SecProbe.io

CORRESP
1
filename1.htm

CORRESP

 100 King Street West, 24th Floor,

Toronto, Ontario, M5X 1A1

 September 13, 2013

 Stephanie J. Ciboroski

 Senior Assistant Chief Accountant

 Security and Exchange Commission

Washington, D. C. 20549

Re:
Bank of Montreal

 Form 40-F for the Fiscal Year Ended October 31, 2012

Filed December 4, 2012

 Response dated
June 21, 2013

 File No. 001-13354

 Dear
Ms. Ciboroski:

 I am writing to respond to your letter dated August 14, 2013. Your questions are set out below in bold.

Form 40-F for the Fiscal Year Ended October 31, 2012

Exhibit 99.2

 Management’s Discussion and Analysis, page 24

 Critical Accounting Estimates, page 70

 Goodwill and
Intangible Assets, page 72

1.
You state in your response to prior comment 4 that the Canadian Mortgage Housing Corporation (CMHC) automated valuation model (AVM) service is used for residential mortgages that are below a certain threshold and
have a loan-to-value (LTV) ratio within a specified range. We further note that you test the output of this AVM model by taking a sample of existing valuation assessments and running them through the AVM model to determine if the output from the
model is consistent with the other valuation process. Please address the following:

•

In future filings, disclose the specific threshold and LTV ratio range that determines your use of the CMHC AVM given that the threshold for use of AVM is also driven by the AVM output because the LTV is based
upon the AVM valuation of the underlying property.

 The CMHC evaluation does not impact the Bank’s LTV calculation; it is
only a factor that may be used in the Bank’s assessment of the need for an external appraisal.

 If a residential mortgage is below a certain dollar threshold and the LTV falls within an acceptably low
range, one of the steps we may undertake is to use the CMHC service as an additional source of information to consider when determining if an external appraisal should be obtained. The CMHC service considers a variety of factors and their
interrelationships, and uses different information sources, including the physical characteristics of the property in question (such as square footage, lot size, age, style of the home, etc.), the municipal property tax assessment, historical and
current sales activity within the local housing market, and prior sales activity of the property being assessed. The output from the service is information which, when processed in the Bank’s risk framework, indicates whether an external
appraisal should be obtained. The CMHC evaluation does not impact the Bank’s LTV calculation; it is only a factor that may be used in the Bank’s assessment of the need for an external appraisal.

We note your request to disclose the LTV range in our future filings. We do not believe specific disclosures of the threshold and LTV ranges would be
useful to investors because 1) we do not use the CMHC service to value properties but rather as one input to the decision to obtain an external appraisal, 2) we do not want to create an expectation with customers that external appraisals will not be
required if their mortgage is outside the specified range as the CMHC service is only one factor in the assessment, 3) this aspect of the loan origination process is used for a very small portion of our loan portfolio, and 4) these specific factors
may change in response to changes in economic factors as we review our overall risk management approach. We will include additional disclosure on how we use AVM, as described above, including that the tool is not used to provide a valuation, but is
used to assist in the risk assessment of whether residential mortgages require a full external appraisal.

•

Confirm that in future filings you will disclose how you validate the valuations provided by the AVM, including a description of the ‘existing valuation assessments’ and ‘alternative means’
that you utilize for determining that the output from the AVM was consistent with your valuation findings.

 The CHMC service,
which is only used for those residential mortgages that are below a specific dollar threshold with an LTV within a specified low range, assists in the risk assessment of whether residential mortgages require a full external appraisal. To validate
the service, we selected a sample of loan applications, processed them through the service, and also ordered full external appraisals. By comparing the actual external appraisal with BMO’s internal assessment we were able to validate the CMHC
recommendation on obtaining an external appraisal.

 In future filings we will clarify the use of the CMHC service.

 Page | 2

 Exhibit 99.3

 Notes
to consolidated financial Statements, page 124

 Note 4: Loans, Customers’ Liability under Acceptances and …page 131

Purchased Loans, page 133

2.
Refer to your response to prior comment 6. We note that the incurred credit mark for purchased performing loans represents the estimate of future cash flows that are not expected to be collected on the acquired
portfolio of purchased performing loans based on incurred but not specifically identified losses inherent in the portfolio. Please tell us the following related to your purchased performing loan portfolio:

•

How you differentiate between loan pools with incurred losses and those pools without incurred losses within this portfolio.

