The Art of Marginal Cost Analysis and Assessing Your True Cost of Funds
Part One: Marginal Cost 101 and Offensive Funding Strategies
“We got to get better on defense, we got to get better on offense, we got to get better everywhere. Simple as that.”
~Steve Spurrier, Head Football Coach, University of South Carolina
At some point in our careers, we’ve all experienced the phenomenon of marginal cost,
whether we’ve known it or not. For some, it was first introduced as a foreign and
nebulous concept in Microeconomics 101—not necessarily something that we thought
might be applicable in our everyday lives! These days, however, the development
of a marginal cost mentality has a direct impact on our profitability as bankers over both the
short and long terms.
Let’s consider our strategy in terms of the offensive and defensive sides of the playing field. After winning the coin toss, let’s bring on the offense.
There are instances in which marginal cost analysis can not only help you avoid
the damage sustained by unnecessarily paying higher rates, but also bring in those
highly coveted new accounts. Many Seattle Bank members report the same frustration:
many of the customers that they perceive as being “highly valued” accounts do not view
the institution as a highly valued business partner. This means that many highly
valued customers have multiple accounts with other financial services firms. This
spells opportunity for you in the form of potential new funding sources—from your
existing customers!
Back to the playing field… Let’s put marginal cost to the test and consider the
implementation of a new, highly segmented, non-maturity deposit account. Remember
that it’s best to increase your deposit customer segments as rates rise, and collapse
them when rates decline. As such, today’s low-yield environment bodes well for introducing
a premium, non-maturity deposit program aimed squarely at money-market mutual
funds. Why? Many of your existing customers have uninsured money market funds. These
funds typically have embedded management fees, perhaps in the range of 50 basis
points. In a low-rate environment, government money market funds are yielding approximately
2.00%. Once you’ve done the math (0.50/2.00), you can see that, in a low-rate environment,
a high proportion of net yield to your customer (yes, the same customer that has
his/her money with other institutions) is lost in the form of management fees.
Want another competitive play to deploy? Of late, some of your customers may have been attracted to “non-money” market funds that were, up until several months ago, advertised as funds that “provide higher yields on your cash with only marginally higher risk [and therefore] could be a smart alternative.” Many of these funds were not money market funds at all, but actually “ultra-short” funds that had invested in short tranches of collateralized debt obligations. Some of these funds have lost in excess of 20% of their respective net asset values so far this year. Chances are that you have a few customers that have sullied relationships with other financial institutions!
Where there is chaos, there is opportunity. So, we’ll march down the field and plot
a funds-driven strategy and introduce a premium non-maturity deposit program aimed
at the money market funds (the real money market funds, as well the masqueraders)!
You might want to consider targeting such an account at your higher-balance customers
with brokerage or money market fund balances. Perhaps you could offer the account
at a spread of 50 basis points (or so) over a brokerage money market fund index.
We’ll look at this offensive strategy in the context of its true cost—its marginal cost.
On the defensive side of the field... Many financial institutions have faced these sorts of situations before. You have had what you thought was a stable core of savings account balances at 1.0%. Then, you sense that the sharks are circling: rates are going up and you’re nervously watching to see how your competitors might respond. You’re hopeful that your deposit market will not react to rising rates immediately, and certainly not to a scenario in which deposit rates increase in lockstep with each rising basis point in the capital markets. Why? The marginal cost of raising rates for existing depositors would be exorbitant. You never want to be in a position in which you are unnecessarily paying up for deposits that were never in jeopardy of leaving your institution in the first place!
Here, you need to play some serious defense. You’ll want to raise your rates on
the savings account by the least amount possible in order to retain the most balances
that you can. You don’t want to pay up unnecessarily for existing accounts that,
at the end of the day, were more attracted to your non-price variables such as
your superior branch locations, outstanding customer service, and brand. It’s not
rates alone that keep customers at your bank; otherwise, why would you invest in
good real estate, exceptional employees, a great marketing program?! Your best way
to hold the line would be to reinforce the concept throughout your organization
that you will offer the best rate (singular), and not the best rates (plural).
We’ll revisit this situation and the defensive aspects of marginal cost in Part Two of this acticle in the October 2008 edition of What Counts.
Marginal Cost 101
Before applying the marginal-cost-of-funds tool to present-day conundrums, let’s
review a textbook example of marginal cost. (For some readers, that may mean going
back to the world of “Star Wars,” “Saturday Night Fever,” and scraggly sideburns.
