The Role of Branch Networks and Information Scale in the Coming Deposit Re-price Cycle

ginna
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

SEE LAST PAGE OF THIS REPORT Howard Mason

FOR IMPORTANT DISCLOSURES 203.901.1635

hmason@ssrllc.com

June 22, 2014

The Role of Branch Networks and Information Scale in the Coming Deposit Reprice Cycle

“The customer who likes to visit a branch is less price sensitive than a customer whose relationship is online-only. Banks haven’t fully understood that if I have a branch in this neighborhood, it reduces price sensitivity[1].” Frank Rohde, CEO of Nomis Solutions

  • Over the next three years, the key driver of bank earnings will be the rate sensitivity of deposits. Applying the 50% re-pricing beta (i.e. ratio of rate-increase on deposits vs. rate-increase on short-term wholesale funds) of the last up-rate cycle to the $10tn of deposits held by US commercial banks and the 2.8% increase in 6-month Libor indicated by the forward markets through mid-2017 suggests deposit franchises will add $100bn to after-tax earnings; this compares with total 2013 earnings for US commercial banks of $144bn.
  • We acknowledge the benefit in this up-rate cycle may be less with the customary shift to high-yielding certificates more pronounced (and reversing a trend that has caused non-interest bearing deposits to rise to 25% of total deposits from 17% in 2008) given lower search-and-switch costs for depositors: services like bankrate.com create have increased pricing transparency and online-only bank platforms (now accounting for 6% of deposits) have empowered consumers to “rate-surf” by shifting balances between banks.
  • However, the bigger issue is likely to be variations in rate-sensitivity across banks: in the last up-rate cycle, the best-performing quartile of the to-50 US banks had a re-pricing beta (i.e. the ratio of change in rate-paid on deposits to change in short-term wholesale funding rates) of 33% vs. 78% for the worst-performing. The hi-lo range of 45% across banks compares with JPM’s adjustment of 5% to its deposit re-pricing beta to account for secular change in customer behavior across time; specifically, JPM uses a go-forward beta of 50% vs. the 45% of the last up-rate cycle.
  • As rates begin to rise and the risk of deposit outflows increase, the challenge for banks will be attracting and retaining balances from rate-sensitive customers while not re-pricing the entire book; we see the trade-off between balance-formation and margin-enhancement as stimulating more sophisticated and finely-segmented deposit pricing than in the past. Winners will execute pricing strategy at the level of individual customers in a massive extension to data-enabled marketing of the “desk drawer” policies of past up-rate cycles (where branch officers would extend a special rate-offer to high-value customers threatening attrition).
  • Key success factors in data-enabled deposit marketing and pricing will be information-scale and return-to-skill: the scale to capture large amounts of data on customer interactions so as to develop statistically reliable insight into behavior, and the skill to monetize these behavioral insights across large numbers of customers. Beyond their favorable impact on customer loyalty, branches play an important role in data-gathering since they remain a key point-of-interaction for large swathes of customers[2], and advantaged pricing for these customers gives a branch-bank more flexibility in competing on price with online platforms for less branch-engaged customers.
  • This thesis rests on banks digitizing in-branch interactions, and we see this (along with reducing the need for tellers and improving customer convenience) as an important driver of large-bank moves to automate in-branch experience through, for example, touchscreen ATMs and iPad-equipped tellers. We expect this will evolve to support phone-enabled interaction of customers with “smart” physical infrastructure over wireless channels such as Bluetooth and NFC. Of course, this is part of a broader move towards the digital engagement of customers at retail outlets to integrate transaction-processing (whether payment at a retail store or deposit-posting at a bank branch) into a digital channel that improves the customer experience and supports data-enabled marketing and messaging.
  • We believe the large chain-banks (so BAC, WFC, and JPM) will be increasingly advantaged with the shift to digitization of in-branch processes and data-enabled customer management because of increasing returns to scale (both processing and information) and to skill:
  • Processing Scale: The payoff for making the fixed-cost investment in digital platforms is lower variable costs-to-serve; JPM, for example, reports that the variable cost of a remote deposit (where a user scans a check, often using a smartphone, and transmits the scanned image or ACH-data to a bank for posting) is 3 cents vs. 60 cents for a teller-assisted transaction and BAC that ~10% of deposits are now remote (vs. 6% in the year-ago quarter and zero in 2012Q1). This swapping of fixed-for-variable costs lifts the return to scale.
  • Information Scale: Large chain-banks can cast a broader net to capture changes and nuances in customer behavior as they emerge (in one region ahead of another, for example) and have larger sample sizes that can increase statistical reliability of data-driven inferences.
  • Research Scale (or return-to-skill): Furthermore, the large chain banks can invest more in test-and-learn research because the findings can be deployed across large customer bases.
  • Given these structural advantages will enable more precise deposit re-pricing, we expect the chain-banks to show margin outperformance relative to the industry, and note that this view is endorsed by specialists with access to in-house data. In a December 2011 study[3], for example, consultant PwC concluded that “the success of the largest banks in maintaining low volatility, low-cost deposits indicates that national banking capabilities (leading products and branch coverage) are most advantageous”.
  • Furthermore, there is already evidence that the national branch networks of the chain-banks confer economic advantage. For example, WFC, BAC, and JPM have a disproportionate share of deposits (with an “impact” ratio of deposit-share to branch-share of over 2x vs. an industry-average, adjusted to exclude the “big-3” of 0.6x – see Exhibit) and hence advantaged leverage of fixed-costs.
  • We use “core” deposits in the analysis (i.e. excluding brokered deposits and time deposits over $250k), and note that the ratio of core-to-total deposits is higher (at 96%, 92%, and 98% for WFC, BAC, and JPM respectively) than the adjusted industry average of 87%.
  • Despite the use of core deposits, there may be a residual distortion from corporate franchises so that, for example, the 5.5x ratio for C is likely not a fair measure of branch efficacy. We exclude C from the analysis along with banks, such as COF and DFS, where branch footprint is not representative of the deposit franchise because of large online-only platforms.
  • Finally, we expect a wave of innovation in deposit products from the large chain-banks as they implement test-and-learn strategies around deposit pricing. As described by Elting Morison in “Gunfire at Sea: A Case Study in Innovation[4]”, this is a frequent consequence of platform innovations improving process-control and feedback, was a feature of the credit card industry in the mid-to-late 1990s (where COF’s information-based strategy led to innovations such as teaser-rates and zero-balance transfers), and is nascent in the deposit business.
  • For example, Russia’s largest private bank, Alfa Bank, has introduced a “healthy is the new wealthy” savings account[5] which ties the rate paid to a customer’s physical activity (tracked through apps such as RunKeeper, Jawbon, and Fitbit). This, in turn, drives increased mobile and online usage (and hence customer engagement with the Alfa brand and data) because customers are curious to see the impact of their latest work-out on their balances and rates.

