February 13, 2009
VIA EDGAR AND OVERNIGHT COURIER
Mark Webb,
Legal Branch Chief
United States Securities and Exchange Commission
Division of Corporation Finance
Mail Stop 4561
100 F Street, N.E.
Washington, D.C. 20549
Re: | Boston Private Financial Holdings, Inc. |
Form 10-K for December 31, 2007 and Related Filings
File Number 0-17089
Dear Mr. Webb:
This letter is submitted on behalf of Boston Private Financial Holdings, Inc. (the “Company” or the “Holding Company”) in response to the comments of the staff of the Division of Corporation Finance (the “Staff”) of the Securities and Exchange Commission (the “Commission”) with respect to our Annual Report on Form 10-K for the year ended December 31, 2007 (the “2007 Form 10-K”), and our Quarterly Report on Form 10-Q for the Quarterly Period ended September 30, 2008 (the “Form 10-Q”) (together the “Filings”), all as set forth in your letter dated February 6, 2009 to David J. Kaye (the “Comment Letter”).
For ease of reference, the relevant portion of the text of the Comment Letter with the Staff’s comments has been reproduced herein with management’s responses set forth below each numbered comment.
Form 10-K for the Fiscal Year Ended December 31, 2007:
Management’s Discussion and Analysis of Financial Condition and Results of Operations
Financial Condition
1. We note your response to comment one of our letter dated December 18, 2008. With respect to your description of appraisal practices, please describe for us the mitigating procedures for ensuring appropriate valuations of property and related valuation allowances prior to and as of December 31, 2007.
Response:
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
Prior to December 31, 2007, First Private Bank & Trust (“First Private”) had stated appraisal practices which called for the frequent review and reappraisal of troubled credits as well as performing credits where the “deterioration of economic conditions in an overbuilt area may prompt the Bank to conduct an overall review of its real estate collateral.” The existing loan policy also required First Private management to assess the completeness of the appraisals and make adjustments to the values in order to account for such factors as the passage of time, the volatility of the local market, the availability of financing, the inventory of competing properties, new improvements to or lack of maintenance on the subject property, and progress on the existing property. First Private was also subject to the annual review by a local, independent third party loan review firm to assess the effectiveness of its appraisal procedures and associated property valuations.
During the fourth quarter of 2007, home prices fell rapidly in Southern California with the Case-Shiller index registering declines of approximately 9% in the Los Angeles area just in that quarter. In conjunction with a routine FDIC safety and soundness examination conducted in January 2008, the Company determined that First Private management had failed to follow established procedures, resulting in an inadequate allowance for loan loss initially being established for the quarter ended December 31, 2007. Three key members of First Private’s management team (the CEO, Chief Credit Officer, and Chief Construction Lending Officer) were terminated, and replaced in the interim by executives from the Company and Borel Private Bank & Trust Company, a subsidiary of the Company located in Northern California.
Recognizing that there was an initial, inadequate allowance for the fourth quarter of 2007, the Company undertook a comprehensive review of First Private’s land and construction portfolio, using the work performed by the FDIC staff in their examination as a starting point. As part of this review, and before the Company filed its Form 10-K for the period ending December 31, 2007, updated appraisals were gathered, and the Company’s Chief Risk Officer, a retired FDIC examiner hired as a consultant by the Company, and the Chief Credit Officer from Borel Private Bank further examined the portfolio and related appraisals as an added layer of compensating controls. Roughly 75% ($190 million of the $260 million) of the construction and land portfolio and 20% of the commercial loan portfolio was reviewed. These loans were selected for review because they represented the largest and riskiest portion (e.g.,raw land) of the portfolio. Approximately $98 million of loans (including the unfunded balance) were downgraded to either substandard or doubtful, with 98% of the downgrades coming from the construction and land portfolio which was concentrated in the Inland Empire.