Still, we survived!) Think of marginal cost analysis simply as determining the cost
of producing a unit—be it a widget or a deposit. Then we’ll talk about that
unit’s cost—whether from a home-grown deposit or wholesale sources—in terms of its
impact on your institution’s funding costs.
Figure 1 details a textbook example of the cost of producing an increasing quantity of a hypothetical good or service. We’ll assume that we are producing widgets—the long-standing example from Econ 101! (After so many years of working with that mythical commodity, interestingly, the word “widget” was recently adopted as an official noun and became, of all things, a third-party item that can be embedded in a Web page.) Here, we are producing up to 10 widgets, and you can see that even without producing a single widget, we still have fixed costs. These fixed costs are prerequisites for producing anything (e.g., factory costs for producing a widget or bank branch costs for producing a deposit).
In our example, if we produce one unit, the marginal or incremental cost of producing that one unit is $8.00. Why? The variable cost (call that raw materials or perhaps labor) is $8.00. Fixed costs are not included at the margin. As more units are produced, economies of scale kick in, the operation becomes more productive, and variable costs drop. After a certain point, the law of diminishing returns would set in, and variable costs might go up, as we may need to hire additional workers, etc.
Pay close attention to what happens to average fixed costs. As production increases,
the average fixed cost per unit of output drops. You can also appreciate the meaning
of marginal cost via the incremental changes in the total cost of a good or service. (See the last column of
Figure 1.)
Figure 1. Hypothetical Summary of Production Costs of a Good or Service
|
Unit Output | Marginal Cost | Average Fixed Cost | Average Variable Cost | Average Total Cost | Total Cost
|
|---|
|
0 | - | 36.00 | - | 36.00 | 36.00 |
|
1 | 8.00 | 36.00 | 8.00 | 44.00 | 44.00 |
|
2 | 4.00 | 18.00 | 6.00 | 24.00 | 48.00 |
|
3 | 3.00 | 12.00 | 5.00 | 17.00 | 51.00 |
|
4 | 5.00 | 9.00 | 5.00 | 14.00 | 56.00 |
|
5 | 7.00 | 7.20 | 5.40 | 12.60 | 63.00 |
|
6 | 9.00 | 6.00 | 6.00 | 12.00 | 72.00 |
|
7 | 10.00 | 5.14 | 5.47 | 11.71 | 81.97 |
|
8 | 19.00 | 4.50 | 8.13 | 12.63 | 101.04 |
|
9 | 25.00 | 4.00 | 10.00 | 14.00 | 126.00 |
|
10 | 40.00 | 3.60 | 13.00 | 16.60 | 166.00 |
Figure 2 gets to the heart of marginal cost in terms of the corporate decision-making
process. A magical event occurs at the intersection of Marginal Cost and Average
Total Cost—the point at which you have the lowest cost of production. In
this case, does this mean that you would stop your production at seven units? Not
necessarily, because that amount of production would represent the lowest cost per
unit with the current production structure. If you added some additional
infrastructure, a higher point of lowest cost production might be possible.
Not-so-good things happen with the intersection of the Marginal Cost and the Average Variable Cost curve. This is your shutdown point of production. Of late, many industries (e.g., airlines and automotive) have necessarily cut capacity. These are industries that have high fixed costs to begin with, and logarithmically increasing variable costs that are associated with the rising cost of fuels, labor, and raw materials. In our Figure 2 example, the shutdown point would appear to be at four units of production.
Figure 2. Hypothetical Impact of Production Unit Quantity vs. Production Costs
Now that we’ve passed Marginal Cost 101, we’ll revert back to our offensive and defensive cases for practical application of what we’ve learned in terms of how we fund our financial institution.
Offensive Strategy: Introducing a New High-Yield Checking Account
Earlier, we mentioned that there’s reason to believe that some of your best customers
are investing in other financial institutions—and that they may not see your institution
as their top bank. (Perceptions can sometimes be asymmetrical, especially if you
spend too much time looking in the mirror!) So, let’s take marginal cost analysis
to task and analyze the impact on the cost of funding a new high-yield checking
account that is aimed squarely at brokerage money market accounts (and their unreasonable
facsimiles). Watch what happens when you benchmark your pricing of this new product
against your competitors, versus benchmarking them against your alternative source
of funds (e.g., Seattle Bank advances). Institutions that have successfully implemented
this strategy regularly use a benchmark curve that represents real-life alternative
sources of funds.
