The first bail-out: NFIP accessed by manipulation of construction price data following Hurricane Katrina

Connect the dots.  Russell left a big one in his recent post, Where insurance finally meets the big bailout – AIG a massive ponzi scheme?

They [AIG]would sign a reinsurance contract with an insurance company, take a certain amount of money, and at the same time form a side agreement that essentially said: “We will not pay you on this policy.” Why would the insurance company do this: to lower their capital requirements. And since AIG did not perform much in the way of the services, they could kick some of the money back to the insurance company via an offshore entity which would allow the insurance company to write off all the money paid off to AIG, and then accept back part of the money back as tax free profits.

CLS has been leaving dots on the ALL Board for some time – here’s one from October 2007, for example.

Connect the dots in any fashion and they lead to a  bottom line “short” on reinsurance in the face of the nation’s largest disaster; a legal storm as claims were denied; and to the scheme for a bail-out from the US Treasury via the NFIP.

In a February 2007 report, Congressional Research Service identified two broad sets of post-Katraina claims-adjustment issues that might be relevant to Congress… One set of these relevant issues is centered on the previously discussed policy language on causation; and, the other…

…is the alleged adverse impact on insureds of computerized claims settlement systems and products. Public interest advocacy groups have alleged that the insurance industry uses computer programs…to systematically underpay homeowners claims. Their argument is that these systems allow insurers to calibrate the amount of savings they want to generate to the detriment of their insureds.

XACT-ly what the Rigsby sisters said in their qui tam claim.

Other lawsuits against insurers make similar allegations.

The Xactware program itself is not at issue in this case but the price lists used to value the damage is at issue.

However, in other litigation, the program itself has been at issue. In July 2005, Xactware filed suit against Symbility Solutions alleging unfair competition.

This case relates to United States Patents No. 6,037,945 (“the ‘945 Patent’) and United States Patent No. 6,810,383 (“the ‘383 Patent). Other patents are pending.! Discovery may reveal the applicability of other patents of Xactware.

The case was dismissed on a jurisdiction issue and followed a few months later by a Complaint for Declaratory Judgment filed by Symbility in Michigan with Xactware, the Defendant, responding with a counterclaim

No doubt an earlier suit filed against Xactware by software developer Corel et al figured in Symbility’s request for the Michigan court to:

enter a judgment or decree declating that the ‘945 Patent, the ‘383 Patent and the ‘819 Patent are invalid and unenforceable… a judgment or decree declaring that the ‘945 Patent, the ‘383 Patent and the ‘819 Patent are not infringed by any products currently or previously made, used, or sold by Symbility Solutions;  a judgment or decree that it is the right of Symbility Solutions and any buyers, sellers, or users of Symbility Solutions to continue to make, use and sell such products, without any threat or other interference whatsoever against them by Xactware or any person or entity in privity with Xactware, based on or arising out of the ownership of the ‘945 Patent, the ‘383 Patent and the ‘819 Patent or any interest herein…

The case ended when the parties reached agreement with each paying its own cost – but not before an Xactware motion and Symbility response showed cost determination is a two-part process.

The disclosure describes a two-phase process including a first phase of updating  the parameters of the default polyhedron or model of the estimated room (by iteratively altering planes). In this first phase, only the dimensions are updated, not the estimated costs and materials.

Only after a query by the user are the desired material and labor estimates obtained.

Nothing in the specification discloses any automatic updating without further action by the user. Accordingly, the specification discloses only that in this first phase, only the dimensions are updated, not the estimated costs and materials.

Updating of the desired material and labor estimates does not happen automatically without further action by the user as proposed by Xactware.

By the end of the month the case was filed, August 2005, it was time to put the two-part process to the ultimate test.

The Rigsby qui tam complaint describes the result.

In addition, when directing adjusters to categorize claims as flood damage and shift responsibility to the government for payment of the claim under flood insurance, State Farm insisted that adjusters “hit limits” when adjusting flood claims.

