Santander Holdings USA, Inc. (NYSE:SOV-C) Files An 8-K Other Events

Santander Holdings USA, Inc. (NYSE:SOV-C) Files An 8-K Other Events

Story continues below

Item 8.01 Other Events

Introduction
Santander Holdings USA, Inc. (the Company or “SHUSA”) is a bank
holding company (“BHC”) in the U.S. with $137.4 billion in assets
as of December 31, 2016. Headquartered in Boston, Massachusetts,
the Company is the parent company of Santander Bank, National
Association (Santander Bank), and owns a majority interest
(approximately 59%) of Santander Consumer USA Holdings Inc.
(“SC”), a specialized consumer finance company focused on vehicle
financing and third-party servicing. Through these entities, the
Company offers a full range of consumer and commercial banking
products and services. SHUSA is a wholly-owned subsidiary of
Banco Santander, S.A. (Santander). The Company has developed
stress testing processes that capture its unique mix of
businesses, risks and geographic footprint.
Dodd-Frank Act Stress Test Requirements
The Board of Governors of the Federal Reserve System (the Federal
Reserve) requires BHCs with total consolidated assets of $50
billion or more to publish a summary of the annual Dodd-Frank Act
Stress Test (DFAST) results based on the supervisory severely
adverse scenario. The results are to reflect a standard set of
capital actions over the nine-quarter forecast horizon (DFAST
Capital Actions).
Forecast Approach
The Federal Reserve provides macroeconomic scenarios which BHCs
are required to employ in their annual stress testing processes.
A complete description of the most stressful scenario, the
supervisory severely adverse scenario, and the corresponding
macroeconomic variables, can be found on the Federal Reserves
website.
The Company considers and evaluates all material risks facing the
Company in its capital adequacy and stress testing processes
including, but not limited to, the following:
Credit risk is the risk of loss arising from a borrowers or
counterpartys failure to perform on an obligation. Credit
risk at the Company is driven by real estate-related
lending, large corporate exposures, individual consumer
lending and automobile finance exposures. Credit risk is
incorporated into our stress results primarily through
credit-sensitive macroeconomic models.
Operational risk is the risk of loss resulting from
inadequate or failed internal processes, people, or
systems, or external events. This risk is incorporated into
our stress test results through macroeconomic models,
increases in legal reserves and through specific adverse
events considered in our forecast.
Market risk is the risk to the Companys financial condition
from changes in interest rates or changes in the market
value of assets and traded financial instruments, both
on-balance sheet and off-balance sheet. Market risk impacts
stress results through impairment of consumer assets as
well as changes in net interest income and changes in the
value of the investment portfolio.
Liquidity risk is the risk arising from an inability to
meet obligations as they come due in the normal course of
business without causing undue hardship or incurring
unacceptable loss. Liquidity risk is incorporated in the
stress forecast through limited funding availability,
increased cost of funds and credit rating downgrades.
Business risk is the risk of decline in new loan volume
originations and thereby the associated income-generating
capacity of assets. The stress scenarios are characterized
by recessionary environments with reduced economic
activity, which results in less demand for loans by
consumers and businesses.
The Company employs various quantitative and qualitative
methodologies in its stress tests for balance sheet, income
statement, and capital projections. Where practical, the Company
utilizes modeling techniques in developing its stress test
estimates, including regression-based modeling, transition
analytics, business analytics, and other objective, quantitative
forecasting methodologies. When models are not used, non-model
forecast tools are leveraged to prepare estimates. Forecasting
methodologies, management overlays and final stress test results
undergo multiple rounds of review and challenge. The Capital Plan
is ultimately approved by the Companys Board of Directors.
Additional methodologies employed in developing the stress
testing include:
Pre-provision Net Revenue (PPNR): PPNR forecasting
methodology consists of multiple models, calculations and
driver-based expert judgment analytical approaches used in
conjunction to forecast individual line items for various
components of the balance sheet and income statement.
Forecasts for term loans and deposit balances as well as
forecasts for yields and pricing generally leverage
contractual terms for the existing portfolio, while new
loans and deposits are forecasted using statistical
regression models and expert judgment approaches.
Non-interest income and expense items are forecasted
primarily using driver-based analytical approaches
utilizing modeled and expert judgment components. Net
interest income is calculated based on contractual and
projected interest rates applied to forecasted loan and
deposit balances.
Credit Loss Estimation:>For the purposes of capital
planning and loss forecasting, the Company has developed
loss forecasting models driven by macroeconomic variables
across the commercial, retail and automobile portfolios.
Credit loss estimations drive changes in reserve levels and
ultimately provisions. The forecast for provisions
incorporates loss projections and target reserve coverage
rates.
Operational Loss Estimation:>The Company utilizes
statistical approaches to explicitly link operational
losses to macroeconomic factors, based on both internal and
external loss data and scenario analysis, to identify and
evaluate the potential impacts from low-frequency,
high-severity loss events linked to the Companys risk
