Tuesday, September 03, 2019

|| The Other Side of the Balance Sheet

  • We have been on a journey to create a credible alternative to branches, centered on mobile
  • We have digitized many things that can be done in a branch
    • Deposit checks
    • Order cashiers check
    • Etc
  • Beyond everyday banking, mobile is also central to our decisioning, messaging, identity and authentication strategy, ie. fraud SMS communication, mobile ID validation
  • Mission
    • Build an obsessively efficient, zero cost to serve business
    • Grow engaged customers and deposits, nationally
    • Differentiate from the competition
  • Agenda - account opening and eligibility, digital adoption, identity and protect, inform and alert
  • Accelerate scale to digital first national bank by building compelling customer experiences that enable zero marginal cost of banking
  • Fraud, key high risk transactions, authentication method, code complexity
    • ATO, internal fraud
    • Cash withdrawal, cashiers check
    • Debit card printing
    • Domestic wires
    • International wires
  • Inspiration outside of the financial services
    • Amazon go - mobile led
    • Aila - simple user interface
    • Biometrics - clear
  • Enabling purchase push notification will accelerate bank’s ability to immediately inform our customers at key moments to drive PBR
  • And we continue to push towards greater fraud controls through 2-way SMS capabilities
  • Differentiation, Predictive communication

Lyft payment - receiving end, fraud has no liability (eg. customers use fraudulent card), more important on customer confidentiality (internal fraud), cyber security (cell phone, account, identify fraud), SHUA DAN (Lyft drivers need volumes to get a higher incentive rates; need to look into driver, passenger, location, matching logics to identify these clusters of transactions), customers’ disputes (not riding but getting charged), 1) revenue 2) loss (transaction fraud, identify fraud) 3) disputes … Financials, valuation, predictability (secondary order impact, customer based valuation … budget, planning, limit loss for xx) … Didi recruiting for data science modeling, look for job description; volume is not as high as uber … not as innovative, more as a follower … Customer and Trust position just released

Head of Deposit Analytics - data insights, correlations (deposit more or less balances, economics, pricing, rates, checking products / savings for cannibalization, customer growth, retention, testing, experimental testing, customer personalization tree … optimized); personalize experiences, interaction, UI, placement, verbage, value proposition

  • Hypothesis test, A or B testing
  • Regression prediction and pattern recognition
    • Linear extrapolation
    • Logistic regression - category 1 or 0, customer direct mail (info and bureau, decision tree, score the customer, closer to 1, approve) 

LA = DC almost; SF = 2.5 owning of DC index (1.6 or 1.8 renting DC rate); lacking exposure, job opportunity

Deposit 
  • Monitor transactions by deposit channel, ATM, RDC, Teller
  • Sign on
  • Enrollment
  • Check image
  • Statement
  • RDC
  • Transfers
  • Billpay
  • P2P
  • Debit card conversion
  • Cum money flow
    • Cann from other deposit products
    • Growth from existing customers
    • New from new customers
    • Total outflow
  • Landing page
  • Product feature: branch access, platform, OD policy, cycle fee, ATM, upmarket add-on, Other fees (stop payment, cashiers check, wire, checkbook, expedited debit card)

Deposit analytics

  • Mission of the deposit business, funding, relationship values, profitability
  • Funding - account level balance stickiness, seasonality, pricing sensitivity, different rate environment savings along, savings and checking, and lending
  • Relationship values - lower losses, debit transactions, fee based services, robo advising investing opportunities
  • Profitability - relationship based pricing between lending and deposit, customer perceived values
  • Product - deposit product build, rotations, rate environment (putting everything together for the customer financial product life cycle)

  • Monthly reviewing balance sheet deposit demand framework, fed rate outlook, competitor (used to do WAMP) but these days big 8 modeling
  • Market value of deposit, deep understanding of the deposit FTP fundamentals, ECR beta (small business), core, valuation of deposit
  • Rate paid, Direct vs the big 8 direct players on savings; monitoring deposit flows, patterns, competitor movements … model scenarios of balance and profitability.  This is the cost of gathering deposit


  • Passionate about checking, eligibility (qualifile … FICO, fraud claims, income, have DDA etc) and approval, transactions, customer protection, debit etc
  • Customer value, by market, by age group, by channel
  • Multi-product relationship

  • Customer segmentation - savings, checking (risk driven), by market

Purpose of managing deposits
  • Balance sheet strength between consumer, commercial, debt and capital market; maturity and concentration
    • Daily cash inflow and outflow monitoring
    • Competitor bank rate, product launches
    • Forecasting accuracy
    • New money, cannibalization, product rotation (if you have multiple products, eg. between  CD and savings)
  • Profitability
    • Pricing decisions for portfolio management - vs Big 8 or WAMP; strategic positioning vs short term earnings power
    • Market decisions, mostly applicable to retail banking with geo pricing capabilities; in combination of distribution cost and etc … this may lead to M&A opportunities
    • Segmentation analysis, ACES, product experiences, customer management and marketing
    • Cost consumption analysis, customer protection cost and call center
    • Relationship, checking agenda (a lot more work, savings), long term values, multi-product analysis
  • Testing agenda
    • Marketing promo
    • Customer beta vs driver rates
    • Moment that matters for NPS
    • Product and pricing capability build
  • Data lake, account level profitability; visual reporting

No comments: