CORRESP 1 filename1.htm SEC Letter - Response 2-15-13


_____________________________________________________________________________________________________







February 15, 2013




Ms. Suzanne Hayes
Assistant Director
Division of Corporation Finance
Securities and Exchange Commission
100 F Street, N.E.
Washington, D.C. 20549

Re:    Zions Bancorporation
Form 10-K for Fiscal Year Ended December 31, 2011
Filed February 29, 2012
Response Dated October 30, 2012
Form 10-Q for Fiscal Quarter Ended September 30, 2012
Filed November 8, 2012
Form 8-K filed December 4, 2012
File No. 001-12307

Dear Ms. Hayes:

The following responds to your letter dated January 18, 2013 regarding your review of the above-referenced filings. Most of our responses in this letter contain language that we expect to include in our 2012 Form 10-K. We are still in the process of preparing this document and completing the necessary review procedures and controls, including SOX-related controls. In addition, our executive management and Board of Directors have yet to finalize their review and approval of the document. As we complete this process, further changes may be made to the 2012 Form 10-K that may have affected our responses contained herein.

We have keyed our responses to your comments, which are repeated below:


Form 10-K for Fiscal Year Ended December 31, 2011

Item 7. Management's Discussion and Analysis of Financial Condition . . . , page 23

Focus on Growth Markets, page 23

1.
Please expand your response to prior comment 2 to address the following:

•
Please clarify your most significant market areas. For example, please identify each state that accounts for more than 10% of your portfolio.





Management Response:

We expect to add the following underlined language to the “Executive Summary/Focus on Growth Markets” section of the MD&A in our 2012 Form 10-K:

The higher economic growth indicators of GDP, population growth, and job creation in our footprint are driven by the economies of Texas, California and Utah where more than 75% of the Company's assets are located.

Texas has a well diversified economy that is the third largest in the United States. Significant drivers of its growth are the energy, health care, manufacturing, real estate, and computer technology sectors. These sectors propel the Texas economy to outperform the nation with employers adding jobs in the past 12 months, resulting in an unemployment rate declining to 6.1% compared to the national rate of 7.8%. Amegy's three primary markets - Houston, Dallas and San Antonio - experienced job growth in 2012. Amegy has approximately $13 billion in assets, which represent 24% of the Company's assets. In addition, the Texas economic environment benefits from business-friendly growth policies, highly affordable housing markets, and a relatively small percentage of homeowner borrowers in a negative equity situation. See page xx under “Business Segment Results” for further discussion of the impact the foregoing may have had on the 2012 performance of Amegy Bank.

California's economy represents about 13% of the nation's GDP and is based on a diverse group of business sectors. The state has been experiencing improvements in residential and commercial real estate values. Additionally, voters recently approved increases in taxes to shore up the state's budget shortfall. CB&T's primary markets - the major metropolitan areas in California (San Francisco Bay area, Los Angeles County, Orange County, and San Diego) - continue to experience economic improvements in 2012 compared to 2011. CB&T has approximately $11 billion in assets, which represent 20% of the Company's assets. Trends in unemployment, home foreclosures, and bank credit problems continue to improve throughout California, resulting in corresponding reductions in problem credits and nonperforming assets at CB&T. The state's unemployment rate declined from a peak of 12.4% in October 2010, to 11.2% in December 2011, to 9.8% in December 2012. Unemployment rates are much lower in CB&T's primary markets compared to the state as a whole. See page xx under “Business Segment Results” for further discussion of the impact the foregoing may have had on the 2012 performance of CB&T.

The Utah economy is primarily based on the energy, agriculture, real estate, computer technology, education, health care, and financial services sectors. During 2012, Utah employment grew at a 2.9% rate compared to the national employment growth rate of 1.4%. This growth decreased Utah's overall unemployment rate to 5.2% in 2012 from 5.8% in 2011. Zions Bank is the third largest commercial bank in the state of Utah and operates in all submarkets of the state. Zions Bank has approximately $18 billion in assets, which represent 32% of the Company's assets. In addition, the Utah state government has been recognized for its policies promoting a business-friendly climate, providing a predictable and stable tax policy, and controlling government spending levels. See page xx under “Business Segment Results” for further discussion of the impact the foregoing may have had on the 2012 performance of Zions Bank.

•
Please describe material risks related to geographic concentration.

Management Response:

Our Form 10-K and 10-Q filings describe material risks related to geographic concentration as they relate to the states where we conduct business. Specifically, we disclose material risks in our risk factors in the Form 10-K and we provide a schedule of concentrations of commercial real estate loans by state in each Form 10-K and 10-Q. Excerpts of certain of the risk factor disclosures are as follows:





•
We have been and could continue to be negatively affected by adverse economic conditions.

