Money · Credit · New Zealand · Part 23
If banks create the principal, where does the money for the interest come from?
The question sounds like an accounting paradox. It disappears once we separate a stock of money from a flow of payments over time.
Here is the argument in its strongest form.
A bank creates a $100 loan.
That creates $100 of new deposit money.
But the borrower might eventually owe $110 — $100 principal plus $10 interest.
So where does the extra $10 come from?
If the bank only created $100, does that mean there can never be enough money to repay the debt?
No.
The mistake is treating all future debt-service payments as though they must exist simultaneously as a separate pile of money on the day the loan is created.
Debt is a stock of obligations. Interest payments are flows through time. The same unit of money can be used more than once.
Start with what the Reserve Bank actually says
The Reserve Bank explains that most broad money in New Zealand is bank deposits, and that those deposits are created largely through bank lending.
It also explains that repayment of bank loans destroys broad money.
Reserve Bank — Money creation in New Zealand ↗
So the starting point is correct:
new bank lending can create new deposit money.
But that does not imply that every dollar of interest due over the life of the loan has to be created by that same loan at origination.
The simplest example
Suppose a bank creates a $100 loan for Alice.
Alice spends the $100 buying goods from Bob.
Bob now holds the $100 deposit.
Alice later performs work for Bob and earns $20.
Bob transfers $20 of the existing deposit money to Alice.
Alice uses $10 to pay interest to the bank.
The same money has circulated.
There did not need to be $110 sitting in the economy at the same moment.
There only needed to be enough money circulating over time for Alice to earn the income required to make the payment.
A $100 money stock can support more than $100 of payments over a year because money can change hands repeatedly.
This is the difference between a stock and a flow
A stock is measured at a point in time.
The money supply on 30 June is a stock.
The amount of debt outstanding on 30 June is a stock.
A flow is measured over a period of time.
Income earned during a month is a flow.
Interest paid during a year is a flow.
GDP is a flow.
Bank interest income is a flow.
Comparing the stock of money today with every interest payment that will occur over the next twenty or thirty years is therefore not an apples-to-apples comparison.
The same dollar can service many transactions
Imagine there is only one $20 note in a small town.
On Monday it pays a mechanic.
On Tuesday the mechanic buys groceries.
On Wednesday the grocer pays a cleaner.
On Thursday the cleaner pays a bill.
The town has made $80 of payments using one $20 note.
No new note had to be created for each transaction.
Bank deposits work differently from physical cash operationally, but the same principle applies:
money can circulate through multiple transactions over time.
This is why the total value of transactions or income during a period can be much larger than the money stock measured at one instant.
What happens when the borrower pays interest to the bank?
Part 22 established the accounting path.
When a borrower pays interest from a deposit:
the customer's deposit falls,
and the bank recognises interest income.
That interest payment does not reduce the principal loan balance dollar-for-dollar.
But the bank does not normally lock all of its interest income in a vault forever.
Banks have expenses and distributions.
They pay:
interest to depositors and wholesale funders,
wages,
technology costs,
rent and property expenses,
professional services,
tax,
suppliers,
and dividends.
Those payments send purchasing power back into the economy.
The current New Zealand numbers show these flows clearly
In the June 2026 quarter, registered banks reported:
$8.427 billion of interest income,
$4.457 billion of interest expense,
$2.197 billion of operating expenses,
and $1.626 billion of profit after tax.
Reserve Bank — Banks: Summary income statement, June 2026 ↗
Those numbers matter because they show that interest received by banks is not simply removed permanently from economic circulation.
A large share is paid back out through banks' own funding costs and operating expenses.
Some remains as profit and retained earnings or can later be distributed to shareholders.
Interest paid to a bank becomes part of the bank's income stream. Banks then have their own outflows.
Interest paid to depositors is particularly obvious
Banks charge interest on loans.
But they also pay interest on many deposits and other funding liabilities.
The Reserve Bank's June 2026 figures show $4.457 billion of interest expense in the quarter.
Reserve Bank — Bank interest expense ↗
When a bank credits interest to a customer's deposit account, the customer's deposit balance increases.