Specifically, adjusters were told that if they initially analyzed a claim and found that the insured had less damage under flood coverage than policy limits allowed, the adjuster was told to go back through the claim a second time to ensure that the flood claim “hit limits.” Adjusters used a computer program designed by State Farm known as “XACT Total” as a shortcut to determine the amount of damage for a claim. XACT Total was used to make determinations of total flood loss even in homes that sustained moderate flood damage.

Graphs are pictures produced from a database (the reverse of the process used to produce construction cost price list from drawing/dimensions of a room(s).
Graphs are pictures produced from a database - the reverse of the process used to produce construction cost price list from drawing/dimensions of a room(s).

In other words, the user can determine the outcome – either “hit the limits” or hit using a specific cost list. A Symbility white paper, Pricing Database, explains the process that works in much the same way as graphs are created in reverse if one thinks of a graph as a picture produced from a database such as the sample shown on the right.

Symbility licenses its pricing database from the Craftsman book company—one of the oldest and most respected construction pricing data providers in North America…The pricing database is updated monthly…

What is perhaps most unique about the Craftsman data that Symbility uses is that the data is published by a completely independent entity. Users of the Symbility data can be assured the data is not skewed or adversely influenced by large customers or contracts. This same impartiality is not typically afforded to companies who develop both a database and software…

Prices in the database do not include markup. Insurance repair markup is almost always 20 percent: 10 percent for overhead and 10 percent for profit…

When a new claim is created, the local region that best suits the location of the claim is selected automatically based on postal or zip code. This will look up the appropriate Regional Modification Factors for each pricing component, and ultimately result in unit pricing tailored pages-from-whitepaper_pricing_database-1cto the local region. These modifiers can be altered for this new claim, allowing you to adjust the pricing at the claim level to account for sudden price changes (such as catastrophe situations). The default Regional Modification Factors can also be altered (use the Claim Defaults tool in the Administration menu). The Symbility database contains regional modification factors for zip codes accurate to five digits, which results in over 22,000 separate regions.

Company administrators can further “tweak” the pricing, down to the subcategorypages-from-whitepaper_pricing_database-2c level using Category Modification Factors. If you find that prices in one or more categories are consistently too high or too low, you can create your own modification factors for materials, labor or equipment to apply to all items within a specific subcategory. The Category Modification Factors can be set as defaults (use the Claim Defaults tool in the Administration menu) or they can be set per claim, which allows for sudden but temporary shifts if pricing for particular item types (ex. lumber, roofing, electrical).

Symbility also provides company administrators control over all components of each individual trade rate, allowing easy control over all labor prices throughout the entire database.

For every electronic Symbility claim file that is created, several pricing database parameters are specified by the company originating the claim (for instance which version of the database, regional modification factors, O&P values, etc.). For any vendor creating an estimate for a claim that was originated by an insurance company, the database pricing that will be used by default will be whatever pricing the insurance company has established, governed by the parameters associated with the claim (anyone creating an estimate can always alter any price that’s in the database when creating an estimate, but Symbility will always record all variations from the claim’s associated pricing parameters). Vendors creating claims for insurance companies not yet using Symbility become the originator of the electronic claim file and therefore control all aspects of the pricing.

Symbility Solutions has not been a party to any Katrina litigation – and a careful reading of the white paper suggests no reason it should be.  Leaving the question, who was involved in bailing out the insurance industry with the NFIP, one for the courts to answer.

2 thoughts on “The first bail-out: NFIP accessed by manipulation of construction price data following Hurricane Katrina”

  1. I wonder if they have an XACT Total program for their adjusters to use on the wind claims? Of course how hard is it to figure out how to get the number 0 to be the total? It too is a two step process. Step one say I don’t see any evidence of wind damage. Step two place the number 0 in the total wind damage section.
    Drive car to next slab and repeat process. OK its a three step process.

  2. Folks I have little time to post today but I thought I’d roll out one of Nowdy’s finer entries on the topic of insurance. We’re read a lot out of academia and this post is an example from our archives that ages well against the test of time via continued quality readership.


Comments are closed.