Approximately $70 million of the $98 million in downgraded loans were subject to FAS 5 analyses and the results added approximately $11 million of specific reserves for this portion of the portfolio. The specific reserve percentage ranged from 15 to 20%, and was based on several factors including competition, recent changes in home prices, home sales, and foreclosure rates. The appraisals indicated an average loan to value of 65%, so with an additional 15% reserve, management felt comfortable that even a 50% decline in property values would leave an adequate allowance for loan loss. Of the loan balances subject to the FAS 5 analyses, 85% had updated appraisals dated between July 2007 and March 2008 when the Company filed its Form 10-K on March 14, 2008. The remaining $28 million of the $98 million
2
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
in downgraded loans were subject to FAS 114 analyses, and specific reserves of approximately $7 million were established based on updated appraisals, discounted cash flows, and other market related information.
In addition, management established unallocated reserves of $4 million to take into account the remaining portion of the construction portfolio that was not reviewed, and the continuing deterioration of economic conditions within Southern California. The Company’s banking affiliates generally have an unallocated reserve policy accounting for 10 to 20 basis points of their overall loan portfolio, and while the First Private unallocated reserve was nearly 80 basis points, management felt this heightened reserve level was justified given the nature of the portfolio and economic conditions. In summary, taking into account the updated appraisals received on the bulk of the construction portfolio, the additional reserves applied as a result of the FAS 5 and FAS 114 analyses, and the increased level of unallocated reserves, management believed that the allowance for loan loss reserve was appropriate. As of December 31, 2007, First Private had an allowance to loan ratio of over 4.3% which was well above industry averages, but reasonable when considering the level of classified loans.
Based on the comprehensive review of the portfolio described above, the FAS 114 analyses that were performed, and the increase to the unallocated portion of the reserves, a significant increase to the allowance was recorded for First Private for the quarter ended December 31, 2007, and the Company filed its Form 10-K on March 14, 2008 with this new analysis in place.
It is important to recognize that while this particular portfolio would incur significant charge-offs subsequently in 2008, these charge-offs were directly related to the continued significant and precipitous deterioration of real estate values experienced during the course of 2008. According to the California Association of Realtors, from the fourth quarter of 2007 to the fourth quarter of 2008, the statewide median price of an existing home fell 42%. The value of First Private’s construction and land portfolio fell even more dramatically for two distinct reasons. First, raw land values fall faster than existing home prices due to the inherent increased risks associated with construction and development. Second, First Private’s portfolio was sold to investors who require significant profit margins to compensate for assuming the risk of the debt. The impact of the profit margins would not have been factored into the value of the portfolio as of December 31, 2007 since management had the intention to hold these assets until maturity at that point.
See our response to comment (4) for a further discussion of First Private’s construction and land portfolio.
2.We note your response to comment one of our letter dated December 18, 2008. We note your statement that a revised allowance for loan loss methodology was adopted at all five banks beginning at September 30, 2008 and fully implemented at December 31, 2008. Please clarify for us and in future filings the actual changes to your methodology and the impact on your allowance for loan losses, quantifying the impact if applicable.
3
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
Response:
In December of 2006, the Federal Financial Institutions Examination Council (“FFIEC”) issued a revised Allowance for Loan Loss (“ALL”) Policy Statement (the “Policy”). Prior to the new Policy, our Banks generally used industry benchmark factors by loan type to support the general allowance. Our Banks had very low amounts of actual loan losses in recent years which made it difficult to rely only on internal data. The new Policy does not allow banks to rely only upon industry benchmarks. As a result of the new Policy, the Banks changed their methodology to start with FDIC historical loss rates by loan type adjusted for the following qualitative factors:
• | Loss and recovery trends |
• | Problem loan (delinquencies) trends |
• | Non accrual trends |
• | Quality of loan review system |
• | Nature and volume of the loan portfolio |
• | Lending policies and procedures |
• | Experience, ability, and depth of management |
• | Concentrations of credit |
• | Other miscellaneous adjustments |
As a result of this change to the methodology, each one of the above qualitative factors is assigned a loss rate (percentage), if applicable. Each Bank makes an independent determination of the applicable loss rate for these factors based on the relevant local market conditions, credit quality, and portfolio mix. Each quarter, all of the Banks review the loss factors to determine if there have been any changes which would result in a change to the current loss rate. In addition, the Banks changed their methodology related to the unallocated portion of the ALL to provide more emphasis on the overall economic conditions within their local market. The most significant changes our Banks made in the methodology were in documenting the above factors to support the loss rates used.