Exhibit: Impact Ratio across US Banks using Core Deposits (see footnote for definition)

Source: SNL Branch Analytics Tool. Core deposits exclude brokered deposits and time deposits over $250k. Ex-Big 3 excludes BAC, WFC, and JPM. Analysis excludes C because of likely distortion from its corporate franchise and banks, such as COF and DFS, whose branch footprint is not representative of their deposit franchise because of large online-only platforms.

Bank Deposit Franchises are Massively Under-Earning

In 2013, the deposit franchises of US commercial banks contributed 10 basis points to aggregate net interest margins vs. a decade-average of 70 basis points (see Exhibit 1). Two reasons for this underperformance are:

  • First, at an average cost of 40 basis points, interest-bearing deposits provided no funding-cost advantage over wholesale funds in the interbank market vs. a decade-average benefit of near 40 basis points (using 6-month Libor as the wholesale funds benchmark – see Exhibit 2).
  • Second, prevailing low interest rates reduce the value of non-interest bearing deposits which contributed 10 basis points to bank net interest margins vs. a decade average of 30 basis points.

Exhibit 1: SSR Net Interest Margin Model for US Commercial Banks

Source: FDIC, SNL, SSR Analysis

Exhibit 2: Basis-Point Difference between 6-month Libor and Rate-Paid on Interest-Bearing Deposits

Source: FDIC, SNL, SSR Analysis

We expect the returns from deposit franchises to improve dramatically in the next up-rate cycle notwithstanding the customary shift by consumers to high-yielding certificates of deposit and the growing importance of price-discovery services such as bankrate.com and online-only platforms, such as ING purchased by COF in Feb 2012, (which, between them, now account for 6% of deposits); the former have increased consumer awareness of relative deposit rates and the latter empowered them to “rate-surf” by making it easier to shift savings balances between banks. Indeed, we believe the key driver of bank earnings over the next 3 years will be the rate-sensitivity of deposits.

In the last up-rate cycle, from 2004Q2 to 2006Q3, deposits re-priced 50% of the increase in short rates of ~400bps (as Fed funds moved from 1% to 5.25% – see Exhibit 3). If we apply this re-pricing “beta” to the 1.3% increase in 6-month Libor or 6ML indicated by the forward markets for mid-2015 (with 6ML at 1.6% vs. 0.3% today) and $10tn of deposits at US commercial banks, it adds $45bn to after-tax earnings or one-third of the 2013 base of $144bn; if we apply it to the 2.8% increase in 6ML indicated for mid-2017 (with 6ML at 3.1%), it adds $100bn to after-tax earnings.

