Setting ROE Targets – What NOT to Do

March 16, 2017 Joel Rosenberg

One of the most important decisions your bank will make when using a pricing platform like PrecisionLender (PL), is the determination of the various loan products’ target Return on Equity (ROE). ROE is what your institution would like to earn on that next loan as a percentage of the capital committed.

Establishing these targets, since they can vary by loan product, can determine the growth in the bank’s net income and capital, its risk profile, and the ability of its bankers to increase the loan portfolio.

Because it’s one of the most important assumptions established within PL, we spend a lot of time on this with our clients during the on-boarding process. But we’re often met with this question (or one like it):

“Our bank’s overall target ROE for this year is 10%, why not set our target for each product type equal to this goal?”

To which we emphatically respond:

“Don’t do it.”

Here’s why.

Several areas of the bank are currently not earning anywhere near the 10% goal. Setting loan targets at the institution’s overall target level ensures you will miss your bank-wide goal, because the loan portfolio generally subsidizes those lower earning areas. Instead, the loan portfolio’s target ROE, needs to be set at a higher level than your overall bank-wide target.

Different Lines of Business/Products Produce Various ROE’s

Let’s take this a step further. A bank has several lines of business (LOB), products and cost areas that account for the overall net income of the institution and to which capital can be assigned. So different ROE’s can be calculated for profit/(loss) centers such as these:

  1. Lending, including both retail and commercial;
  2. Deposits, including any brokered activity;
  3. Investment Portfolio, including taxable and tax exempt;
  4. Non-Interest Income such as trust, brokerage, insurance, mortgage banking, etc.;
  5. Internal Cash Management, typically short term investments and borrowings; and
  6. Administrative (Other).

The last item is a “catch all” and would include any other activities that cannot be directly attributed to one of the other categories. Also, most banks have excess capital (generally a good thing), which is usually parked in this category until needed.

Example of ROE Targets

Let’s look at an example of the various ROE’s by profit centers at a bank. We will call our institution, First Example Bank and although fictional, it is loosely based on an actual bank. The following is the financial statement for First Example Bank:

Here’s a breakdown:

  • The Deposits at FRB and Banks are short-term interest-bearing accounts, with an earnings rate near the current fed funds level.
  • They have a number of municipal securities, and about 60% of the total securities held are tax exempt.
  • The loans secured by 1-4 family mortgages includes HELOCs, as well as closed-end first and second lien residential mortgages. About 71% of these loans are in a first lien position.
  • The bank has a very small non-mortgage consumer loan portfolio and only a few agricultural and municipal loans.
  • The other borrowings represent short-term debt, typically borrowed on an overnight basis.

For this example, the bank expects a fiscal year net income after taxes of about $9.9 million and an ROE of about 8%. Its marginal tax rate (both federal and state) is 37.1%, but due to the tax-exempt securities and loans, as well as some timing issues, the effective tax rate is about 34.3%.

The bank would like to achieve a 10% ROE over the next several years. But as we noted earlier, we urge our clients not to set their targets at this level for all loans, for these two reasons:

  1. There are a number of existing assets and liabilities that are set at a certain earning level and until they mature, or rates can be repriced, their contribution to the bank’s ROE goal cannot be changed.
  2. The bank has a number of different profit/(loss) centers; if the targets for all of these can be set to 10% that would work well. However, in most cases these areas barely break even and often show a loss.

Let’s look at the returns for these centers.

Profit/(Loss) Center Analysis

The following analysis is based on a number of assumptions, several of which could be modified with different information. However the overall result should not be greatly affected. ROE equals net income over capital. In the PL platform, the numerator and denominator are adjusted for risk, or what is often referred to as a Risk-Adjusted Return Over Risk-Adjusted Capital (RAROC) analysis.

Let’s start with the denominator or capital. First Example Bank has $124 million of capital, which is a ratio of 12.4% to assets. You might simply use this ratio for all of the bank’s assets, but this ignores that different assets have significantly different risk profiles.

For example, cash has very little risk to the bank, while a lower rated High Volatility Commercial Real Estate loan may carry a great deal of concern. In addition, certain liabilities, such as deposits, carry operational, interest rate, and reputation risk and should be allocated some capital. The bank may also carry excess capital, which is generally good and can allow the institution to continue to increase its assets.

The following table shows the allocation of the capital. In general, we allocated 2% to Deposits, Portfolio, Internal Cash Management. All of these typically represent little risk and two percent is reasonable for a low risk asset or liability. For the loan portfolio, we allocated 7% for first lien 1-4 family mortgages and 10% for second lien and consumer loans. For the commercial and agricultural loans, we used the ratios developed for the underlying bank, based on their loans in the Relationship Awareness module of PL. We combined the non-interest income areas with the Administrative area, where we placed all unallocated capital, which amounted to about 25% of the total capital. It is not uncommon to have “excess capital,” and while from an ROE point of view it’s not great, from a safety and growth point of view, it is preferable.


