RTW CORRESP Correspondence with SEC

February 23, 2007

Attn: Jim B. Rosenberg
Senior Assistant Chief Accountant
Mail Stop 6010
Division of Corporation Finance
Securities and Exchange Commission
Washington, DC 20549-6010

Re: RTW, Inc. Form 10-K for the fiscal year ended December 31, 2005

Dear Mr. Rosenberg:

RTW, Inc. (“RTW”) received your comment letter dated August 11, 2006 related to the filing listed above and responded to that comment letter on August 31, 2006. This letter will respond to further telephonic comments from the Securities and Exchange Commission (SEC) staff on October 26, 2006. As we discussed in our telephonic conversations on October 26, 2006, January 10, 2007, February 13, 2007 and February 22, 2007, the staff has requested that RTW provide this information to the SEC supplementally. We are doing that through today’s response. Based upon our conversation, we anticipate that the disclosure that we are providing you will serve as the foundation for additional disclosures in future SEC filings.

Telephonic Comment 1

Please revise your proposed disclosure provided in response to Comment 2(a)(1) to specifically identify the actuarial methods you use to determine your reserve for unpaid claim and claim settlement expenses for the 250 or so reserve cells you identify, and please describe the nature and characteristics of the reserving cells for each actuarial method and explain why that method is used.

RTW Response to Telephonic Comment 1

The attached exhibit C has been revised to reflect the information requested.

Telephonic Comment 2

Please revise your proposed disclosure provided in response to Comment 2(a)(2) to clarify when the three internal actuarial studies are performed and clarify whether the year-end independent actuarial studies were performed for statutory compliance purposes and whether it is used as a reasonableness check.

RTW Response to Telephonic Comment 2

The attached exhibit C has been revised to reflect the information requested.

Telephonic Comment 3

Please revise your proposed disclosure provided in response to Comment 2(b) to quantify the incremental differences from the actuarial point estimate recorded in each period. Describe the method you use to determine it and the extent which that method and the amounts differ from period to period.

RTW Response to Telephonic Comment 3

We have quantified the incremental differences as a percentage of net reserves in the attached exhibit C. We acknowledge our responsibility in future Exchange Act filings (Form 10-K or Form 10-Q) to disclose and explain any material changes in these incremental differences from the actuarial point estimates, including the effect of any such change on our “roll forward” of prior year net reserve balances.

Telephonic Comment 4

Please revise your proposed disclosure provided in response to Comment 2(c) to quantify the impact of each identified factor prompting your reserve revisions and specifically clarify why the adjustment was not warranted in prior periods. In addition, please revise your proposed disclosure to explain what you meant by “our actuaries were appropriately conservative in their reserve estimates.”

RTW Response to Telephonic Comment 4

The comment “our actuaries were appropriately conservative in their reserve estimates”, was included in our initial response related only to the closing of two of our regional offices in 2002. We understand that this comment may create unintended “color” for the entire discussion of unpaid claim and claim settlement expenses. We propose to remove that specific comment and feel that the remaining proposed disclosure “Our claims management processes traditionally have relied upon being in close proximity to the claimant. The lack of proximity to the claimant and recent significant adverse development created significant uncertainty in how these claims would emerge” adequately conveys our intended meaning.

The attached exhibit C has been revised to reflect the following additional information:

	•		We feel that it is impractical, if not impossible, to separate and quantify the effect of any of the identified factors during any period. Events, such as a significant law change that directly affects reserves, will be identified and quantified as they occur, and

	•		The reserve changes recognized during 2005, 2004 and 2003 were the result of claim emergence pattern changes occurring and recognized during 2005, 2004 and 2003 respectively. These changes were reflected by our actuaries in their 2005, 2004 and 2003 analyses respectively.

Telephonic Comment 5

Notwithstanding your proposed disclosure item in response to comment 2(d), please revise your disclosure to identify your key reserving assumptions. In addition, for each identified key assumption, please quantify and explain what caused them to change from the assumptions used in the immediately preceding period. Explicitly, identify and discuss the key assumptions as of December 2005 that are premised on future emergence that are inconsistent with historical loss patterns and explain why these assumptions are now appropriate given the inconsistency identified.

RTW Response to Telephonic Comment 5

The attached exhibit C has been revised to reflect the following:

•

Our basic assumptions that i) ultimate losses are determinable by extrapolation of claim emergence and settlement patterns observed in the past and ii) that ultimate loss ratios in the current and more recent accident years can be projected from ultimate loss ratios of prior years have not changed during the periods being reported. Our data inputs into the actuarial methods have consistently been incurred claim dollars, paid claim dollars and claim counts. Our actuarial methodologies used have been consistent during the periods reported. The volatility that we have experienced is a direct result of the data emergence patterns being different than what was actuarially expected. New actuarial judgments are made at each evaluation and reflect the changes in the data emergence patterns.

Telephonic Comment 6

We acknowledge your response to comment 2(e). FR-72 requires “the disclosure of reasonably likely changes in the critical accounting estimates and their resulting impact on results of operations and liquidity”. Please quantify and present, preferably in tabular format, the impact that reasonably likely changes in key assumptions identified may have on reported results. Explain why you believe this scenario is quantified or reasonably likely.

RTW Response to Telephonic Comment 6

The attached exhibit C has been revised to reflect the following:

•

We believe that favorable or unfavorable reserve development of $5.6 million is reasonably likely. During the period 1997 to 2005, seven years had favorable or unfavorable reserve development that exceeded $5.6 million, one year had favorable reserve development that approximated $5.6 million and one year had favorable reserve development that was 55% of $5.6 million. Favorable (unfavorable) reserve development directly increases (decreases) pre-tax income from operations. Other than the change in income taxes, reserve development has minimal effect on liquidity.

