partner login

BMS News

2017 losses, long-term weather trends and their insurance consequences

Catastrophic weather events in 2017 left the re/insurance business with a conundrum. Record-breaking insured catastrophe losses exceeded any previous year. However, the impact of those unprecedented losses on re/insurance pricing has been less than many had hoped for or expected. Claims of nearly $140 billion have not delivered a traditional hard market.

To solve the conundrum and learn the year’s lessons, we must view the storms, fires, and floods of 2017 in a long-term historical perspective. We should not use loss data alone to define weather and climate trends; we need to also understand the impact of the growth of global wealth, as well as patterns of weather extremes.

When we do so, we find that the record-breaking losses were not so enormous, and could have been worse.  About $92 billion of 2017’s insured catastrophe losses arose from three major hurricanes according to the latest Swiss Re Sigma Report. When indexed to 2017 values, Hurricanes Katrina, Rita, and Wilma in 2005 caused insured losses of $112 billion.

We also need to consider that extreme weather memories are typically short. Catastrophe models that simulate the impact of historic storms in today’s built-up environments show that  events such as 1926’s Miami Hurricane or the 1928 Okeechobee Hurricane would cause insured losses that surpass those of 2005 and 2017 hurricane losses. When other named storms and natural catastrophes that occurred worldwide during those years are included, the insured impact on our modern built environment would easily exceeds $140 billion.

Bay Shore Drive in Miami then 1926 after Hurricane and Now. Source NOAA and Google Street View

Meanwhile there is a perception that extreme weather patterns may be changing.  In some cases the data supports this perception.  It is likely that, since 1951, the number of statistically significant regional increases in heavy precipitation events is greater than the number of declines. Strong regional and sub-regional variations modify the trend, but in short, when it rains it often rains more. However, despite the increase in these extreme precipitation events, little evidence suggests the rainwater has led to an increase in floods, with lots of regional variations.

The U.S. Climate Extreme Index quantifies the observed changes in one-day precipitation extremes across the U.S. Since 1990, there has been an increase in the area of the U.S. that has experienced a one-day extreme rainfall.

Tropical cyclones seem not to be changing. After accounting for past shifts in observational capabilities over time, the best records that started in the 1970’s suggest, globally, there are currently no significant observed trends in the number of named tropical cyclones, but the data is suggesting that, when named storms form, they have become stronger in recent years.

2017 had 18 total landfalling named storms. 14 of these had a category 1 or 2 and four of the three major categories were in North America.

As for severe weather, poor data quality makes conclusions about long-term trends problematic. Observed trends in small spatial-scale phenomena such as tornadoes and hail can be made with only low confidence levels, but more recently the data is suggesting clearly a downward trend in the number of occurrences of major tornadoes.

With all this in mind, we should learn lessons from the 2017 experience.

  • Whether it was the major wildfires in California, flooding of Harvey, the extreme winds of Irma over south Florida, or one of the many damaging hail storms; exposure management is very important and there needs to be a refocus on this simple task to limit losses.
  • Valuable data will be collected from the 2017 events, especially from the high wind speeds from HIM. This data will fill critical gaps in the historical record and will be valuable both meteorologically and financially, as it will help insurers and customers by providing better loss models, and lessons for loss mitigation and resilience.
  • Building codes matter. When wind-speed data is combined with claims data, we get critical knowledge of how high winds damage different types of construction. We now have validation for some wind-tunnel testing which has been lacking due to the lack of hurricane landfalls over the last decade.
  • Hurricane forecast track and intensity still have a great deal of uncertainty. The 2017 landfalls raised questions about the value of early landfall loss projections, and how they could be used.
  • Effectively, the people who need flood insurance do not have it.  How many more flood events like Harvey and Katrina do we need before more action is taken?
  • Companies with the best business interruption (BI) insurance came out on top after an event, but few small businesses don’t have enough BI.
  • Demand for certain post-event services seriously tested supply. The lack of adjusters, for example, created a major bottleneck that led to very high loss adjustment expenses.
  • Given the long 10 years without major hurricane landfalling activity, 2017 was a great systems test for InsurTech.  It would appear no major lapses in these new technologies have occurred which should further expedite system enhancements in InsurTech.