Starting in 2007 the Banks began implementing these changes. Since 2007, as our Banks were subject to their routine FDIC or OTS examinations, they further revised their ALL policies and procedures based on guidance and feedback received from their respective regulators. In summary, changes were made over the course of late 2007 through 2008 to further enhance the individual bank’s compliance with this new Policy as indicated above.
Further, the Company believes that the financial impact of the new Policy on the allowance for loan loss was not material. In each of the quarters that the five Banks adopted the new Policy, there was no material change in the allowance to reflect the updated factors. The material changes to the allowance for loan losses recorded during 2008 were the result of loan deterioration and not the new Policy being implemented as required by the FFIEC.
The Company will disclose the methodology changes in future filings.
4
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
3.We note your response to comment one of our letter dated December 18, 2008. Please describe for us and disclose in future filings your new OREO policy that was adopted. Describe the critical changes in your OREO policy and the impact on your financial statements, quantifying the impact if applicable.
Response:
The new OREO policy at First Private is a three page document which covers the following key points:
• | Determination of the appropriate recording of OREO |
• | Identification of the minimum standards for an appraisal of an OREO property |
• | Identification of the maximum length of time OREO can be held. |
• | Identification of the proper accounting procedures of OREO |
The critical changes include the definition of fair value of the OREO as the cash price that might reasonably be anticipated in a current sale (within 12 months), or should a current sale be unlikely, all cash flows generated by the property that must be discounted to obtain the estimate of fair value. Another critical change states the requirement that all OREO valuations must be reviewed and approved by the Chief Credit Officer and reported to the board. The final critical change covers the requirement that any charges resulting from the differences in fair value and carrying value must be approved by the Chief Financial Officer. The former OREO policy was silent on each of these matters, so quantifying a specific impact on the financial statements is not applicable. However, as of December 31, 2007, First Private had only one OREO property with a carrying value of approximately $700,000, so any policy change would not have had a material impact on our financial statements.
The Company will disclose the policy in future filings.
Form 10-Q for the Quarterly Period ended September 30, 2008:
Consolidated Financial Statements
Note 7 – Allowance for loan Losses, Page 17
4.We note your response to comment ten of our letter dated December 18, 2008. Please tell us the following additional information regarding the portfolio of loans that was transferred to the Loans Held for Sale Category at September 30, 2008:
a) | quantify the amount of “general” allowance related to the portfolio for loans that were not determined to be impaired as of June 30, 2008; |
b) | describe how management considered the portfolio’s risk profile, deteriorating local economic trends, and the increasing level of classified loans in determining the amount of the “general” allowance related to the portfolio; |
c) | describe any adjustments made by management to appraisals, separately addressing appraisals obtained in 2008 and those obtained prior to 2008; and |
5
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
d) | describe how you evaluated the collateral for properties for which you did not obtain an updated appraisal in 2008, separately addressing performing and non-performing loans. |
Response:
a) The portfolio transferred was largely impaired loans. There were only seven loans totaling $39 million that were not impaired. The general allowance for those loans was approximately $3.3 million (8% coverage) as of June 30, 2008.
b) In assessing the portfolio’s overall risk profile, management assessed the adequacy of the allowance for loan losses on a monthly basis. Loans that were not impaired were assigned general reserve percentages based on the bank’s historical and projected losses. The reserve percentage was determined by first examining actual loss history for each type of loan, then adjusting for a variety of factors including changes in lending policies; changes in experience, ability and depth of lending management and staff; changes in trends of past due and classified loans; and concentrations of credit.
In determining the general allowance noted above, management also considered the continuing housing slump and deterioration in the Southern California real estate market which had unfavorably impacted our homebuilder borrowers and the value of the real properties securing their loans. According to the California Association of Realtors, existing home prices were dropping 3 – 5% per month and, on a statewide basis, dropped 42% on average from November 2007 to November 2008. The reduced values affected the value of land most severely, as both the housing values declined and the value of land as a percent of the overall value of the house also declined.