In order to properly assess the cost of this new funding product, you need to appreciate that the type of deposit account influences the degree of rate sensitivity. High-tiered MMDAs have historically been quite sensitive to changes in interest rates, unlike lower-balance DDAs, which characteristically have lower rate sensitivity. Also, higher-tier MMDAs tend to have shorter average lives and durations. All of these variables, however, are subject to the unique characteristics of each local market. It’s best if we fine-tune our rate sensitivity assumptions with good data tracking provided either by outside consultants or good internal record-keeping over multiple economic cycles.
Figure 3. Premium Non-Maturity Deposit Program
|
Objective: | Attract new funds at the lowest possible marginal cost |
|
Market: | Affluent, high-balance customers with brokerage balances at other institutions |
|
Target: | Rate-sensitive deposits |
|
Features: | Complimentary direct deposit, bill pay, wire transfer to brokerages and other locations of customer funds |
|
Action Step: | Benchmarked vs. wholesale borrowing levels |
Remembering that marginal cost analysis incorporates measuring the cost of your
entire current deposit base against a proposed deposit base, you’ll
have to consider the direct cost of raising new funds, as well as the indirect
cost of cannibalizing existing funds. This is effectively a two-part analysis: (1) the interest
expense of initiating the new deposit program, and (2) the expense associated with
the amount of existing funds on deposit with your institution in lower yielding
accounts that would be attracted to the new product. The latter component of marginal
cost frequently derails a decision to launch a new product for fear of unnecessarily
arousing the attention of lower-yielding depositors.
Figure 4, 5, and 6 depict hypothetical current and proposed deposit structures, assuming that a new $5-million premium DDA account is introduced and that 50% of the program will be derived from new funds and the remaining 50% will be cannibalized from existing savings and DDA accounts.
Figure 4 assumes that our hypothetical institution currently supports a deposit base composed of 51% savings, 28% DDAs, and 21% CDs and that the weighted average life of this deposit base is 2.73 years.
Figure 4. Hypothetical Current Deposit Base
|
Current Deposit Funding Structure | Rate | Amount($000) | % Total | % Retained | Expense($000) | Assumed Average Life
|
|---|
|
Savings | 1.70 | 37,000 | 51.4 | 100 | 629.0 | 2.0 |
|
DDA | .50 | 20,000 | 27.8 | 100 | 100.0 | 5.0 |
|
CD | 3.50 | 15,000 | 20.8 | 100 | 525.0 | 1.5 |
|
Total Balances/ Expense | | 72,000 | 100.0 | 100 | 1,254.0 | 2.73 |
Figure 5 depicts the sources of funding that will be applied to the new high-tier DDA product that will be priced at 2.50% (that’s roughly 50 basis points higher than government money market account yields) and assumes that the total amount of funds going into the product will be $5 million. We will assume, just for the moment, that 50% of the funds come from new sources and that the remaining $2.5 million will be sourced equally from existing DDA and savings accounts.
Figure 5. Hypothetical Sources of Funds for New High-Tier DDA Product
Assuming 50% of Funds from New, Non-Cannibalized Sources
|
Source of New DDA Balances | Amount | Percentage |
|---|
|
New Funds | $2,500,000 | 50% |
|
DDA | $1,250,000 | 25% |
|
Savings | $1,250,000 | 25% |
Figure 6 completes the “before and after” picture. Total deposits have grown by $2.5 million and represent the incremental change in balances outstanding. Also, the average life of the deposit portfolio has dropped from 2.73 years to 2.66 years, as the funding mix has shifted to shorter-term sources of funds.
Figure 6. Hypothetical Proposed Deposit Base
|
Current Deposit Funding Structure | Rate | Amount($000) | % Total | % Retained | Expense($000) | Assumed Average Life
|
|---|
|
Savings | 1.70 | 37,000 | 51.4 | 100 | 629.0 | 2.0 |
|
DDA | .50 | 20,000 | 27.8 | 100 | 100.0 | 5.0 |
|
CD | 3.50 | 15,000 | 20.8 | 100 | 525.0 | 1.5 |
|
New High Balance DDA | 2.50 | 5,000 | 6.7 | new | 125.00 | 2.0 |
|
Total Balances/ Expense | | 74,500 | 100.0 | 100 | 1,351.5 | 2.66 |
With this information, it’s now possible to consider the cost of this forecast incremental growth in funding. Remember there are two components of marginal cost: (1) the interest expense on the new account; and (2) the opportunity cost of having to pay a higher-than-original rate on existing balances. Marginal cost is simply the difference in interest expense before and after the inclusion of the new account promotion, divided by incremental change in funding balances.
