Tip:
Highlight text to annotate it
X
- Hi, I'm Wade Walker, a Supervisory Examiner
in the Salt Lake City Field Office.
- And I'm Jaclyn Valderrama, a Case Manager
in the San Francisco Regional Office.
In this segment, we're going to discuss estimating credit
losses under Accounting Standards Codification
or ASC 450, including segmenting the portfolio, calculating
historical loss rates, and applying Qualitative
or Q-Factors.
We will also review a sample board report that summarizes
the ASC 450 analysis and aggregates those results
with the ASC 310 results.
As with all aspects of the ALLL analysis, please remember
that maintaining appropriate documentation is imperative
when estimating credit losses according to ASC 450.
In particular, management should develop and maintain
documentation that:
1) Supports the techniques
used to develop historical loss rates;
2) Supports Q-Factor adjustments to the historical loss rates;
3) And, ultimately, supports the estimated loss rate calculated
for each segment of the portfolio.
The first step in the ASC 450 analysis is to group loans
with similar risk characteristics -
in other words, segment the portfolio.
Some of the typical group characteristics that banks use
include:
1) Loan Type, for example: secured commercial loans,
residential mortgage loans, or consumer installment loans,
2) Past Due Status, and
3) Risk Grade.
Segmenting the portfolio based on other risk characteristics
may be useful as well.
Some of the Other Risk Characteristics might include
geographic location of the collateral or industry
concentrations.
For smaller, less complex banks, Call Report Schedule
RC-C may provide an initial segmentation.
- Less complex portfolios require fewer segmentation
categories than more complex portfolios.
In fact, more complex portfolios may need to break
down a category like Loan Type into several subgroups,
such as product line segments, geographic location,
collateral protection, or originating branch.
The whole purpose of segmenting the portfolio
for ASC 450 analysis is to group loans with similar
characteristics that will most likely have similar loss rates.
By doing this segmentation, management can provide
for a group of loans without having to assess the estimated
loss in each individual loan.
However, without a meaningful segmentation of the portfolio,
the application of loss rates yields little value.
- So, let's look at an example of portfolio segmentation.
See if you notice any concerns.
Hopefully, you noticed that the segmentation includes both
some effective and less effective attributes.
As you can see, the portfolio is segmented by loan type
and risk grade.
However, the loan type breakdown lacks detail.
For instance, real estate loans represent most
of the entire loan portfolio but are only broken down into two
categories.
Even if the bank had used a simplistic approach of
segmenting the real estate loans according to Call Report
Schedule RC-C, the bank would have had at least five
categories for real estate loans.
Another potential concern is that Grades 6 and 7
are combined in one group for analysis despite the fact
that Grade 6 and Grade 7 probably have different loss
characteristics.
- The next step in the ASC 450 analysis is to apply loss rates
against the various portfolio segments identified in step one.
The 2006 Interagency Policy Statement on the ALLL states,
"Methodologies for determining the historical net charge-off
rate on a group of loans with similar risk
characteristics,
can range from the simple average of a bank's
annual net charge-off experience to more complex
techniques, such as migration analysis and models
that estimate credit losses."
- So, to be clear, the regulators don't prescribe
any particular method a bank must use in determining
its historical loss rates?
- That's correct, Jackie.
Banks may choose from a wide variety of methods
to determine historical loss rates and estimate credit
losses inherent in the current portfolio.
Some of the more common methods include:
1) Portfolio Loss History Analysis,
which utilizes an average of losses over time,
2) Loss Migration Analysis, which incorporates
grade changes over time, and
3) Probability of Default/Loss
Given Default Modeling, which requires more complex modeling
and assumptions.
Many banks are now using a Loss Migration Analysis
as their preferred method of analyzing historical losses
and estimating current losses.
We provide additional information on using this type
of analysis in another segment.
- Well, what if a bank is new or has gotten into a new lending
product for which it has no loss experience of its own?
How does the bank estimate loss rates?
- That's a good question, Jackie.
The 2006 policy statement indicates that "if a bank
has no experience of its own for a loan group, it should
reference the experience of other enterprises in the same
lending business ...on a short-term basis until
it has developed its own loss experience..."
Now, let me ask you a question.
As a bank looks back over its historical losses and tries
to estimate credit losses in the current portfolio, what kind
of look-back period should a bank use in its analysis?
- Well, the 2006 policy statement notes that annual
charge-off rates are calculated over a specified
time period - for example, three years or five years -
which can vary based on a number of factors, including
the relevance of past experience to the current
point in the credit cycle.
