Partners (Winter 1996)

Why Banks Don't Make Every Loan You Think They Should Make
The following article is excerpted from a paper prepared by Ron Zimmerman entitled, "Banking for Nonbankers: Why Banks Don't Make Every Loan You Think They Should Make". The paper, originally written in September 1989, was updated in December 1995 and presents a simple discussion of important banking concepts that help explain why below market rate loans and higher default rates are difficult for banks to absorb. For a copy of the paper in its entirety, please contact this Reserve Bank.

The key to understanding how banks work is in knowing that banking is a highly leveraged, low margin, high volume business and gaining an appreciation for the constraints that these characteristics place on banks. In addition, to truly understand banking one must realize that banks are regulated entities and face keen competition from both within and outside the industry.


When a bank loan officer lends money, he or she must be mindful that most of the money belongs not to the owners, but primarily to depositors. As a rule of thumb, a bank in sound condition would be considered to be adequately capitalized if its capital amounted to about 7 percent of the bank's total assets. This means that for each dollar loaned out, only 7 cents is the bank owners' money and the remaining 93 cents belongs to someone else.

While federal deposit insurance has eliminated many concerns, the high leverage in banking continues to be both the boon and the bane of bank investors. On the one hand, leverage can mean that a seemingly insignificant profit on the bank's assets can yield a nice return on the amount the bank owner has invested. On the other hand, even a small loss on the bank's assets can mean a sizeable loss to the bank stockholder.

For instance, assume a one dollar loan. If only 7 cents of that dollar belongs to the bank investor, then even if the bank nets only a 1 cent return on that dollar, the investor's return is slightly in excess of 14.25 percent (i.e. $.01 divided by $.07 times 100). By the same token, if the bank loses 1 cent of the dollar, the investor's loss is 14.25 percent. Note that depositors do not share in the profits or losses because they do not share in the risk of the investments. In effect the depositor has opted for the comparative safety of an insured but lower yielding deposit account rather than an equity investment which brings the possibility of a higher return but also a greater risk of loss.


Bankers use two primary measures of bank profitability: return on assets (ROA) and return on equity (ROE). ROA is net income divided by total bank assets expressed as percentage. ROE is net income divided by total stockholders' equity expressed as a percentage.

ROA and ROE have different uses, but both are important. ROA is used to compare one bank with another. ROE allows analysts and investors to compare a bank's performance to not only other banks but to companies operating in other industries as well. One must realize that a bank's ROA and ROE has to be competitive in the marketplace. Otherwise, the bank cannot attract the investment capital it needs to grow.

A bank is regarded as doing reasonably well if its ROA is 1 percent or better. In 1995, the average ROA for all banks in the U.S. was 1.3 percent. Large money center banks in particular would be very pleased with a 1 percent rate of return, since competition is so keen among these banks and with other competitors in the nonbank financial service industries.

Low Margin

One might reasonably ask why banks cannot make much more than 1 cent on each dollar of assets. The answer is because banking is a low margin business. Banks' costs greatly offset the gross yields received on their investments.

Earning and Nonearning Assets

Bank assets may be divided into two broad categories: earning and nonearning. Earning assets for the most part consist of loans and securities. Nonearning assets might include actual cash on hand, the bank building, other real estate owned (which primarily consists of properties acquired through foreclosure) and loans that are not being repaid. One would logically conclude that the higher the percentage of earning assets, the more income a bank might expect to generate. For this reason, banks monitor the relationship between earning assets and nonearning assets very closely.

Hence, the volume of foreclosed properties is particularly critical since, not only are these assets not earning interest but the bank typically incurs costs to maintain and sometimes improve the property until it can be sold. In addition, the bank must pay interest on the deposits used to fund a foreclosed asset despite the fact that the bank is receiving no income. This is why one often hears bankers say they do not lend solely based on collateral value. Absent a borrower's reasonably reliable source of cash flow, a bank generally will not make a loan no matter how much the collateral is worth in relation to the requested loan amount.

