We are writing this piece shortly after the bombings in Brussels. Once again terror has struck its ugly hand in a major city resulting in the death of innocent people.  The videos of blood spattered survivors being led by rescuers through familiar airport concourses and Metro platforms that we ourselves have traversed in the past have the effect desired by the perpetrators. Outrage, sadness and, yes, fear too, which is the ultimate goal of the killers.

It seems as if there is no safe harbor from this emotional disruption. The UK "Brexit" vote, with the possible future of Europe teetering in the balance, looms in less than three months. The US Presidential campaign, with seven more long months until the election, has already devolved into a spectacle leaving most watchers recoiling in embarrassment. Military tension in the South China Sea seems to be increasing with each passing week and the global economy has been shifting up and down like a yo-yo since the start of the year.

In the midst of this carnage and disruption we encountered Warren Buffett's annual letter to Berkshire Hathaway shareholders. We will discuss it later but anything Buffett writes is worth reading and particularly so from the insurance perspective. In this case though we derived the most satisfaction simply from the man's unquenchable optimism about the future. It was a good anchor to rely on while the TV screens were flashing images of Brussels.

We can't predict the future but this next bit of news might be something to remember. A regional German far-right neo fascist political leader, who had campaigned for office calling the influx of refugees entering Europe "lawless primates" and "benefit-scrounging tourists", was involved in a car accident on March 16th. He was seriously injured. 

The first people to arrive at the accident scene were two vans filled with 16 Syrian refugees who were passing by.  The vans stopped. Two of the refugees ran to the car and dragged the man out and administered first aid until the ambulance arrived. The political party represented by the injured man issued a statement saying that the Syrians had "likely performed a very good, humane deed."

It's possible that the actions of the rescuers can show us that extremists on either side of the fence, who paint enemies as entire cultural, religious, national or racial groups, are dead wrong.  In this case it was proven to us by disenfranchised, homeless refugees who had everything to lose by aiding the man and staying until authorities arrived.  They stopped, stayed and helped anyway.  

There was a lot of movement on several levels this month.  In one month it was announced that the heads of Munich Re, Swiss Re and Axa were retiring.  Munich Re's CEO Nikolaus von Bomhard, Swiss Re CEO Michel Lies and Axa CEO Henri du Castries all announced their retirements.

At both Munich Re and Swiss Re the newly promoted CEO's, Joachim Wenning and Christian Mumenthaler respectively came from reinsurance backgrounds.  Munich's new CEO Joachim Wenning comes from the life reinsurance division. It's no secret that large reinsurers are looking more carefully at life reinsurance and longetivity coverage as sources of increased revenue. By naming one of the world's top life reinsurance operators as its new CEO the largest reinsurer in the world is signaling that it does not intend to step back from that strategy.

Insurers are increasingly attracted to lower reinsurance premiums as a way to move liabilities off their books.  This would seem a basic principle of reinsurance but remember that in the rush to "move down the food chain", to get closer to the risk and keep more of the premium, it's possible that some insurers ended up keeping too much of the premium --and ended up holding more of the risk at higher retention points.  

Events like Tianjin, or the trio of UK storms, are like the gifts that never stop giving. The insured losses just keep increasing.  You can assume that reinsurance purchasing decisions are back to focusing on capping losses at an acceptable level --and having someone else pay for losses above that level.  Picking up revenue down the food chain by retaining exposure can be profitable but very costly if things go wrong.

One company that makes no secret about its determination to move down that food chain is Berkshire Hathaway Specialty Insurance.  In fact when it was formed both Warren Buffett and Ajit Jain were clear that they intended to make Berkshire's enviable balance sheet available to commercial insurance buyers right down at the lowest level of the chain.

In Berkshire's case you can bet that BHSI isn't ceding any coverage out to anyone other than a Berkshire reinsurer.  Berkshire will get the whole premium, but it will cover the whole risk. Few companies can risk doing that. But as Buffett said in his annual letter "if the industry should experience a $250 billion loss from some mega-catastrophe (Tianjin, Eva, Frank and Desmond take note: you have a long way to go to reach this level! Editor's insert.) --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 earning." 

Buffett went on to say "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." Just in case you didn't get the message he concluded with "Meanwhile other major insurers and reinsurers would be swimming in red ink, if not facing insolvency."

Last July, in a Wall Street Journal story, Ajit Jain signaled that Berkshire was going to begin to cut back its appetite for CAT exposure and "move down the food chain to get closer to the risk and keep more of the premium.  A big part of that new appetite is going to satisfied by BHSI which has developed more than $1 billion in premium volume in less than three years.

Remember that Buffett's view on insurance is slightly different than anyone else's. He views the insurance subsidiaries as "float producers", that generate cash that can be used for investments. The float is the money the P&C operations receive in premiums upfront that may be needed to pay claims later on.

Float production is a big business at Berkshire. By 2015 a total of nearly $88 billion in float had been amassed.  Berkshire paid out about $24.5 billion in claims in 2015, but it was replaced by new premium. You can see why Buffett regards the float as a revolving fund.  As he said in his letter "Daily, we pay old claims and related expenses and that reduces the float." "Just as surely," he continued "we each day write new business that will soon generate its own claims, adding to float."

But don't think that the Berkshire underwriting operation is only a giant scheme to generate cash for investment purposes.  Far from it. In fact in the annual letter Buffett sets out the four disciplines that a sound insurance operation must adhere to. He emphasizes that in addition to serving as a float generating machine, his underwriting operations are finely tuned toward creating underwriting profits.  That means, at essence, that the underwriting takes in more money in premium than it pays out in claims, operating costs and claims expenses.

