Core Depostit Retention: Transparency and Accountability via Flow of Funds

Portfolio growth is the single most important revenue driver in our industry, yet achieving transparency of performance metrics and accountability for portfolio management has always been elusive. We believe that a sea-change is about to occur, as banks adopt flow of funds metrics as a basis for understanding

We will first review the strengths and weaknesses of the common approaches to measuring deposit retention to provide a backdrop and perspective for our case. You might want to use this review to assess how comfortable you are with your bank's present approach. We will then discuss using flow of funds as retention metrics, and show how this can provide you with entirely new insights into the performance of your branches, products and customer segments.

Core Deposit Retention

Portfolio growth is a primarily a function of acquiring and retaining business. We spend millions managing the "front door" sales and marketing efforts within our banks, but always seem to pay less attention to the "back door" where business slips quietly out of our book.

This isn't because we don't think it matters. On the contrary the industry average defection rate for deposits is believed to be around 15% in the USA every year. Deposit defection typically erodes more than 90% of the good that results from all our sales and marketing efforts every year. So why don't we do more about it?

The problem, we assert, is not that we have been unaware of the business issue but that we have been unable to get a solid handle on measuring and understanding it. What we don't measure we don't manage, and this is as true in banking as any industry. The bottom line is that we have a big transparency and accountability weakness in our basic performance stats.

So why is measuring deposit retention so difficult? The problem is rooted in the fact that deposit account products are to some extent substitutes for one another. Money can be transformed from one product to another relatively freely by any customer, and when they do "switch" money around, our banking systems have no way of tracking these flows. Let's look more closely at what we do in many banks today.

Metric 1: Net Portfolio Change

That we measure portfolio change is pretty obvious but nonetheless important. When we report to our shareholders there is inevitably discussion in the MD&A about how this product went up and that product went down over the course of the quarter / month or whatever. Everyone does this and that's fine, it does give some indication of whether we are winning or losing overall.

Unfortunately this approach to portfolio analysis reveals nothing of cause, symptom or effect of management action. If we want to achieve accountability for portfolio growth we need the flows in and out of each portfolio to be transparent, and to do that we need to better metrics than net portfolio change.

Metric 2: Lost Households

With the advent of customer databases and MCIF systems, many banks have adopted customer-centric metrics, looking at lost households as a basis for monitoring and managing retention performance. While it is true that lost households really do represent lost business to your bank and a serious form of it, lost households account for only a small fraction of lost business volume - about 15% of core deposit balance defection.

The lost household metric fails to take into account what is going on with ongoing customers - the vast majority who remain with the bank from period to period. These ongoing customers close accounts and reduce balances in ongoing accounts, accounting for 85% of deposit defection. Household analysis cannot explain the change we see in our portfolios and cannot be reconciled to your financial reports because they are incomplete. To establish verifiable reporting and accountability for retention we simply can't ignore these large flows.

Another major shortcoming of the household analysis is that it does not provide any insight into the specific products or branches that contributed to the lost business. The information is too summarized to be useful for measuring performance and accountability of branch and product management. The bottom line is that analysis of customers won and lost cannot give us the information we need.

Metric 3: Lost Accounts

Counting new and lost accounts and the balances related to them is one of the more common methods of measuring sales and retention performance. It is useful as an indicator of activity levels associated with new and lost business, but falls seriously short of explaining what is really going on.

We know that lost accounts are not a perfect barometer of lost business because money may leave accounts over time, so there will often be a drain on balances before an account actually hits zero or closes. We also know that we can lose part of an account without losing all of the business.

Our research1 has shown that only 35% of deposit balance losses occur in lost accounts (of which less than half were accounts of lost households). The majority - fully 65% of all deposit balance losses happen in accounts that continue to be open, with some balance. Clearly getting a handle on these "partial defections" is imperative to have a proper understanding of retention rates. Looking at lost accounts cannot explain the change we see in our portfolios, or be reconciled to your financial reports because the information is incomplete. New and lost account analysis simply cannot give us the information we need.

Metric 4: Transaction Analysis

Another widely used portfolio flow analysis method involves classifying transactions into categories such as deposit, withdrawal, renewal, redemption, interest credit, fee debit and the like. This analysis does produce a holistic view of the flows within a product portfolio, "explaining" how the portfolio rolled forward from one period to the next.

Unfortunately this approach to portfolio analysis overlooks the flows that occur between products. When a customer shifts money between accounts the transaction analysis identifies the flows as lost business from one account and a sale in the other.

Does this really matter? Indeed it does: our research shows that over 1/3 of the value of account outflows actually flow into other accounts of the same customers. The implication is that the transaction analysis method can overstate sales and lost business by more than 30%. If we want to achieve accountability we need transparent and verifiable information... this method fails the test.

Comment

We can see that each of the popular methods of measuring retention and it's complement, lost business, have significant drawbacks in terms of being sufficiently verifiable, transparent or actionable to be used as a basis for accountability and performance measurement. If your bank is presently using one or more of the methods outlined above, you are in good company. This is what most banks are doing today.

On the other hand, should we be settling for the status quo? Obviously this depends on how comfortable we are with not knowing what is really going on. Our sense is that many banks have become complacent with their portfolio metrics because they have always seemed "good enough". We suggest that these metrics appear to be good enough only because internal movements of money do not change the overall deposits of the bank. We believe this point of view overlooks some persuasive arguments for better information:

1. If sales are overstated by 30% so is performance based sales compensation

2. Marketing and retention interventions are aimed at the wrong customers a third of the time

3. Product performance measures are distorted by product cannibalization and substitutions

4. Branch performance is distorted by internal flows of business that are simply relocating within the bank.

We could go on and on with reasons to justify better information, but let's stop there on the assumption you agree that these are sufficient. The question now turns to what we can do to fix this problem.

Flow of Funds

We need a better way to look at deposit retention. What we require is a flow of funds at the account level that accounts for different types of flows that happen within each individual customer's portfolio. We need to be able to tell if an account balance has declined because it transferred to a different deposit product or rolled over into a new account of the same product. We should also be able to tell if deposit proceeds were used to pay off loans, or were actually taken out of the bank.

A well designed flow of funds approach to measuring and analyzing portfolio change will enable you to precisely identify and measure:

1. Internal versus external flows (i.e. outside the bank)

2. Flows between locations

3. Flows between products

4. Flows between accounts of the same product

When these flows are measured properly they will "explain" portfolio change totally and reconcile 100% to your overall portfolio by branch, product and customer.

This can be done today, using information that already exists in the MCIF files of most banks. Shouldn't you be considering the benefits of managing business flows at your bank?