The Company hereby acknowledges that:

	•		it is responsible for the adequacy and accuracy of disclosures contained in the Company’s filings;

	•		staff comments or changes to disclosure in response to staff comments in the filings reviewed by the staff do not foreclose the Commission from taking any action with respect to the filing; and

	•		the Company may not assert staff comments as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States.

Please contact the undersigned at (952) 897-6147 if you have any questions or need additional information.

Very truly yours,

RTW, Inc.

/s/ Alfred L. LaTendresse
Alfred L. LaTendresse
Executive Vice President and Chief Financial Officer


cc:		Thomas G. Lovett, IV, Lindquist & Vennum PLLP
		Gregory D. Koschinska, Audit Committee Chairman
		Daniel R. Bauer, Ernst & Young LLP

RTW, Inc.
Response to Securities and Exchange Commission Letter Dated August 11, 2006
Comment 2
Exhibit C

UNPAID CLAIM AND CLAIM SETTLEMENT EXPENSES

Unpaid claim and claim settlement expenses relate solely to our insurance operations. Our Absentia division and service business do not bear claim risk for its customers and carry no unpaid claim and claim settlement expenses.

At December 31, 2005, net reserves totaled $76.8 million and included the liability for unpaid claim and claim settlement expenses of $160.1 million net of reinsurance recoverables on unpaid claim and claim settlement expenses of $83.3 million. The net reserve at December 31, 2004 totaled $78.4 million and included the liability for unpaid claim and claim settlement expenses totaling $156.1 million net of reinsurance recoverables on unpaid claim and claim settlement expenses of $77.7 million.

Accounting for workers’ compensation insurance operations requires us to estimate the liability for unpaid claim and claim settlement expenses (reserves) and the related reinsurance recoverables, (together, the “net reserves”) at each balance sheet date. Our reserves at December 31, 2005 represent the estimated total unpaid cost of claim and claim settlement expenses that cover events that occurred in 2005 and prior years. These reserves reflect our estimates of the total costs of claims that were reported to us, but not yet paid, and the cost of claims that had occurred but had not yet been reported to us. For reported claims, we establish case reserves. For workers’ compensation coverage, case reserves alone are an insufficient measure of the ultimate cost due in part to the nature of the settlement process, the potential for protracted litigation, the ongoing effect of inflation over the long period which medical care will be provided, the potential for workers’ compensation claims to close and subsequently reopen and the incompleteness of facts available at the time the case reserve is established. We establish additional incurred but not reported (“IBNR”) reserves by first projecting ultimate claim and claim settlement expenses for all claims as determined by using recognized standard loss development methods and techniques. The difference between our projected ultimate cost of claim and claim settlement expenses, reduced by payments, and case-basis reserves is IBNR. IBNR reserves reflect our estimated cost of claims that have not been reported to us and our estimate of the cost of claims that have been reported to us but where the remaining case reserves may be insufficient to cover the remaining cost of the claim and related claim settlement expenses.

The amount by which estimated net reserves, measured subsequently by reference to payments and additional estimates, differ from those originally reported for a period is known as “development.” Development is unfavorable (deficient) when losses ultimately settle for more than the levels at which they were reserved or subsequent estimates indicate a basis for reserve increases on open claims. Development is favorable (redundant) when losses ultimately settle for less than the amount reserved or subsequent estimates indicate a basis for reducing loss reserves on open claims. Favorable or unfavorable development of loss reserves is reflected in earnings in the period realized.

The level of reserves we maintain represents our best estimate, as of a particular point in time, of the ultimate cost of settlement and administration of claims based on our assessment of facts and circumstances known at that time. Workers’ compensation claims can remain open for many years. Claims that are open and claims that re-open will continue to evolve and change based on new information. It can be decades before all claims related to an accident year are reported and settled. Industry experience in the more recent accident years of long-tailed casualty lines, including workers’ compensation, shows limited statistical credibility because a relatively low proportion of ultimate claim and claim settlement expenses will be known and an even smaller portion will have been paid.

Reserves are not an exact calculation of liability, but instead are complex estimates that we derive from historical data and our expectations about future events, both internal and external, many of which are highly uncertain. There are two significant assumptions in determining our loss reserves. First, we assume that ultimate losses are determinable by extrapolating claim emergence and settlement patterns observed in the past that can reasonably be expected to persist into the future. This assumption implies that historical claim reporting, handling and settlement patterns may predict future activity and can be used to forecast ultimate liabilities on unpaid and unreported claims. Since the many factors that influence claim activity can change over time and are often difficult to isolate or quantify, the rate at which claims arose in the past and the costs to settle them may not always be representative of what will occur in the future. Our objectives in developing estimates of ultimate losses are to identify aberrations and systemic changes occurring within historical experience and to adjust for them so that the future can be projected reliably.

Our second significant assumption assumes that ultimate loss ratios (ultimate losses divided by earned premiums) in the current and more recent accident years can be projected from ultimate loss ratios of prior years after adjusting for factors including trends and pricing changes, to the extent that those factors can be quantified. This assumption implies consistency in the loss ratio, after adjusting for inflationary factors and other trends that may be affecting losses or premiums. We use various diagnostic tools and maintain close communication among our actuarial, claims and underwriting departments to continually monitor and assess the validity of these assumptions.

