Meadowbrook Insurance (MIG) CORRESP Correspondence with SEC

January 30, 2008

Securities and Exchange Commission
100 F Street, N.E.
Mail Stop 6010
Washington, DC 20549
Attention: Ms. Dana Hartz

Re: Meadowbrook Insurance Group, Inc.
Form 10-K for the Fiscal Year Ended December 31, 2006
Filed March 13, 2007
File Number: 001-14094

Securities and Exchange Commission, Division of Corporation Finance:

          On behalf of Meadowbrook Insurance Group, Inc. (the “Company”), we provide the following information verbally requested by Dana Hartz on January 14, 2008 and have indicated, where applicable, our intent to provide information within our future Form 10-Ks, Form 10-Qs, or other applicable SEC filings. Our response acknowledges that:

•	the Company is responsible for the adequacy and accuracy of the disclosure in the filing;
•	staff comments and changes to disclosure in response to staff comments 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 and any person under the federal securities laws of the United States.
1.	Please expand your disclosure to clarify that the actuarial methods are consistently applied each year.

Management’s Response:

The proposed disclosure in its entirety is included in this letter following question 2 below. The changes from our response letter dated December 17, 2007 are noted in bold, italics and underlined with references to the corresponding verbal question.

2. a) We received your response to comment 1. Please clarify in your disclosure how a 1% annual inflation variance was determined to be reasonably likely, specifically addressing whether this information was based on historical experience. In addition, please disclose what impact the 3% change in reserves adequacy would have on net income.

b) It also appears that additional key assumptions you discussed on page 7 of your response could have a material impact on your loss reserves, specifically your expected loss ratios. Please provide an enhanced sensitivity analysis which depicts changes in your reserves by line of business given reasonably likely changes in your initial expected loss ratios.

c) Lastly, it appears that the ranges you disclose on page 9 of your response resulted from the different actuarial methods run and not from performing sensitivity around key loss reserve assumptions. Therefore, the table does not appear to reference reasonably likely variances in your future loss payment and cash reserves. Please revise your disclosure accordingly to clarify this fact.

Management’s Response:

Losses and Loss Adjustment Expenses

     Significant periods of time can elapse between the occurrence of a loss, the reporting of the loss to the insurer, and the insurer’s payment of that loss. To recognize liabilities for unpaid losses and loss adjustment expenses (“LAE”), insurers establish reserves as balance sheet liabilities representing estimates of amounts needed to pay reported and unreported net losses and LAE.

     We establish a liability for losses and LAE, which represent case based estimates of reported unpaid losses and LAE and actuarial estimates of incurred but not reported losses (“IBNR”) and LAE. Such liabilities, by necessity, are based upon estimates and, while we believe the amount of our reserves is adequate, the ultimate liability may be greater or less than the estimate. As of December 31, 2006 and 2005, we have accrued $501.1 million and $458.7 million of gross loss and LAE reserves, respectively.

     Components of Losses and Loss Adjustment Expense (new heading)

     The following table sets forth our gross and net reserves for losses and LAE based upon an underlying source of data, at December 31, 2006 (in thousands):

	Case		IBNR		Total
Direct	$	181,884	$	227,864	$	409,748
Assumed-Directly Managed (1)		17,777		41,264		59,041
Assumed-Residual Markets (2)		9,242		16,855		26,097
Assumed-Retroceded		1,281		227		1,508
Assumed-Other		3,031		1,652		4,683
Gross		213,215		287,862		501,077
Less Ceded		90,038		108,384		198,422
Net	$	123,177	$	179,478	$	302,655

(1)	Directly managed represents business managed and processed by our underwriting, claims, and loss control departments, utilizing our internal systems and related controls.
(2)	Residual markets represent mandatory pooled workers’ compensation business allocated to individual insurance company writers based on the insurer’s market share in a given state.

     The reserves referenced in the above table related to our direct business and assumed business, which we directly manage (see footnote 1 above) and are established through transactions processed through our internal systems and related controls. Accordingly, the case reserves are established on a current basis, therefore there is no delay or lag in reporting of losses from a ceding company, and IBNR is determined utilizing various actuarial methods based upon historical data. Ultimate reserve estimates related to assumed business from residual markets are provided by

individual states on a two quarter lag between the date of the evaluation and the receipt of the estimate from NCCI, and include an estimated reserve based upon actuarial methods for this lag. Assumed business which is subsequently 100% retroceded to participating reinsurers relates to business previously discontinued and now is in run-off. Lastly, in relation to assumed business from other sources, we receive case and paid loss data within a forty-five day reporting period and develop our estimates for IBNR based on both current and historical data.

