XML 72 R58.htm IDEA: XBRL DOCUMENT v2.4.0.6
Fair Value Measurements and Disclosures (Narrative) (Details) (USD $)
In Thousands, unless otherwise specified
9 Months Ended
Sep. 30, 2012
Dec. 31, 2011
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Fair Value Measurements, Valuation Processes, Description

The Level 3 securities consist of seven collateralized debt obligation securities, PreTSL securities, which are backed by trust preferred securities issued by banks, thrifts, and insurance companies. The market for these securities at September 30, 2012 is not active and markets for similar securities also are not active. The inactivity was evidenced first by a significant widening of the bid-ask spread in the brokered markets in which PreTSLs trade and then by a significant decrease in the volume of trades relative to historical levels. The new issue market is also inactive and there are currently very few market participants who are willing and or able to transact for these securities.

 

Given conditions in the debt markets today and the absence of observable transactions in the secondary and new issue markets, we determined:

·         The few observable transactions and market quotations that are available are not reliable for purposes of determining fair value at September 30, 2012;

·         An income valuation approach technique (present value technique) that maximizes the use of relevant observable inputs and minimizes the use of unobservable inputs will be equally or more representative of fair value than the market approach valuation technique used at prior measurement dates; and

·         PreTSLs will be classified within Level 3 of the fair value hierarchy because significant adjustments are required to determine fair value at the measurement date.

 

The Bank is aware of several factors indicating that recent transactions of PreTSL securities are not orderly including an increased spread between bid/ask prices, lower sales transaction volumes for these types of securities, and a lack of new issuances. As a result, the Bank engaged an independent third party to value the securities using a discounted cash flow analysis. The estimated cash flows are based on specific assumptions about defaults, deferrals and prepayments of the trust preferred securities underlying each PreTSL. The resulting collateral cash flows are allocated to the bond waterfall using the INTEX desktop valuation model.

 

The estimates for the conditional default rates (CDR) are based on the payment characteristics of the trust preferred securities themselves (e.g. current, deferred, or defaulted) as well as the financial condition of the trust preferred issuers in the pool. A near-term CDR for each issuer in the pool is estimated based on their financial condition using key financial ratios relating to the financial institution's capitalization, asset quality, profitability and liquidity. In addition to the specific bank default assumptions, overall deal default rates are modeled. In 2013 and beyond, the CDR rate is calculated based upon a comparison of key financial ratios of active individual issuers without a short-term probability of default compared to all FDIC insured banks. To derive this long-term default rate, a comparison of certain key financial ratios of the active

issuers in the security to all FDIC insured banks is reviewed. The active issuers are summarized by creating a weighted average based on issue size, then divided into categories based upon their status of deferral and whether or not a specific default assumption has been assigned to the issuer. To ensure an accurate comparison, the standard deviation across the issuers for each ratio is calculated and any issuer that falls more than three standard deviations above or below the average for that ratio is removed.

 

The base loss severity assumption and long-term loss severity assumptions are modeled at 95%. The severity factor for near-term CDRs is vectored to reflect the relative expected performance of the institutions modeled to default, with lower forecasted severities used for the higher quality institutions.

 

Prepayments are modeled to take into account the disruption in the asset-backed securities marketplace and the lack of new pooled trust preferred issuances. For purposes of the cash flow analysis, relatively modest rates of prepayment were forecasted (ranging from 0-1%). In addition to the base prepayment assumption, due to the recent enactment of the Dodd-Frank financial legislation additional prepayment analysis was performed. First, all fixed rate trust preferred securities issued by banks with more than $15 billion in total assets at December 31, 2009 were identified. The current credit rating of these institutions was reviewed and it was assumed that any issuer with an investment grade credit rating would prepay their issuance on January 1, 2013 or July 1, 2015 for bank holding company subsidiaries of foreign banking organizations that have relied on Supervision and Regulation Letter SR-01-1. For those institutions rated below investment grade the holding companies' approximate cost of long-term funding given their rating and marketplace interest rate was estimated. The following assumption was made; any holding company that could refinance for a cost savings of more than 2% will refinance and will do so on January 1, 2013, or July 1, 2015. Finally, for issuers not impacted by the Tier 1 regulatory capital legislation enacted by the Dodd-Frank Act, the issuers that have shown a recent history of prepayment of both floating rate and fixed rate issues were identified and it was assumed these issuers will prepay as soon as possible.

 

The internal rate of return is the pre-tax yield used to discount the best estimate of future cash flows after credit losses. The cash flows have been discounted using estimated market discount rates of 3-month LIBOR plus spreads ranging from 3.98% to 9.16%. The determination of appropriate market discount rates involved the consideration of the following:

·         the time value of money

·         the price for bearing uncertainty in cash flows

·         other factors that would be considered by market participants

 

The analysis of discount rates involved the review of corporate bond spreads for banks, U.S. Treasury yields, credit default swap rates for financial companies (utilized as a proxy for credit), the swap/LIBOR yield curve and the characteristics of the individual securities being valued.

 
Fair Value Inputs, Loss Severity 95.00%  
Recorded investment (after charge-offs): with no specific allowance recorded $ 26,491 $ 21,822
Maximum [Member]
   
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Fair Value Inputs, Prepayment Rate 0.00%  
Fair Value Inputs, Discount Rate 9.16%  
Minimum [Member]
   
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Fair Value Inputs, Prepayment Rate 1.00%  
Fair Value Inputs, Discount Rate 3.98%  
Pooled trust preferred securities [Member]
   
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Available-for-sale Securities, number seven  
Partial Charge-offs [Member]
   
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Recorded investment (after charge-offs): with no specific allowance recorded   $ 1,327
Significant unobservable inputs (Level 3) [Member] | Pooled trust preferred securities [Member]
   
Fair Value, Assets and Liabilities Measured on Recurring and Nonrecurring Basis [Line Items]    
Available-for-sale Securities, number seven