Competitive Pressure on Earnings

If a bank were able to earn an average 8 percent on its assets and paid an average 4 percent for deposits, its net interest income would be 4 percent. Net interest income (NII) is the difference between the interest earned by banks on their loans and other assets, and the interest paid by banks for the use of depositors' funds. NII is the largest component of a bank's earnings. Other sources of revenue, called noninterest income, includes earnings from bank services such as fees for safekeeping services and trust accounts, and service charges on deposit accounts.

Overall, a bank averages about 1 cent in noninterest income for each dollar of assets on its books. In our example, if we add this amount to the 4 cents in NII, the bank's earnings before expenses amounts to about a nickel on the dollar of assets. Out of this, the bank must pay for its losses on loans that are not repaid, and pay its overhead expenses and taxes.

"...Banking is a high volume, low margin, highly leveraged business"


A bank's overhead expenses typically include salaries and employee benefits, rent on the bank buildings, furniture and equipment, data processing systems, marketing expenses, insurance, federal assessment for deposit insurance, stationery, postage, telephone, etc. Because of the high volume of transactions banks complete, large staffs and correspondingly large amounts of office space, equipment and supplies are needed. In addition, its "back office" functions (e.g. bookkeeping, data processing, marketing, and the like) are not readily apparent to the public. However, these functions along with the more obvious expenditures result in a large overhead expense relative to many other industries. A representative figure might be around 3 cents on each dollar of assets, although inflation, salary competition to attract and retain good employees, and other factors are constantly straining overhead costs.

If we subtract the 3 cents from our nickel above, we are left with 2 cents before taxes and loan losses. Overall, a bank will be doing reasonably well if it is netting about 1 cent on each dollar of assets after taxes.

That concludes a simplified description of how banks make money and how much money they make. In reality, the process is enormously complex with little room for errors in judgment or faulty execution. The example above used whole percentage points for illustration. In actuality, bankers measure success or failure in fractions of a percentage point, or so-called "basis points" (100 basis points equal 1 percent). A few basis points swing in cost or income can mean a lot of money to a bank. For this reason, bankers are known to have some of the "sharpest pencils" around, figuring their costs to the fraction of a penny.


What constitutes high risk in a credit decision is a matter of opinion. However, perhaps it can be put into perspective by examining in a general sense how much a bank can afford to lose on loans. Let's take our $1 in assets again, remembering that the bank has about 7 cents in capital and nets about 1 cent on the $1 in assets.

Assume, for example, that earning assets average 75 percent of total assets or 75 cents in loans. If only 1 percent of our loans are lost, that's 75 percent of the 1 cent in net income the bank would have made. If the bank netted 1 percent on the remaining 74.25 cents in loans (75 cents minus .75 cents lost), the actual "profit" would amount to .7425 cents in interest (74.25 cents times .01) less .75 cents in loan losses or a net loss of .0075 cents. So losing 1 percent of loans in this example equates to an overall loss of .0075 cents.

As a practical matter, a bank may be able to absorb more in loan losses, perhaps as much as 2 percent, before the bank sustains an overall loss. This is because some banks' cost structures allow them to net a higher amount of interest income and some generate more noninterest income. However, losses of 1 percent on loans would surely have the stockholders howling since the return on their investment would in any case be below normal for the bank (if not negative). Of course, even 2 percent is not much of a margin for error. It should also be apparent from these calculations that a 10 percent loan loss would render the bank insolvent! This helps explain why the highly leveraged nature of banking compels bankers to be so conservative in their credit judgments.


In conclusion, it should be noted that the inherent nature of banking severely restricts how conventional banking products, in the absence of public or private enhancements, can be modified to make them more affordable for low- and moderate-income people. The fundamentals cannot be altered long term without undermining the competitiveness of the banking industry and seriously jeopardizing banks' safety and soundness. There is a limit to the concessions that banks alone can make. That limit is far below the level needed to make long-term progress in addressing the needs of low- and moderate-income people. If one fails to recognize this fact, one will be forever trying to pound a square peg into a round hole.

Fortunately, there is better way: the public/private partnership. Government, charities, and private corporations can work with the banks to leverage their funds in ways that are affordable and effective. In this way, each party can play to its strengths and through enlightened self interest, everyone involved can "win".

Return to Index