Further, management focused on the deteriorating local economic trends, given that its loan portfolio was concentrated in the Inland Empire and in land and construction loans. This focus was driven by the fact that a prolonged or deeper decline in the housing market would impact First Private’s homebuilder borrowers, and would both increase the likelihood of defaults and reduce the value of the real property collateral.
In light of the credit and mortgage crisis affecting the entire financial industry and its impact on First Private’s borrowers, management took a very proactive approach to assessing the risk profile of the specific credits within the portfolio, which resulted in increased levels of classified loans. The existence of certain warning signs indicating possible collectability issues warranted a more careful scrutiny of particular loans for potential impairment and an increase in the general allowance. Specifically, management identified loans that exhibited a number of characteristics including, but not limited to, the following:
• | diminishing or adverse changes in cash flows; |
• | adverse changes in the financial position or net worth of guarantors; |
• | adverse changes and trends in collateral values; |
• | declining or adverse changes in inventory levels securing commercial businesses; |
• | failure to meet financial covenants; |
6
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
Accordingly, consistent with falling property values, increasing classified loans, and worsening economic conditions, management added approximately $36 million of additional provision and charged-off an additional $22 million, which combined, covered over 20% of First Private’s construction and land portfolio during the first half of 2008.
c) Most adjustments to appraised values were made regardless of the appraisal date (i.e., pre- or post-January 2008) and included a variety of factors such as:
• | the costs of disposition and exercise of foreclosure; |
• | reduced absorption rates as housing sales slowed or, in many cases, stopped; |
• | increased discounts rates as risk increased reducing projected cash flows; |
• | the age of comparables included in the appraisal; |
• | known changes in the market and in the collateral, or observable facts that management deemed were not supportable. |
For appraisals obtained prior to 2008 or early in 2008, management also allowed for adjustments that reflected observed home prices dropping at a rate of 3 – 5% per month since the date of the most recent appraisal.
d) When evaluating the collateral values where appraisals were obtained prior to 2008, management specifically adjusted both performing and non-performing loans when observable facts suggested the legacy value was inaccurate or a trend analysis resulted in supporting a different value. In order to discover observable facts, management obtained research, studied market data, spoke to third party professionals regarding trends, analyzed these trends, reviewed comparables, analyzed pending and completed transactions, and utilized management’s own experience in evaluating collateral values.
Performing loans, which represented 96% of the balances for pre-2008 appraisals, were evaluated with the expectation that repayment would come from the sale of properties as construction was completed or the refinancing of property once adequately leased. Therefore, when evaluating the collateral, management considered falling home prices, reduced absorption rates, increased discount rates, and other factors listed above. When evaluating the collateral of the non-performing loans, management also considered the costs of disposition and foreclosure.
Conclusion
Despite the fact that the assumptions listed above drove large increases in the loan loss provision as of June 30, 2008, this portfolio then incurred significant additional write-downs in the subsequent quarter as the direct result of two primary factors. First and foremost, management’s decision in the third quarter to immediately sell the portfolio during a volatile and illiquid market meant that the portfolio would essentially be valued at “fire-sale” prices, rather than the inherent long-term value assumptions used at June 30, 2008. Second, the overall market conditions in the third quarter dramatically deteriorated due to the FDIC seizure of IndyMac, the bankruptcy filing of Lehman Brothers, and the flood of loans put up for sale by banks such as Washington Mutual and Vineyard Bank. These events had a major and direct impact on the construction and land portfolios in Southern California because each of these institutions had significant concentrations of holdings in that particular region.
7
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
Finally, it should be noted that management successfully sold off the majority of the portfolio during the fourth quarter of 2008. The estimates of fair value recorded as of September 30, 2008 proved to be highly accurate as the subsequent fourth quarter adjustment was less than $2 million. To further illustrate the continued and extreme volatility of the market conditions, the third party agent that led the sale of the loans noted in January 2009 that had the Company delayed the sale of the First Private portfolio by just two months, we likely would have received sales proceeds of 20 – 25% less, overall.