Beyond the impact on aggregate bank earnings, margin shifts will create winners and losers among banks. In the last up-rate cycle, the top-performing banks (first quartile among the largest-50 banks in the US) had a re-pricing beta of 33% while the bottom-quartile had a re-pricing beta of 78%. Differences of this magnitude will have a profound impact on reported earnings and the unit economics of branches, and our thesis is that the large chain-banks (WFC, JPM, and BAC) will meaningfully outperform the industry because of the impact of national branch networks and the ability to employ sophisticated customer-level pricing strategies.

Exhibit 3: Differences in Deposit Re-pricing Among Top-50 US Banks from 2004-2006

Source: Novantas Inc[6]

Importance of Branch Networks

Investors are focused on the generational shift from branches to online and mobile channels (see Exhibit 4), but it is important not to attribute the current lows in branch profitability, which are largely associated with the rate environment, to these secular trends. In particular, calls for sweeping branch cuts are likely be an over-response to cyclical downturn, typical of decision-making in capacity-intensive industries, generating not only a false economy but also, given high start-up costs and lags (particularly in branch markets where suitable real-estate is limited), a competitive blunder particularly as rising rates will create a more difficult operating environment for deposit franchises.

Exhibit 4: The Consumer Shift from Branch- to MobileBanking

Source: Capital One Investor Presentation

The strategic value of a branch network is in creating pricing power – that is the ability to pay below-market rates on deposits – and this value will increase as rising rates lift deposit outflow risk. Indeed, the key success factor for banks as rates normalize will be the ability to attract rate-sensitive deposits without re-pricing the entire book; winners in managing this trade-off between balance formation and margin enhancement will have structural advantages, including branch networks, and effective pricing models that are able to create different and dynamic rate-offers by geography, product-market, and individual customer.

Quantifying the benefit of branches on deposit pricing, and particularly through the reduce rate-sensitivity of deposits as rates rise, is not easy. However, the evidence for the benefit from consultants is strong. Frank Rohde, CEO of Nomis Solutions which specializes in advising banks on deposit pricing strategies and includes in its staff a number of executives who have run large deposit franchises, is unequivocal: “The customer who likes to visit a branch is less price sensitive than a customer whose relationship is online-only. Banks haven’t fully understood that if I have a branch in this neighborhood, it reduces price sensitivity[7].” Furthermore, in a December 2011 study, “When Cash Isn’t King: Driving Deposit Value in a World of Excess Liquidity[8]”, consultant PwC concluded that banks with large branch networks are particularly advantaged: “the success of the largest banks in maintaining low volatility, low-cost deposits indicates that national banking capabilities (leading products and branch coverage) are most advantageous”.

We can demonstrate empirically that the branch networks of some large banks are more effective than those of small banks through examining the impact ratio – that is the ratio of deposit share to branch share. This is an important ratio since higher deposits-per-branch creates advantaged fixed-cost leverage and hence raises pricing power and/or profitability. Impact ratios vary meaningfully across banks (see Exhibit 5) and suggest a particular advantage for the national networks of WFC, BAC, and JPM, and some advantage for the large regional players. We exclude from the analysis banks such as COF and DFS whose large online platforms mean that their branch footprints are not representative of their deposit franchises; we have also excluded C (with an impact ratio of 5.5x) because of likely distortion from its corporate franchise and particularly the large transaction-processing business.

Exhibit 5: Impact Ratio across US Banks using Core Deposits (see footnote for definition)

Source: SNL Branch Analytics Tool. Core deposits exclude brokered deposits and time deposits over $250k. Ex-Big 3 excludes BAC, WFC, and JPM. Analysis excludes C because of likely distortion from its corporate franchise and banks, such as COF and DFS, whose branch footprint is not representative of their deposit franchise because of large online-only platforms.

Effect of Customer-Level Pricing Strategies

As the rate environment normalizes and the risk of deposit outflow increases, banks are going to need to attract balances from rate-sensitive customers without re-pricing the existing book of business; this trade-off between balance-formation and margin-enhancement will require more sophisticated and customer-level pricing strategies. This is conceptually no different from the practice, in the last up-rate environment, whereby branch staff would use “desk drawer” or “back pocket” rate offers to retain high-value customers threatening attrition. However, implementation in the next up-rate cycle will be centralized and based on data-enabled quantification of attrition risk.

The key success factor in data-enabled deposit pricing and marketing will be information-scale: the ability to capture and analyze large amounts of data so as to improve the reliability of statistical inferences about the behavior patterns of customers down to finely-defined segments and individuals. Of course, a similar data-enabled strategy transformed the credit card industry from the mid-1990s as information-players such as COF were able to dynamically-adjust pricing and line-sizes at the individual customer level as behavior patterns indicated changing default risk; the CARD Act of 2009 reduced the efficacy of the approach by limiting the ability to re-price outstanding balances, but we believe the card industry provides a precedent for the likely evolution of the deposit business.

