Title loss reserves
Our most critical accounting estimate is providing for title loss reserves. Our liability for estimated title losses at December 31, 2008 comprises both known claims ($135.1 million) and our estimate of claims that may be reported in the future ($326.4 million). The amount of the reserve represents the aggregate future payments (net of recoveries) that we expect to incur on policy and escrow losses and in costs to settle claims.
Provisions for title losses, as a percentage of title operating revenues, were 11.1%, 8.5% and 6.0% for the years ended December 31, 2008, 2007 and 2006, respectively. Actual loss payment experience, including the impact of large losses, is the primary reason for increases or decreases in our loss provision. A change of 100 basis points in this percentage, a reasonably likely scenario based on our historical loss experience, would have changed our provision for title losses and pretax earnings approximately $15.1 million for the year ended December 31, 2008.
Our method for recording the reserves for title losses on both an interim and annual basis begins with the calculation of our current loss provision rate, which is applied to our current premiums resulting in a title loss expense for the period. This loss provision rate is set to provide for losses on current year policies and is determined using moving average ratios of recent actual policy loss payment experience (net of recoveries) to premium revenues. Large losses (those exceeding $1.0 million on a single claim) are analyzed and reserved for separately due to the higher severity of loss, lower volume of claims reported and sporadic reporting of such claims.
At each quarter end, our recorded reserve for title losses is initially the result of beginning with the prior period's reserve balance for claim losses, adding the current period provision to that balance and subtracting actual paid claims, resulting in an amount that our management compares to its actuarially-based calculation of the ending reserve balance. The actuarilly-based calculation is a paid loss development calculation where loss development factors are selected based on company data and input from our third-party actuaries. We also obtain input from third-party actuaries in the form of a reserve analysis utilizing generally accepted actuarial methods. While we are responsible for determining our loss reserves, we utilize this actuarial input to assess the overall reasonableness of our reserve estimation. If our recorded reserve amount is within a reasonable range of our actuarially-based reserve calculation and the actuary's point estimate (+/- 3.0%), but not at the point estimate, our management assesses the major factors contributing to the different reserve estimates in order to determine the overall reasonableness of our recorded reserve, as well as the position of the recorded reserves relative to the point estimate and the estimated range of reserves. The major factors considered can change from period to period and include items such as current trends in the real estate industry (which management can assess although there is a time lag in the development of this data for use by the actuary), the size and types of claims reported and changes in our claims management process. If the recorded amount is not within a reasonable range of our third-party actuary's point estimate, we will adjust the recorded reserves in the current period and reassess the provision rate on a prospective basis. Due to the inherent uncertainty in predicting future title policy losses, significant judgment is required by both our management and our third party actuaries in estimating reserves. As a consequence, our ultimate liability may be materially greater or less than current reserves and/or our third party actuary's calculation.
Agency revenues
We recognize revenues on title insurance policies written by independent agencies (agencies) when the policies are reported to us. In addition, where reasonable estimates can be made, we accrue for revenues on policies issued but not reported until after period end. We believe that reasonable estimates can be made when recent and consistent policy issuance information is available. Our estimates are based on historical reporting patterns and other information about our agencies. We also consider current trends in our direct operations and in the title industry. In this accrual, we are not estimating future transactions. We are estimating revenues on policies that have already been issued by agencies but not yet reported to or received by us. We have consistently followed the same basic method of estimating unreported policy revenues for more than 10 years.