BMS Seasonal Outlook April 2017


  • An El Niño is forecasted to emerge for late this summer, but weather patterns suggest that it has already arrived.
  • Late spring-early summer warming will occur over the central and eastern U.S.; then, temperatures will trend cooler into the summer for the northern plains.
  • Heavy spring rains across the Gulf States and into the Great Plains will accompany storm systems, but nothing like past record years (like 1993).
  • Best matching analog years: 1982, 1986, 2004, 2006 and 2014

The Pattern

There has been no shortage for the insurance industry of severe weather during the first quarter of 2017. A warmer-than-normal start to the year, aided by record Gulf of Mexico sea surface temperatures, has produced a relatively active period of severe weather. Will this weather pattern continue into the next few months, or are we on the road to a quieter-than-normal weather pattern?

Severe weather reports from Jan 1 – April 20, 2017. Bar graph shows the number of reports per day so far this year.

In reviewing the first three months of 2017, the weather patterns have been dominated by Pacific influences, with storm after storm pounding the West Coast. As these storm systems moved across the U.S. and into the central U.S., they were enhanced by tapping a large amount of warm air from the Gulf of Mexico, which also produced warmth across the eastern U.S. This weather pattern has provided plenty of opportunity for severe weather, and resulted in several significant severe weather outbreaks that impacted the insurance industry. Some of the bigger events occurred on January 2, 21-23, February 28-March 1, March 6-7, and most recently, April 2-3.

So far in April, it would appear this weather pattern has continued the pattern established in March, with a series of infrequent, but very energetic storm systems digging into the western U.S. before lifting up into the mid-west and northeastern U.S.  This has meant that much of the U.S. should continue to experience similar above-normal severe weather activity; but these storm systems should start traversing a more northern track across the northern-tier states.  This pattern indicates that June-like weather might appear in April and May.

Long-Range Forecasts

Many long-range climate forcer signals can provide seasonal forecasters clues about what weather to expect over the next few months.  The El Niño-Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), and Atlantic Multidecadal Oscillation (AMO) are some of the most common signals.  Although ENSO seems the most popular climate forcer in recent seasonal forecasts, a few words of caution are required:  fickle spring weather is notoriously misleading and results in difficult forecasts (this phenomenon is so familiar that it has its own name, the “spring predictability barrier”).   Second, we are technically coming out of a weak La Niña (which has a traditionally predictable outcome), but the atmosphere isn’t following its typical pattern following a weak La Niña.  Instead, the weather pattern over the continental U.S. reflected more of an El Niño pattern, with widespread warmth across much of the country and above-normal precipitation across parts of the southern tier and West Coast states.

The PDO has been positive for a record 39 months (during a positive phase, the west Pacific becomes cooler and part of the eastern ocean warms), but the weather pattern over the western U.S. has been anything but reflective of a positive PDO pattern.  Usually a positive PDO will lead to warmer and drier conditions in the western U.S., but this has not been the case over the last few months.

Weekly U.S. Drought Monitor, which uses a composite index on the level of drought that is occurring

Seasonal forecasting can also help us understand current areas of drought.  Large areas of moderate-to-severe drought can create a positive feedback loop, meaning that already-dry regions are more susceptible to warm and dry weather than non-drought areas.  As illustrated by the above map, patches of drought are currently scattered throughout the country, but these scattered patches are too insignificant to raise red flags for long-term warm and dry conditions over next few months.

Research suggests that the continued warm water temperatures in the Gulf of Mexico will likely mean severe weather will continue to erupt over the southern and central U.S. into late spring.  Historically, when the Gulf of Mexico trends warmer than normal, there is more energy to fuel severe storms and tornadoes.

Convective energy is needed for storm development and correlates to storm activity.