Note 8 – Goodwill and Intangible Assets, page 19
5. We note your response to comment three of our letter dated December 18, 2008. Please provide us with your goodwill impairment analysis at your Investment Management and Wealth Advisory segments as of December 31, 2008. In your analysis please address the following:
(a) | provide the actual multiples and market comparables used in your analyses; |
(b) | explain how those multiples and market comparables are similar to your reporting unit, in regards to nature, scope and size of operations, as contemplated in paragraph 25 of SFAS 142; |
(c) | provide the reporting unit estimated fair value as determined by the comparable market method; |
(d) | provide the reporting unit estimated fair value as determined by the discounted cash flow method; and |
(e) | whether you use a simple, weighted, or some other type of average when determining the reporting unit estimated fair value; |
(f) | explain how you evaluate and weight the results of the comparable market method and the discounted cash flow method in considering the reasonableness of the range indicated by those results; and |
(g) | explain how you determine the estimated fair value of the reporting unit as the point within that range that is most representative of fair value in the circumstances. |
Response:
Background
In the third quarter of 2008, the Company determined that there was a triggering event under FASB Statement No. 142Goodwill and Other Intangible Assets(“FAS 142”) due to a significant adverse change in the business climate and performed interim impairment testing, which resulted in an estimated impairment loss of approximately $34.5 million in the investment management and wealth advisory segments. Management then conducted its annual test of goodwill for impairment of the investment management and wealth advisory segments in the fourth quarter of 2008. As a result of this testing, the Company recorded an additional $14 million of goodwill impairment in these segments. The tests were performed at the affiliate level, which were determined to be reporting units under the Statement.
8
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
To begin, management performed an impairment scoping exercise (e.g. Step 1 of FAS 142) as part of its annual testing that was used to determine which affiliates would undergo a third-party income and market approach valuation. For this scoping exercise, management used a very conservative (25th percentile) market multiple and applied it to the trailing three months of EBITDA to calculate fair values. Management decided to use the trailing three months, rather than the trailing twelve month Price/EBITDA multiples used in previous testing, in order to use a very conservative estimate for the initial review, as that period best reflected the current negative economic market conditions.
As a result of this initial testing, if the fair value of a particular affiliate was significantly greater than its book value, management concluded that no further testing was necessary because the initial testing had been based on decidedly conservative multiples. Using a forward EBITDA multiple of 5.9 times (the 25th percentile market multiple), Anchor Holdings, LLC (“Anchor”), KLS Professional Advisors Group LLC (“KLS”) and Boston Private Value Investors, Inc. (“BPVI”) had sizeable cushions where fair value exceeded book value by factors ranging from 10 – 30% without any control premium, and thus were deemed not to be impaired.
If the fair value was close to or less than book value, management concluded that a third-party valuation firm should be retained to conduct income and market approach testing as of October 31, 2008. Accordingly, management, utilizing a third party valuation firm, then performed a Step 1 impairment valuation on the following affiliates that had estimated fair values less than their book values as of October 31, 2008 (testing date):
¨ | Dalton Greiner Hartman Maher & Co LLC (“DGHM”) |
¨ | Davidson Trust Company (“DTC”) |
¨ | Bingham Osborn & Scarborough LLC (“BOS”) |
¨ | Sand Hill Advisors, LLC (“SHA”) |
¨ | Coldstream Capital Management, Inc. (“CCM”) |
Management and the third party valuation firm utilized both the market and income approaches and calculated the present value of future cash flows in determining the fair value of each reporting unit. The resulting overall valuation was then compared to multiples of recent acquisitions to assess the reasonableness of the overall valuation.