The advantage of branch networks in these information-based strategies is illustrated by the desk-drawer offers of the last up-rate cycle; branch officers acted on the proprietary information generated from a branch interaction to adjust pricing. While customers are migrating to other channels, there is still significant branch interaction among large swathes of the customer base and advantaged deposit-pricing for these customers, based on the insight gained from branch interaction, will afford branch-banks more pricing flexibility among less branch-engaged customers where there is less of an edge versus online-only platforms.

In other words, our view of digital banking is that it is not just about shifting customers to digital channels (and arguing that COF is ahead because two-thirds of its customers are online or mobile vs. just under one half at, say, BAC[9]) but also about digitizing the customer communication at point-of-interaction whether that occurs online, mobile, or in-branch. To this extent, our view is more aligned with that of WFC which frames customer “intensity” around depth (how many transactions a customer executes) and breadth (how many channels, including the branch channel, a customer uses). Indeed, we believe data-capture, to move the “desk-drawer” approach to data-enabled deposit pricing, is an important and under-appreciated aspect of investment by the large chain-banks to equip branches with self-service facilities such as touchscreen ATM’s and tablet-enabled tellers.

Of course, above and beyond their contributions to data-strategy, the opportunity of these technologies to reduce teller costs, allow for smaller branches, and improve customer convenience are also important. The combination of these factors drives competitive advantage to large chain banks for three reasons – return-to-scale, from both processing and information standpoints, and return-to-skill:

  • Processing Scale Economies: The platforms supporting digital customer interactions are fixed-cost and tend to reduce the variable costs to process transactions; the swapping of variable for fixed costs increases the return-to-scale.
  • Information Scale Economies: Large chain-banks can cast a broader net to capture changes and nuances in customer behavior as they emerge (in one region ahead of another, for example) and have larger sample sizes that increase statistical confidence in data-driven inferences.
  • Skill Economies: Furthermore, the large chain banks can invest more in test-and-learn research because the findings can be deployed across large customer bases.

Expect a Wave of Innovation in the Deposit Business

We expect return-to-skill to be a defining feature of the deposit industry over the next decade because the more consistent capture of data, and hence feedback to changes in deposit pricing and product design, will drive a wave of innovation. As described by Elting Morison in “Gunfire at Sea: A Case Study in Innovation”, this is a generic consequence of platform innovations improving process-control and feedback[10], was a feature of the credit card industry in the mid-to-late 1990s (where COF’s information-based strategy led to innovations such as teaser-rates and zero-balance transfers), and is nascent in the deposit business.

For example, Russia’s largest private bank, Alfa Bank, has introduced a “healthy is the new wealthy” savings account[11] which ties the rate paid to a customer’s physical activity (tracked through apps such as RunKeeper, Jawbon, and Fitbit). This, in turn, drives increased mobile and online usage (and hence customer engagement with the Alfa brand and data) because customers are curious to see the impact of their latest work-out on their balances and rates.

  1. http://www.forbes.com/sites/tomgroenfeldt/2013/12/17/rising-interest-rates-could-wreak-havoc-on-bank-balance-sheets/
  2. At JPM, for example, over 55% of commercial customers use a branch each quarter. And, at WFC, 85% of new products are sold through branches even while 85% of routine transactions are conducted over digital channels.
  3. http://www.pwc.com/us/en/financial-services/publications/viewpoints/assets/fs-viewpoint-driving-deposit-value-in-excess-liquidity.pdf
  4. http://cs.gmu.edu/cne/pjd/TT/Sims/Sims.pdf
  5. http://csbcorrespondent.com/blog/banks-deposit-innovation-should-spur-your-banks-creativity
  6. http://novantas.com/deposit-strategy-preparing-for-rising-rates/
  7. http://www.forbes.com/sites/tomgroenfeldt/2013/12/17/rising-interest-rates-could-wreak-havoc-on-bank-balance-sheets/
  8. http://www.pwc.com/us/en/financial-services/publications/viewpoints/assets/fs-viewpoint-driving-deposit-value-in-excess-liquidity.pdf
  9. http://www.marketratesinsight.com/Blog/post/2013/12/20/Capital-One-Taking-the-Lead-in-the-Race-for-Digital-Banking.aspx
  10. http://cs.gmu.edu/cne/pjd/TT/Sims/Sims.pdf
  11. http://csbcorrespondent.com/blog/banks-deposit-innovation-should-spur-your-banks-creativity
Print Friendly