SecProbe.io

CORRESP
1
filename1.htm

corresp

    Commerce Bancshares, Inc.

1000 Walnut Street

Post Office Box 419248

Kansas City Missouri 64141-6248

(816) 234-2000

July 8, 2011

By Edgar Filing

Ms. Suzanne Hayes

Assistant Director

United States Securities and Exchange Commission

Division of Corporation Finance

100 F Street, N.E.

Washington, D.C. 20549-4561

Dear Ms. Hayes:

We are in receipt of your letter dated June 30, 2011 wherein you have reviewed our Form 10-K and
Definitive Proxy Statement for the year ended December 31, 2010 and our Form 10-Q for the three
months ended March 31, 2011 and have noted several areas where you are requesting additional
information. In response we submit the following:

Form 10-K for the Fiscal Year Ended December 31, 2010

1a. Risk Factors, page 9

You stated that many of our risk factors omit a discussion as to how the presented risk ultimately
impacts our business. Accordingly we have reviewed all the risks presented and where appropriate
have added additional information about the impact to our business. Specifically we added
information which addresses the Dodd-Frank regulations you mentioned. We have also deleted the
2nd and 3rd sentences in the introductory paragraph of this section. We have
attached a red-lined version of this section to this letter as Exhibit A and similar information
will be included in future filings as applicable.

Loan Portfolio Analysis, page 29

You ask that in future filings, we should revise our tabular disclosure on page 30 to include
principal payment due dates for our consumer loans as required in Part III(B) of Guide 3. As we
read and interpret Guide 3 requirements, there appears to be an exception referenced in Part III(B)
that allows exclusion of this information for installment loans to individuals (category 4 listed
in Part III(A)). The Consumer loans listed in the referenced table are in fact installment loans
to individuals for which an exception to presenting the data is permitted as per above. Please
advise if we are not interpreting this section correctly.

Loans with Special Risk Characteristics, page 38

You ask that we clarify how we monitor current collateral values and trends experienced for the
various categories of real estate loans. You further ask that we provide refreshed LTV’s in future
filings. You also ask us to explain the extent to which we are able to monitor for delinquency and
loan balance of the first liens in evaluating related second lien loans. If we are only able to
access such data for home equity loans where we hold or service the first mortgage, then you ask us
to disclose that fact and quantify the amount of such loans.

In regards to monitoring current collateral values, while we obtain new appraisals at the time we
originate new personal real estate loans (residential first mortgages) and home equity loans (both
fixed and revolving), we do not seek to obtain updated appraisals or valuations unless the loan is
significantly delinquent or in process of being foreclosed upon. While our real estate department
stays informed of real estate values and trends in our local and national markets, our main method
for monitoring credit in these portfolios

1

remains delinquency monitoring. Additionally, for residential first mortgages and home equity
loans, FICO scores are updated quarterly while for home equity loans only, line utilization and
credit bureau information are also regularly used to identify specific borrower issues. This has
remained an effective way to evaluate credit trends and identify problem loans partly because our
Company mostly offers standard, conservative lending products. Because the LTV ratio is not our
primary method of monitoring credit, we do not maintain refreshed LTV data, and the LTV disclosures
for personal real estate and home equity loans on Form 10-K page 38 were intended to provide a
general indication of the risk in these portfolios. As such, we are not able to provide refreshed
LTV information in future filings.

You request information regarding our ability to monitor delinquency and loan balances of first
lien residential mortgage loans which coincide with our second lien mortgage loans (mainly fixed
and revolving home equity loans). Our systems do not allow us to identify and monitor the balances
and delinquency status of first lien mortgage loans which relate to our second lien mortgage loans.
As such we do not have information on first lien mortgage loans which are related to our second
lien home equity loans. As noted above, home equity loans regularly go through a review process
for credit line usage (revolving loans), changes in FICO scores and delinquencies. If specific
loans exhibit severe weakness, further investigation of first lien positions is performed.
Furthermore, the Company is notified of all foreclosures of related first lien positions. Given
the relatively low loan losses and delinquency rates experienced for these loans over the last few
years, including the recent period of financial crisis, the Company views these credit monitoring
processes as prudent measures for credit risk purposes.

Attached as Exhibit B is a revision of the original Form 10-K page 38 paragraph. This revised
disclosure provides more clarity to our monitoring process.

