asbestos losses of US$85 billion in the US today. (97) By comparison, catastrophe losses from Superstorm Sandy (2012) in the US are estimated at US$20-25 billion. (98)

Buffett’s argument about limited periods of exposure doesn’t apply to liability risk. Unlike most other forms of insurance, claims can be made against an insurer long after the alleged transgression occurred. So the insurer’s exposure to liability risk hangs around for many many years.

Again, Buffett has other things to talk about. But don’t take his comments as “climate change doesn’t affect insurers”. He’s not even been very convincing about how Berkshire Hathaway is managing climate risk.

He’s correct about insurers writing 12-month exposure. But saying that climate change isn’t a concern for big insurers is an extreme overreach.