Sometimes you hear a comment that crystallizes many trends that are converging at once.  We think we heard such a comment when, at the A.M. Best annual conference in Arizona in March, Hamilton Insurance Chairman and CEO Brian Duperreault said data analysis could be either a threat or an opportunity for brokers in a world where advanced analytics contributed to the intermediary's value. 

As to whether the broker could actually bring such a level of value he said "I think the answer is a qualified yes-if the broker or agent brings a level of expertise and counsel that far surpasses what the carrier offers or the client can determine himself. This means setting the gold standard for manipulating and interpreting data."

There are several implications to his comments.  Mr. Duperreault clearly notes that both insurers and their clients are rapidly gaining better data analysis capabilities of their own data.  This means that the insurance buyer is aiming at better understanding his own data prior to going to the market.  And anyone reading the industry trade press is aware of the herculean efforts currently exerted by insurers to better understand the insured's data once they receive it.

What Mr. Duperreault is telling the brokers is that if they want to continue to bring value to a transaction they need to be able to understand, interpret and manipulate that data better than the owner of the data and better than the insurer who is setting rates based on that data.

This is a high bar for an intermediary to meet. It explains the vast sums invested in data analytics by the broker community.  By any measure of effort Mr. Duperreault's message, it would seem, has already been received by the broker industry.

As all of us know there's a big difference between "getting the message" and actually doing something about it.  In this case well capitalized insurers with armies of actuaries and analysts have long since jumped into this race with both feet.  They've been joined by players further down the risk chain including larger MGA's and risk management units of large commercial insurance purchasers.  Everyone is seeking a better understanding of the risks represented by the data.

At CATEX we began to hear this data drumbeat several years ago. Our Pivot Point System software provides transaction and claim detail down to the lowest granular risk level. We had assumed that underwriters and analysts would be thrilled with that level of extreme granularity so that virtually "atomic" accounting and underwriting could be applied to each specific risk processed through the system.

They were, in fact delighted.  But it developed that clients generally did not have the granular risk data to load into our systems. All too often our systems were presented with only aggregate risk data by clients with the explanation that they simply had no more detailed information.

 

But when we raised our noses from the insurance and reinsurance grindstones it seemed that everywhere we looked disaggregated, granular data was available to consumers. Banking, online retailers, reservation websites, Amazon, etc. were force-feeding as much granular data to consumers as they could ingest!  If the start of the insurance rate building process is based on understanding the potential perils to which any one risk is exposed, then surely it would only be a matter of time until insurance demanded that granularity?

As Brian Duperreault alluded to that time has now arrived.  If we needed more confirmation we read remarks by Robert De Souza, CEO of Aon Benfield Australia and New Zealand who said the "majority of insurers now price risk at address level, and the availability of high-resolution hazard data allows rates to be calculated to each specific property."

The CATEX Pivot Point software is designed to import and process risks at the most granular level and currently churns through nearly $7 billion annually in premiums and claims. The challenge for our clients has always been to obtain risk information at the most granular level to use Pivot Point to its fullest extent.  And a significant percentage of our transaction activity is Delegated Underwriting Authority business so you can guess where this is going.

Converting placement, claim or settlement bordereau to reporting templates approved by regulators is actually the easy part.  The hard part is ensuring that the data included in those bordereaux is accurate.  Since the hinge of our system's success swings on being able to load the most granular data into our transaction system, we had no choice but to build a better way to import disparate data; cleanse it; complete missing fields; provide a full audit log of all changes and then be able to export data to any report template without the need to import the same data all over again for a different destination template.

Our Data Vera application can do all this without the need for laborious and often mistake prone mapping of fields to output templates.  Data Vera is able to learn as it goes and recognizes previous choices made in regards to a specific contract, facility or even system-wide to suggest an appropriate data value for a non-conforming data value.

The Data Vera application imports, cleanses and exports data to any desired format and is now the preferred means to import data on to our Pivot Point systems for brokers, MGAs and markets.  Our clients can be assured that the data has been cleansed to an enterprise standard level and they can direct the export of that data to any destination they desire.

To revert to Messrs. Duperreault and De Souza, there is yet another obvious use for cleansed and complete data.  CAT models, or models of any type for that matter, have been brutally exposed to the gaps caused by poor and incomplete data quality over the years. 

Data Vera can fix this problem too.  To Data Vera any model template is simply another export destination, just like the ACORD 3001 standard, XML or legacy template. A client can add a specific data template for any model as a data destination easily and quickly. 

Data Vera maintains all the data it takes in, including fields not needed by clients.  If the data is in the incoming data source Data Vera imports it, cleans it, validates it and stores it in case it's ever needed.  All the data is converted to structured data and maintained in a "data mart" ready for export to any template.

In our view having accurate data is the key to just about everything.  In fact it's so important that it's one of those things that has often been overlooked.  It's like saying you need to stop smoking and lose weight.  Everyone knows it but when bad things happen they are surprised.  We won't be surprised again.

 

Data exported from Data Vera into either a CATEX Pivot Point system, a legacy system, a competitor's system, a regulatory report, or a CAT model will be of unparalleled accuracy with a complete audit log reflecting every single data cell change that Data Vera has made. 

We've found that clients are sending that audit log, with a copy of the original data received, and the cleansed and validated data, back to the sender demanding a "sign off" so that they affirm the data is now accurate. Not only has no one responded saying the Data Vera cleansed data is inaccurate but several have now engaged our Data Vera clients to clean more of their data for business they previously weren't seeing. 

The need to have confidence in your data seems to extend pretty far down the "food chain" of risk - and why not?  Even the smallest coverholder or broker wants to have a better understanding of the business they are processing.  Neither the metal filing cabinets packed with spreadsheets, nor the desktop file with the multiple formatted Excels, can provide any help on the road to "setting the gold standard for manipulating and interpreting data" but Data Vera does.