     The completeness and accuracy of data received from cedants on assumed business that we do not manage directly is verified through monthly reconciliations to detailed statements, inception to date rollforwards of claim data, actuarial estimates of historical trends, field audits, and a series of management oversight reports on a program basis.

     The following table sets forth our net case and IBNR reserves for losses and LAE by line of business at December 31, 2006 (in thousands):

	Net Case		Net IBNR		Total
Workers’ Compensation	$	60,882	$	76,231	$	137,113
Residual Markets		9,242		16,856		26,098
Commercial Multiple Peril/General Liability		21,340		41,716		63,056
Commercial Automobile		24,555		30,087		54,642
Other		7,158		14,588		21,746
Total	$	123,177	$	179,478	$	302,655

     Claim Reserving Process and Methodology (new heading)

     When a claim is reported to one of our Insurance Company Subsidiaries, for the majority of claims, our claims personnel within our risk management subsidiary will establish a case reserve for the estimated amount of the ultimate payment. The amount of the reserve is primarily based upon a case-by-case evaluation of the type of claim involved, the circumstances surrounding each claim, and the policy provisions relating to the type of losses. The estimate reflects the informed judgment of such personnel based on general insurance reserving practices, which focus on the ultimate probable cost of each reported claim, as well as the experience and knowledge of the claims person. Until the claim is resolved, these estimates are revised as deemed necessary by the responsible claims personnel based on subsequent developments, new information or periodic reviews of the claims.

     In addition to case reserves and in accordance with industry practice, we maintain estimates of reserves for losses and LAE incurred but not yet reported. We project an estimate of ultimate losses and LAE at each reporting date. The difference between the projected ultimate loss and LAE reserves and the case loss reserves and LAE reserves, is carried as IBNR reserves. By using both estimates of reported claims and IBNR determined using generally accepted actuarial reserving techniques, we estimate the ultimate liability for losses and LAE, net of reinsurance recoverables.

     (response to 1.a.2 and 1.a.4)

     In developing claim and claim adjustment expense (“loss” or “losses”) reserve estimates, the Company performs a complete and detailed reserve analyses each quarter. To perform this

analysis the data is organized at a “reserve category” level. A reserve category can be a line of business such as commercial automobile liability, or it may be a particular geographical area within a line of business such as Nevada workers’ compensation. The reserves within a reserve category level are characterized as either short-tail or long-tail. About 98% of our reserves can be characterized as coming from long-tail lines of business. For long-tail business, several years may lapse between the time the business is written and the time when all claims are settled. Our long-tail exposures include workers’ compensation, commercial automobile liability, general liability, professional liability, products liability, excess, and umbrella. Short-tail exposures include property, commercial automobile physical damage, and inland marine. The analyses generally review losses both gross of reinsurance and net of reinsurance.

     (response to 1.a.1)

     The standard actuarial methods that we use to project ultimate losses for both long-tail and short-tail exposures include, but are not limited to, the following:

Paid Development Method,
Incurred Development Method,
Paid Bornhuetter-Ferguson Method,
Reported Bornhuetter-Ferguson Method,
Initial Expected Loss Method,
Paid Roll-forward Method, and
Incurred Roll-forward Method

     (response to 1.a.3) (Verbal question number 1 – dated January 14, 2008)

     All of these methods areconsistentlyapplied to every reserve category where they are applicable and they create indications for each accident year. We use judgment selecting the best estimate from within these estimates or adjusted estimates. As such, no one method or group of methods is strictly used for any line of business or reserve category within a line of business. The individual selections by year are our best judgments based on the strengths and weaknesses of the method indications, the inherent variability in the data and the specific modifications to selections for data characteristics.

A brief description of the methods and some discussion of their inherent strengths and weaknesses are as follows:

Paid Development Method.    This method uses historical, cumulative paid losses by accident year and develops those actual losses to estimated ultimate losses based upon the assumption that each accident year will develop to estimated ultimate cost in a manner that is analogous to prior years, adjusted as deemed appropriate for the expected effects of known changes in the claim payment environment, and to the extent necessary supplemented by analyses of the development of broader industry data.