Both internal and independent external actuaries review our reserves for adequacy. These actuaries estimate and evaluate unpaid claim and claim settlement expenses using recognized standard loss development methods and techniques. We conduct internal reserve studies each year at May 31, August 31 and November 30. The November 30 study forms the basis for our year-end analysis. We engage external actuaries to complete their study at year-end. The external study is completed for statutory compliance and to provide management with a second independent actuarial determination of gross and net reserves. Both internal and external actuaries estimate reserves based on a ground-up approach. Case reserve, payment and claim count data is classified into segments (reserve cells), primarily based on accident year (currently 14), region (up to 5) and type of liability (medical, indemnity and settlement expenses) and analyzed over time. These aggregations are called loss triangles and are the primary basis for the projected ultimate cost of claim and claim settlement expenses. Reserve cells are reviewed using one or more of the following actuarial methods:

	•		Incurred Loss Development Method.The incurred loss development method is based on the assumption that the relative change in a given year’s reported loss estimates from one evaluation point to the next is similar to the relative change in prior years’ reported loss estimates at similar evaluation points. This method evaluates actual historical loss data. Successive years can be arranged to form a triangle of data. Report-to-report (RTR) development factors are calculated to measure the change in cumulative reported costs from one evaluation point to the next. These historical RTR factors and comparable industry benchmark factors form the basis for selecting the RTR factors used in projecting the current valuation of losses to an ultimate basis. In addition, a tail factor is selected to account for loss development beyond the observed experience. The tail factor is based on trends shown in the data and consideration of external benchmarks. The incurred loss development method’s implicit assumption is that the relative adequacy of case reserves has been consistent over time, and that there have been no material changes in the rate at which claims have been reported;

	•		Paid Loss Development Method.The paid loss development method is similar to the incurred loss development method, but excludes case reserves in the analysis. While this method has the disadvantage of not recognizing information provided by current case reserves, it has the advantage of avoiding potential distortions in the data due to changes in case reserving methodology. This method’s implicit assumption is that the rate of payment of claims has been relatively consistent over time;

	•		Average Claim Cost Method.The average claim cost (or frequency/severity) method calculates ultimate losses by separately projecting ultimate claim frequency (claims per exposure) and ultimate claim severity (cost per claim) for each accident period. Typically, the incurred and paid loss development methods are used to project ultimate frequency and severity based on historical data and ultimate losses are calculated as the product of the two. The average claim cost method is intended to avoid distortions that may exist with the other methods for the most recent years as the result of changes in case reserve levels, settlement rates or other factors for prior years. In addition, this method may provide insight into the drivers of the loss experience;

	•		Average Hindsight Outstanding Method.The average hindsight outstanding method estimates of ultimate losses for more mature accident years are used to estimate what the average cost per outstanding and IBNR claim would have been at various maturities, if the ultimate cost had been known. The average hindsight outstanding method is an iterative process.

	•		Loss Ratio Method.The loss ratio method assumes a loss ratio is appropriate for a particular accident year, and multiplies this selected loss ratio by the earned premiums for that accident year in order to obtain an estimate of ultimate losses. The loss ratio selected for any accident year is based upon older accident years, which are adjusted to conditions for the current accident year.

Our internal and external actuaries give different consideration to the results of these methods based on the accident year being reviewed and the type of data. For more mature or older accident years, our actuaries generally use a combination of the incurred development, paid development and average claim cost methods. For less mature or more recent accident years, excluding the current accident year, the incurred development method is relied upon more heavily, the paid development method is relied upon less heavily (because of the high leverage due to the magnitude of the development factors) with some consideration given to the average claim cost method and average hindsight outstanding. For the current accident year, our actuaries rely more on the incurred development method, the average hindsight outstanding method and the average claim cost method, as the paid development method is extremely leveraged at that point and is given very little weight. In addition for the current accident year, the resulting loss ratio is compared with the loss ratios for the more recent years (adjusted for trend, benefits and rate changes to current accident year conditions) for reasonableness, as these loss ratios factor into the actuary’s selections. The loss ratio method is relied upon more for the current accident year for the May and the August studies and is less important for the November study as the estimates from the other methods become more reliable as the current accident year matures. In general, methods other than the loss ratio method are relied upon more for reserve cells that have greater history and greater amounts of data. In general, loss development factors are much larger for indemnity costs and settlement expenses than medical costs. Our actuaries will take into consideration the inherent leverage of larger development factors in forming their judgments concerning ultimate claim and claim settlement expense values.

Loss triangles are used to determine the expected case loss emergence. In general, loss development factors are selected by a retrospective analysis of the overall adequacy of historical case reserves. The determination of the expected loss emergence pattern is not strictly a mechanical process. Factors affecting loss development patterns include, but are not limited to, the following: (i) trends in general economic conditions, including the effects of medical and wage inflation; (ii) estimates of trends in claims frequency and severity; (iii) our and industry historical loss experience; (iv) legislative enactments, legal developments and changes in social and political attitudes; (v) the lag time between the occurrence of an insured event and the time it is ultimately settled, referred to in the insurance industry as the “tail”; (vi) changes in how we adjudicate claims; (vii) changes in our personnel; (viii) unanticipated claim development; and (ix) statistical variability. These items influence the selection of the expected loss emergence patterns. Many of these items are not directly quantifiable, particularly on a prospective basis. There is no precise method for subsequently evaluating the effect of any specific factor on the adequacy of reserves because the eventual redundancy or deficiency is affected by many factors. Due to our commencing operations in 1992, we have limited historical data to estimate our reserves for unpaid claim and claim settlement expenses and reinsurance recoverables on unpaid claim and claim settlement expenses. Accordingly, we supplement our experience with external industry data related to loss development patterns and large losses, as adjusted, to reflect anticipated differences between our results and the industry.