“…may directly or indirectly adversely impact economic conditions faced by us and our customers. Any increase in the severity or duration of adverse economic conditions, including a double-dip recession or delay in a full economic recovery, would adversely affect us…”

•
Weakness in the economy and in the real estate market, including specific weakness within the markets where our subsidiary banks do business and within certain of our loan products, has adversely affected us and may continue to adversely affect us.

“…If the strength of the U.S. economy in general and the strength of the local economies in which we and our subsidiary banks conduct operations decline further, this could result in, among other things, further deterioration in credit quality and/or continued reduced demand for credit, including a resultant adverse effect on the income from our loan portfolio…”

These disclosures provide readers of the Form 10-K and 10-Q an understanding of the risk factors associated with our business strategy to operate in the western states of the United States.

Also, on page xx of the MD&A, the following schedule provides details of the commercial real estate loan exposures by state/geographical area:





















COMMERCIAL REAL ESTATE PORTFOLIO BY LOAN TYPE AND COLLATERAL LOCATION
(Amounts in millions)
 
 
Collateral Location
 
 
 
 
Loan Type
As of
Date
 
Arizona
 
Northern
California
 
Southern
California
 
Nevada
 
Colorado
 
Texas
 
Utah/
Idaho
 
Washing-ton
 
Other 1
 
Total
 
% of total
CRE
Commercial term
Balance outstanding
12/31/2012
 
$
1,180

 
$
604

 
$
2,155

 
$
605

 
$
518

 
$
934

 
$
1,013

 
$
205

 
$
849

 
$
8,063

 
80.6
%
% of loan type
 
 
14.7
%
 
7.5
%
 
26.7
%
 
7.5
%
 
6.4
%
 
11.6
%
 
12.6
%
 
2.5
%
 
10.5
%
 
100.0
%
 
 
Delinquency rates 2:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30-89 days
12/31/2012
 
0.2
%
 
0.1
%
 
0.1
%
 
0.2
%
 
—
%
 
0.1
%
 
0.2
%
 
—

 
1.6
%
 
0.3
%
 
 
 
12/31/2011
 
0.6
%
 
0.4
%
 
1.2
%
 
0.5
%
 
0.5
%
 
1.6
%
 
0.5
%
 
—

 
1.1
%
 
0.9
%
 
 
≥ 90 days
12/31/2012
 
0.3
%
 
1.3
%
 
0.5
%
 
0.8
%
 
0.7
%
 
0.5
%
 
0.1
%
 
—
%
 
2.1
%
 
0.7
%
 
 
 
12/31/2011
 
0.9
%
 
0.3
%
 
0.3
%
 
0.4
%
 
—

 
1.7
%
 
0.6
%
 
—

 
2.1
%
 
0.8
%
 
 
Accruing loans past due 90 days or more
12/31/2012
 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
 
 
12/31/2011
 
—

 
—

 
—

 
—

 
—

 
3

 
—

 
—

 
1

 
4

 
 
Nonaccrual loans
12/31/2012
 
10

 
9

 
19

 
14

 
11

 
8

 
4

 
3

 
47

 
125

 
 
 
12/31/2011
 
14

 
3

 
27

 
37

 
14

 
23

 
9

 
—

 
29

 
156

 
 
Residential construction and land development
Balance outstanding
12/31/2012
 
$
96

 
$
42

 
$
156

 
$
1

 
$
35

 
$
234

 
$
111

 
$
4

 
$
35

 
$
714

 
7.1
%
% of loan type
 
 
13.4
%
 
5.9
%
 
21.8
%
 
0.1
%
 
5.0
%
 
32.8
%
 
15.5
%
 
0.6
%
 
4.9
%
 
100.0
%
 
 
Delinquency rates 2:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30-89 days
12/31/2012
 
0.6
%
 
1.0
%
 
—

 
0.1

 
4.9
%
 
7.9
%
 
0.2
%
 
—

 
—

 
3.1
%
 
 
 
12/31/2011
 
0.6
%
 
14.1
%
 
—

 
0.8
%
 
13.8
%
 
0.4
%
 
0.2
%
 
—

 
—

 
1.3
%
 
 
≥ 90 days
12/31/2012
 
0.7
%
 
—

 
0.2
%
 
—
%
 
0.5
%
 
6.7
%
 
—
%
 
—

 
—

 
2.4
%
 
 
 
12/31/2011
 
2.7
%
 
—

 
3.9
%
 
6.8
%
 
5.3
%
 
11.6
%
 
4.5
%
 
24.1
%
 
—

 
6.7
%
 
 
Accruing loans past due 90 days or more
12/31/2012
 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
1

 
$
—

 
$
—

 
$
—

 
$
1

 
 