The Verdict

Evaluating recent weather patterns and various climate forcers produces the following analog years: 1982, 1986, 2004, 2006, and 2014.  These years suggest an active storm pattern that may result in:

  • Much of the U.S. experiencing near-normal temperatures between April-August
  • The western U.S. and southeast warmer then normal
  • The northern states slightly colder than normal

Combining the analog years yields the following temperature and precipitation anomalies. Note the temperature scale is less than a 1 degree +/- long term average.

However, above-normal national temperatures now and into early summer should give way to more normal temperatures or cooling temperature patterns in the Great Plains later in the year.  Hit-or-miss precipitation across much of the country will be a by-product of severe weather and will provide rain in some areas but not others.  As a result, dry conditions in the southeast could progress into summer.  Severe weather should remain active until May, with activity waning to more normal levels as summer progresses; but overall, the insurance industry should expect to see much higher levels of insured loss than in the last few years.


50th Anniversary – Insurance Retrospective of May 6, 1965 MSP Tornadoes

50 years ago today, the upper Midwest was devastated by an outbreak of severe weather. A similar storm today would result in what might be one of the largest-ever severe thunderstorm losses to the insurance industry. For four consecutive days in May 1965, severe weather, including 37 significant tornadoes, of which at least nine were major tornadoes, affected much of the central United States. The most devastating part of the outbreak occurred when six tornadoes swept across the western and northern portions of Minneapolis/St. Paul. Of the six tornadoes that swept the seven-county region, four were rated F4, one was rated F3, and the other produced F2 damage.


Shown is a photo of a tornado crossing to the west of the junction of highways 7 and 101 on May 6, 1965 (taken by Minnetonka resident. H. B. Milligan). It is believed that this was the tornado that touched down in Chanhassen at 6:27 p.m. and dissipated in Deephaven at 6:43 pm.

The outbreak was significant in several ways. First, it occurred just three weeks after the historic Palm Sunday tornado outbrea, which no doubt left local residents feeling a heightened sense of alarm about storms. Second, it was the first time in Minneapolis/St. Paul history that civil defense sirens, now used nation-wide, were used to warn people about severe weather. The U.S. Weather Bureau issued quick and successful warnings that were widely communicated by local radio and television stations. These significant new communication methods likely saved many lives, despite the fact that the storms killed 13 and injured 683. Finally, the tornadoes served as yet another example that helped dispel the myth that tornadoes won’t hit a city, a lake or a river.

Historical accounts have focused mostly on the tornadoes during the outbreak because of their remarkable pattern: at a single point in time three tornadoes were on the ground, and two tornadoes hit the same location just hours apart. Some of the hardest-hit cities included Chanhassen, Deephaven, Fridley, Mounds View and the Spring Lake Park suburbs, all of which have undergone tremendous growth since 1965. It is estimated that 600 homes were destroyed during the series of storms, and over a thousand more were damaged. Today, the affected suburbs have expanded from what was farmland 50 years ago, to what is now primarily residential.

The increased number of homes and the increased value of homes in Minneapolis/St. Paul suburbs also increase the potential for catastrophic damage from a similar future event. In some cities, the type of housing exposed to storm perils affects the potential loss severity. Two of the F4 tornadoes touched down in the west suburbs, and these suburbs have seen explosive growth since 1965. For example, one of the 1965 F4 tornadoes reportedly impacted 35 homes on Lotus Lake and 50 homes on Christmas Lake. This area now hosts many multi-million dollar lake homes. Home values in other nearby neighborhoods have grown significantly as well, including multi-million dollar homes that line the expansive shores of Lake Minnetonka. Fridley, Mounds View and Spring Lake Park were impacted by two F4 tornadoes that touched down just hours apart and crossed paths. Those cities may not include as many high-value homes, but the number of residential properties has grown, and now includes not only more residential properties, but high-value commercial properties as well. For example, one of the 1965 tornadoes tracked less than a mile from what is now the Medtronic world headquarters campus.

Despite the loss of life and property damage caused by the May, 1965 storms, it’s important to note that the historical data includes only limited reports of hail and damaging winds. These perils most likely also impacted the area 50 years ago, and arguably, including damage from these lesser-reported perils would increase the financial impact of the storms.