Our accruals for revenues on unreported policies from agencies were not material to our consolidated assets or stockholders' equity at December 31, 2008 and 2007. The differences between the amounts our agencies have subsequently reported to us compared to our estimated accruals are substantially offset by any differences arising from prior years' accruals and have been immaterial to consolidated assets and stockholders' equity during each of the three prior years. We believe our process provides the most reliable estimation of the unreported revenues on policies and appropriately reflects the trends in agency policy activity.
Goodwill and other long-lived assets
Our evaluation of goodwill is completed annually in the third quarter using June 30 balances or when events may indicate impairment. This evaluation is based on a combination of a discounted cash flow analysis (DCF) and market approaches that incorporate market multiples of comparable companies and our own market capitalization. The DCF model utilizes historical and projected operating results and cash flows, initially driven by estimates of changes in future revenue levels, and risk-adjusted discount rates. Our projected operating results are primarily driven by anticipated mortgage originations, which we obtain from projections by industry experts. Fluctuations in revenues, followed by our ability to appropriately adjust our headcount and other operating expenses, are the primary reasons for increases or decreases in our projected operating results. For example, in our current projections, a change of 5% in revenues in year one of our projections would result in an estimated $4.1 million change in pretax operating results and could potentially change the fair value of our title reporting unit by $23.2 million. Our market-based valuation methodologies utilize (i) market multiples of earnings and/or other operating metrics of comparable companies and (ii) our market capitalization and a control premium based on market data and factors specific to our corporate governance structure. As a result of current market conditions, overall market volatility, including our market capitalization and our operating results, we updated our evaluation of goodwill through December 31, 2008. Based upon our updated evaluation, we have concluded that our goodwill is not impaired as of December 31, 2008. However, to the extent that our future operating results are below our projections, or in the event of continued adverse market conditions, a future review for impairment may be required.
We evaluate goodwill based on two reporting units (Title and REI). Goodwill is assigned to these reporting units at the time the goodwill is initially recorded. Once assigned to a reporting unit, the goodwill is pooled and no longer attributable to a specific acquisition. All activities within a reporting unit are available to support the carrying value of the goodwill. The following reflects our conclusions relating to our goodwill reporting units at December 31, 2008:

                                                       Title       REI
                                                         ($ millions)

              Fair value (1)                           672.8       66.7

              Carrying value (1)                       616.3       27.5

              Goodwill, included in carrying value     196.7       14.1

(1) comprised of debt and equity values

In addition, the Company considered the carrying value of the Company's stockholders' equity as compared with the Company's market capitalization at year end and the implied control premium to reconcile these amounts:

                                                                                 Title
                                                                             ($ millions)

Total stockholders' equity                                                         493.8

Market capitalization based on 5-day average stock price at
December 31, 2008                                                                  398.9

Implied control premium                                                             23.8 %

We believe this implied control premium is reasonable based on average premiums for transactions in the insurance industry, as well as the unique characteristics of our Class B Common Stock.
We also evaluate the carrying values of title plants and other long-lived assets when events occur that may indicate impairment. The process of determining impairment for our goodwill and other long-lived assets relies on projections of future cash flows, operating results, discount rates and overall market conditions, including our market capitalization. Uncertainties exist in these projections and are subject to changes relating to factors such as interest rates and overall real estate and financial market conditions, our market capitalization and overall stock market performance. Actual market conditions and operating results may vary materially from our projections.
Based on these evaluations, we estimate and expense to current operations any loss in value of these assets. As part of our process, we obtain input from third-party appraisers regarding the fair value of our reporting units. While we are responsible for assessing whether an impairment of goodwill exists, we utilize the input from third-party appraisers to assess the overall reasonableness of our conclusions. In June 2008, the Company's REI segment incurred an impairment charge of $6.0 million, included in depreciation and amortization in our consolidated statement of earnings, relating to its internally developed software that we subsequently determined will not be deployed into production. There were no impairment write-offs of goodwill or other long-lived assets during 2007 or 2006.
Operations. Our business has two main operating segments: title insurance-related services and real estate information (REI). These segments are closely related due to the nature of their operations and common customers. Our primary business is title insurance and settlement-related services. We close transactions and issue title policies on homes and commercial and other real properties located in all 50 states, the District of Columbia and international markets through policy-issuing offices and agencies. We also provide post-closing lender services, automated county clerk land records, property ownership mapping, geographic information systems, property information reports, document preparation, background checks and expertise in Internal Revenue Code Section 1031 tax-deferred property exchanges.
Factors affecting revenues. The principal factors that contribute to changes in operating revenues for our title and REI segments include:
• mortgage interest rates;

• ratio of purchase transactions compared with refinance transactions;

• ratio of closed orders to open orders;

• home prices;

• consumer confidence;

• demand by buyers;

• number of households;

• availability of loans for borrowers;

• premium rates;

• market share;

• opening of new offices and acquisitions; and

• number of commercial transactions, which typically yield higher premiums.

To the extent inflation causes increases in the prices of homes and other real estate, premium revenues are also increased. Premiums are determined in part by the insured values of the transactions we handle. These factors may override the seasonal nature of the title insurance business. Generally, our first quarter is the least active and our fourth quarter is the most active in terms of title insurance revenues.

