Money · Credit · New Zealand · Part 26

Does housing lending starve productive businesses of capital?

Housing absorbs most New Zealand bank lending. But that does not mean every mortgage dollar mechanically removes a dollar that could have financed a business.

The numbers are striking.

At June 2026, registered banks had approximately $615.1 billion of gross loans outstanding.

Of that:

$395.4 billion was housing lending,

$133.4 billion was business lending,

and $63.6 billion was agriculture lending.

Reserve Bank — Banks: Assets — Loans by purpose ↗

Housing therefore represented about 64.3% of total bank lending.

Business lending was about 21.7%.

Agriculture was about 10.3%.

That naturally raises a serious question.

If so much bank credit is tied to housing, is productive enterprise being starved of capital?

The answer is:

sometimes credit allocation can favour property in ways that matter for productivity — but the mechanism is more complicated than a fixed pot of money being divided between houses and businesses.

The problem is not that banks run out of dollars. The problem is that they have finite capital, liquidity, risk appetite and management capacity — and some forms of lending are easier to underwrite than others.

There is no fixed bucket of loanable deposits

We should remove the first misconception immediately.

A bank does not receive $1 million of deposits and then choose whether that exact $1 million becomes a mortgage or a business loan.

As Parts 1 to 3 established, bank lending can create new deposits.

So housing lending does not mechanically crowd out business lending because a finite pile of customer deposits has been “used up”.

But banks are still constrained.

They need:

shareholder capital,

stable funding,

liquidity,

profitable spreads,

creditworthy borrowers,

and risk capacity.

Those constraints make the composition of lending economically important.

Housing is attractive to banks for understandable reasons

A residential mortgage usually comes with a large, identifiable physical asset as security.

Housing markets have extensive valuation data.

Mortgage products are relatively standardised.

Borrower income can be assessed against servicing requirements.

There are large numbers of similar loans, allowing risk to be diversified across households and locations.

And if a borrower defaults, the lender has recourse to property collateral, subject to law and the actual value realised.

None of this means mortgages are risk-free.

House prices can fall.

Borrowers can lose income.

Interest rates can rise.

Natural disasters can damage collateral.

But mortgage lending is often easier to standardise, price and secure than lending to a small company whose future value depends on management, customers, intellectual property and uncertain cash flows.

Business lending is much less uniform

A small software company is not the same credit as a supermarket.

A manufacturer is not the same as a hospitality operator.

An exporter is not the same as a property developer.

Banks may need to assess:

cash flow,

industry risk,

management experience,

contracts,

customer concentration,

inventory,

equipment,

receivables,

and business history.

The Reserve Bank's May 2026 Financial Stability Report notes that lending to other businesses tends to be more expensive than lending against commercial property or agriculture because other businesses generally have more diverse risk profiles and less security coverage.

Reserve Bank — Financial Stability Report May 2026 ↗

A house gives the lender an asset it can see. A young business may be asking the lender to believe in cash flows that do not exist yet.

That does not make housing more productive

This distinction is crucial.

A loan can be relatively safe for a bank without creating much new productive capacity for the economy.

Consider a mortgage used to buy an existing house.

The loan facilitates a transfer of ownership.

It does not, by itself, build another dwelling.

The bank may have made a perfectly rational, well-secured loan.

But the economy has not necessarily gained:

another factory,

another export product,

new technology,

or additional housing supply.

That is why bank profitability and national productivity are not the same objective.

But not all housing lending is economically unproductive either

Mortgage credit can support new construction.

Residential development finance can create new homes.

Housing transactions can enable labour mobility.

Renovations can improve the housing stock.

And a functioning mortgage market is essential to household formation and ownership.

So the argument should not become:

“housing credit bad, business credit good.”

That would be another oversimplification.

Business credit is not automatically productive

At June 2026, the Reserve Bank classified $48.7 billion of the $133.4 billion of business lending as commercial property lending.

That included approximately $43.2 billion of commercial property investment, $2.2 billion of commercial property development and $3.3 billion of residential property development.

The remaining $84.7 billion was classified as other business lending.

Reserve Bank — Business lending by purpose, June 2026 ↗

Even within “other business”, a loan might finance:

productive machinery,

working capital,

an acquisition,

a share buyback,

inventory,

or simply a business surviving a temporary cash-flow squeeze.

The label tells us something.

It does not tell us the full economic return.

The SME numbers reveal the deeper problem

At June 2026, banks reported approximately $84.3 billion of SME lending within the business-lending category.

Reserve Bank — Banks: Assets — Loans by business size ↗

The Reserve Bank says small firms make up a disproportionately low share of bank business lending relative to their contribution to economic activity and employment.

It also identifies an important bridge between the housing market and business finance.

At the smaller end of the SME market, business owners often borrow against their own homes.