 
12/31/2011
 
1

 
—

 
—

 
—

 
—

 
—

 
—

 
—

 
—

 
1

 
 
Nonaccrual loans
12/31/2012
 
6

 
—

 
—

 
—

 
—

 
29

 
4

 
—

 
—

 
39

 
 
 
12/31/2011
 
13

 
—

 
6

 
5

 
2

 
50

 
15

 
—

 
—

 
91

 
 
Commercial construction and land development
Balance outstanding
12/31/2012
 
$
86

 
$
49

 
$
194

 
$
68

 
$
85

 
$
452

 
$
254

 
$
12

 
$
25

 
$
1,225

 
12.3
%
% of loan type
 
 
7.0
%
 
4.0
%
 
15.8
%
 
5.6
%
 
6.9
%
 
36.9
%
 
20.7
%
 
1.0
%
 
2.1
%
 
100.0
%
 
 
Delinquency rates 2:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30-89 days
12/31/2012
 
2.4
%
 
—
%
 
—

 
0.3

 
0.4
%
 
2.0
%
 
2.3
%
 
—

 
0.1

 
3.1
%
 
 
 
12/31/2011
 
1.6
%
 
—

 
—

 
—

 
4.4
%
 
1.7
%
 
—

 
—

 
—

 
1.2
%
 
 
≥ 90 days
12/31/2012
 
—
%
 
2.6
%
 
0.1
%
 
—

 
—

 
4.0
%
 
—
%
 
—

 
—

 
1.6
%
 
 
 
12/31/2011
 
2.1
%
 
—

 
1.1
%
 
5.6
%
 
5.5
%
 
6.0
%
 
1.7
%
 
—

 
—

 
3.6
%
 
 
Accruing loans past due 90 days or more
12/31/2012
 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
$
—

 
 
 
12/31/2011
 
—

 
—

 
2

 
—

 
—

 
—

 
—

 
—

 
—

 
2

 
 
Nonaccrual loans
12/31/2012
 
—

 
1

 
—

 
22

 
—

 
29

 
14

 
3

 
—

 
69

 
 
 
12/31/2011
 
6

 
—

 
—

 
12

 
9

 
82

 
20

 
—

 
—

 
129

 
 
Total construction and land development
12/31/2012
 
$
182

 
$
91

 
$
350

 
$
69

 
$
120

 
$
686

 
$
365

 
$
16

 
$
60

 
$
1,939

 
 
Total commercial real estate
12/31/2012
 
$
1,362

 
$
695

 
$
2,505

 
$
674

 
$
638

 
$
1,620

 
$
1,378

 
$
221

 
$
909

 
$
10,002

 
100.0
%
1 No other geography exceeded $104 million for all three loan types.
2 Delinquency rates include nonaccrual loans.

We respectfully submit that we believe the schedule and the risk factor disclosures provide a robust summary of the geographic risk concentrations within our business footprint.

•
Please clarify the states in your footprint that have experienced a significant slowing in economic activity and how these trends impact your business.

Management Response:





We closely monitor our business performance and related economic activity in our footprint. The economic activity in the states of our footprint has not been significantly slowing. In fact, the economic activity has improved at various rates over the last three years in each state where we conduct business. Specifically, economic activity measured by state for new job creation, gross domestic product, unemployment rate, and real estate prices (FHFA Purchase-Only Home Price Index) demonstrate that each state has improved from the depths of the economic crisis in 2009. The tables below show this information for 2010, 2011 and 2012:

Job Growth (In thousands)
2010
2011
2012
 
GDP (In billions)
2010
2011
2012
 
 
 
 
 
 
 
 
 
Arizona
(46.8
)
24.5

52.0

 
Arizona
224

227

236

California
(153.4
)
136.4

236.8

 
California
1,702

1,735

1,788

Colorado
(22.8
)
32.5

42.5

 
Colorado
230

234

240

Idaho
(6.2
)
3.6

10.1

 
Idaho
51

51

53

Nevada
(30.9
)
8.6

9.2

 
Nevada
111

113

115

Oregon
(10.7
)
17.4

14.0

 
Oregon
178

186

194

Texas
36.9

215.0

251.9

 
Texas
1,113

1,150

1,192

Utah
(6.3
)
25.6

29.1

 
Utah
106

108

112

Washington
(36.1
)
35.0

49.7

 
Washington
305

311

320

 
 
 
 
 
 
 
 
 
U3 Unemployment Rate (%)
2010
2011
2012
 
HPI (Index, 1991=100) *
2010
2011
2012
 
 
 
 
 
 
 
 
 