That is one route through which money paid to banks as interest can return to the public's deposit balances.
Wages and supplier payments do the same thing
Suppose the bank earns $1,000 of interest income.
It then pays an employee $700 and a software supplier $300.
If those recipients hold bank accounts, their deposits are credited.
The purchasing power has returned to the non-bank economy.
The exact accounting depends on whether payments remain within the same bank or settle across banks, but the system-wide point is unchanged.
The money can circulate again and be used by someone else to service debt.
What about profit?
Profit is not necessarily paid back into circulation immediately.
A bank can retain some earnings as equity.
Retained earnings strengthen the bank's capital position and can support future balance-sheet growth.
Some profit may later be paid as dividends.
Those dividends become income to shareholders.
So not every dollar follows the same timing.
But there is no accounting rule requiring the banking system to permanently hoard every dollar of interest it receives.
New lending also changes the stock of money
The money stock is not fixed while old loans are being serviced.
New mortgages are created.
Businesses draw new credit.
Credit-card balances change.
Existing loans are refinanced.
Some loans are repaid.
Some are written off.
Banks pay expenses and interest.
Government and private financial transactions also change deposits and other balance-sheet positions.
The quantity of broad money therefore evolves continuously.
The Reserve Bank says the amount of broad money depends on interactions between banks and their customers rather than being a fixed quantity mechanically controlled by the central bank.
Reserve Bank — Broad money creation ↗
So does new debt have to be created forever just to pay old interest?
No.
This is another version of the same mistake.
A borrower can service interest from:
wages,
business revenue,
investment income,
asset-sale proceeds,
transfers,
or other receipts.
Those receipts are transfers of existing purchasing power between participants in the economy.
The borrower does not need a new loan every time an interest payment falls due.
Of course, some borrowers do refinance or borrow more to meet obligations.
That can happen.
But it is a financing choice or a sign of cash-flow stress — not a mathematical necessity created by the existence of interest.
A business shows this more clearly
Suppose a company borrows $1 million to buy productive machinery.
The loan creates financing.
The machinery helps the company produce goods.
Customers buy those goods.
The company earns revenue.
Part of that revenue pays wages.
Part pays suppliers.
Part services the loan.
The money used to pay interest comes from the company's income stream, which depends on the circulation of money through the wider economy.
The crucial question is not whether the bank originally created the interest amount.
It is whether the borrower created enough real economic value and cash flow to earn the money required to service the debt.
Interest is paid from income. The banking system does not have to pre-create every future interest payment at loan origination.
Now consider the whole economy
The economy is not one borrower and one frozen bank account.
Millions of payments occur.
Businesses pay households.
Households pay businesses.
Banks pay depositors.
Banks pay staff and suppliers.
Companies pay dividends.
Governments tax and spend.
Investors buy and sell assets.
New bank loans are created while old loans are repaid.
Money circulates across all of those transactions.
So the question “where is the extra interest money?” assumes an economy that does not actually exist — one in which the only transaction after a loan is made is repayment to the bank.
But there is a genuine problem hiding underneath the bad argument
The fact that the arithmetic paradox is wrong does not mean debt burdens can never become unsustainable.
They can.
The real problem is debt-service capacity.
A borrower can fail even when plenty of money exists elsewhere in the economy.
Why?
Because the borrower may not be earning enough of it.
This is a distribution and cash-flow problem.
Not an aggregate accounting impossibility.
This distinction matters enormously
Suppose there are $500 billion of deposits in the economy.
That does not mean every household can service its mortgage.
The deposits might be concentrated elsewhere.
A particular borrower may lose their job.
A business may lose customers.
Interest rates may rise.
An asset may fall in value.
Refinancing may become difficult.
The system can contain plenty of money while individual borrowers still default.
Aggregate money can exist while the wrong borrower has too little income at the wrong time.
Higher interest rates change the distribution of cash flow
When lending rates rise, borrowers transfer a larger share of their income to lenders.
Depositors and other bank funders may also receive higher interest income.
Bank margins can change.
Spending patterns change.
That is one channel through which monetary policy affects demand.
The problem is not that the interest money literally does not exist.