Expected ultimate claim and claim settlement expenses are judgmentally determined by reserve cell based upon reviewing the selected loss emergence pattern. If the selected emergence pattern is not accurate, then the indicated ultimate claim and claim settlement expenses will not be correct. The expected loss emergence patterns are critical judgments and are updated with each reserve study. Once the year-end IBNR reserves are determined, our actuary calculates the expected case loss emergence and the expected claim payment activity for the upcoming calendar year. This calculation does not involve new judgments and uses the prior year-end expected loss emergence patterns. The expected losses are then allocated into interim estimates that are compared to actual reported claim and claim settlement case reserves and payments in the subsequent year. This comparison provides a test of the adequacy of prior year-end IBNR reserves and forms the basis for recognizing any reserve development during the course of the year.

The reserve analyses performed by the internal and external actuaries result in point estimates. We use these point estimates and other internal and external factors in determining carried reserves. These factors include, but are not limited to the historical pattern and volatility of the actuarial indications, the sensitivity of the actuarial indications to changes in paid and incurred loss patterns, the consistency of claims handling processes, the consistency of claim reserving practices, changes in our pricing and underwriting and overall pricing and underwriting trends in our markets. There is no standard checklist or formula that is or can be applied to each situation. We recorded net reserves approximately 6% to 7% in excess of the external actuary’s point estimates at December 31, 2005, 2004 and 2003. Our overall net reserves

were $78.6 million, $78.4 million and $76.8 million at December 31, 2003, 2004 and 2005 respectively. At December 31, 2005 we established net reserves in excess of the external actuary’s point estimate due to: (i) the significant adverse development experienced during 2000 and 2001; (ii) the significant volatility experienced in our reserves during the last 5 and 10 year periods; and (iii) the overall net deficiency shown in our reserves over a 10 year period offset by net redundancy achieved during the last 5 years. At December 31, 2004 we established net reserves in excess of the external actuary’s point estimate due to: (i) the significant adverse development experienced during 1998, 2000 and 2001; (ii) the significant volatility experienced in our reserves during the last 5 and 10 year periods; and (iii) the overall net deficiency shown in our reserves over a 10 year period. At December 31, 2003 we established net reserves in excess of the external actuary’s point estimate due to: (i) the significant adverse development experienced during 1998, 2000 and 2001; (ii) the significant volatility experienced in our reserves during the last 5 and 10 year periods; and (iii) the overall net deficiency shown in our reserves over a 10 year period. Our analysis of changes in our net reserves excluded the one-time $6.1 million favorable effect from a change in Minnesota statutes in 2002. Our reserve patterns are beginning to show greater stability, the amount of our history is increasing and the amount of data available is increasing. If the reserve patterns continue their trend of stabilizing, and no other significant internal or external factors emerge in the near future, we expect to place greater reliance on the actuarial point estimates in determining our net reserves. As a result, we would expect net reserves to trend towards the actuarial point estimates in the near future. Though not anticipated at present, significant factors that could influence this trend include, but are not limited to, a significant change in the composition of our business, significant changes in the geographic distribution of our business, significant legislation affecting workers compensation coverage or significant changes in claim emergence patterns. See the discussion concerning volatility later in this section.

Our reserves are primarily undiscounted; however, we discounted reserves for selected claims that have fixed and determinable future payments at rates ranging from 3.5% to 8.0% in 2005 and 2004. The discount rates are subject to change as market interest rates change. We also reduce the unpaid claim and claim settlement expenses for estimated amounts of subrogation.

We continually monitor loss development trends and data to determine reasonable reserve estimates. Reserves are based on estimates that are inherently uncertain and represent a significant risk to the business. We attempt to mitigate this risk by continually improving and refining our workers’ compensation claims processing practices and by continual monitoring through actuarial estimation methods.

After taking into account all relevant factors, we believe our reserves for unpaid claim and claim settlement expenses and reinsurance recoverables on unpaid claim and claim settlement expenses at December 31, 2005 are adequate to cover the ultimate net costs of claim and claim settlement expenses at that date. The ultimate cost of claim and claim settlement expenses may differ materially from the established reserves, particularly when claims may not be settled for many years. Establishing appropriate reserves is an inherently uncertain process and there can be no certainty that currently established reserves will prove adequate in light of subsequent actual experience. See Notes 1 and 5 in the accompanying Consolidated Financial Statements. The following two tables reconcile the beginning and ending insurance reserves, displayed individually for each of the last three years.

The following table sets forth reserves on a gross (before reinsurance) basis (000’s):


	*Year Ended December 31,*
	2005			2004			2003
Gross Reserves for Claim and Claim Settlement Expenses:
Gross reserves for claim and claim settlement expenses, beginning of year	$	156,123		$	150,044		$	181,262
Provision increases (decreases) for claim and claim settlement expenses:
Current year		43,936			53,563			42,777
Prior years		(3,231	)		(4,654	)		(21,846	)

Total provision		40,705			48,909			20,931

Payments for claim and claim settlement expenses:
Current year		11,748			12,666			11,077
Prior years		24,939			30,164			41,072

Total payments		36,687			42,830			52,149


Gross reserves for claim and claim settlement expenses, end of year	$	160,141		$	156,123		$	150,044

During 2003, we re-evaluated and lowered our estimate for excess of loss unpaid claim and claim settlement expenses by $15.1 million, which inured to the benefit of our reinsurers. Excluding the effect of this benefit, our ceded losses to reinsurers would have been $8.8 million.