You also request us to revise our discussion of troubled debt restructures on page 38 to quantify
our re-default rates and revise, in the appropriate section of our Form 10-K, our discussion of how
modifications and re-defaults are considered in our allowance methodology. We submit the following
to consider. For credit card modification programs, as of 12/31/10 such loans totaled $18.8
million, or approximately 2.4% of total average credit cards for the 4th quarter of
2010. Amounts representing re-default (internally defined as 90 days delinquent) totaled $3.6
million, or 19% of total outstanding credit card modifications. Given that total modifications of
credit card balances are very small, we do not monitor re-default rates on these specific balances
or specifically consider these amounts in our allowance methodology unless the balance become 120
days delinquent after modification. In that case, the credit card balances are charged off, in
accordance with regulatory requirements. We have not chosen to further disclose information about
these balances due to their immateriality. Should these balances ever grow to be material, we
would of course make these types of further disclosures.

Exclusive of the credit card modifications noted above, the Company identified $75.8 million of
loans, which primarily were commercial loans, classified as restructured. Of these loans, $34.5
million were on non-accrual, while the remaining $41.3 million were performing loans on accrual
basis. Personal real estate loans classified as TDRs are all on non-accrual and totaled only $877
thousand. As noted on Form 10-K page 76, these restructured loans are considered impaired loans,
and on Form 10-K pages 70-71, our summary of significant accounting policies explains how these
impaired loans are evaluated. We do not monitor or attempt to quantify re-default rates on either
restructured non-accrual loans or performing TDRs. Once a loan is put on non-accrual status, we
are in a work-out mode, and past due status is usually not as relevant to us. Furthermore, when a
loan is put on non-accrual, unpaid interest is reversed, and loans are either charged down to fair
value or comparable loan loss reserves are established based on specific evaluation. Performing
TDR loans, as disclosed on Form 10-K page 38, are classified as TDRs solely because at renewal, the
renewal rate was not determined to be market and the borrower was judged to be having financial
difficulties. To stay on accruing basis however, the borrower cannot be delinquent, which is why
we do not monitor for re-default rates. Furthermore, for performing TDRs, no other concessions
were granted (e.g. forgiven principal or accrued interest) except that the renewed interest rate
was not judged to be market. Loans in this performing TDR status could ultimately become
delinquent and/or migrate to a non-accrual status, however, we do not track this currently due to
the nature of our TDRs, as noted above. We understand that new rules effective in the third
quarter of 2011 may require us to modify our tracking of TDRs going forward.

We attached Exhibit C representing our revised Form 10-K page 38 disclosure in which we have
attempted to clarify how TDR’s impact our allowance methodology.

Consolidated Statements of Income, page 66

You ask that for future filings, we revise certain labels concerning net income attributable to
Commerce Bancshares and consolidated net income. We will make these suggested changes and will
change the lines “Net income before non-controlling interest” to “Net

2

income” and “Net income” to “Net income attributable to Commerce Bancshares, Inc.” The
Consolidated Statements of Cash Flows will also be adjusted to reflect this change. These changes
are labeling changes and have no impact on our financial results.

Footnote 3. Allowance for Loan Losses, Credit quality, page 79

You request that we tell you whether we monitor loan-to-collateral ratios for all classes that have
collateral liens attached, including our consumer loans and real estate classes for construction,
business and personal loans. If so you request that in future filings we include these loan
classes by weighted average LTV ratios on a disaggregated basis (e.g., multiple weighted average
LTV buckets for each loan class).

Please be advised that as noted above, we obtain new appraisals only on personal real estate loans
and home equity loans. We do not obtain updated appraisals unless the loan is significantly
delinquent or is in the process of being foreclosed upon. LTV rates are not considered reflective
of current fair value of the underlying collateral but are only indications of collateral support
since the appraised values are at origination. As such, we rely on delinquency trend information
to identify potential problem personal real estate loans, and our credit collection department
pro-actively works with borrowers that become delinquent. Furthermore, delinquencies and loan
charge-offs have remained very low, as evidenced on page 35 of Form 10K, throughout the recent
severe credit cycle, reflecting the less volatile markets in which we operate and conservative
products and underwriting standards offered by our Company. Given these circumstances, we do not
believe that providing weighted average LTVs on these products, which reflect dated appraised
values, would be helpful to readers of our financial statements.