The issue is that higher debt service can reduce the disposable cash flow available for other spending and increase default risk.
What if every borrower tried to repay every bank loan at once?
Then the banking system's deposit money would contract dramatically as principal was extinguished.
That would be highly disruptive.
But this is not because interest created an arithmetic impossibility.
It is because modern banking is built around continuously circulating deposits, staggered maturities, rolling credit relationships and an economy in which debts are not all settled simultaneously.
A mass attempt to pay down debt at once can create a powerful contraction in spending and credit.
This is sometimes called a deleveraging process.
It can be economically painful precisely because repayment destroys deposits and reduces purchasing power.
This is why credit cycles matter
During a credit boom:
new lending may exceed repayments,
deposits expand,
spending capacity rises,
and asset prices can be supported.
During deleveraging:
repayments can exceed new lending,
credit growth slows or contracts,
deposit creation weakens,
and demand can fall.
The monetary problem is therefore not “interest creates an impossible shortage”.
The more useful question is whether the growth and distribution of income, credit and deposits are consistent with sustainable debt service.
There is another misconception: “all debt must be repaid at once”
It does not.
A thirty-year mortgage is repaid over thirty years.
A revolving business facility may be renewed.
A government bond may mature and be refinanced.
Companies continuously roll working-capital facilities.
The financial system is built around different maturities.
So comparing total outstanding debt today with the money stock today and demanding that the entire debt plus all future interest could be paid tomorrow is not a meaningful solvency test.
The same distinction applies to government debt
The Crown pays interest over time from tax revenue and other receipts.
Government bonds mature at different dates.
Some principal is refinanced.
The Crown does not need every dollar of principal plus every future coupon payment sitting in a bank account today.
What matters is its capacity to service and refinance obligations as they fall due without destabilising the fiscal position.
This is exactly the distinction Parts 18 to 20 developed.
Where the “missing interest” argument goes wrong
I would reduce the error to four points.
1. It compares a money stock with future payment flows.
2. It assumes money can be used only once.
3. It ignores bank expenses, interest payments, wages, taxes and distributions that return income to the wider economy.
4. It ignores continual new credit creation, repayment and refinancing across the banking system.
But we should not swing to the other extreme
Rejecting the missing-interest argument does not mean debt is harmless.
Interest is a real transfer of income.
High leverage can make households and businesses fragile.
Rapid credit creation can amplify asset-price cycles.
High interest rates can push borrowers into default.
Debt service can reduce consumption and investment.
And if credit grows much faster than productive capacity, the financial system can become increasingly vulnerable.
Those are serious problems.
They simply need to be described accurately.
A better debt-service test
Instead of asking “Was the interest money created?”, I would ask:
What income does the borrower have?
How stable is that income?
What proportion goes to interest and principal?
What happens if rates rise?
Can the borrower refinance?
Does the financed asset generate income or productive value?
Is the debt growing faster than the borrower's capacity to service it?
Those are the questions that determine whether debt is sustainable.
My conclusion
Banks do not need to create a separate dollar of principal and a separate dollar of future interest at the moment a loan is written.
They create the principal deposit when the loan is advanced.
Interest is then paid over time from income and money already circulating through the economy.
Banks themselves return substantial flows to the economy through interest expense, wages, suppliers, tax and distributions.
The stock of deposits also changes continually as new loans are made and old loans are repaid.
So there is no built-in accounting law requiring perpetual new debt merely because loans charge interest.
The real danger is different.
Borrowers can owe more each period than their income can support.
And an economy can build a debt structure that becomes fragile when income falls, rates rise or refinancing disappears.
The problem is not that the interest money is mathematically missing. The problem begins when debt service outruns the income available to pay it.
Next
That gives us the next question:
If banks can create credit, why do interest rates matter so much?
Part 24 will trace the Official Cash Rate through settlement balances, wholesale funding, deposit rates, mortgage pricing, business lending, asset prices and borrower demand — and show how the Reserve Bank influences credit creation without directly deciding how many loans banks make.
Primary sources
Reserve Bank — Money Creation in New Zealand ↗
Reserve Bank — New Zealand's Monetary Policy Implementation Framework ↗