profile.
Capital and Risk-Weighted Assets: The methodologies
described above translate identified risks into potential
revenue and loss projections, which are aggregated into
consolidated net income (loss) estimates over the
nine-quarter planning horizon. These estimates then feed
the Companys regulatory capital estimation process.
Regulatory capital and risk-weighted asset calculations are
based on existing regulatory guidance, including Basel III
transition provisions. The resulting regulatory capital
ratios are continuously compared to managements capital
target levels, a process which is a key factor in the firms
capital adequacy framework.
Summary of Results
The results presented below contain forward-looking projections
that represent estimates based on the supervisory severely
adverse scenario, a hypothetical set of conditions provided by
the Federal Reserve that involve an economic outcome that is more
adverse than expected. These estimates are not forecasts of
expected losses, revenues, net income or capital ratios.
Under the supervisory severely adverse scenario, the most
significant drivers of the Companys regulatory capital ratios and
those of Santander Bank are reductions in capital, driven by
higher credit losses and lower PPNR, and declines in total assets
and risk-weighted assets. For both the Company and Santander
Bank, in the supervisory severely adverse scenario, capital
depletion has a more significant impact on capital ratios than
the reduction in assets, resulting in a decline across all
capital ratios.
The supervisory severely adverse scenario also leads to the
following results:
Increases in credit losses and provisions that are
associated with the more significant portfolios, such as
retail mortgage, commercial and industrial, and commercial
real estate at Santander Bank and automotive at SC;
Stress impacts on PPNR that are driven by various factors
including lower net interest income and fees caused by
lower asset balances, net interest margin compression, and
higher operational risk losses. Stress impact also includes
higher depreciation and impairment expenses on operating
lease assets.
The decline in assets at the Company and Santander Bank are
largely due to lower loan balances, while Santander Bank
also forecasts reductions in automobile leases as well as
cash and securities.
The chart below shows the impact of these changes to SHUSAs
common equity tier 1 (CET1) ratio over the nine-quarter forecast
horizon. The CET1 ratio decreases approximately 320 basis points
from December 31, 2016 to March 31, 2019. The primary driver for
this decrease is the $8.9 billion provision expense versus PPNR
of $3.1 billion. Risk-weighted assets also decrease by $15
billion during the forecast, which offsets some of the decline in
the CET1 ratio.
chart02.jpg
(1)>Other includes the impact of intangible assets,
as well as income allocated to minority interests
(2)>Capital Actions include contributions
associated with planned transactions in compliance with
Regulation YY.
SHUSA
The estimates below reflect capital ratios for SHUSA under the
DFAST Capital Actions described previously.
Table 1: SHUSA Actual and Projected Capital Ratios through Q1
2019
Actual Q4 2016
Severely Adverse
Ending
Minimum
Common equity tier 1 capital (%)
14.5
%
11.3
%
11.3
%
Tier 1 risk-based capital ratio (%)
16.1
%
12.7
%
12.7
%
Total risk-based capital ratio (%)
18.0
%
14.5
%
14.5
%
Tier 1 leverage ratio (%)
12.5
%
10.4
%
10.4
%
Table 2: SHUSA Actual and Projected Risk-Weighted Assets through
Q1 2019
Actual Q4 2016
Q1 2019
Risk-weighted assets ($bn)
$
104.3
$
89.3
(1) Risk-weighted assets are calculated under the
transition rules for the Basel III standardized approach.
Table 3: SHUSA Projected Losses, Revenues, and Net Income through
Q1 2019
Severely Adverse
$bn
% of average assets
Pre-provision net revenue
$
3.1
2.5
%
Other revenue
Less
Provisions
8.9
Realized gains (losses) on securities
available-for-sale (“AFS”) / held to maturity (“HTM”)
Total trading and counterparty losses
Total other losses
0.0
Equals
Net income (loss) before taxes, extraordinary items and
minority interest
$
(5.8
)
(4.6
)%
(1)>Average assets is the nine-quarter average of
total assets.
(2)>PPNR includes losses from operational risk
events, mortgage repurchase expenses, and other real estate owned
costs.
(3)>Total may not sum due to rounding of line
items.
Table 4: SHUSA Projected Loan Losses by Loan Type Q1 2017 through
Q1 2019
Severely Adverse
$bn
Portfolio Loss Rates (%)1
Loan losses
$
9.3
11.6
%
First lien mortgages
0.2
3.5
%
Junior liens and home equity lines of credit (“HELOCs”)
0.2
2.5
%
Commercial and industrial
0.8
4.9
%
Commercial real estate
0.9
5.2
%
Credit cards
0.1
21.7
%
Other consumer
7.1
26.0
%
Other loans
0.1
1.0
%
(1)>Average loan balances used to calculate
portfolio loss rates exclude loans held-for-sale and loans
held-for-investment under the fair value option, and are
calculated over nine quarters.
(2) >Total may not sum due to rounding of line
items.
(3)>Commercial and industrial loans include small-
and medium-enterprise loans.
(4)>Other consumer loans include student loans and
automobile loans.
Santander Bank
The estimates below reflect capital ratios for Santander Bank
under the DFAST Capital Actions described previously.
Table 5: Santander Bank Actual and Projected Capital Ratios
through Q1 2019
Actual Q4 2016
Severely Adverse
Ending
Minimum
Common equity tier 1 capital (%)
16.2
%
13.4
%
13.4
%
Tier 1 risk-based capital ratio (%)
16.2
%
13.4
%
13.4
%
Total risk-based capital ratio (%)
17.4
%
14.3
%
14.3
%
Tier 1 leverage ratio (%)
12.3
%
11.3
%
11.3
%
Table 6: Santander Bank Actual and Projected Risk-Weighted Assets
through Q1 2019
Actual Q4 2016
Q1 2019
Risk-weighted assets ($bn)
$
61.9
$
55.1
(1)>Risk-weighted assets are calculated under the
transition rules for the Basel III standardized approach.
An ad to help with our costs