When we acquired the loan portfolio in the acquisition of Marshall and Ilsley (“M&I”), we assessed all individually significant purchased
loans on an individual basis to determine whether there was objective evidence of impairment as at the acquisition date in view of the existing facts and circumstances as at that date. Those individually significant loans identified as impaired were
classified as purchased credit impaired (“PCI”) loans. Loans that were not individually significant, such as the retail loan portfolio, were assessed for impairment based on characteristics such as delinquency status, FICO / risk rating
scores, and collateral type, with certain of the non-individually significant loans classified as PCI based on their characteristics. An incurred credit mark was determined for all PCI loans and is accounted for as discussed in our response to
question 8 of our prior response letter.

 We classified the remaining loans, for which there was no objective evidence of impairment, as the
purchased performing loan portfolio. For loans in the purchased performing loan portfolio, we determined an incurred credit mark on a pool-by-pool basis representing incurred but not yet specifically identified losses inherent in the portfolio as at
the acquisition date.

 Our methodology for determining the incurred credit mark on a pool-by-pool basis for the purchased performing loans was
consistent with our collective allowance methodology which incorporates both quantitative and qualitative factors. We grouped purchased performing loans on the basis of similarities in credit risk characteristics following a detailed review of the
loans at the acquisition date. Based on observable data obtained from our portfolio review upon acquisition, quantitative factors, such as probabilities of default, historical loss rates and loss given default, and qualitative factors, such as
current macroeconomic and business conditions, portfolio specific considerations, collateral value and model factors (as discussed in our response to the next question), were considered in determining the risk ratings for these loans. As discussed
in our response to question 6 of our prior response letter, the incurred credit mark for the purchased performing loan portfolio is re-measured each reporting period consistent with our collective allowance methodology.

 Page | 3

•

We note your response to prior comment 8 that you performed a detailed review of the risk rating of the purchased performing loans at acquisition and based on this review, you determined the incurred losses in
the portfolio that have not been identified within individual performing loans. Tell us the other characteristics, such as past due status, you relied upon in determining the loan pools with incurred losses.

As noted above, in order to determine the incurred credit mark on a pool-by-pool basis within the purchased performing loan portfolio at the date of
acquisition, we grouped loans into pools with similar credit risk characteristics to appropriately estimate the credit losses in each pool that had not been identified on an individual loan basis. The purchased performing loan portfolio was divided
into commercial, residential and consumer portfolios. Within each of these categories, the loans were further sub-divided by credit risk characteristics such as borrower type, collateral type, risk rating, past due status and FICO scores as
applicable, to determine the risk rating at the acquisition date. As noted above, we then determined the incurred credit mark based on the risk ratings assigned and the loan amount for each group using a methodology consistent with our collective
allowance methodology.

•

Discuss whether the loans with incurred losses were purchased at a discount due to credit quality, and clarify how the incurred credit losses are reflected in the price.

We recorded all acquired M&I loans at fair value on the date that we acquired them. The fair value of the acquired purchased performing loan
portfolio was determined by discounting expected cash flows using current market interest rates. The difference between the contractual cash flows discounted at the contractual interest rate and the fair value represented (i) an estimate of the
interest rate premium or discount on the loans calculated as the difference between the contractual rate of interest on the loans and prevailing interest rates (i.e., the interest rate mark), and (ii) an estimate of the cash flows not expected
to be collected in comparison to contractual cash flows (i.e., the credit mark). We disaggregated the credit mark into two components: an estimate of the amount of losses that existed in the acquired loan portfolio on the acquisition date but that
had not been specifically identified at an individual loan level on that date (i.e., the incurred credit mark, which was determined as noted above) and the remainder which represented future expected losses (that is, the future credit mark).

•

Tell us whether the effective interest rate you calculated on the purchased performing loans with incurred losses is equivalent or similar to a market yield. To the extent that the effective interest rate is
not, tell us, why, and how you concluded that the effective interest rate determined for these loans is consistent with the guidance in paragraph BC 32 of IAS 39.