About $5 billion of SME lending is secured against owners' houses

The May 2026 Financial Stability Report estimates residential mortgage-secured business lending at around $5 billion, or about 11% of SME bank lending excluding agriculture and commercial property.

The Reserve Bank warns that the actual relationship may be larger because some lending to sole traders can be classified as residential lending — for example where a mortgage revolving-credit facility is used to fund business operations.

Reserve Bank — SME access to finance ↗

This is a major structural issue.

A person may have a viable business idea.

But access to bank finance can partly depend on whether they already own residential property.

An economy should care when access to entrepreneurial capital is partly determined by whether the entrepreneur already owns a house.

That can reinforce wealth inequality

Two entrepreneurs may have equally strong skills and equally promising businesses.

One owns a $1 million house with substantial equity.

The other rents.

The first may be able to offer residential collateral.

The second cannot.

The bank's behaviour may be rational from a credit-risk perspective.

But the economy can still end up allocating opportunities according to pre-existing asset ownership.

This is an example of how individually rational lending decisions can produce broader structural consequences.

Collateral can influence the price of credit too

The Reserve Bank's 2026 work on SME finance finds that collateral materially affects business lending spreads.

Commercial-property investment tends to have the lowest spread over wholesale interest rates because security is strong and typical LVRs are below 65%.

Agriculture is also well secured.

Other businesses generally face higher spreads because collateral coverage is more diverse and cash-flow risk can be greater.

Reserve Bank — Business lending costs and collateral ↗

So the issue is not merely whether credit is approved.

It is also:

at what price?

Capital requirements influence the economics of lending

Banks must fund part of their risk-weighted exposures with loss-absorbing capital supplied by their owners.

The Reserve Bank says the capital framework determines exposures and assigns risk weightings as part of calculating required capital ratios.

Reserve Bank — Capital requirements for banks ↗

Different lending risks therefore have different balance-sheet economics.

But we should be careful before claiming that one regulatory number alone explains New Zealand's housing-heavy banking system.

Credit allocation is the combined result of:

risk weights,

collateral,

historical losses,

funding costs,

operational costs,

competition,

borrower demand,

bank expertise,

and expected returns.

The capital rules are changing

The Reserve Bank finalised updated Banking Prudential Requirements in July 2026 following its review of key capital settings.

The transition is important:

from 1 October 2026, banks may begin using any of the new standardised credit risk weights;

by 1 November 2026, they must use the new residential-mortgage risk weights;

and by 1 April 2027, all new standardised credit risk weights must be in use.

Reserve Bank — 2026 Banking Prudential Requirements decisions ↗

That means any discussion of regulatory incentives needs to distinguish current settings from the transition now underway.

Does mortgage lending literally crowd out a business loan?

Not necessarily.

If a bank writes a $700,000 mortgage today, there is no accounting law saying it must reject a $700,000 business loan tomorrow.

If it has:

enough capital,

adequate liquidity,

stable funding,

a creditworthy business applicant,

and acceptable risk-adjusted return,

it can make both.

This is why the one-for-one crowding-out argument fails.

But portfolio crowding can still occur

Banks do not have infinite balance sheets.

Management has risk limits.

Capital is finite.

Funding has a cost.

Sector concentrations are monitored.

Staff and underwriting expertise are finite.

If mortgages offer an attractive risk-adjusted return relative to difficult SME credit, a bank may rationally allocate more balance-sheet capacity toward housing.

That can occur without any single mortgage directly “using up” the dollars required for a business loan.

The crowding mechanism is not a shortage of money. It is competition for balance-sheet capacity and risk-adjusted return.

Demand matters too

We should not assume every business that does not borrow was rejected by a bank.

The Reserve Bank's May 2026 report says SME borrowing demand had been muted during the period of high interest rates and subdued economic conditions, although applications had started to recover from mid-2024.

Banks reported more SMEs seeking finance for acquisitions, asset purchases, expansion and working capital.

Reserve Bank — SME credit demand ↗

So weak business lending can reflect:

banks refusing credit,

businesses not wanting debt,

investment opportunities being unattractive,

or borrowers deciding that the offered terms are too expensive.

Those are different problems and require different responses.

Loan rejection rates alone do not settle the issue

The Reserve Bank notes that business loan rejection rates tend to be low.

But that does not prove all viable SMEs have excellent access to capital.

Some firms may never apply because they expect rejection.

Others may refuse the terms offered.

Others may be constrained by collateral requirements before a formal application ever reaches a credit committee.

The Bank also says data limitations make it difficult to assess how competitive SME finance really is, particularly because pricing is much less transparent than in the residential mortgage market.

This is a measurement problem worth fixing.

New Zealand's wider productivity problem makes the question more important

Treasury research published in 2025 describes New Zealand as relatively capital shallow compared with other advanced economies.

It identifies the high cost of capital as one contributor and argues that greater capital intensity and innovation are both potential routes to stronger productivity growth.