Arizona
10.5

9.5

8.2

 
Arizona
182.0

163.9

185.2

California
12.3

11.8

10.5

 
California
164.3

152.7

159.1

Colorado
8.9

8.3

7.9

 
Colorado
266.3

259.6

272.1

Idaho
8.8

8.7

7.5

 
Idaho
201.4

182.3

197.9

Nevada
13.7

13.6

11.7

 
Nevada
129.9

114.3

118.7

Oregon
10.6

9.5

8.6

 
Oregon
268.5

248.2

257.7

Texas
8.2

7.9

6.9

 
Texas
190.3

188.3

196.6

Utah
8.0

6.7

5.6

 
Utah
254.7

238.0

255.3

Washington
9.9

9.2

8.2

 
Washington
231.8

211.0

214.2

Source: Moody's Analytics (Economic & Consumer Credit Analytics)
*HPI is the FHFA Purchase-Only Home Price Index

These positive economics have resulted in improved credit quality of the loan portfolio at the affiliate banks in our footprint and increased lending to credit-worthy borrowers. Both concepts have been well disclosed and discussed in 2010, 2011 and 2012 Form 10-K and 10-Q filings.

Business Segment Results, page 47

2.
We note your response to prior comment 3. In addition to your proposed revision to Note 22, in future filings please expand this section in the MD&A to describe significant components of income and expense for your “Other” segment as well as the reason for material changes within this segment from prior periods. Provide us with this disclosure in your response.

Management Response:

We expect to add the following underlined language to the end of the “Business Segment Results” section of MD&A in our 2012 Form 10-K:

Other Segment

Operating components in the “Other” segment, as shown in Notes 21 and 23 of the Notes to Consolidated Financial Statements, relate primarily to the Parent. The major components include (1) net interest income,




which includes interest expense on other borrowed funds and (2) net impairment losses on investment securities. The Other segment also includes eliminations of transactions between segments.

Significant changes from 2011 to 2012 related to the Parent included a $75.7 million improvement in net interest income primarily related to lower interest expense on other borrowed funds. This was due to reduced discount amortization and reduced accelerated discount amortization on convertible subordinated debt, as discussed in “Net Interest Income, Margin and Interest Rate Spreads.” Also included is a $68.2 million increase in net impairment losses on investment securities, as discussed in “Investment Securities Portfolio.” Significant changes from 2010 to 2011 included these same items with the same directional changes of $73.2 million and $51.3 million, respectively.

Item 9A. Controls and Procedures, page 185

3. Please confirm that there were no changes (as opposed to “material changes”) in your internal control over financial reporting that materially affected, or are reasonably likely to materially affect, your internal control over financial reporting. Please also confirm this with respect to the periods covered by each of your Forms 10-Q filed in 2012 and revise future filings accordingly.

Management Response:

We will include the following underlined language in our 2012 Form 10-K and similar language in future Form 10-K and 10-Q filings regarding changes in internal control over financial reporting:

…There were no changes in the Company's internal control over financial reporting during the fourth quarter of 2012 that have materially affected, or are reasonably likely to materially affect, the Company's internal control over financial reporting…

We also confirm that this wording applied in each of our Forms 10-Q filed in 2012.

Item 11. Executive Compensation, page 186

Definitive Proxy Statement on Schedule 14A

Restricted Shares, page 25

4. We note your response to prior comment 15. Please provide us with your proposed disclosure as it would have appeared in this year's proxy statement.

Management Response:

The following underlined language is our proposed disclosure that would have appeared in the “Long-Term Incentive Compensation/Restricted Shares” section of the April 2012 proxy statement:

In January 2012, the Committee approved CPP-compliant restricted stock bonus awards for EMC members, including NEOs, based on a subjective evaluation of a variety of factors, including, but not limited to, the following:

•
compensation paid to senior managers with similar qualifications, experience, and responsibilities at other institutions,
•
individual job performance,
•
local market conditions,
•
internal equity considerations,
•
recommendations of the Company's CEO (for other NEOs), and




•
the Committee's assessment of the overall financial performance (particularly operating results) of the Company and its operating units.

Financial performance results reviewed in the determination of the 2012 awards included: (i) 2011 operating profits; (ii) 2011 loan loss provisions; and (iii) 2011 charge-offs. Since both the level and trend of these measures were positive in 2011 and many other asset quality measures are stronger than industry averages, the Committee awarded to Messrs. Simmons, Arnold, McLean, Blackford, and Anderson a number of restricted shares that would have had a weighted average value at the time of the award equal to 44 percent of each executive's total salary, just under the maximum permitted under the Interim Final Rule. On an individual basis, the restricted shares had a grant value ranging from 37.5 percent to 50 percent. The percentages differed based on the results of the subjective evaluation of the factors listed above for each individual executive and after considering the financial performance of the enterprise for Messrs. Simmons and Arnold and the financial performance (operating results) of the applicable affiliate bank organization for Messrs. McLean, Blackford and Anderson. As a result of these awards, the 2011 total compensation, in aggregate, for Messrs. Simmons, Arnold, McLean, Blackford, and Anderson was equal to the market median total compensation for similarly situated executives at peer firms in the 2011 Custom Peer Group.