Here’s a breakdown:

  • In the case of deposits, the interest income allocation uses the funding curve at the time of the analysis and assumes a weighted average three-year duration.
  • For the other borrowings, we simply assume that the interest allocation equals the interest expense.
  • The Admin center is the reverse of the deposit and borrowing allocation. It also includes a very small amount of other interest income.
  • For the loans, we use the strategic cost of funds, as shown by the Relationship Awareness.
  • In the case of Internal Cash Management, we use the short end of the funding curve and for the Investment Portfolio we use the seven-year point on the funding curve, which mirrors the duration of this portfolio.
  • On all the investment and loan interest expense allocation, we did not include capital as a possible funding source.
  • The Admin represents the offset from the amounts allocated on the assets side so the total equals the deposit and borrowing actual level. (Note: In the table that follows that the Interest Income and Expense for Admin shows some differences, which mostly represents interest rate risk mismatch.)

The loan provision is based on the Relationship Awareness information for the loan products. Non-interest income is allocated based on its associated product. Part of this represents deposit service and overdraft fees, while another part is based on the bank’s credit card portfolio. All the remainder is placed in the Admin category and represents various fee based products.

For the sake of this analysis we combined the non-interest fee center with the Admin. Allocation of expenses can be a tricky area. We simply assumed 2% of the total should each go to Internal Cash Management and Investment Portfolio. These usually don’t have a large cost basis. Based on the cost studies we publish every year, about 17-19% of expenses are so called General Overhead and cannot be specified to a particular area, see: . We allocated this amount to the Admin section. We also assumed that costs equal 80% of the non-interest income are allocated to Admin. The remainder was allocated equally to Deposits and Loans.

The following table shows our allocations and projected net income (loss) for each category. It should be noted that we used the bank’s marginal tax rate to determine net income, although we made adjustments for tax exempt securities and municipal loans.

The final table shows the ROE by Profit(Loss) Center.

Several conclusions can be developed from the above table:

  1. About 42.7% of the capital is associated with non-loan related centers. The combination of all these centers amount to only a minimal positive net income. The weighted combined ROE on these centers is basically zero. However, the Loan Center ROE is 13.78%. Therefore, lowering the target to 10% on the loan portfolio would result in the bank’s overall ROE declining to about 5.3%, from the current of almost 8%. In order to hit an overall 10% for the bank, the loans instead need to be close to 17.5% ROE.
  2. The bank has excess capital (assuming there are no significant regulatory issues). Growing loans, with their higher ROE, will be beneficial to the bank. For example, if $56.4 million of internal cash management funds were converted from deposits at the Federal Reserve to CRE loans with the same characteristics of the existing loans, assuming no other changes, the bank’s overall ROE would increase to 9.17%. These new loans would have an ROE of 14.66% compared to the negative 12.24% as an internal deposit. Almost $5 million of capital would be moved to the Loan Center from the Admin Center. However, if the new loans are targeted at 10%, the total increase in bank overall ROE would be about 1%, or approximately 20 basis points less than using the current ROE. Moving funds to commercial loans is helpful, but if the overall goal is to be at 10%, new loans need to have a target above the overall bank goal.
  3. Higher market interest rates will likely have a positive impact on the Deposit and Internal Cash Management centers, although it will likely be negative to the Investment Portfolio. However, a moderate rise in rates is not likely to meet the bank’s overall 10% target without adding commercial loans with individual targets above that of the bank’s goal.
  4. Finally, except for general overhead non-interest expense, the commercial loans were fully allocated bank costs in this analysis. At many of our clients’ banks, the amounts allocated for origination and servicing expense may not equal the fully allocated costs. If this is the case, a higher ROE is required to account for this shortfall.

Every bank will have different results and assumptions will differ by financial institution. However, it’s clear there are a number of Profit/(Loss) Centers at a bank that will produce ROE’s below the overall bank’s target. In order to reach the bank’s goal, the commercial loan centers need to, and usually can, produce superior ROEs. Setting them to the bank’s overall goal will make it impossible for most institutions to reach their objective.


The post Setting ROE Targets – What NOT to Do appeared first on PrecisionLender.

Previous Article
5 Articles on AI to Help You Sleep at Night
5 Articles on AI to Help You Sleep at Night

Each month we like to compile a list of articles on Artificial Intelligence you must read. This month's edi...

Next Article
The Seven Deadly Sins of Pricing
The Seven Deadly Sins of Pricing

Things are looking up in the banking industry. But as you look at reshaping the experience for your commerc...