Fifth Third Bank 2014 Annual Report Download - page 74

MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS
72 Fifth Third Bancorp
The ALLL attributable to the portion of the residential
mortgage and consumer loan and lease portfolio that has not been
restructured is determined on a pooled basis with the segmentation
based on the similarity of credit risk characteristics. Loss factors for
real estate backed consumer loans are developed for each pool
based on the trailing twelve month historical loss rate, as adjusted
for certain prescriptive loss rate factors and certain qualitative
adjustment factors. The prescriptive loss rate factors and qualitative
adjustments are designed to reflect risks associated with current
conditions and trends which are not believed to be fully reflected in
the trailing twelve month historical loss rate. For real estate backed
consumer loans, the prescriptive loss rate factors include
adjustments for delinquency trends, LTV trends, refreshed FICO
score trends and product mix, and the qualitative factors include
adjustments for credit administration and portfolio management
practices, credit policy and underwriting practices and the national
and local economy. The Bancorp considers home price index trends
in its footprint when determining the national and local economy
qualitative factor. The Bancorp also considers the volatility of
collateral valuation trends when determining the unallocated
component of the ALLL.
The Bancorp’s determination of the ALLL for commercial
loans is sensitive to the risk grades it assigns to these loans. In the
event that 10% of commercial loans in each risk category would
experience a downgrade of one risk category, the allowance for
commercial loans would increase by approximately $162 million at
December 31, 2014. In addition, the Bancorp’s determination of the
allowance for residential and consumer loans is sensitive to changes
in estimated loss rates. In the event that estimated loss rates would
increase by 10%, the allowance for residential and consumer loans
would increase by approximately $34 million at December 31, 2014.
As several qualitative and quantitative factors are considered in
determining the ALLL, these sensitivity analyses do not necessarily
reflect the nature and extent of future changes in the ALLL. They
are intended to provide insights into the impact of adverse changes
to risk grades and estimated loss rates and do not imply any
expectation of future deterioration in the risk ratings or loss rates.
Given current processes employed by the Bancorp, management
believes the risk grades and estimated loss rates currently assigned
are appropriate.
TABLE 54: CHANGES IN ALLOWANCE FOR CREDIT LOSSES
For the years ended December 31 ($ in millions) 2014 2013 2012 2011 2010
A
LLL:
Balance, beginning of period $ 1,582 1,854 2,255 3,004 3,749
Impact of change in accounting principle - - - - 45
Losses charged-off (679) (637) (837) (1,314) (2,485)
Recoveries of losses previously charged-off 104 136 133 142 157
Provision for loan and lease losses 315 229 303 423 1,538
Balance, end of period $ 1,322 1,582 1,854 2,255 3,004
Reserve for unfunded commitments:
Balance, beginning of period $ 162 179 181 227 294
Impact of change in accounting principle - - - - (43)
Benefit from the reserve for unfunded commitments (27) (17) (2) (46) (24)
Balance, end of period $ 135 162 179 181 227
Certain inherent, but unconfirmed losses are probable within the
loan and lease portfolio. The Bancorp’s current methodology for
determining the level of losses is based on historical loss rates,
current credit grades, specific allocation on impaired commercial
credits above specified thresholds and restructured loans and other
qualitative adjustments. Due to the heavy reliance on realized
historical losses and the credit grade rating process, the model-
derived estimate of ALLL tends to slightly lag behind the
deterioration in the portfolio, in a stable or deteriorating credit
environment, and tend not to be as responsive when improved
conditions have presented themselves. Given these model
limitations, the qualitative adjustment factors may be incremental or
decremental to the quantitative model results.
An unallocated component to the ALLL is maintained to
recognize the imprecision in estimating and measuring loss. The
unallocated allowance as a percent of total portfolio loans and leases
at both December 31, 2014 and 2013 was 0.12%. The unallocated
allowance was eight percent of the total allowance as of December
31, 2014 compared to seven percent as of December 31, 2013.
As shown in Table 55, the ALLL as a percent of portfolio
loans and leases was 1.47% at December 31, 2014, compared to
1.79% at December 31, 2013. The ALLL was $1.3 billion as of
December 31, 2014, compared to $1.6 billion at December 31, 2013.
The decrease was reflective of decreases in nonperforming loans
and leases and improved delinquency metrics in commercial and
consumer loans and leases.