We will add similar language appropriately adapted to our April 2013 proxy statement.

Compensation Tables, page 35

5.
Please note that restricted stock awards should be disclosed in the summary compensation table and grant of plan based awards table in the year granted. See pages 23-24 of Proxy Disclosure Enhancements, Release No. 33-9089, and Regulation S-K Item 402(d)(1). Please also note that additional explanation of restricted stock awards that were earned during the fiscal year but granted in the following fiscal year may be included in your CD&A or the narrative disclosure accompanying your compensation tables. Revise future filings accordingly.

Management Response:

Our future disclosures will only display restricted stock and restricted stock unit grants made during the fiscal year in the Summary Compensation Table and Grant of Plan Based Awards Table. We will describe in the Compensation Discussion and Analysis any grants earned during the fiscal year, but made in the following year. Accordingly, the 2011 stock awards made on January 15, 2012 that were disclosed in the April 2012 proxy statement will reappear in the April 2013 proxy statement as 2012 awards.

Form 10-Q for Fiscal Quarter Ended September 30, 2012

Item 2. Management's Discussion and Analysis of Financial Condition . . . , page 59

Risk Elements - Credit Risk Management, page 78

Allowance and Reserve for Credit Losses, page 86

6.
We note that you recorded a negative provision for loan losses for the three months ended September 30, 2012. We further note you attribute this reduction in loan losses to an improvement in credit quality metrics and a decrease in the portion of your allowance for loan losses related to qualitative and environmental factors. Please revise your disclosure in future filings to describe in more detail the reasons for such reductions in your allowance for loan losses, including a discussion of both the positive and negative credit trends experienced in each portfolio segment as well as any significant changes made to your qualitative and environmental factors.

Management Response:





We expect to add the following underlined language to the “Allowance and Reserve for Credit Losses” section of MD&A in our 2012 Form 10-K:

…The total ALLL declined during 2012 due to the positive credit trends experienced in our major loan portfolio segments and to somewhat improving economic conditions in some of our markets. See Note 6 of the Notes to Consolidated Financial Statements for tables showing positive and negative credit trends experienced in each portfolio segment.

The Company decreased the portion of the ALLL related to qualitative and environmental factors to reflect improving credit quality trends and stabilizing economic conditions in some of our markets. Otherwise, there were no significant changes to our qualitative and environmental factors.

Credit trends experienced in the FDIC-supported portfolio are also described in Note 6 of the Notes to Consolidated Financial Statements.

The total ALLL declined during 2011…

Interest Rate and Market Risk Management, page 87

Interest Rate Risk, page 88

7.
We note your response to prior comment 6 as well as your disclosures beginning on page 88. In your response you stated that you do not believe your disclosures need to educate investors on different types of interest rate risk measures. While an in-depth discussion of the model mechanics may not be beneficial to investors, we do believe that it is important to provide more context around the outputs of such models, particularly with respect to your duration of equity model and how it interacts with your market value of equity (MVE) model. For example, you disclose that the MVE model is used as a complement to your duration of equity model and that both models measure the expected changes in the fair values of equity in response to changes in interest rates. However, we are unclear as to how the outputs of your duration of equity model should be evaluated in isolation as well as when compared to the outputs of the MVE model under the same interest rate scenario(s). In this regard, please address the following in your response as well as in your disclosure in future filings:

Management Response:

Effective December 31, 2012, we no longer use duration of equity as our primary model to monitor interest rate risk. We made this change primarily to conform to prevailing reporting practices in the banking industry. See the “Interest Rate and Market Risk Management” section in our last response to this comment that describes our use of the MVE model. We adjusted prior year information and tables to maintain comparability.

•
Clarify whether your duration of equity model is attempting to measure the interest rate elasticity of your equity.

Management Response:

Yes, the duration of equity is the change in value of equity as a result of a change in interest rates.

•
More clearly explain what the base case duration of equity scenario is trying to show. For example, given your disclosure on page 88 that a negative value implies that an increase in interest rates increases the dollar value of equity, does a -3.9% duration of equity (assuming slower deposit re-pricing speeds) imply that a 100 bps increase in interest rates would increase the dollar value of equity by 3.9%?





Management Response:

See the “Interest Rate and Market Risk Management” section in our last response to this comment that describes our use of the MVE model. You are correct with regard to your example.

•
Provide a better linkage between the duration of equity and MVE model outputs. Your current disclosure appears to imply that these are separate and distinct models, while it is our understanding that duration of equity is actually an input to the MVE model. Thus, your disclosure should more clearly explain how these models are interrelated.