Regarding other lending products such as construction, business and other consumer loans which have
real estate or other collateral as security, we do not monitor loan to collateral values on a class
basis but instead do so individually on a case by case basis and usually only at the time a loan is
originated or renewed. For consumer loans secured by vehicles, while we have standards which
dictate how much we are willing to loan at origination on a particular vehicle, we monitor the loan
on an on-going basis using delinquency information and FICO scores and do not store current LTV
information in our systems. Other business and construction loans can be individually large with
diverse types of collateral such that LTV information would not be useful. Again, this information
is not readily available in our loan systems.

Valuation methods for instruments measured at fair value on a non recurring basis, page 109

You asked why the write-down of $4.0 million was such a significant portion of the foreclosed asset
balance of $8.5 million. You ask that in future filings we describe the steps taken to verify the
deterioration in fair value on these properties subsequent to the loans being categorized as other
real estate owned. Also in future filings we should enhance our disclosure to discuss the factors
driving the large impairment post-foreclosure.

The Company has standard policies and procedures that require, at the time of foreclosure or when
foreclosure is probable, the loan be written down to fair value less estimated selling costs, based
on written appraisals or other similar valuation methods. In accordance with Company policies and
regulatory requirements, subsequent adjustments to fair value on foreclosed real estate within 90
days of foreclosure are considered additional loan charge-offs, while after this period, fair value
adjustments are treated as impairment losses and classified as foreclosed property expense. For
other non-real estate foreclosed property we record additional loan charge-offs within 60 days of
foreclosure. The losses reported on Form 10-K page 111 relate only to balances of foreclosed
property held at year end for which impairment was recorded. This also excludes $7.1 million in
foreclosed real estate on hand at year end which did not have any impairment losses in 2010. The
amount of foreclosed property reported in this table is comprised of 27 real estate properties with
a total balance of $4.9 million and one repossessed airplane with a balance of $3.6 million.

As an example, two properties with a total fair value of $4.6 million represented $2.3 million of
the entire write-down noted in the table. One foreclosed asset was an airplane which was
determined to have a fair value of $4.9 million at the foreclosure date of 12/30/09, based on
airplane market data at that time. Due to deteriorating market conditions in 2010, the discovery
of other similar aircraft for sale, and after our 60 day policy noted above, further write-downs of
$1.3 million were required. Note that this airplane was sold in early 2011 for its carrying value
of $3.6 million. Another foreclosed property consisting of undeveloped land for residential lots
was appraised at foreclosure on 3/31/09 at $2.4 million, and the property was written down to this
appraised value through loan charge-off at that time. However, updated appraisal information
obtained in both late 2009 and 2010 showed the property had further deteriorated mainly due to more
stringent appraisal assumptions. As a result, it was appropriate to further write-down this
property, classifying it as foreclosed property expense ($350 thousand in 2009 after the 90 day
period and $1.0 million in 2010).

3

The severe downturn in housing and related slow economy has resulted in much lower values for
foreclosed properties and thus has driven our expenses and losses much higher than our previous
experience four or five years ago. Also it now takes longer to sell such properties than in the
past. The Company’s practice is to obtain updated appraisals on all foreclosed property at least
annually and more often when believed necessary. For the year ended 12/31/10, total realized and
unrealized losses on foreclosed real estate totaled $3.4 million. This represents 27.9% of the
total foreclosed real estate balance at year end.

As stated above, prior to foreclosure, the Company obtains appraisals on loan collateral and
charges down the related loan to fair value less estimated selling costs. Further, in determining
adequate loan loss reserves, the Company uses historical experience, economic factors, experience
with appraisals, and other information suggesting inherent losses within a particular portfolio to
help determine allowance levels. However, the allowance for loan losses is not impacted by future
write-downs in foreclosed property after the stated look back periods of 60 and 90 days, noted
above.

Commitments, Contingencies and Guarantees, page 116

Regarding your request to revise in future filings our disclosure of the two overdraft lawsuits,
attached as Exhibit D is a revised disclosure we believe satisfies the criteria of ASC 450-20-50.

Definitive Proxy Statements on Schedule 14A

Benchmarks, page 18

Yo