In other words, banks should consider their historical
losses and make sure the time period they're looking at
is relevant to the current environment.
More specifically, banks should not estimate losses
in a recession based on loss experience during an economic
boom.
Make sure the time periods used are relevant.
- Okay, so let's say a bank is using a three-year look-back
period in an improving environment.
Wouldn't the most recent year's loss experience be more
meaningful to the bank than the loss experience from three
years ago?
- Absolutely.
That's why many banks choose to weight the time periods
differently.
In the example you laid out, the bank may choose to assign
varying weights to the different years
in the look-back period.
So the most recent year may be assigned a greater weighting,
while the third year back may be assigned a smaller
weighting.
By doing this, the bank doesn't have to constantly
change the number of years in the look-back period but can
still account for changes in the economic environment
or other pertinent factors.
- Speaking of other pertinent factors, historical losses,
or even recent trends in losses, do not by themselves form
a sufficient basis to determine the appropriate level
of the Allowance - which brings us to the final step
of the ASC 450 analysis - Qualitative factor adjustments.
Qualitative factors, or Q-Factors, are qualitative
or environmental factors that are likely to cause estimated
losses to differ from historical losses.
The 2006 Policy Statement indicates that management
should consider the following factors:
1) Changes in lending policies and procedures, including
underwriting standards,
2) Changes in economic and business conditions,
3) Changes in the nature and volume of the portfolio,
4) Changes in the experience,
ability, and depth of lending management,
5) Changes in the volume and severity of past due loans,
nonaccrual loans, and adversely classified loans,
6) Changes in the quality of the loan review system,
7) Changes in collateral values,
8) Existence of, and changes in,
any concentrations of credit, and
9) Changes in competition and legal or regulatory
requirements.
- One of the most difficult and imprecise components
of the ALLL analysis is trying to convert these Q-Factors
into actual numerical adjustments.
In our experience, unless there are significantly
changing economic conditions, Q-Factor adjustments
are usually relatively small adjustments to the overall
estimate of credit losses.
Q-Factors that are overly aggressive, large,
or unsupported, may call into question the appropriateness
of the adjustments.
As examiners, we recognize the difficulty of converting
Q-Factors into numerical adjustments and generally will
accept management's estimates when supported by appropriate
documentation.
- Well, that wraps up our ASC 450 analysis.
But, before we conclude this segment, we'd like to review
an example of what a Board report might look like
that aggregates the ASC 310 and ASC 450 analyses.
First, the sample board report lists the individual loans
considered under ASC 310, their entire loan balance,
and the measurement of impairment derived during
the individual impairment analysis.
Next, the report lists the ASC 450 calculation.
For purposes of this example, a Loss Migration Analysis
was used to determine the ASC 450 calculation.
As you can see, the portfolio is broken down by loan type
and risk grade - and then a historical loss factor
and a Q-Factor adjustment are applied against each segment's
loan balance to arrive at an appropriate Allowance
allocation.
Finally, the ASC 310 portion is added to the ASC 450
portion to arrive at the Total ALLL balance.
- One important thing to note is that in order to avoid double
counting or "layering," the loan balances associated
with the ASC 310 impaired loans are not included in their
respective categories in the ASC 450 breakdown.
If those balances are included, then potential
losses are allocated for those loans twice -
once in the individual ASC 310 impairment calculation,
and again in the group ASC 450 calculation.
In the example that we just covered, the loan
to WidgetMaker, Inc., which is listed in the ASC 310 section,
is a Commercial and industrial, or C&I, loan.
If that loan were included in any of the C&I categories
in the ASC 450 analysis, then that loan would have been
accounted for twice.
However, management must consider ASC 310 losses when
calculating historical loss rates for the ASC 450
segmentation; otherwise, the loss rates for the ASC 450
analysis will be understated.
- Thanks, Jackie.
In this segment, we reviewed the steps in the ASC 450
Allowance analysis.
The first step is to segment the portfolio into groups
with similar risk characteristics.
Typical group characteristics include loan type, risk rating,
and past due status.
The second step is to calculate historical loss rates.
Several methods are available to calculate historical loss
rates, including portfolio loss history analysis,
loss migration analysis, and probability of default/loss
given default modeling.
And, finally, the third step is to apply qualitative
factors, or Q-Factors, that are likely to cause estimated
losses to differ from historical losses.
We concluded by reviewing a sample board report
which aggregated the ASC 310 calculation with the ASC 450
calculation to arrive at the total Allowance amount.
We hope this segment has been beneficial in understanding
the process for estimating losses under ASC 450.