Industry data. Published mortgage interest rates and other selected residential data for the years ended December 31, 2008, 2007 and 2006 follow (amounts shown for 2008 are preliminary and subject to revision). The amounts below may not relate directly to or provide accurate data for forecasting our operating revenues or order counts.
Our statements on home sales, mortgage interest rates and loan activity are based on published industry data from sources including Fannie Mae, the National Association of Realtorsฎ, the Mortgage Bankers Association and Freddie Mac.

                                                            2008        2007        2006

Mortgage interest rates (30-year, fixed-rate) - %
Averages for the year                                        6.04        6.34        6.41
First quarter                                                5.88        6.22        6.24
Second quarter                                               6.09        6.37        6.60
Third quarter                                                6.32        6.55        6.56
Fourth quarter                                               5.87        6.23        6.25

Mortgage originations - in $ billions                       1,877       2,543       2,761
Refinancings - % of originations                             50.3        50.8        47.6

New home sales - in thousands                                 483         776       1,051
Existing home sales - in thousands                          4,913       5,652       6,478
Existing home sales - median sales price in $ thousands     198.1       219.0       221.9

Most industry experts project mortgage interest rates to remain low in 2009. Refinancing mortgage originations increased late in 2008 and early 2009 due to lower interest rates and government efforts to strengthen credit markets. Industry experts agree that mortgage originations will increase moderately in 2009 mainly due to refinancing originations, but decrease in 2010.
Although mortgage originations are expected to increase in 2009, Fannie Mae and The Mortgage Bankers Association anticipate continued difficulties for the real estate market due to the tightened credit market, decreased housing prices and low housing starts. Therefore, our financial condition and results of operations will continue to be adversely affected by the current market conditions. Trends and order counts. For the three years ending 2008, mortgage interest rates (30-year, fixed-rate) have fluctuated from a monthly high of 6.8% in July 2006 to a monthly low of 5.3% in December 2008. As a result of the above, and the effects of the collapsed subprime mortgage lending market and limited availability of credit, mortgage originations decreased 32.0% from 2006 through 2008. Sales of new and existing homes decreased 54.1% and 24.2%, respectively, in this same two-year period.
As a result of the above trends, our direct order levels have decreased significantly from 2006 to 2008, which is consistent with the decline of the U.S. real estate market.
The number of direct title orders we opened follows (in thousands):

                                          2008      2007      2006
                        First quarter      151       173       193
                        Second quarter     130       179       202
                        Third quarter      110       152       183
                        Fourth quarter     101       128       162

                                           492       632       740

The number of direct title orders we closed follows (in thousands):

                                          2008      2007      2006

                        First quarter       90       110       132
                        Second quarter      93       125       147
                        Third quarter       79       107       136
                        Fourth quarter      66        93       124

                                           328       435       539

Regulatory and legal developments. In June 2008, the California Department of Insurance released for public notice and comment revised regulations that place certain limits on payments by title insurance marketing representatives to agents and brokers, eliminate a previously proposed interim rate reduction and a maximum rate formula, and substantially scale back the proposed financial data requirements on title insurance companies. The final regulations are expected to be issued by July 1, 2009.
We cannot predict the outcome of proposed regulations. However, to the extent that rate decreases are mandated in the future, the outcome could materially affect our consolidated financial condition or results of operations. We are subject to administrative actions and litigation relating to the basis on which premium taxes are paid in certain states. Additionally, we have received various other inquiries from governmental regulators concerning practices in the insurance industry. Many of these practices do not concern title insurance and we do not anticipate that the outcome of these inquiries will materially affect our consolidated financial condition or results of operations.
We are also subject to various other administrative actions and inquiries into our conduct of business in certain of the states in which we operate. While we cannot predict the outcome of the various regulatory and administrative matters referenced above, we believe that we have adequately reserved for these matters and that any outcome will not materially affect our consolidated financial condition or results of operations.
Stewart Title of California, Inc., our subsidiary, is a defendant in four putative class action lawsuits filed in California state and federal courts. These lawsuits are commonly referred to as "wage and hour" lawsuits. These lawsuits generally claim, among other things, that (i) the plaintiffs were misclassified as exempt employees and were not paid overtime, (ii) the overtime payments made to non-exempt employees were miscalculated and (iii) the plaintiffs worked overtime hours, but were not paid. The plaintiffs seek compensatory damages, statutory compensation, penalties and restitution, exemplary and punitive damages, declaratory relief, interest and attorneys fees. We are seeking to consolidate the two federal court cases. All of these cases are in the discovery stage, and their outcomes cannot be predicted with certainty at this time. We intend to vigorously defend ourselves against the allegations. We do not believe that the outcomes will materially affect our consolidated financial condition or results of operations.
On January 12, 2009, a lawsuit was filed against four defendants, including our subsidiaries, Stewart Title Guaranty Company and Stewart Title of California, Inc., entitled Wooldridge, et al v. Stewart Title Guaranty Co., et al, Superior Court of the State of California for the County of San Luis Obispo. The plaintiffs allege that they suffered damages relating to loans or interests in loans they made in connection with several properties in San Luis Obispo County. The plaintiffs assert causes of action against our subsidiaries for breach of contract, negligence, and violation of fair claims settlement practices and breach of covenants of good faith and fair dealings. Although we cannot predict the outcome of this action, we intend to vigorously defend ourselves against the allegations and do not believe that the outcome will materially affect our consolidated financial condition or results of operations.