Management Response:

Actually, the MVE model is an input to the duration of equity. The MVE model measures the market value of assets minus the market value of liabilities. The calculation is done for a range of interest rate shocks including parallel shocks of -200bp, -100bp, -10bp, +10bp, +100bp, +200bp, +500bp.

The duration of equity is calculated as follows: [MVE(10bp) - MVE(-10bp)] / (0.20) / MVE(0bp) = % change in MVE scaled per 100 bp.

The factor 0.20 represents the difference between +10bp and -10bp as a fraction of 100bp (20/100).

We expect to add the following underlined language to the “Interest Rate and Market Risk Management” section of the MD&A in our 2012 Form 10-K that cover all of the responses above:

We attempt to minimize the impact of changing interest rates on net interest income primarily through the use of interest rate floors on variable rate loans, interest rate swaps, interest rate futures, and by avoiding large exposures to long-term fixed rate interest-earning assets that have significant negative convexity. Our earning assets are largely tied to the shorter end of the interest rate curve. The prime lending rate and the LIBOR curves are the primary indices used for pricing the Company's loans. The interest rates paid on deposit accounts are set by individual banks so as to be competitive in each local market.

We monitor interest rate risk through the use of two complementary measurement methods: Market Value of Equity (“MVE”) and income simulation. In the MVE method, we measure the expected changes in the fair values of equity in response to changes in interest rates. In the income simulation method, we analyze the expected changes in income in response to changes in interest rates.

MVE is calculated as the fair value of all assets and derivative instruments minus the fair value of liabilities. We report changes in the dollar amount of MVE for parallel shifts in interest rates.

The Company's policy is generally to limit declines in MVE to 3% per 100bp movement in interest rates in either direction. The following table demonstrates the formal limits adopted:

Parallel change in interest rates
 
Allowable decline in MVE
 
 
 
 
 
 
 
 
 +/- 100bp
 
 
 
3%
 
 
 +/- 200bp
 
 
 
6%
 
 
 +/- 300bp
 
 
 
9%
 
 
 +/- 400bp
 
 
 
12%
 
 
 +/- 500bp
 
 
 
15%
 





These MVE limits, adopted in late 2012, are a change from previous policy which imposed limits on duration of equity. We had been calculating both measures in parallel for several quarters prior to the adoption of MVE as a primary policy limit, and no significant change in the Company's interest rate risk position resulted from this policy change. Further, we still calculate and monitor both measures. Duration measures changes in MVE for small changes in interest rates only. The changes to the policy to limit declines in MVE over a wider range of rate movements enhance the interest rate risk management practice of the Company. Changes to the MVE limits are subject to notification and approval by the Company's Board of Directors. Due to embedded optionality and asymmetric rate risk, changes in MVE can be useful in quantifying risks not apparent for small rate changes. Examples of such risks may include out of the money caps on loans which have little effect for small rate movements but may become important if larger rate shocks were to occur, or substantial prepayment deceleration for low rate mortgages in a higher rate environment.

Income simulation is an estimate of the net interest income and total rate sensitive income that would be recognized under different rate environments. Net interest income and total rate sensitive income are measured under several parallel and nonparallel interest rate environments and deposit repricing assumptions, taking into account an estimate of the possible exercise of options within the portfolio. For income simulation, Company policy requires that interest sensitive income from a static balance sheet be limited to a decline of no more than 10% during one year if rates were to immediately rise or fall in parallel by 200 basis points.

Each of these measurement methods requires that we assess a number of variables and make various assumptions in managing the Company's exposure to changes in interest rates. The assessments address loan and security prepayments, early deposit withdrawals, and other embedded options and noncontrollable events. As a result of uncertainty about the maturity and re-pricing characteristics of both deposits and loans, the Company estimates ranges of MVE and income simulation under a variety of assumptions and scenarios. The Company's interest rate risk position changes as the interest rate environment changes and is actively managed to maintain a slightly asset-sensitive position. However, positions at the end of any period may not be reflective of the Company's position in any subsequent period.

The estimated MVE and the income simulation results are highly sensitive to the assumptions used for deposits that do not have specific maturities, such as checking, savings, and money market accounts and also to prepayment assumptions used for loans with prepayment options. Given the uncertainty of these estimates, we view both the MVE and the income simulation results as falling within a wide range of possibilities.

As of the dates indicated, the following schedule shows the Company's percentage change in interest sensitive income, based on a static balance sheet, in the first year after the rate change if interest rates were to sustain an immediate parallel change of 200 basis points. The Company estimates interest rate risk with two sets of deposit re-pricing scenarios. The first scenario assumes that administered-rate deposits (money market, interest-earning checking, and savings) reprice at a faster speed in response to changes in interest rates. The second scenario assumes that those deposits re-price at a slower speed. The following schedule includes changes in the MVE from -200 bps to +500 bps parallel rate moves for both 'fast' and 'slow' scenarios.





Schedule XX
INCOME SIMULATION - CHANGE IN INTEREST INCOME

As of December 31, 2012
(In millions)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Repricing scenario
 
  -200 bps
 
 -100 bps
 
 +100 bps
 
 +200 bps
 
 +300 bps
 
 +400 bps
 
 
 
 +500 bps
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Fast
 
 
(1.8)%
 
 
 
(1.8)%
 
 
 
3.9%
 
 
 
 
9.8%
 
 
 
 
16.7%
 
 
 
 
23.6%
 
 
 
 
30.6%
 
 
Slow
 
(0.02)%
 
 
(0.02)%
 
 
5%
 
 
 
12.1%
 
 
 
20.2%
 
 
 
28.4%
 
 
 
36.5%
 
 

As of December 31, 2011
(In millions)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Repricing scenario
 
 -200 bps
 
 -100 bps
 
 +100 bps
 
 +200 bps
 
 +300 bps
 
 +400 bps
 
 
 
 +500 bps
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Fast
 
 
(1.9)%
 
 
 
(1.8)%
 
 
 
2.8%
 
 
 
 
7.4%
 
 
 
 
13.1%
 
 
 
 
18.9%
 
 
 
 
24.8%
 
 
Slow
 
(2.3)%
 
 
(2.1)%
 
 
4.1%
 
 
 
10%
 
 
 
17%
 
 
 
24.1%
 
 
 
31.3%
 
 

Schedule XX
CHANGES IN MARKET VALUE OF EQUITY

As of December 31, 2012
(In millions)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Repricing scenario
 
 -200 bps
 
 -100 bps
 
 +100 bps
 
 +200 bps
 
 
 
 +300 bps
 
 +400 bps
 
 +500 bps
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Fast
 
 
5.3%
 
 
 
 
0.7%
 
 
 
 
1.7%
 
 
 
 
3.9%
 
 
 
 
6.3%
 
 
 
 
9.4%
 
 
 
 
12%
 
 
Slow
 
(0.5)%
 
 
 
(2.8)%
 
 
 
4.9%
 
 
 
10.6%
 
 
 
16%
 
 
 
25%
 
 
 
30.8%
 
 

As of December 31, 2011
(In millions)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Repricing scenario
 
 -200 bps
 -100 bps
 
 
 
 +100 bps
 
 +200 bps
 
 +300 bps
 
 +400 bps
 +500 bps
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Fast
 
 
2.9%
 
 
 
0.1%
 
 
 
 
 
0.7%
 
 
 
 
1.6%
 
 
 
 
2.3%
 
 
 
 
2.2%
 
 
 
 
4.8%
 
 
Slow
 
(2.1)%
 
 
(2.6)%
 
 
 
 
3.1%
 
 
6.7%
 
 
9.8%
 
 
15.3%
 
 
18.9%
 

During 2012, changes in interest rate sensitivity were primarily driven by a 16% year-over-year increase in noninterest-bearing DDA liabilities, with the majority of that increase being invested in short-term money market assets. Other factors which marginally affected the changes in interest rate sensitivity include prepayment of TARP from cash reserves, changes in the modeling assumptions for capturing balloon payments, and changes to prepayment assumptions.

Form 8-K filed December 4, 2012

8.
We note that you expect to record a significant OTTI charge in your CDO portfolio during the fourth quarter which is partly attributable to changes in modeling assumptions for certain bank holding company trust preferred securities that are currently deferring distributions and nearing the end of their deferral periods. We also note from your disclosures in previous annual and quarterly reports that the weighted average assumed loss rate on deferring collateral has steadily decreased as follows: 44% (December 31, 2009), 30% (December 31, 2010), 26% (December 31, 2011), 24% (June 30, 2012) and 21% (September 30, 2012). Please tell us, and revise your future filings to disclose, the reasons for the




decrease in the weighted average assumed loss rates despite the uncertainty around deferring banks' ability to become current on their deferring collateral.

Management Response:

We expect to add the following underlined language in a separate subsection of MD&A in our 2012 Form 10-K as follows:

Investment Securities Portfolio…

Significant Assumption Changes for 2012

Probability of Default of Deferring Bank Holding Company Trust Preferred Collateral

Historically, our ratio-based model assessed both performing and deferring issuers. Ratios predictive of bank failure were used in our model to identify the PD of bank holding company issuers of trust preferred securities. The reduction of the weighted average assumed loss rate on deferring collateral from 2009 onward was due to a combination of factors. These included the increased predictive ability of the model during that period in identifying which holding companies would have its sole bank or primary bank seized and which did not appear to be in danger of seizure. Bank seizures were largely the driver of bank holding company defaults on trust preferred during the 2009 to late 2012 period. As the highest PD institutions defaulted, and relatively stronger new deferrals were added, the averages declined.

