Core Deposit Analytics Session 1: Determining Core The Basics

Transcription

1 Core Deposit Analytics Session 1: Determining Core The Basics Thomas A. Farin Chairman of the Board 1

2 Building Blocks for Deposit Analysis Pricing Betas Decay Rates Surge Balance Identification Segmentation Strategies Benchmark Rates Marginal vs. Average Cost Rate/Volume Tracking 2

3 Typical Lagging Reaction Rate Trends by Sector Mkt Range Mkt Low Average FHLB Advance - 12 Our High Our Low 33 Pct 3 10/28/ /4/ /11/ /18/ /25/ /2/ /9/ /16/ /23/ /30/2007 1/6/2008 1/13/2008 1/20/2008 1/27/2008 2/3/2008 2/10/2008 2/17/2008 2/24/2008 3/2/2008 3/9/2008 3/16/2008 3/23/2008 3/30/2008 4/6/2008 4/13/2008 4/20/2008 4/27/2008 5/4/2008 5/11/2008 5/18/2008 5/25/2008 6/1/2008 6/8/2008 6/15/2008 6/22/2008 6/29/2008 7/6/2008 7/13/2008 7/20/2008 7/27/2008 8/3/2008 8/10/2008 8/17/2008 8/24/2008 8/31/2008 9/7/2008 9/14/2008 9/21/2008 9/28/ /5/ /12/ /19/ /26/ /2/ /9/ /16/ /23/ /30/ /7/ /14/ /21/ /28/2008 1/4/2009 1/11/2009 1/18/2009 1/25/2009 9/30/ /7/ /14/ /21/2007

4 Pricing Betas Definition A pricing beta is a variable that is used to predict the effect of an increase in market rates on rates paid on a product. For example, if market rates increase 200 bp and your beta for MMDAs is 0.5 (50%) then the beta would predict you will raise MMDA rates by 100 bp 200 Bp X 0.5 = 100 Bp Betas can be: SWAG d Derived statistically from historic data. Examiners prefer the latter Betas can also be modified by use of segmentation strategies. 4

5 How Betas Play Out in Industry 1.0 Lag Premium MMDA & CDs Regular Savings 5 5

6 Pricing Betas Y = a + bx = Rate Paid a = Current Rate Note: b is the pricing beta 6

7 Pricing Betas a = 0.4% b =.75 Y = a + bx = Rate Paid 7

8 Pricing Betas Questions: 1. Which of these cost of funds profiles would you prefer to have with rising rates? 2. What would that information allow you to do with asset and funding allocation? 8

9 Ranked by Correlation Treasuries, Agencies, FHLB Advances 3, 6, 12, 24, 36, 60 months lagged 0-12 months = 234 combinations 9

10 Rate Paid vs. Equation (Back Test) Passbook Savings 10

11 Forecast Rates 12 Month Gradual 11

12 Pricing Beta Conclusions Applications A/L Modeling Income at Risk Value at Risk (core deposit intangible) Branch Valuation Credibility Issues Do you really price MMDAs off 2 year Treasury? What issues rise when historical betas are used to predict the future? Betas affected by Growth/Shrinkage Strategies Tier Structure Betas for different tiers will be different Segmentation Strategies Account Groupings Consumer Preferences Competitive/economic environment Betas Apply to CDs Non-maturity deposits 12

13 Non-Maturity Deposit Life Key Concept Non-maturity deposits don t all mature at the same time. Instead, balances in accounts decay off the books over time. Decay rates can be statistically measured. Once measured, decay rates can be used to forecast cash flows coming off pools of non-maturity deposits For all these reasons, your decay rates could be dramatically different than national averages. Cash Flow Decay rates affected by: Life events death, divorce, population turnover Satisfaction with the institution Movements in market rates and your pricing strategy. Economic events local (plant closings), and national (911, stock market health, economic outlook, etc.) Technology Interaction between CSRs and customers Flight to quality (Surge) Relationship between CD and NMD rates (Surge) 13

14 But before we calculate decay rates, we need to deal with surges Non-Maturity Deposits as a Percent of Assets All US banks on UBPR Stock CD Surge Fast Decay Core Slow Decay Relative pricing of non-maturity deposits vs. CDs 14 14

15 Decay Rate Methods Begin with a set of accounts and balances as of a certain date. Most recently we ve used a date at the top of the rate cycle Track what happens to balances of these accounts over time No new money considered Decay rate is the speed at which money disappears over time Must be adjusted for surge balances as they represent short-term funding sources. 15

16 Rolling Decays as Surge Indicators Decay rate on outstanding study balances Decay rate on outstanding study accounts Marginal (12 month rolling) decay rate 16

17 After Surge Adjustment After surge adjustment (20% surge) the marginal decay rate tracks along the cumulative average decay rate. 17

18 Average Balances Before and After Surge 18

19 Surge-Adjusted Decays 19

20 Decay Rates Surge Balance Burnout Truncation Assumption: 20% of balances are surge balances 20

21 Farin Core Study - Betas You don t want to end up applying these rate sensitive betas, to NRS balances. 21

22 Farin Core Studies Surges Surges Most don t represent a liquidity threat, instead they are a cost of funds threat as they will move back into CDs with rising rates. Need to knock surges out of low beta funding sources. 22

23 Farin Core Studies Decay Rates Remaining low beta funding sources (after surge knock outs) are long-term funding at a relatively fixed rate. 23

24 Farin Core Study Surges vs Core Betas In rising rate environment, current surge balances are likely to return to their source. We don t want to pay up for this portion to keep them in NMDs 24

25 Farin Core Study Surge vs. Core Betas LT FR Funding We want to increase rates on accounts with low betas by 25 those betas, which means we need to segment.