Although there are large uncertainties on exact track, width, and strength of all the tornadoes that occurred 50 years ago across the seven-county metro region, BMS Analytics digitized these historic tracks to create a deterministic scenario in an attempt to understand the potential impact of a similar storm today. The scenario suggests an insurance industry loss between $10 – $14 billion dollars. This is equivalent to a return period of 2,500 – 7,500 years according to various catastrophe models.


BMS Analytics digitized deterministic tornado scenarios with four perturbation to account for track uncertainty

Recently we have seen images of single tornadoes hitting cities like Tuscaloosa, AL; Joplin, MO; and Moore, OK and causing billions of dollars worth of insured damages ($6 Billion in total insured loss). Recalling the events of early May, 1965, is a reminder that several major tornadoes can hit a large metropolitan area such as Minneapolis/St. Paul on the same day. Although unlikely, a similar event would be extremely damaging due to the ongoing population and wealth growth described above. And if such an event occurred, it is likely a similar outbreak would have a large impact on the insurance industry.

Northeast snowfalls and insured losses

Over a decade ago, there were several claims that “snowfalls are now just a thing of the past.”  Even as recently as February of last year, The New York Times published an article titled “The End of Snow.” These commentaries predicted that snow would soon be a distant memory, and our children and grandchildren would never see it, except in photographs.  These claims may alarm people, but the data suggests otherwise, particularly along the East Coast of the United States.

One of the most common methods to examine the impact of a winter storm is NOAA’s National Climatic Data Center Regional Snowfall Index (RSI), which uses the area of snowfall, the amounts of snowfall and the number of people living in the snowfall area to quantify the societal impact of a snowstorm.  In the Northeast, the RSI is also known as the Northeast Snowfall Impact Scale (NESIS), and the values are both a raw index value and a categorical value from 0 through 5, much like the Saffir-Simpson Hurricane Scale or Fujita Tornado Scale (i.e., the more snow that falls over a large populated area, the greater the impact on a population and the larger the number.)  This data shows that, over the last decade, there has been a trend toward an increase in high-impact snowstorms along the East Coast.

NESIS Events

Northeast Snowfall Impact Scale (NESIS) Events

The New York City metropolitan area has the largest population in the U.S., and this area has experienced five of the top 10 snowfalls on record dating back to 1869.  However, not all large East Coast snowstorms result in large insured losses.  For example, New York City’s biggest snowfall occurred on February 11 and 12 of 2006, dumping a total of 26.9 inches.  Yet this event was never declared a PCS event (insured losses over $25 million).  Long Island’s biggest snowstorm in history was the result of a Nor’easter on February 8, 2013 that dropped 33.5 inches of snow in Medford, NY, but this storm did not receive a PCS designation either.  Of the 84 documented NESIS events since 1960, 34 have been a Category 3 (Major), 4 (Crippling) or 5 (Extreme), but only 50% have resulted in PCS losses.  The extent of the losses from the latest snowstorms that have impacted the Northeast, which were preliminarily rated between a Category 2 (Significant) and 3 on the NESIS scale, have yet to be determined.  However, we do know that winter losses are a growing area of concern for the insurance industry due to the fact that, typically, the first quarter is a stable period with little catastrophe loss.

There appears to be a continued lack of understanding around winter storm losses, which is why I am speaking on the subject at the RAA Cat Risk Management Conference in Orlando, FL.  In this presentation, I will dive deeper into the topic and provide insight into the trends and hidden issues that often result when winter storm losses fall below the retention of a normal catastrophe program.  Some of these same issues might emerge due to the latest Nor’easter/blizzard of January 26 – 27, which presented cases for business interruption insurance, CBI coverage loss, and insured loss around ingress/egress and civil authority actions due to shutdowns.

Unusual Weather we’re Having, Ain’t It?