In February 2008, an antitrust class action was filed in the United States District Court for the Eastern District of New York against Stewart Title Insurance Company, Monroe Title Insurance Corporation, Stewart Information Services Corporation (SISCO), several other unaffiliated title insurance companies and the Title Insurance Rate Service Association, Inc. (TIRSA). The complaint alleges that the defendants violated Section 1 of the Sherman Act by collectively filing proposed rates for title insurance in New York through TIRSA, a state-authorized and licensed rate service organization. Complaints were subsequently filed in the federal district courts for the Eastern and Southern Districts of New York and federal district courts in Pennsylvania, New Jersey, Ohio, Florida (since dismissed), Massachusetts, Arkansas, California, Washington, West Virginia, Texas and Delaware. All of the complaints make similar allegations, except that certain of the complaints also allege violations of RESPA statutes and various state consumer protection laws. The complaints generally request treble damages in unspecified amounts, declaratory and injunctive relief, and attorneys' fees. At least 72 such complaints are currently pending (and have been consolidated in the aforementioned jurisdictions), each of which names SISCO and/or one or more of our affiliates as a defendant. Although we cannot predict the outcome of these actions, we intend to vigorously defend ourselves against the allegations and do not believe that the outcome will materially affect our consolidated financial condition or results of operations.
We are also subject to lawsuits incidental to our business, most of which involve disputed policy claims. In many of these lawsuits, the plaintiff seeks exemplary or treble damages in excess of policy limits based on the alleged malfeasance of an issuing agency. We do not expect that any of these proceedings will have a material adverse effect on our consolidated financial condition or results of operations. Along with the other major title insurance companies, we are party to a number of class action lawsuits concerning the title insurance industry. We believe that we have adequate reserves for the various litigation matters and contingencies discussed above and that the likely resolution of these matters will not materially affect our consolidated financial condition or results of operations.
Results of Operations
A comparison of our results of operations for 2008 with 2007 and 2007 with 2006 follows. Factors contributing to fluctuations in our results of operations are presented in the order of their monetary significance and we have quantified, when necessary, significant changes. Results from our REI segment are included in our year-to-year discussions as those amounts are immaterial in relation to consolidated totals. When relevant, we have discussed our REI segment's results separately.
Title revenues. Our revenues from direct title operations decreased $240.6 million, or 25.4%, in 2008 and $81.3 million, or 7.9%, in 2007. The largest revenue decreases in 2008 were in California, Texas, New York, Florida, and Washington, partially offset by increases in New Jersey and Pennsylvania. The largest revenue decreases in 2007 were in California, Florida, Nevada and Arizona, partially offset by increases in Canada, New York and Texas. Acquisitions increased revenues $3.5 million and $22.1 million in 2008 and 2007, respectively. Revenues from commercial and large transactions decreased $53.9 million to $117.0 million in 2008 after increasing $22.6 million to $170.9 million in 2007. In 2008, the decreases in residential and commercial title revenues were a result of the dramatic decline in the U.S. real estate market.
The number of our direct closings, excluding large commercial policies, decreased 24.6% in 2008 and 19.2% in 2007. However, the average revenue per closing, excluding large commercial policies, was comparable in 2008 and increased 9.4% in 2007. . . .