26 Using Results of Core Study Modeling Pricing betas drive cost of funds on non-maturity deposits as rates change in A/L Models. In other words they drive the interest cash flows. Decay rates drive the principal cash flows in A/L models. Crucial to Economic Value of Equity (EVE) calculations. Strategy Development Surge balances are a relatively short-term funding source. Beta will be in the range of 1.0 because funds will move back into CDs with a beta of 1.0 or leave for the markets and need to be replaced at market rates High beta non-surge are a close to variable rate funding source. Low beta non-surge are a long-term fixed-rate funding source. 26

27 Learning How To Effectively Segment Segmentation Goals Pay Up for Rate Sensitive Funds Don t Pay Up for Non-Rate Sensitive Funds Use Barriers to Entry to Block Cannibalization into Premium Products Reducing Combined Beta Geographic Minimum Balance or Tiering New Money Transaction Barriers Channel Barriers Stealth Products 27

28 Problem with Premium Accounts Issue Created to attract rate sensitive customers But over time lobby behavior loads with more and more non-rate sensitive accounts We don t want to pay up for non-rate sensitive funds in premium service lines Solution bottom of rate cycle Merge service lines to get everyone back in same pool. Inexpensive because Tiers are compressed Premium account rates pushed down against regular account rates. 28

29 Solution Rising Rates Merging service lines puts both rate sensitive and non-rate sensitive funds in same pool. Introduce new premium product in rising rate environment. No balances in premium account when introduced Rate sensitive funds move to new service lines which is priced using premium beta from study (+/-) Merged service line priced using regular account beta (+/-) 29

30 How To Segment Effectively Segmentation Goals Typical CD Pay Up for Rate Sensitive Funds Don t Pay Up for Non-Rate Sensitive Funds Use Barriers to Entry to Block Cannibalization into Premium Products Reducing Combined Beta Geographic Minimum Balance or Tiering New Money Relationship Pricing Exception Pricing Transaction Barriers Channel Barriers Stealth Products Non-Maturity Deposits Except New Money 30

31 New Product - Defensive New product with $50K barrier to entry priced defensively cut existing rates Saves $133 thousand per year, 8bp on funding cost Is it worth it? many would answer No and keep their new product powder dry 31 31

32 New Product Defensive (+200) Weighted cost cut by 28 bp saving $453,000. Beta cut from.51 to.37 32

33 New Product Offensive (+200) Betas same as previous slides but New MMDA is more aggressive Marginal cost of new funds, 1.21%, 107 bp below benchmark 33 33

34 New Product Offensive Teaser (+200) Rates rolled back to 70 th percentile after 6 months 15% of new money leaves Marginal savings on runoff is 7.79% 551 bp above replacement funding cost 34 34

35 Surge Balance Strategies 20% Surge Returns to CDs $43.1 K New money raised at MC of 1.2% 35

36 Effective Tracking Systems Price / Demand Pricing Consistent With Strategy? Hitting Demand Goals? Cost Effectiveness of Strategy Cannibalization New Money 36

37 Rate Trends Month CD 11/15/ /22/ /29/2006 7/11/ /6/ /13/ /20/ /27/2006 1/3/2007 1/10/2007 1/17/2007 1/24/2007 1/31/2007 2/7/2007 2/14/2007 2/21/2007 2/28/2007 3/7/2007 3/14/2007 3/21/2007 3/28/2007 4/4/2007 4/11/2007 4/18/2007 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/18/2007 7/25/2007 8/1/2007 Rate Trends Mkt Range Mkt Low 1 Year FHLB Advance Atlanta Our High Median Our Low ISR IstQSR 2 nd QSR 37

38 Volume Trends Balance Trends by Sector 38 8/29/ /1/ /13/ /20/ /31/2006 1/3/2007 1/10/2007 1/17/2007 1/24/2007 1/31/2007 2/7/2007 2/14/2007 2/21/2007 2/28/2007 3/7/2007 3/14/2007 3/21/2007 3/28/2007 4/4/2007 4/11/2007 4/18/2007 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/11/2007 7/18/2007 7/25/2007 8/1/2007 8/8/2007 8/15/2007 8/22/2007 Sum of balance ISR IstQSR Month CD sector date 2 nd QSR type Actual

39 Cannibalization Trends Balance Trends by Sector Sum of balance Balances On Off 39 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/11/2007 7/18/2007 7/25/2007 8/1/2007 8/8/2007 8/15/2007 8/22/2007 8/29/2007 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/11/2007 7/18/2007 7/25/2007 8/1/2007 8/8/2007 8/15/2007 8/22/2007 8/29/2007 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/11/2007 7/18/2007 7/25/2007 8/1/2007 8/8/2007 8/15/2007 8/22/2007 8/29/2007 4/25/2007 5/2/2007 5/9/2007 5/16/2007 5/23/2007 5/30/2007 6/6/2007 6/13/2007 6/20/2007 6/27/2007 7/4/2007 7/11/2007 7/18/2007 7/25/2007 8/1/2007 8/8/2007 8/15/2007 8/22/2007 8/29/ Mo CD 12 Mo CD Jumbo 14 Mo Special 14 Mo Special Jumbo Month CD sector account date type Actual

40 QUESTIONS? 40