9
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
(a) | In response to the first sub-question you have noted, the following chart shows the third party valuation multiples used as of October 31, 2008 |
Third-Party Valuation Multiples used as of October 31, 2008
MV/ REVENUE | MV/ EBITDA | |||||||
Forward | Forward | |||||||
LTM | FY1 | LTM | FY1 | |||||
Blackrock, Inc. | 3.3 | 3.4 | 9.3 | 9.2 | ||||
EatonVance Corp. | 2.5 | 2.6 | 7.0 | 7.0 | ||||
Federated Investors, Inc. | 2.1 | 2.1 | 6.7 | 6.7 | ||||
Gamco Investors, Inc. | 3.4 | 4.0 | 8.9 | 10.9 | ||||
Waddell & Reed Financial | 1.2 | 1.3 | 5.2 | 5.2 | ||||
Janus Capital Group, Inc | 2.2 | 2.4 | 6.5 | 6.9 | ||||
Affiliate Managers Group, | 2.5 | 2.8 | 6.1 | 8.7 | ||||
Franklin Resources Inc. | 1.9 | 2.2 | 4.9 | 5.1 | ||||
T.Rowe Price Group, Inc | 4.3 | 4.6 | 9.5 | 10.6 | ||||
High | 4.3 | 4.6 | 9.5 | 10.9 | ||||
Median | 2.5 | 2.6 | 6.7 | 7.0 | ||||
Average | 2.6 | 2.8 | 7.1 | 7.8 | ||||
Low | 1.2 | 1.3 | 4.9 | 5.1 | ||||
Market Multiples used by Management: | ||||||||
DGHM | n/m | 1.3 | n/m | 5.0 | ||||
DTC | 1.1 | 1.1 | n/m | n/m | ||||
BOS | 2.5 | 3.0 | 7.5 | 8.0 | ||||
SHA | 2.0 | 2.0 | n/m | n/m | ||||
CCM | 0.75 | 1.0 | n/m | n/m |
(b) In applying the Guideline Company method, we compared each affiliate to publicly traded companies in industries that have similar growth prospects and risks to that affiliate. To search for appropriate guideline companies, we screened various databases for publicly traded companies engaged in similar activities, including those in the standard industrial classification of “investment services.” We refined our search by reviewing the business descriptions of each selected company, selecting those closest to the asset management business that were classified under investment advice by SIC code 6282. Based on this process, we selected nine publicly traded companies that we believe are similar to the investment managers and wealth advisors in the Company’s segments to be used for comparative purposes.
Consistent with paragraph 25 of SFAS 142, management believes that our reporting units have comparable operations and similar economic characteristics to the selected market
10
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
multiples. We noted that the nature and scope of the market multiples (e.g. investment managers with a concentration of domestic equities, both institutional and high net worth clients, etc.) were especially similar. However, we noted that our reporting units are smaller in size than the selected market comparable companies. Given the small variance between the income approach values and market comparables, we believe the market comparables support the income approach values. In addition, the comparable company analysis provided a wide range of multiples and we selected low to average multiples within the range. Based on the foregoing and after further comparable comparison, we believe that these market multiples represent a reasonable approach, as they are the most observable data points available.
(c) and (d) The reporting unit fair values as determined by the comparable market and income (discounted cash flow) approaches are as follows:
DGHM | BOS | SHA | DTC | CCM | |||||||||||
BPFH ownership | 80 | % | 70 | % | 76 | % | 80 | % | 44 | % | |||||
STEP 1 Market Approach | 12,300 | 32,500 | 13,400 | 6,200 | 2,800 | ||||||||||
STEP 1 Income Approach | 11,200 | 31,900 | 15,400 | 6,300 | 2,300 | ||||||||||
STEP 1 Transaction Approach | none | none | 5,800 | none | none | ||||||||||
STEP 1 FMV at 100% | 11,500 | 32,000 | 10,100 | 6,300 | 2,494 | ||||||||||
BPFH Carrying value at 100% | 49,758 | 26,468 | 17,507 | 5,503 | 6,439 | ||||||||||
STEP 1 | FAIL | PASS | FAIL | PASS | FAIL | ||||||||||
Step 2 — GW fair value @ ownership % | zero | n/a | 2,487 | n/a | zero | ||||||||||
BPFH GW carrying value | zero | 16,492 | 13,327 | 2,446 | 3,456 | ||||||||||
Impairment charge | fully impaired | none | 10,840 | none | 3,456 |
11
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
(e), (f ) and (g ) The following are discussions of the weighting used to determine the reporting unit fair values and the rationale for the conclusions based on the approaches described above.
DGHM, BOS, DTC, CCM
For DGHM, BOS, DTC, and CCM, we placed the majority of the weight on the income approach in concluding the fair value. Management felt the income approach was more reflective of the intrinsic value of these companies due to the volatility in recent and projected earnings. However, both the market and income approaches yielded a narrow range of values, so changes in the relative weighting would have had no impact on the overall results of the test.