I have been saving this title for awhile, and with the recent 75th anniversary of the release of The Wizard of Oz, in which the cowardly lion says this line as he notices the fallen snow on the poppy field, I find it a fitting start to a discussion about extreme weather. Interestingly, this might also be the first case where a blockbuster movie promotes the idea that average weather can manifest into “extreme weather,” such as a garden-variety tornado in Kansas turning ugly and transporting people to alternate universes.

Images courtesy of Warner Bros. Entertainment

As a meteorologist, I often run into self-proclaimed armchair meteorologists all the time. It has never been easier to get weather information via a blog, Twitter, or on television, which now has at least four cable channels devoted solely to weather. Because weather impacts almost everyone on a daily basis and changes often, it is closely watched. However, with this accessibility of information, one can easily become brainwashed with the idea that normal weather is somehow extreme.

The Media Research Center has just released what I think is fascinating research. The Center analyzed broadcast television network transcripts for morning and evening shows looking for stories using the phrase “extreme weather” between July 1, 2004 and July 1, 2005, and also between July 1, 2013 and July 1, 2014. Ten years ago, ABC, CBS, and NBC barely used the phrase. Now, its use is prolific, despite scientific disagreement regarding extreme weather trends, as discussed in the most recent Intergovernmental Panel on Climate Change Fifth Assessment Report (Chapter 2).

According to the Media Research Center, between July, 2004 and July, 2005, the three networks only used the phrase “extreme weather” in 18 stories on morning and evening news shows in that entire year, even though there were several opportunities to use the phrase when reporting on the 13 named storms that impacted the U.S. during that period.

Now, the familiar phrase of the networks, “if it bleeds, it leads” has taken a backseat to “extreme weather.” In the past year (July, 2013 through July, 2014), the same network news shows discussed extreme weather 988 percent more often, in a whopping 196 stories. That is more than enough stories to see, on average, one every other day. Here is a short video montage to illustrate:

This is despite lower occurrences of severe weather (e.g., hail, wind, tornado) and hurricanes than were observed during the same period 10 years ago.

The Media Research Center study states that “extreme weather” was frequently used by the networks to describe fairly normal weather events, such as heat waves, droughts, tornadoes, hurricanes and winter storms, and they often included the phrase in onscreen graphics or chyrons during weather stories. ABC even has an extreme weather team, dedicated to covering such events. We also get footage from storm chasers that make a living driving into the worst weather.

Since some people still read the old-fashioned newspaper, let’s analyze the 162-year history of the New York Times, which can be done using a tool for graphing the frequency of use of certain words and phrases called the Chronicle.

It is interesting to note that the 1933 hurricane on Long Island or a major drought in 1988 were not considered extreme weather events. The disproportionally high use of the phrase “extreme weather” started after 2005.

The publishing of news is inherently an ephemeral act. A big story will consume public attention for a day, a month or a year, only to fade from memory as quickly as it erupted. There is no doubt that weather events get more attention in this day and age of instant communication and technology, and the speed with which this information is shared certainly has an influence on how people think. It is important to remember that extreme weather is completely natural and there will always be extreme weather somewhere, as the atmosphere is in a constant battle to reach equilibrium. In fact, it is less likely to have a day that is perfectly average than to have one that is one or two standard deviations above or below the average. However, the use of the phrase “extreme weather” in the media occurs with alarming regularity and is undoubtedly influencing the insurance industry.

2013 Half Year Review – U.S. Extreme Weather Events

Andy Siffert, BMS’ resident Meteorologist, reviews the first 6 months of 2013 in terms of U.S. extreme weather events and their impact on the industry.