The following table sets forth reserves on a net (after reinsurance) basis (000’s):


	*Year Ended December 31,*
	2005				2004			2003
Net Reserves for Claim and Claim Settlement Expenses:
Net reserves for claim and claim settlement expenses, beginning of year	$	78,401		$		78,578		$	89,440
Plus: Deferred retrospective reinsurance gain, beginning of year		49				49			49
Provision increases (decreases) for claim and claim settlement expenses:
Current year		36,264				42,583			33,954
Prior years		(5,457	)			(7,047	)		(6,698	)

Total provision		30,807				35,536			27,256

Payments for claim and claim settlement expenses:
Current year		11,748				12,666			10,761
Prior years		20,637				23,047			27,357

Total payments		32,385				35,713			38,118

Less: Deferred retrospective reinsurance gain, end of year		(49	)			(49	)		(49	)

Net reserves for claim and claim settlement expenses, end of year	$	76,823			$	78,401		$	78,578

The following gross loss reserve development table sets forth the change, over time, of gross reserves established for claim and claim settlement expenses at the end of the last ten years. The table is cumulative and, therefore, ending balances should not be added since the amount at the end of each calendar year includes activity for both current and prior years (000’s):


	*December 31,*
	*(000’s)*
	2005		2004		2003		2002		2001		2000		1999		1998		1997		1996		1995
Loss Reserve Development:
Gross reserves for unpaid claim and claim settlement expenses	$	160,141	$	156,123	$	150,044	$	181,262	$	181,310	$	128,841	$	99,831	$	97,269	$	61,069	$	49,256	$	37,138

Paid (cumulative) as of:
One year later			$	24,937	$	30,164	$	41,072	$	46,043	$	49,241	$	45,933	$	37,062	$	28,315	$	20,529	$	10,032
Two years later						43,436		61,363		17,086		74,681		67,442		56,031		42,889		29,841		15,306
Three years later								71,132		86,710		90,484		78,244		65,664		50,558		35,370		18,415
Four years later										93,348		98,689		85,754		70,631		54,835		38,880		19,964
Five years later												103,391		89,956		73,979		57,261		41,029		21,289
Six years later														93,045		76,311		59,012		41,980		22,117
Seven years later																78,476		60,352		42,728		22,702
Eight years later																		61,637		43,511		23,020
Nine years later																				44,208		23,448
Ten years later																						23,809

Reserves re-estimated as of:



	*December 31,*
	*(000’s)*
	2005		2004			2003			2002			2001			2000			1999			1998			1997			1996			1995
End of year	$	160,141	$	156,123		$	150,044		$	181,262		$	181,310		$	128,841		$	99,831		$	97,269		$	61,069		$	49,256		$	37,138
One year later				152,892			145,389			159,415			183,923			160,065			118,205			85,384			72,443			44,862			26,086
Two years later							147,117			160,147			166,738			168,222			130,120			95,696			64,499			48,233			22,295
Three years later										163,724			168,892			157,251			137,002			101,893			73,031			44,587			24,111
Four years later													174,451			161,905			129,819			107,522			75,554			50,552			23,054
Five years later																168,190			132,813			103,064			79,398			52,063			26,485
Six years later																			138,782			105,705			76,610			54,327			27,237
Seven years later																						110,136			77,921			53,047			28,411
Eight years later																									80,855			54,250			28,533
Nine years later																												55,491			29,107
Ten years later																															29,627
Initial reserves in excess of (less than) re-estimated reserves
Amount			$	3,231		$	2,927		$	17,538		$	6,859		$	(39,349	)	$	(38,951	)	$	(12,867	)	$	(19,786	)	$	(6,235	)	$	7,511
Percent				2.1	%		2.0	%		9.7	%		3.8	%		(30.5)	%		(39.0)	%		(13.2)	%		(32.4	)%		(12.7	)%		20.2	%

The table above represents the development of balance sheet gross reserves for 1995 through 2005. The upper portion of the table shows the cumulative amount paid with respect to the previously recorded reserves as of the end of each succeeding year. The lower portion of the table shows the re-estimated amount of the previously recorded gross reserves, based on experience as of the end of each succeeding year. The estimate is either increased or decreased as more information becomes known about the frequency and severity of claims for individual years.

In evaluating this information, it should be noted that each amount includes the total of all changes in amounts for prior periods. For example, the amount of redundancy (deficiency) to losses settled in 2002, but incurred in 1999, is included in the cumulative redundancy (deficiency) amounts in 1999, 2000 and 2001. This table does not present accident or policy year development data, which certain readers may be more accustomed to analyzing. Conditions and trends that have affected development of the reserves in the past may not necessarily occur in the future. Accordingly, it may not be appropriate to extrapolate redundancies or deficiencies based on this table.