SHA
During the fourth quarter, the Company was in the midst of negotiations involving the sale of this particular affiliate’s interests with a willing and able third party buyer. Management felt that this potential transaction represented a very credible “fair value” estimate and thus weighted the offer at 50%. Thus, the discounted cash flow method and the guideline public company method were each weighted at 25.0%. This resulted in an indicated fair value of $10.1 million.
Carry-forward of Fair Value from Impairment Test
The Company updated the October 31, 2008 market comparables based on the December 31, 2008 equity prices in order to determine if a roll-forward of fair value from the impairment test to December 31, 2008 would be accurate and appropriate. The results of the December 31, 2008 update to the market approach supported the conclusions drawn from the October 31, 2008 testing. Furthermore, management noted the absence of any impairment review triggers during the fourth quarter of 2008 besides general market conditions. Based on these facts, management concluded that our October 31, 2008 testing conclusions and the October 31, 2008 fair values obtained during testing could be carried forward to December 31, 2008.
In conclusion, Anchor, BOS and KLS account for 92% of the Company’s $109 million of remaining goodwill. As noted above, BOS was subjected to a rigorous fair value assessment conducted by an independent third party valuation firm. Based on that testing, its fair value exceeded its carrying value by over 20% and thus had no impairment to its goodwill. While management determined that no third party evaluation was needed at either KLS or Anchor, a relevant point of comparison is that KLS has roughly the same carrying value as BOS, yet has twice the EBITDA of BOS. Further, Anchor has approximately 2.5 times the carrying value as BOS, yet has four times the BOS EBITDA. Accordingly, management was very comfortable in its conclusion that neither Anchor nor KLS pose any risk of goodwill impairment as of December 31, 2008.
12
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
6. As a related matter, please tell us how you validate the reasonableness of the fair values determined. For example, tell us whether you reconcile the fair values determined in your analyses to your total market capitalization. In this regard, we note that paragraph 23 of SFAS 142 indicates that the fair value of a reporting unit refers to the price that would be received to sell the unit as a whole in an orderly transaction between market participants at the measurement date, and that quoted market prices in active markets are the best evidence and should be used in available. In circumstances where an entity has multiple reporting units and all reporting units have goodwill that are tested for impairment, a tool that could be used to validate the reasonableness of the fair values determined for the individual reporting units is a reconciliation of the total fair values of the reporting units, to the market capitalization of the company, adjusted for any control premium as deemed appropriate. If you do not perform such an analysis, please indicate whether this is due to the fact that such an analysis cannot be easily performed (for example, because not all of your reporting units contain goodwill or there were not triggers of impairment at each of your reporting units at each of the times you tested goodwill for impairment), or whether you don’t believe such an analysis is meaningful. Additionally, if such an analysis is not performed, please tell us whether you perform any other procedures to evaluate the reasonableness of the fair value of the reporting units. If such an analysis is performed, please tell us the results of your procedures, including a discussion of any control premium assumed in the analysis, and how the reasonableness of that premium was evaluated.
Response:
As noted in response to comment (5) above, in the fourth quarter of 2008, management conducted its annual test of goodwill for impairment and as a result of this test, the Company recorded an additional $39 million of goodwill and intangible assets impairment for all segments. The tests were performed at the affiliate level, which were determined to be reporting units under Statement 142.
In response to your comment regarding where there are circumstances in which an entity has multiple reporting units,and all reporting units have goodwill, you note that a tool that could be used to validate the reasonableness of the fair values determined for the individual reporting units is a reconciliation of the total fair values of the reporting units, to the market capitalization of the company. With respect to the Company, we note that although we do have multiple reporting units, only aminority of our reporting units have remaining goodwill. Notwithstanding this, the Company still performed a reconciliation of aggregate fair values of reporting units to the Company’s overall market capitalization. Management believes that the Company’s overall market capitalization should reconcile, within a reasonable range, to the sum of the fair values of the individual reporting units used for impairment testing. In performing this reconciliation the Company notes that it often reviewed both qualitative and quantitative assessments.