     Selection of the paid loss pattern requires analysis of several factors including the impact of inflation on claims costs, the rate at which claims professionals make claim payments and close claims, the impact of judicial decisions, the impact of underwriting changes, the impact of large claim payments and other factors. Claim cost inflation itself requires evaluation of changes in the

cost of repairing or replacing property, changes in the cost of medical care, changes in the cost of wage replacement, judicial decisions, legislative changes and other factors. Because this method assumes that losses are paid at a consistent rate, changes in any of these factors can impact the results. Since the method does not rely on case reserves, it is not directly influenced by changes in the adequacy of case reserves.

Incurred Development Method.    This method uses historical, cumulative reported loss dollars by accident year and develops those actual losses to estimated ultimate losses based upon the assumption that each accident year will develop to estimated ultimate cost in a manner that is analogous to prior years, adjusted as deemed appropriate for the expected effects of known changes in the claim payment and case reserving environment, and to the extent necessary supplemented by analyses of the development of broader industry data.

     Since the method uses more data (case reserves in addition to paid losses) than the paid development method, the incurred development patterns may be less variable than paid patterns. However, selection of the incurred loss pattern requires analysis of all of the factors listed in the description of the paid development method. In addition, the inclusion of case reserves can lead to distortions if changes in case reserving practices have taken place and the use of case incurred losses may not eliminate the issues associated with estimating the incurred loss pattern subsequent to the most mature point available.

Paid Bornhuetter-Ferguson Method.    This is a method that assigns partial weight to initial expected losses for each accident year and partial weight to observed paid losses. The weights assigned to the initial expected losses decrease as the accident year matures.

     The method assumes that only future losses will develop at the expected loss ratio level. The percent of paid loss to ultimate loss implied from the paid development method is used to determine what percentage of ultimate loss is yet to be paid. The use of the pattern from the paid development method requires consideration of all factors listed in the description of the paid development method. The estimate of losses yet to be paid is added to current paid losses to estimate the ultimate loss for each year. This method will react very slowly if actual ultimate loss ratios are different from expectations due to changes not accounted for by the expected loss ratio calculation.

Reported Bornhuetter-Ferguson Method.    This is a method that assigns partial weight to the initial expected losses and partial weight to observed reported loss dollars (paid losses plus case reserves). The weights assigned to the initial expected losses decrease as the accident year matures.

     The use of case incurred losses instead of paid losses can result in development patterns that are less variable than paid patterns. However, the inclusion of case reserves can lead to distortions if changes in case reserving have taken place, and the method requires analysis of all the factors that need to be reviewed for the expected loss ratio and incurred development methods.

Initial Expected Loss Method.    This method is used directly, and as an input to the Bornhuetter-Ferguson methods. Initial expected losses for an accident year are based on adjusting prior accident year projections to the current accident year levels using underlying loss trends, rate changes, benefit changes, reinsurance structure and cost changes and other pertinent adjustments specific to the line of business.

     This method may be useful if loss development patterns are inconsistent, losses emerge very slowly, or there is relatively little loss history from which to estimate future losses. The selection of the expected loss ratio requires analysis of loss ratios from earlier accident years or pricing studies and analysis of inflationary trends, frequency trends, rate changes, underwriting changes, and other applicable factors.

Paid Roll-forward Method.    This method adjusts prior estimates of ultimate losses based on the actual paid loss emergence in the quarter compared to the expected emergence. It is useful in determining reserves that avoid overreacting to ordinary fluctuations in the development patterns.

Incurred Roll-forward Method.    This method adjusts prior estimates of ultimate losses based on the actual case incurred loss emergence in the quarter compared to the expected emergence. It is also useful in determining reserves that avoid overreacting to ordinary fluctuations in the development patterns and generally reacts faster than the paid roll-forward method.

     (response to 1.a.1 and 1.a.3)

     Claims for short-tail lines of business settle more quickly than long-tail lines of business, and in general, loss development factors for short-tail lines are smaller than long-tail lines. For long-tail lines, we tend to rely on initial expected loss methods throughout the current accident year then move to development factor based methods for older accident years. Development methods on short-tail lines are generally reliable in the third and fourth quarter of the initial accident year and recorded loss ratios reflect a blend of the development and forecast methods. Short-tail lines represent less than 2% of our total reserves at December 31, 2006.