The following net loss reserve development table sets forth the change, over time, of net reserves established for claim and claim settlement expenses at the end of the last ten years. The table is cumulative and, therefore, ending balances should not be added since the amount at the end of each calendar year includes activity for both current and prior years (000’s):


	*December 31,*
	2005		2004			2003			2002			2001			2000			1999			1998			1997			1996			1995
Loss Reserve Development:
Gross reserves for unpaid claim and claim settlement expenses	$	160,141	$	156,123		$	150,044		$	181,262		$	181,310		$	128,841		$	99,831		$	97,269		$	61,069		$	49,256		$	37,138
Reinsurance recoverables		83,318		77,722			71,466			91,822			90,115			61,845			41,179			21,403			5,374			6,183			8,312

Net reserves for unpaid claim and claim settlement expenses	$	76,823	$	78,401		$	78,578		$	89,440		$	91,195		$	66,996		$	58,652		$	75,866		$	55,695		$	43,073		$	28,826

Paid (cumulative) as of:
One year later				20,637			23,047		$	27,357		$	30,285		$	32,028		$	35,932		$	34,380		$	27,737		$	19,439		$	8,595
Two years later							32,068			40,956			43,825			43,823			48,069			49,958			42,046			28,173			12,894
Three years later										46,673			50,871			49,531			54,360			56,376			49,671			33,438			15,521
Four years later													54,078			53,130			58,113			60,453			53,814			36,904			16,869
Five years later																55,342			60,690			63,278			56,140			38,919			18,020
Six years later																			62,465			65,347			57,903			39,770			18,714
Seven years later																						66,674			59,219			40,530			19,200
Eight years later																									60,442			41,289			19,530
Nine years later																												41,924			19,934
Ten years later																															20,232
Reserves re-estimated as of:
End of year		76,823		78,401		$	78,578		$	89,440		$	91,195		$	66,996		$	58,652		$	75,866		$	55,695		$	43,073		$	28,826
One year later				72,944			71,531			82,742			82,839			74,727			74,181			67,753			66,674			39,988			20,751
Two years later							69,831			78,482			76,545			71,202			76,502			77,205			61,075			43,484			18,469
Three years later										78,037			77,055			71,911			75,321			78,391			68,065			41,451			19,796
Four years later													78,054			71,177			77,443			78,772			69,474			45,959			19,389
Five years later																73,660			75,588			80,522			69,595			47,147			21,254
Six years later																			79,204			78,878			69,926			47,126			22,568
Seven years later																						82,676			69,095			46,969			22,388
Eight years later																									71,540			47,039			22,342
Nine years later																												48,288			22,727
Ten years later																															22,950
Initial reserves in excess of (less than) re-estimated reserves
Amount			$	5,457		$	8,747		$	11,403		$	13,141		$	(6,664	)	$	(20,552	)	$	(6,810	)	$	(15,845	)	$	(5,215	)	$	5,876
Percent				7.0	%		11.1	%		12.7	%		14.4	%		(9.9	)%		(35.0	%)		(9.0	)%		(28.4	)%		(12.1	)%		20.4	%

The table above represents the development of balance sheet net reserves for 1995 through 2005. The top three rows of the table reconcile gross reserves to net reserves for unpaid claim and claim settlement expenses recorded at the balance sheet date for each of the indicated years.

The upper portion of the table shows the cumulative amount paid with respect to the previously recorded reserves as of the end of each succeeding year.

The lower portion of the table shows the re-estimated amount of the previously recorded net reserves, based on experience as of the end of each succeeding year. The estimate is either increased or decreased as more information becomes known about the frequency and severity of claims for individual years. For example, for the 1998 calendar year valued as of December 31, 2005, we paid $66.7 million of the currently estimated $82.7 million of claim and claim settlement expenses that were incurred through the end of 1998. Thus, the difference, an estimated $15.7 million of claim and claim settlement expenses incurred through 1998, remained unpaid as of December 31, 2005.

The “initial reserves in excess of (less than) re-estimated reserves” (Amount and Percent rows) represent the aggregate change in the estimates over all prior years. For example, the 1998 reserve developed a $6.8 million net deficiency over the course of the succeeding years. The net amount has been included in income and the changes have been recorded in the period identified. The cumulative net deficiencies in 2000 and 1999 are the result of reserve development inherent in the uncertainty in establishing reserves and anticipated loss trends. As discussed above, due to our relatively limited historical claim data and small claim population, our estimate of the liability for net reserves is difficult and volatile. As discussed further below, the reserve redundancy in 2001 is the result of accrual reversals resulting from changes in methods of assessing second injury funds, lower frequency in claims reported from the estimate at December 31, 2001, and reductions in amounts expected to be incurred for our participation in mandatory state and national assigned risk pools.

In evaluating this information, it should be noted that each amount includes the total of all changes in amounts for prior periods. For example, the amount of redundancy to losses settled in 2002, but incurred in 1999, will be included in the cumulative redundancy (deficiency) amounts in 1999, 2000, and 2001. This table does not present accident or policy year development data, which certain readers may be more accustomed to analyzing. Conditions and trends that have affected development of the reserves in the past may not necessarily occur in the future. Accordingly, it may not be appropriate to extrapolate redundancies or deficiencies based on this table.

The following table is derived from the net loss reserve development table and summarizes the effect of reserve re-estimates, net of reinsurance, on calendar year operations for the same ten-year period ended December 31, 2005. The total of each column details the amount of reserve re-estimates made in the indicated calendar year and shows the accident years to which the re-estimates are applicable. The amounts in the total accident year column represent the cumulative reserve re-estimate (increase) decrease for the indicated accident year (000’s):