Effective in the fourth quarter of 2012, we developed and began using a new probability of default (“PD”) overlay model for deferring bank trust preferred issuers, which increased the PD of many such issuers. The majority of these deferring issuers face a reperformance-or-default deadline within the next two and half years. Increased regulatory and restructuring risk as evidenced by several fourth quarter 2012 events and developments led to our modeling change.

Prior to the fourth quarter of 2012, we had observed only a few bankruptcies of holding company issuers of trust preferred securities. These were generally confined to the weakest institutions. Thus we assumed no recovery.

In the fourth quarter of 2012, several regulatory and restructuring events were observed that negatively affected our assessment of current conditions for those issuers still in deferral and our forecast for these deferrals:

1)
We observed the first Section 363 bankruptcy of a bank holding company whose bank did not appear to be in danger of regulatory seizure. While the bank had a 2.18% modeled PD, the effective loss will be all of the back interest or approximately 8% of the par amount. In this case, the relative strength of the bank did not bring a sale price for the bank stock held by the holding company sufficient to fully pay off the trust preferred security issued by the holding company.

2)
We observed the uncertainty of recovery amounts for trust preferred securities issued by bank holding companies when those holding companies chose to file Section 363 bankruptcies. The extent to which a recovery is achieved for the trust preferred security is highly dependent on multiple factors including the active participation by CDO trustees under the direction of collateral managers or sufficient noteholders. We cannot assume this level of participation will be achieved in every future case.

3)
We observed what appeared to be the first instance of regulatory involvement in support of a bank holding company bankruptcy.





4)
We observed bank holding companies starting into the last year of the allowable five-year deferral period with a well capitalized and in some cases also a profitable bank and several years of improved financial performance and related ratios. In particular, we observed maintenance of regulatory orders precluding or requiring permission for holding companies to come current and subsidiary banks from making dividend payments to the holding company which, if left in place, will trigger default of the trust preferred securities issued by the holding company and collateralizing the CDOs. After discussion with regulators, we are unable to rule out that such regulatory orders may cause default.

5)
We observed a collateral manager for a CDO allowing the restructuring of a holding company trust preferred security from an issuer with a relatively strong underlying bank such that the loss exceeded our modeled PD.

The earliest deferring issuers to which we are exposed will face the end of the allowable deferral period starting in 2013. Our fourth quarter 2012 experiences and observations led us to identify greater than previously assessed risk to the collectability of deferring bank holding company issuers, even when the underlying bank is not at risk of regulatory seizure. Events seen in the fourth quarter of 2012 indicated that our previous assumption that a well capitalized and profitable bank has a low probability of its bank holding company defaulting should be supplemented with additional analysis. Said differently, the subsidiary of a deferring bank holding company may be in decent shape, but because cash may be trapped at the subsidiary due to regulatory constraints, the bank holding company may simply be unable to bring the trust preferred obligation current once its five-year deferral period is up. Alternatively, debt restructurings approved by a CDO collateral manager and Section 363 bankruptcies approved by relevant courts may result in recoveries below our previously modeled probabilities of default.

The deferral PD overlay model addresses these risk by sorting remaining deferrals within our CDO pools into four “buckets” based on four factors indicative of bank holding company strength at the start of their deferral period. We then assume that the historical failure rate we have observed within our CDO pool for collateral in each bucket will be the future default rate of current deferrals in each bucket. From this we can create an expected default curve by vintage of deferrals. Where the overlay PD of a deferral is higher than the PD identified by our traditional ratio-based PD model, we use the overlay PD. We expect to recalibrate the expected default curve for each bucket periodically as the outcome for more deferring issuers becomes known over time.

Adoption and utilization of the PD overlay model for deferrals effective December 31, 2012 raised our weighted average loss assumption on deferring collateral from 20% under our base model to 49% under the overlay model. The assumption change generated $50.4 million pretax of OTTI.


We acknowledge that:

•
the Company is responsible for the adequacy and accuracy of the disclosure in the filing;
•
staff comments or 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 or any person under the federal securities laws of the United States.

We trust the above adequately responds to the comments in your letter. We look forward to discussing our responses and disclosures with you, if necessary, at your convenience.






 
 
 
Sincerely,
 
 
 
 
 
 
 
 /s/ Alexander J. Hume
 
 
 
Alexander J. Hume
 
 
 
Senior Vice President and Corporate Controller