     The reserve categories where the above methods are not applicable are few. The largest of these is our workers’ compensation residual market reserve category, where we utilize detailed reserve analyses performed by the industry statistical agency NCCI in making our estimates. We adjust these estimates for timing differences in the reporting of the data. The other reserve categories that deviate from the above methods are small, together constituting less than 3% of the total reserves.

     Each of the methods listed above requires the selection and application of parameters and assumptions. For all but the initial expected loss method, the key assumptions are the patterns with which our aggregate claims data will be paid or will emerge over time (“development patterns”). These patterns incorporate inherent assumptions of claims cost inflation rates, and trends in the frequency of claims, both overall and by severity of claim. These are affected by underlying loss trends, rate changes, benefit changes, reinsurance structure and cost changes and other pertinent adjustments which are explicit key assumptions underlying the initial expected loss method. Each of these key assumptions is discussed in the following paragraphs.

     To analyze the development patterns, we compile, to the extent available, long-term and short-term historical data for our insurance subsidiaries, organized in a manner which provides an indication of the historical development patterns. To the extent that the historical data may not provide sufficient information about future patterns—whether due to environmental changes such as legislation or due to the small volume or short history of data for some segments of our business—benchmarks based on industry data, and forecasts made by industry rate bureaus regarding the effect of legislative benefit changes on such patterns, may be used to supplement, adjust, or replace patterns based on our insurance companies’ historical data.

     (response to 1.a.2)

     Actuarial judgment is required in selecting the patterns to apply to each segment of data being analyzed, and our views regarding current and future claim patterns are among the factors that enter into our establishment of the losses and LAE reserves at each balance sheet date. When short-term averages or external rate bureau analyses indicate that the claims patterns are changing from historical company or industry patterns, the new or forecasted information typically is factored into the methodologies. When new claims emergence or payment patterns have appeared in the actual data repeatedly over multiple evaluations, those new patterns are given greater weight in the selection process.

     Because some claims are paid over many years, the selection of claim emergence and payment patterns involves judgmentally estimating the manner in which recently occurring claims will develop many years and at times, decades in the future. When it is likely that the actual development will occur in the distant future, the potential for actual development to differ substantially from historical patterns or current projections is increased.

     This process assumes that past experience, adjusted for the effects of current developments and anticipated trends, is an appropriate basis for predicting future events. In particular, the development factor based methods all have as a key assumption that the development of losses in the future will follow a pattern similar to those measured by past experience and as adjusted either explicitly or by actuarial judgment. There is no precise method for subsequently evaluating the impact of any specific factor on the adequacy of reserves, because the eventual deficiency or redundancy is affected by multiple and varied factors. (response to 1.b.1 and 1.b.2) With respect to the ultimate estimates for losses and LAE, the key assumptions remained consistent for the years ended December 31, 2006 and 2005.

     Variability of Claim Reserve Estimates

     (response to 1.c)(Clarification for verbal question 2.c – dated January 14, 2008)

     By its nature, the estimate of ultimate loss and LAE is subject to variability due to differences between our assumptions and actual events in the future. Although many factors influence the actual cost of claims and our corresponding reserve estimates, we do not measure and estimate values for all of these variables individually. This is due to the fact that many of the factors that are known to impact the cost of claims cannot be measured directly, such as the impact on claim costs due to economic inflation, coverage interpretations and jury determinations. In most instances, we rely on our historical experience or industry information to estimate the

values for the variables that are explicitly used in our reserve analyses. We assume that the historical effect of these unmeasured factors, which is embedded in our experience or industry experience, is representative of the future effects of these factors. Where we have reason to expect a change in the effect of one of these factors, we perform analyses to perform the necessary adjustments.

     One implicit assumption underlying development patterns is that the claims inflation trends will continue into the future similar to their past patterns. To estimate the sensitivity of the estimated ultimate loss and settlement expense payments to an unexpected change in inflationary trends, the actuarial department derived expected payment patterns separately for each major line of business. These patterns were applied to the December 31, 2006 loss and settlement expense reserves to generate estimated annual incremental loss and settlement expense payments for each subsequent calendar year. Then, for the purpose of sensitivity testing, an explicit annual inflationary variance of one percent was added to the inflationary trend that is implicitly embedded in the estimated payment pattern, and revised incremental loss and settlement expense payments were calculated.