																															Cumulative
																															Re-estimates
																															*for each*
	*Effect of Reserve Re-estimates on Calendar Year Operations:*																														*Accident*
	2005			2004			2003			2002			2001			2000			1999			1998			1997			Pre-1997			*Year*
Accident Year:
1992	$	(21	)	$	25		$	(30	)	$	2		$	(6	)	$	(19	)	$	38		$	(15	)	$	(37	)	$	(110	)	$	(173	)
1993		12			(98	)		14			261			(706	)		(160	)		7			(96	)		(42	)		741			(67	)
1994		3			(174	)		(77	)		178			(633	)		(459	)		(68	)		(574	)		413			3,943			2,552
1995		(217	)		(138	)		139			(261	)		31			(1,227	)		430			(642	)		1,948			4,604			4,667
1996		(1,026	)		315			111			(159	)		126			(2,643	)		1,626			(2,169	)		803						(3,016	)
1997		(1,196	)		901			(488	)		(142	)		(221	)		(2,482	)		3,566			(7,483	)								(7,545	)
1998		(1,353	)		813			(1,419	)		(320	)		283			(2,462	)		2,514												(1,944	)
1999		182			211			(372	)		1,622			(1,195	)		(6,077	)														(5,629	)
2000		1,133			(1,121	)		1,413			2,344			(5,410	)																	(1,641	)
2001		1,484			1,756			4,003			4,831																					12,074
2002		1,444			1,770			3,404																								6,618
2003		1,255			2,787																											4,042
2004		3,757																														3,757

Total	$	5,457		$	7,047		$	6,698		$	8,356		$	(7,731	)	$	(15,529	)	$	8,113		$	(10,979	)	$	3,085		$	9,178		$	13,695

Volatility in net carried loss reserves: It is inherent in a long-tailed line of business that current year estimates of ultimate liability for claim and claim settlement expenses will differ from prior estimates, sometimes materially. Volatility in our estimates can be viewed on a retrospective basis (that is, what changes have occurred) and a prospective basis (what could occur). The primary characteristics influencing the level of volatility are the length of the claim settlement period, the potential for changes in medical, wage and other claim costs, changes in the level of litigation or other dispute resolution processes, the potential for different types of injuries emerging and changes in the legal environment.

Our basic assumptions that i) ultimate losses are determinable by extrapolating claim emergence and settlement patterns observed in the past and ii) that ultimate loss ratios in the current and more recent accident years can be projected from ultimate loss ratios of prior years, have not changed during the periods being reported. Our data inputs into the actuarial methods have consistently been incurred claim dollars, paid claim dollars and claim counts. The actuarial methodologies used have been consistent during the periods reported. The volatility that we have incurred is a direct result of the data emergence patterns being different than what was actuarially expected. New actuarial judgments are made at each evaluation and reflect the changes in the data emergence patterns.

We believe that the historical unpaid claim and claim settlement expenses development recognized into income illustrates the potential variability of our estimate of net unpaid claim and claim settlement expenses. The “Effect of Reserve Re-estimates on Calendar Year Operations” table contains 85 entries for accident year re-estimates within the years shown: 43 of these are favorable adjustments and 42 are unfavorable. The 13 accident years (1992 to 2004) indicate that seven cumulatively have developed unfavorably at December 31, 2005 and six have developed favorably. The magnitude of the calendar re-estimates has been as favorable as $8.4 million (calendar year 2002) and as unfavorable as $15.5 million (calendar year 2000). The “net loss reserve development” table also provides insight into the potential variability of our estimates for net unpaid claim and claim settlement expenses. On average, our beginning of the year net reserve estimate for 1995 to 2004 has developed unfavorably by 2.9%. The largest favorable development has been $13.1 million (calendar year 2001) and the greatest unfavorable change has been $20.6 million (calendar year 1999). In 2002, Minnesota changed the method of assessing companies for its Special Compensation Fund. This one-time law change benefited us $6.1 million in 2002 and has not been excluded from the results presented in these tables.

To illustrate the sensitivity of our net reserves to changes in loss development factors, our internal actuary estimated the effect of increases in the development factors used in the incurred loss development methodology, one of several different estimation methodologies used. We increased each year’s development factor by 15%. Historically, it is not unusual for these factors to change by 15% to 20%, in either direction. Assuming that our net reserves were based solely on the incurred loss development methodology and the preceding changes to the loss development factors, we estimate that our net unpaid claim and claim settlement expense reserve could increase by approximately $5.6 million at December 31, 2005 for all accident years combined. We believe that favorable or unfavorable reserve development of $5.6 million is reasonably likely. Workers’ compensation is a long-tailed line of insurance, in which the period from initial report of a claim to ultimate settlement of that claim may take many years or even decades to complete. During calendar year 2005, five accident years had total unfavorable development of $3.6 million while eight accident years had total favorable development of $9.1 million. In addition, calendar year 2005 had eight accident years with either favorable or unfavorable development in excess of $1 million. Although we have experienced net favorable accident year development during each calendar year in our recent history (2002 to 2005), this will not always be the case. During the period 1997 to 2005, seven years had favorable (1999, 2002, 2003, 2004) or unfavorable (1998, 2000, 2001) reserve development that exceeded $5.6 million. Note that while we present this sensitivity analyses to comply with the rules of the Securities and Exchange Commission, it is not integral to our process of establishing net reserves. We have previously presented our discussion of qualitative factors that we considered in establishing our net reserves.

We previously indicated that we make two significant assumptions concerning reserves: (i) that ultimate losses are determinable by extrapolating claim emergence and settlement patterns observed in the past; and (ii) that ultimate loss ratios in the current and more recent accident years can be projected from ultimate loss ratios of prior years. These assumptions do not lend themselves to a granular sensitivity analysis as presented by others in this industry. Additionally, workers’ compensation insurance, being a long-tailed line of insurance business, is affected by previously enumerated factors, most of which are not subject to strict quantification. We do not believe that a more “sophisticated” sensitivity analysis would be cost beneficial or provide credible information to management or the reader.