 Page | 4

 In connection with the purchase accounting process, we reviewed the effective interest rates calculated on
the purchased performing loans and determined that they were similar to market yields on comparable loans, based on the market conditions prevailing at the time of acquisition.

We are accreting the purchase date fair value of the performing loan portfolio up to the contractual amount, less the incurred credit mark. In
determining the effective interest rates, we referred to the guidance in IAS 39.AG5 and IAS 39.BC32 which indicates that incurred credit losses should be considered in the expected cash flows when determining the effective interest rate.

3.
Refer to your response to prior comment 6 regarding revolving credit facilities and where you state that since you have committed to lend at any time, amortizing the total interest rate mark and credit mark to income
on a straight line basis over the term of the facility best reflects your standby obligation. Please respond to the following:

•

Tell us how you concluded that it was unlikely that specific lending arrangements will be entered into and the loan commitment is outside the scope of IAS 39 for all acquired revolving balances. In this
regard, if it was probable that you will enter into a specific loan arrangement, and the loan commitment is not within the scope of IAS 39, then the commitment fee is deferred and recognized as an adjustment to the effective interest rate. In your
response, provide quantitative information regarding the undrawn facilities for which you think usage is probable versus unlikely to occur.

We considered the acquired revolving loan facilities to be within the scope of IAS 39 and accordingly looked to the guidance in IAS 39.47(d) and IAS
18.

 As noted in our previous response, over the term of a revolving facility subsequent to acquisition, borrowers may repay the amounts that were
drawn on the acquisition date, make additional draws under the facility, repay all or a portion of those drawn amounts, and subsequently draw and repay numerous times in varying amounts. In a revolving credit facility, we are uncertain as to when,
and for how much, any drawdowns will be on any individual facility as such drawdowns are not within our control and therefore there is significant difficulty in predicting the timing and amount of the draws and repayments on each individual
facility.

 In view of the inherent uncertainty in estimating the amounts and timing of draws on each facility, and also considering the amounts
involved, we determined that an appropriate approach would be to amortize the total interest rate mark and credit mark for both the funded and unfunded revolving facilities on a straight line basis over the term of the facilities. We also believe
that such accounting is consistent with our view that the fair value mark on the revolving facilities (which comprises the aggregate of the interest rate mark, and the incurred and future credit marks) represents the amount a market participant
would expect as compensation for standing ready to provide credit for the full amount of the revolving facility over the term. As we have committed to lend at any time, amortizing the total interest rate mark and credit mark to income on a straight
line basis over the term of the facility best reflects our stand by obligation and the revenue earned for that service.

 Page | 5

 For those revolving facilities where amounts had been drawn as of the acquisition date, we determined the
fair value of the drawn amounts which, consistent with our discussion above, reflected the impact of credit losses inherent in the portfolio and the impact of interest rates. We determined the fair value of the undrawn portion of the revolving
facilities considering the impact of credit losses inherent in the portfolio, the impact of interest rates and the probability of funding the facility based on the borrower type. This probability ranged from 60% for portfolios such as agriculture to
100% for portfolios such as developer residential construction.

•

For revolving loans that had an existing balance at acquisition, tell us why you did not use the effective interest rate guidance outlined in IAS 39.9.

As noted above, for revolving loans that had a drawn amount as at the acquisition date, the borrower has the ability under the facility to repay the
drawn amount the next day, make further draws shortly thereafter, repay all or a portion of those drawn amounts, and subsequently draw and repay numerous times in varying amounts. Accordingly, for purposes of revenue recognition, we believe that a
reasonable approach is to associate the total interest rate mark and credit mark in respect of the drawn amounts at the acquisition date with the revolving facility as a whole, rather than with the drawn portion alone. For the reasons noted above,
we amortize the total interest rate mark and credit marks for the revolving facilities on a straight line basis over the term of the facilities.

•

Separately quantify the amount of the discount related to credit and interest rates on the revolving loans.

The credit mark of $835 million at the acquisition date represented less than 4% of th