     (Verbal Question 2.a dated January 14, 2008)

     General inflation trends have been fairly stable over the past several years but experienced fluctuations of 1% to 2% over the past ten years and therefore we used a 1% annual inflation variance factor. The effect differed by line of business but overall was less than a 3% change in reserve adequacyor approximately $5.5 million effect on after tax net income. A variance of this type would typically be recognized in loss and settlement expense reserves and, accordingly, would not have a material effect on liquidity because the claims have not been paid.

     (Verbal Question 2.b dated January 14, 2008)

     An explicit assumption used in the analysis is the set of initial expected loss ratios (IELRs) used in the current accident year reserve projections and in some of the prior accident year ultimate loss indications. To estimate the sensitivity of the estimated ultimate loss to a change in IELRs, the actuarial department recasted the loss reserve indications using a set of IELRs all one percent higher than the final IELRs. The effect differed by line of business but overall was less than a 1% change in reserve adequacy or less than $2.0 million effect on after tax net income. Often the loss ratios by line of business will vary from the IELR in different directions causing them to partially offset each other. A variance of this type would typically be recognized in loss and settlement expense reserves and, accordingly, would not have a material effect on liquidity because the claims have not been paid.

     (Verbal Question 2.c dated January 14, 2008)

     The other factors having influence on the loss and LAE reserve levels are too numerous and interdependent to efficiently model and test for sensitivity. Likewise, the development factors by reserve category and age are too numerous to model and test for sensitivity. Instead, ranges are estimated by reserve category considering past history, patterns, recent developments, trends and other factors. The ranges are compiled and the total range is estimated considering the sensitivity to all of the underlying factors together. The resulting range is our best estimate of the expected ongoing variability in the loss reserves.

     Historic development as shown on the page 12 “Analysis of Loss and Loss Adjustment Expense Development” table has been 1% or less in the last two years but was 11% to 35% in the years prior to the underwriting and reserving shift in 2002. At that time, we concentrated our efforts on eliminating underwriting relationships where we had substantial liabilities above an aggregate exposure retained by risk sharing associations and captives. We also accelerated the pace at which we brought the claim administration to our Company employees and away from outside third party administrators.

     Our range of estimates on page 42 shows that presently we evaluate it as going from favorable development of 7.4% to unfavorable of about 5.1%. The range was evaluated as described above.

Pre-tax Impact on Earnings from a variance in Future Loss Payments
and Case Reserves (in Thousands)
	Minimum Reserve					Maximum
Line of Business	Range					Reserve Range
Workers’ Compensation (including Residual Markets)	$	(9,066	)	-5.6	%	$	7,174	4.4	%
Commercial Multiple Peril / General Liability	$	(5,806	)	-9.2	%	$	5,673	9.0	%
Commercial Automobile	$	(4,924	)	-9.0	%	$	1,407	2.6	%
Other	$	(2,572	)	-11.8	%	$	1,194	5.5	%
Total	$	(22,368	)	-7.4	%	$	15,448	5.1	%

     The sensitivity around our workers’ compensation reserves primarily reflects the size and the maturity of the underlying book of business. Our workers’ compensation reserves represent 53.9% percent of our total reserves at December 31, 2006. Workers’ compensation was our first line of business and is still our largest.

     The sensitivity around our Commercial Multiple Peril / General Liability reserves primarily reflect the longer duration of reserves relating to our liability excess program which started in 2003 and currently represents 40% of the $63.1 million reserves in this line of business, as of December 31, 2006.

     The sensitivity around our commercial automobile reserves primarily reflects the speed of reporting of the underlying losses, as well as the maturity of the case law surrounding automobile liability.

     The sensitivity around the other lines of business primarily reflects the size of the underlying book of business. Our other reserves represent 7.2% of total reserves at December 31, 2006. The majority of these reserves represent professional liability programs which tend to be claims-made and reinsured at lower limits, therefore reducing the volatility that is inherent in a smaller book of business.

     All of our reserves are sensitive to changes in the underlying claim payment and case reserving practices as well as the other sources of variations mentioned above.

     We hope you find the responses complete and sufficient in relation to the above comments. Please direct any questions or comments to Karen M. Spaun (248) 204-8178 or Robert S. Cubbin (248) 204-8031.

Sincerely,

Robert S. Cubbin	Karen M. Spaun
Chief Executive Officer	Chief Financial Officer