Favorable (unfavorable) reserve development directly increases (decreases) pre-tax income from operations. Other than the change in income taxes currently payable, reserve development has minimal effect on current year liquidity and will affect liquidity only as claims are paid.

In addition to the factors described above, other factors may also affect claim and claim settlement expenses reserve development in future periods. These factors include: (i) governmental actions, including court decisions interpreting existing laws, regulations or policy provisions; (ii) adverse or favorable outcomes in pending claims litigation; (iii) the number and severity of claims; (iv) the effect of inflation on claims; and (v) the effect of residual market assessments. Although the actuaries do not make specific numerical assumptions about these factors, changes in these factors from past patterns will affect historical loss development factors and in turn, future loss reserve development. Significant

positive changes in one or more factors will lead to positive future loss reserve development, which could result in the actual losses developing closer to, or even below, the lower end of the our estimated reserve variability. Significant negative changes in one or more factors will lead to negative loss reserve development, which could result in the actual loss developing closer to, or even above, the higher end of the our estimated reserve variability. Accordingly, due to these factors and the other factors enumerated throughout this Management’s Discussion and Analysis of Operations and Financial Condition and the inherent limitations of the loss reserving methodologies, the estimated and illustrated reserve variability may not necessarily be indicative of our future reserve variability, which could ultimately be greater than the estimated and illustrated variability. Accordingly, it is possible that our future reserve variability could ultimately be greater than the illustrated variability.

Reserve development recognized: The reserve development for 2005, 2004 and 2003 has been affected by the following factors:

	•		Re-underwriting our book of business. Beginning in 2000 and continuing today, we deliberately eliminated certain classes of business and industries from our book of business. Additionally, we re-emphasized to our customers that they must return workers to light-duty positions when approved by treating physicians and that failure to do so results in cancellation or non-renewal of their account. An employer’s failure to return workers to light-duty positions results in higher claim and claim settlement expenses for us. These shifts changed the nature and frequency of claims that occurred. The resulting data shift continues to emerge in our historical results.

	•		Office closure. During 2002, we stopped writing business in our Missouri and New England regions and closed our offices there. We are still responsible for managing and paying for all accidents related to policies that we underwrote while we operated there. We also experienced significant losses and adverse development in these markets during 2000 and 2001. Our claims management processes traditionally have relied upon being in close proximity to the claimant. The lack of proximity to the claimant and recent significant adverse development created significant uncertainty in how these claims would emerge. Actual claim emergence has been more favorable than anticipated.

	•		Procedural changes. We regularly review and adjust internal processes and procedures with the intent of becoming more effective and efficient in our claims adjudication. The effect of these changes will not be discernable in our data patterns for a minimum of two years, if not longer. Favorable and unfavorable claim results will be recognized in the short-run prior to the actuaries being comfortable in adjusting the long-term development patterns.

	•		Effect of variability of results on estimation process. The frequency and severity of claims will vary by year. Actual emergence will vary by year. Adverse frequency, severity and emergence affect actuarial judgments more rapidly than a corresponding favorable event.

We believe that it is impractical, if not impossible, to separate and quantify the effect of any of the above mentioned factors during any period. Additionally, workers’ compensation insurance, being a long-tailed line of insurance business, is affected by additional, previously enumerated factors, most of which are not subject to strict quantification. We do not believe that a more “sophisticated” analysis would be cost beneficial or provide credible information to management or the reader. Events, such as a significant law change that directly affects reserves, will be identified and quantified as they occur.

The 2005 results include a $5.5 million decrease in prior years’ reserves for unpaid claim and claim settlement expenses. Our estimate of net unpaid indemnity expense decreased significantly as a result of procedural changes introduced earlier and a changing nature of the book of business from our re-underwriting initiative. Our estimate of net unpaid medical expenses moved adversely as the anticipated benefits of changes did not emerge. Accident years 1996 to 1998 suffered adverse claim development which was further affected by an adverse movement in industry excess loss development patterns that we use to supplement our analysis. More recent accident years, where we have lower retention limits were unaffected by this and moved favorably. The reserve changes recognized during 2005 were the result of claim emergence pattern changes occurring and recognized during 2005. These changes were reflected by our actuaries in their 2005 analysis.

The 2004 results include a $7.0 million decrease in prior years’ reserves for unpaid claim and claim settlement expenses. More recent accident years (2001 to 2003) favorably developed as the result of procedural changes and re-underwriting emerged more rapidly than anticipated in the reserve estimation process. Indemnity results improved as we were more effective in adjudicating claims and returning injured workers to work sooner than anticipated. Our expected costs for residual market charges moved favorably as our exposure to states with residual market charges, which are adjusted each year, decreased. The reserve changes recognized during 2004 were the result of claim emergence pattern changes occurring and recognized during 2004. These changes were reflected by our actuaries in their 2004 analysis.

The 2003 results include a $6.7 million decrease in prior years’ reserves for unpaid claim and claim settlement expenses. More recent accident years (2000 to 2002) favorably developed as the result of procedural changes and re-underwriting emerged more rapidly than anticipated in the reserve estimation process. Indemnity results improved as we were more effective in adjudicating claims and returning injured workers to work sooner than anticipated. Procedural changes began in 2002 that favorably affected the claims for the offices that we closed. The reserve changes recognized during 2003 were the result of claim emergence pattern changes occurring and recognized during 2003. These changes were reflected by our actuaries in their 2003 analysis.

RTW (RTWI) Inactive CORRESPCorrespondence with SEC