Market Capitalization Analysis
As discussed in comment (5), fair values for each of the affiliates with goodwill were
13
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
estimated using either the income approach, the market approach, or both. For those affiliates without goodwill, management estimated their fair values using the market approach. The sum of these fair values, along with the fair value of the holding company (estimated using a discounted cash flow approach), was approximately $454 million without any control premiums being applied at the affiliate level. Using the ten day average share price at December 31, 2008 of $6.14, multiplied by approximately 64 million shares outstanding, and adding the fair value of the Series B Convertible Preferred shares (approximately $45 million) yields a total market capitalization of approximately $437 million. This market capitalization is within 4% of the value derived from summing the individual fair values of the affiliates and the holding company, which thus reconciles within a reasonable range.
When applying control premiums to select affiliates (discussed below), aggregating all of the affiliate fair values at December 31, 2008 produces a combined fair value of $621 million. Management noted that the December 31, 2008 estimated fair value is approximately 7% or $48 million lower than the comparable market capitalization of $669 million (calculated using the method described above, plus a 20% control premium, plus $154 million of TARP), and $68 million or 10% lower than the book value of $689 million. Note that in order to obtain an apples-to-apples comparison between book value and market capitalization, since $154 million of TARP capital is included in our book value, we also included $154 million of TARP in the market capitalization calculation.
Control Premium
When the Company’s (total of all affiliates) fair value as of December 31, 2008 was calculated, control premiums were used in order to capture the premium that management believes these affiliates would command if divested in the current market. These control premiums ranged from 0% to 30% depending on the individual affiliates, and had an overall weighted average of 19%. Affiliates assigned a 30% control premium have a strong history of earnings growth and positive cash flow, while those assigned a 0% control premium have experienced weak growth, poor cash flow and/or credit quality issues.
A 20% control premium (at the Company level) was used in the market capitalization analysis and a 19% weighted control premium was used in the December 31, 2008 affiliate-level analysis. Management noted that control premiums can vary considerably depending on the nature of the business, industry and other market conditions. Management believes that control premiums may be stable or rise when there are broad market price decreases.
We believe that although a 30% overall control premium could be justified, management took a conservative approach and used a 20% control premium (1% beyond the 19% used at the total of affiliate analysis). Additionally, a 20% control premium on the Company stock as of December 31, 2008 results in an implied stock price of $7.37 per share ($6.14 * 20% control premium) which underestimates the intrinsic value of the Company as calculated by management’s internal projections.
To further substantiate the control premium, we obtained purchase and sale data of banks,
14
Mark Webb, | February 13, 2009 | |
Legal Branch Chief |
investment managers, and wealth advisors comparable to the Company’s affiliates. The listing population was all banking transactions that were announced in 2008 in the United States with a transaction value of greater than $100 million. Recent transactions were selected due to recent turmoil in the banking market. We selected comparable transactions in similar regions of the country that operate in similar industries and have observable transaction multiples since we believe that observed transactions multiples provide more objective evidence than subjective inputs to a discounted cash flow model. From the list of 18 transactions, we eliminated the transactions over $1 billion as we did not believe that these transactions were comparable to the Company. Of the remaining seven transactions, we calculated a mean premium of 22.2%. Management felt that this data fully supported the 19% fair value of affiliates weighted average control premium assumption and the 20% market capitalization control premium.
Conclusion
Based on the analysis described above, the December 31, 2008 total fair value of affiliates reconciles within a reasonable range as compared to the December 31, 2008 market capitalization of the company and management does not believe that further impairment charges are appropriate. As noted in our response to comment (5), the Company performed thorough and comprehensive testing at the reporting unit level as of December 31, 2008. This testing substantiated the $109 million of goodwill and $70 million of intangible assets. Accordingly, although management will continue to monitor its goodwill and intangible balances as required by FAS 142 and 144, the Company believes that the retention of these balances is reasonable and supportable.
We hope this information is responsive to your questions. If you should have any questions concerning the enclosed matters, please contact the undersigned at (617)912-3949.
Sincerely,
/s/ David J. Kaye
David J. Kaye
cc: | Margaret W. Chambers, |
General Counsel and Executive Vice President
Paul Lee, Esq.
Goodwin Procter LLP
15