KeyBank 2015 Annual Report Download - page 118

others is deteriorating, we may decide to change the level of the allowance for one segment of the portfolio
without changing it for any other segment.
In addition to adjusting the ALLL to reflect market conditions, we also may adjust the allowance because of
unique events that are likely to cause actual losses to vary abruptly and significantly from expected losses. For
example, class action lawsuits brought against an industry segment (e.g., one that used asbestos in its product)
can cause a precipitous deterioration in the risk profile of borrowers doing business in that segment. Conversely,
the dismissal of such lawsuits can improve the risk profile. In either case, historical loss rates for that industry
segment would not have provided a precise basis for determining the appropriate level of allowance.
Even minor changes in the level of estimated losses can significantly affect management’s determination of the
appropriate allowance because those changes must be applied across a large portfolio. To illustrate, an increase in
estimated losses equal to one-tenth of one percent of our consumer loan portfolio as of December 31, 2015,
would indicate the need for a $16 million increase in the allowance. The same increase in estimated losses for the
commercial loan portfolio would result in a $44 million increase in the allowance. Such adjustments to the ALLL
can materially affect financial results. Following the above examples, a $16 million increase in the consumer
loan portfolio allowance would have reduced our earnings on an after-tax basis by approximately $10 million, or
$.01 per common share; a $44 million increase in the commercial loan portfolio allowance would have reduced
earnings on an after-tax basis by approximately $28 million, or $.03 per common share.
As we make decisions regarding the allowance, we benefit from a lengthy organizational history and experience
with credit evaluations and related outcomes. Nonetheless, if our underlying assumptions later prove to be
inaccurate, the ALLL would likely need to be adjusted, possibly having an adverse effect on our results of
operations.
In the third quarter of 2015, we enhanced the approach used to determine the commercial reserve factors used in
estimating the commercial ALLL, which had the effect of capturing certain elements in the commercial
quantitative reserve component that had formerly been included in the commercial qualitative component. Under
the enhanced methodology, we began utilizing more refined commercial estimated loss rates that represent
cumulative losses over the estimated average time period from the onset of credit deterioration loss to the initial
loss recorded for an individual loan. In addition, we began utilizing an enhanced framework to quantify
commercial ALLL adjustments resulting from qualitative factors that may not be fully captured within the
statistical analysis of incurred loss. The enhanced framework utilizes the nine qualitative factors recommended
within the OCC’s Interagency Policy Statement on the Allowance for Loan and Lease Losses (issued 12/13/2006)
to quantify the commercial qualitative component. The impact of these changes was largely neutral to the total
ALLL. However, because the quantitative reserve is allocated to the business segments at a loan level, while the
qualitative portion is allocated at the portfolio level, the impact of the methodology enhancements on the
allowance for each business segment and each portfolio caused the business segment and commercial portfolio
reserves to increase or decrease accordingly. The impact of the increases and decreases on the business segment
and commercial portfolio reserves was not significant.
Our accounting policy related to the allowance is disclosed in Note 1 under the heading “Allowance for Loan and
Lease Losses.”
Valuation methodologies
We follow the applicable accounting guidance for fair value measurements and disclosures, which defines fair
value, establishes a framework for measuring fair value, and expands disclosures about fair value measurements.
In the absence of quoted market prices, we determine the fair value of our assets and liabilities using internally
developed models, which are based on third-party data as well as our judgment, assumptions and estimates
regarding credit quality, liquidity, interest rates and other relevant market available inputs. We describe our
104