The need for a more risk-sharing alternative to traditional mortgage debt is the key lesson to be drawn from the Great Recession.

At least that is the conclusion of Atif Mian, an economics professor at Princeton and co-author of House of Debt, who visited the Reserve Bank last week.

The book analyses the role that the build-up of household debt in the United States played in triggering what became a global financial crisis and, crucially, in deepening and prolonging the recession which followed.

"Economic disasters are almost always preceded by a large increase in household debt," Mian and his co-author Amir Sufi write. "In fact the correlation is so robust that it is as close to an empirical law as it gets in macro-economics."


Between 2000 and 2007, US household debt doubled to $14 trillion and the household debt-to-income ratio rose from 140 to 210 per cent.

In New Zealand, the ratio currently stands at 168 per cent, but unlike the United States, owner-occupiers' mortgage interest payments are not tax deductible.

When housing markets around the US started to crash, ultimately wiping out $5.5 trillion of household wealth, the collateral damage to the broader US, and ultimately global, economy was amplified, they argue, by the nature of debt itself.

Even though a mortgage is a contract between two willing parties, there is a big asymmetry in the downside risk.

If the value of the house against which the loan is secured falls, the borrower's equity takes the hit and it is not until all of that equity is wiped out that the lender is exposed to loss.

To be fair, that same leverage amplifies the homeowner's gains in a rising market which we, mistakenly, tend to think is the only kind there is.

Economic disasters are almost always preceded by a large increase in household debt


Because borrowers tend to spend more of their income than savers do, when they have to tighten their belts the spillover effect on consumption and economic activity is greater than it would be if the downside risk was more symmetrical.

And when they can no longer support their debt and foreclosure ensues — as it did for more than 4 million American families — the rash of mortgagee sales further depresses house prices in their localities and puts further pressure on other indebted households.

Traditional mortgages facilitate bubbles, Mian and Sufi argue, by convincing lenders their money is safe and leading them to lend to optimists who bid prices higher and higher.

"The vicious levered losses cycle can be broken only if we fundamentally alter the way households finance themselves." When someone finances the purchase of a home, the contract they sign must allow for some sharing of the downside risk. It must resemble equity more than debt.

When house prices rise, both lender and borrower would benefit. Likewise, when house prices crash, both would share the burden.

Mian and Sufi propose a "shared responsibility mortgage" which differs from the familiar kind in offering downside protection for the borrower in exchange for giving the lender a share — they reckon 5 per cent — of the capital gain when a mortgaged property is sold in normal times.

Say the house cost $500,000 and was paid for by $100,000 of the buyer's money and $400,000 of the bank's. If house prices in the agreed locality rise, the borrower's payments operate as normal.

But if house prices, reflected in some index, fall below their level when the loan was taken out, the amount the borrower owes is proportionately reduced, until prices recover to the original level.

The borrower's equity falls in dollar terms but is not wiped out, and the impact on spending and the broader economy is mitigated. Mortgagee sales would be much rarer.

Clearly, lenders have to be compensated for holding that additional risk. When Mian and Sufi do the sums, based on historical averages for house price inflation and assuming adequate pooling of risk, they calculate that a 5 per cent share of the capital gain upon sale, whenever that occurs, would adequately compensate the lending institution for its increased risk.

The risk sharing model would reduce the amplitude of the housing cycle and its associated risk.

And it would redistribute risk away from the people least able to bear it.

When Mian and Sufi looked at the US data — and it is unlikely the picture here would be radically different — they found that leverage was inversely correlated with household wealth.

The poor man's debt, in short, is the rich man's asset. And while there is nothing sinister, they say, in the rich financing the poor, when that financing takes the form of incompressible debt, when downturns happen the effect is to deepen inequality.

And when things got really ugly, as in the US when house prices fell 30 per cent, it was banks and not borrowers who got bailed out by the taxpayer, moral hazard notwithstanding.

A purist might argue that it is the creditors' responsibility to choose their debtors. If debtors cannot repay, creditors should bear the losses.

The trouble with that is that "creditors" in this case include depositors who have entrusted their hard-earned savings to a bank.

The prospect of taking a haircut on their deposits, in the event of the Reserve Bank's "bail-in" model of open bank resolution to a bank in trouble, is liable to come as an unwelcome surprise.

Mian and Sufi's risk-sharing approach would also include a shift in the way banks are funded, with more equity capital and less debt. Changes to the Basel rules which give a lower risk weighting to housing mortgages, allowing the banks to hold less capital against that part of their lending, would also be likely to be needed.