As the formula below demonstrates, you come out ahead if you’re using new money to grow your funding base, rather than old money.
Calculation of Marginal Cost Assuming 50% of Funds from New, Non-Cannibalized Sources:
Marginal Cost = (Difference in Current vs. Proposed Interest Expense) / (Difference in Current vs. Proposed Funding Balances); or: (1,351.50-1,254.00) / (2,500) = 3.90%
Now that we’re armed with the marginal cost of the promotional strategy, how do we make the managerial decision to proceed or not proceed? The answer lies in selecting a benchmark rate that represents an alternative source of procuring the incremental funds. In this case, we would select the two-year Seattle Bank advance rate. Why two years? The maturity of the advance correlates with the assumed average life of the new deposit account. Again, the average life assignment of a particular deposit product can be a localized phenomenon, varying from one market to another.
With a two-year Seattle Bank bullet advance rate of 3.12% and an assumed cannibalization rate of 50%, the promotion’s marginal cost of 3.90% would prompt management to reconsider the promotion within the existing parameters. The question would then become: “At what cannibalization rate would it make sense to proceed with the new deposit promotion?” Aside from the trial-and-error method, there are simple marginal-cost-of-funds models that may be applied to solving this conundrum. With the Seattle Bank’s Funding Desk just a phone call away, our staff can walk you through various assumptions and help you model and make the “go/no-go” decision.
It turns out, as illustrated in Figure 7, that in order to reach the break-even cost of 3.90%, two things must happen with the deposit promotion: (1) it must be accepted in the market at a lower rate (the marginal cost of the promotion rate at 2.50% may be too high relative to the wholesale funding benchmark), and/or (2) it must rely less on sourcing from existing depositors (a cannibalization rate of 50% is too high).
Figure 7. Comparison of Marginal Cost of Adding a $5-million High-Tier DDA Promotion at 2.50%, 50% Assumed Cannibalization, and a Two-Year Advance
Promotional Summary Details
|
Deposit Promotion: | 2.50% |
|
Net New Funds: | 50% of Promotion |
|
Repriced Funds: | 1.10% (existing deposits repriced up 140 basis points) |
|
Two-year Seattle Bank Advance: | 3.12% |
|
Marginal Cost: | 3.90% |
Retail Strategy – Marginal Cost: 3.90%
|
Balance | Increase | Cost
|
|---|
|
$2.5 million (existing accounts) | 1.40% | $35,000 |
|
$2.5 million (new money) | 2.50% | $62,500 |
|
Cost of Raising Rates | | $97,500 |
Seattle Bank Advance Strategy – Marginal Cost: 3.12%
|
Balance | Increase | Cost
|
|---|
|
$2.5 million (net liability increase) | 3.12% | $78,000 |
|
Cost of Seattle Bank Advance | | $97,500 |
Break-Even Result:
-
Must raise 69% of funds in the form of new money at 2.50%.
- If 50% of funds are in the form of new money, the promotion rate cannot exceed 2.11%
After modeling a quick break-even analysis and seeing how we might get to the break-even rate of 3.12%, we can determine that if we keep the rate on the new product at 2.50%, we would need to source at least 69% of the $5 million in proceeds from new money. That translates into a cannibalization rate of no more than 31%. Alternatively, if we elect to stay firm on our original cannibalization rate assumption of 50%, we would need to offer the new deposit product at a rate of 2.11% (or, 39 basis points below our originally intended level). It would remain to be seen if the market would let us get away with that level.
With this information in hand, we’re now prepared to see how we might reduce our cannibalization rate. Rather than wishing the cannibalization rate away, we could consider a geographic segmentation strategy, wherein we would target the promotion in new geographies that are not populated by existing depositors. Alternatively, if we believe that we can’t do better than a cannibalization rate of 50%, we may need to find markets in which the product would be accepted at the lower rate of 2.11%. That rate is lower than the 2.50% yield that we targeted, but, to the customer, it’s still a better net yield than many government money market funds.
In the next issue of What Counts, we’ll outline a defensive marginal cost strategy that you can use in the face of potentially rising rates. Remember that it’s all about producing the optimal amount of production units (i.e., funding for your institution) when your marginal costs intersect with your total costs. How you turned theory into practice just might be the hit topic of your 30-year college reunion!
by John P. Biestman CFA, Vice President/Director of Business Development