As we round the corner into the second half of 2013 we can now put into perspective some of the U.S. extreme weather events that occurred during the first half of the year. With the tally of some of these disasters still being assessed, the U.S. insurance losses estimated by Property Claims Services (PCS) will continue to rise. As of July 1, 2013 the U.S. has seen $6.8 billion in PCS claimed losses from weather events across the U.S. Considering the expected upward adjustment of claimed weather events, losses reported thus far would fall below the five-year average for first- and second-quarter weather-related losses, which total $13.1 billion. This below-average loss is primarily connected to the current “Tornado Drought” that has been ongoing since the second half of 2012. Severe convective storm outbreaks in May 2013 produced major tornadoes causing widespread damage to properties in Texas, Oklahoma, and other states. But as of July 1, the tornado count is 42% below the five-year average, with a major portion of the tornado activity occurring in the lower Mississippi and Tennessee River valleys. Given that May is peak tornado season in the Central Plains, it should be no surprise that strong and violent tornadoes formed and caused damage there. In Tornado Alley this typically occurs during the second quarter of the year, but the number of tornadic weather events in the Central Plains and Midwest regions has been below normal again this year.

The overall lower PCS loss numbers could also be a result of fewer hail events, which, according to Storm Prediction Center (SPC) storm report data, are currently 21% below normal (with only 3,714 hail reports). With the main drivers of severe convective storm losses resulting from the May 20 tornado in Moore, OK and overall hail reports below the five-year normal trend, it seems that derecho or straight-line wind events are the likely driver of most U.S. weather-related losses to-date. These events appear to be trending with the five-year SPC severe wind reports, which as of July 1 stand at 7,360 vs the five-year mid-year average of 7,369 severe wind reports.

The Black Forest wildfire in Colorado appears to be one of the most destructive fires in Colorado’s history. Because of this, wildfires have been getting a lot of media attention lately and it might be interesting to put the current wildfire season into perspective.

According to the National Interagency Fire Center, the U.S. is about a million acres below the 10-year running mean of 2.4 million acres burned in the 22,050 wildfires that have been reported. This is also 15,000 fires fewer than the 10-year running mean. In fact, in 2013 there have been fewer fires than in any of the last 10 years, and the year stands next to last in terms of acres burned.

Like the tornado season, so far the fire season has been well below normal. The Black Forest wildfire in Colorado and the recent deaths of 19 fire fighters in the Yarnell Hill, Arizona wildfire are examples of fires that stick out like a sore thumb in a below-normal wildfire season – just like the two late-May tornadoes which were exceptions to the trend of the overall tornado season.

It is my understanding that in both the Black Forest and Yarnell wildfires, areas burned that had not burned in the previous 40 years – which has to be a major factor contributing to the wildfire catastrophe. The media would say the fires are due to dry conditions, which definitely exist and in some cases are extreme. But if it had been a wet spring, then more fuel would have been available as the summers always see drier conditions in the southwest. The old saying, “Pay me now or pay me later” applies here: If it’s wet, the resulting new growth will eventually dry out and die. And if it’s dry and dead, it will eventually burn.

Worldwide, recent catastrophes seem to be focused largely on flooding-related events, with the notable events originating from the remnants of Tropical Cyclone Oswald that triggered severe flooding in Queensland and New South Wales in Australia. More recently, flood losses that impacted a large area along the Elbe river basin in Europe will likely surpass the 2002 European flood losses. In North America, heavy rainfall provoked catastrophic flooding in southern Alberta, Canada – which will likely go down as the largest flood-related loss ever experienced in Canada. However, with the 7th-latest start to the typhoon season, few typhoons have resulted in flooding or the kind of disasters typically seen in Asia. In fact, global Accumulated Cyclone Energy (ACE) is still stuck in the lowest range – which began in 2007 and is similar to the 1980s. Before Super Typhoon Soulik was upgraded on July 10 to a major 96+ knots tropical cyclone, the last major tropical cyclone, Sandra hit just east of Australia on March 11. And the clock is still ticking on the 2,811 days since the U.S. was last hit by a Cat3+ hurricane – the longest such period since 1900, if not before.

Overall it would appear there is a silver lining – because extreme weather events could be worse based on past years, plus you can’t control nature. Most often, catastrophic events like the wildfires, tornadoes and floods of 2013 can be tied to events of similar magnitude that occurred in the past. We are building bigger towns in locations where catastrophic events have occurred in the past, and the understanding of changes in population, income and housing units can often explain the increase in loss.