Treasury — Why not both? The effects of innovation and capital on productivity in New Zealand ↗

This does not prove that bank mortgage lending caused New Zealand's productivity performance.

That would be far too strong.

Productivity is influenced by:

skills,

management,

competition,

innovation,

infrastructure,

market size,

regulation,

technology diffusion,

and capital investment.

Finance is one part of that system.

So what would “more productive credit” actually mean?

It should not mean politicians instructing banks which private companies to lend to.

That would create obvious political and credit risks.

It should mean improving the conditions under which commercially viable productive investment can obtain finance.

That could include:

better competition in SME finance,

more transparent business-loan pricing,

stronger equity and venture-capital markets,

credit guarantees where a genuine market gap is proven,

co-investment structures,

specialist lenders,

better use of movable-asset and cash-flow lending,

and public investment institutions operating under the safeguards discussed in Parts 9 to 12.

Equity matters because not every business should be debt-financed

A young high-growth company may be a poor candidate for an ordinary bank loan.

It may have:

limited collateral,

volatile early revenue,

high research costs,

and a long path before positive cash flow.

That does not necessarily mean the financial system has failed.

It may mean the company needs equity rather than debt.

A mature manufacturer buying proven machinery may be highly suitable for bank finance.

A pre-revenue technology company may need venture capital.

Matching the instrument to the risk is part of good capital allocation.

Public guarantees need discipline

A government guarantee can lower the lender's expected loss and encourage credit where collateral is weak.

But the risk has not vanished.

It has moved partly to the Crown.

Any guarantee programme should therefore have:

a defined market failure,

eligibility rules,

risk sharing,

pricing,

portfolio limits,

loss reporting,

and periodic evaluation.

Otherwise “supporting productive businesses” can become a subsidy to poor underwriting.

Could a national investment institution help?

Possibly — if the evidence shows a genuine financing gap.

Parts 9 to 12 already set the conditions.

It should not lend merely because a private bank said no.

It should ask:

is the project economically valuable?

is private finance genuinely unavailable or structurally unsuitable?

can public capital crowd private capital in?

what risk is the taxpayer taking?

and what measurable additionality is being created?

A public institution should complement a functioning financial market, not replace credit discipline.

We should also look beyond bank lending

A country's capital system includes much more than registered banks.

Businesses can be financed through:

retained earnings,

share issues,

private equity,

venture capital,

corporate bonds,

finance companies,

trade credit,

leasing,

invoice finance,

and government or co-investment programmes.

If New Zealand wants deeper productive investment, the relevant question is not simply:

“How do we make banks lend less on houses?”

It is:

“How do we build a financial system with more credible pathways for productive projects to obtain the right kind of capital?”

A better credit-allocation dashboard

I would monitor:

Housing lending as a share of total bank credit.

New lending for existing property versus new construction.

Business lending excluding commercial property.

SME lending by firm size and sector.

Business loan approval and rejection rates.

The pricing spread between SME loans and wholesale funding.

The amount of business credit secured by residential property.

Equity and venture-capital investment.

Investment in machinery, technology and intellectual property.

Productivity and capital intensity after the finance is deployed.

The purpose of a credit system is not to maximise loans. It is to connect savings, balance sheets and risk-taking with activities capable of creating durable value.

My conclusion

Housing lending does not mechanically starve businesses of money.

Banks can create deposits when they lend, and there is no fixed pool of customer deposits that mortgages permanently exhaust.

But the composition of credit still matters.

Banks have finite capital, liquidity, risk appetite and operational capacity.

Residential property offers strong and standardised collateral.

Small businesses are harder to assess and often pay more for credit.

Some entrepreneurs even need to own a house before they can efficiently finance a business.

That is not proof that mortgages caused New Zealand's productivity problem.

But it is enough to justify a serious policy question about whether the financial system gives productive investment enough pathways to capital.

The question is not whether New Zealand should stop financing homes. It is whether owning property has become too important a gateway to financing everything else.

A parallel discussion starts here

The collateral question is important enough that Part 27 will return to it directly.

But Part 26 also exposes a wider question about Māori capital access, productive ownership and the role of transaction infrastructure.

I have opened a parallel writing subject — Māori Economy & Productive Capital — to examine that problem separately rather than forcing it into the banking series.

We will return to Part 27 after establishing that framework.

Primary sources

Reserve Bank — Banks: Assets — Loans by Purpose (S31) ↗

Reserve Bank — Banks: Assets — Loans by Business Size (S35) ↗

Reserve Bank — Financial Stability Report May 2026 ↗

Reserve Bank — Capital Requirements for Banks ↗

Reserve Bank — 2026 Banking Prudential Requirements Decisions ↗

Treasury — Why Not Both? The Effects of Innovation and Capital on Productivity in New Zealand ↗