Breaking the link between housing cycles, banking crises, and the recession

Avinash Persaud 14 April 2016



Housing booms and busts lie behind the biggest economic and financial crises in recent decades.

  • Between 2006 and 2009, the OECD countries that suffered the largest declines in household spending (Denmark, Ireland, Norway, Portugal, Spain, the UK, and the US) were those that had the greatest increase in household debt over the preceding ten years (Glick and Lansing 2010).

Most of this debt was collateralised on residential homes in one way or another. When house prices collapsed, the net worth of these indebted consumers and their banks followed suit, credit lines were cut, consumption fell and a fire sale of assets ensued.

  • Personal consumption fell by 20% or four times the national average in the one fifth of US counties that suffered the highest decline in housing net worth during the same period.

The collapse of house prices deepened the recession, not the other way around.1

Nexus of housing, banking and business cycles

In a recent paper (Persaud 2016), I explain how a nexus of housing boom-bust, banking crises, and economic cycles is not unique to the last crisis or its foremost features like sub-prime mortgages or credit derivatives. It has been increasingly present in each of the major banking crises since the breakup of Bretton Woods in the early 1970s.

  • Banking crises triggered by real estate collapse were foretold by previous booms which occurred in the UK in 1973, Spain in 1977, the US in 1986, Norway in 1987, Finland and Sweden in 1991, and Japan in 1992.

Real estate booms begetting financial crises are not unique to advanced economies.

  • Boom and bust in commercial real estate played an important role in the Asian Financial Crisis of 1997-98, especially in Thailand, Malaysia, Hong Kong and Singapore.2

Sharp declines in house prices undermine household wealth and bank balance sheets. In response, households reduce consumption and banks stop lending. Economic activity drops and businesses lay off workers. Unemployment rises. This leads to more non-performing mortgages and weaker house prices. The essential link between housing and the economy comes from foreclosure of non-performing mortgages, fear of foreclosure, and also by the risk aversion of lenders triggered by the transformation of their biggest assets from a seemingly liquid, safe asset to one that is illiquid and risky. Both of these links are connected to mortgages being held by those with funding that could fly out of the door the next day.

All assets that are purchased with overnight deposits, money market funds, or other short-term funding have to be valued on the basis of the price they will fetch if sold tomorrow. This is called marking-to-market. This isn’t just a modern legal, regulatory, or accounting requirement, but a sensible risk management one. Tomorrow these funders can ask for their money back and the asset may have to be sold immediately. Insured deposits are traditionally sticky, but internet-depositors, money market funds, and inter-bank markets are far less so. If a holder of assets has the capacity to hold on to them as a result of long-term funding, it makes sense to use a more stable, long-term valuation.3 But if this capacity does not exist, not marking to market will only be a cause of speculation over the real solvency of an institution. That speculation will further undermine a bank’s access to funding.

In the context of a downturn in economic activity, when a mortgage becomes non-performing, the value of the asset starts a downward spiral amid concern for fire sales of homes, rising unemployment, and the deteriorating saleability of foreclosed and abandoned homes. In anticipation of that, the best course of action is to try and recover as much of the non-performing loan as soon as possible. This decision is not substantially altered if the bank had more capital. It is not in the interests of a well-capitalised bank to let undercapitalised banks foreclose on non-performing mortgages first. If asset prices are falling and funding is drying up, slower sellers end up reporting bigger losses than fast sellers. One of the first tremors of the 2008 financial crisis was the announcement in February 2007 by HSBC – one of the better-capitalised institutions – that it was taking early action in response to a rise in non-performing mortgages (see Mollenkamp 2007).

Financial crises are often triggered by assets – previously considered safe by regulators, rating agencies and investors – turning bad at the same time. The abrupt overturning of previously strongly-held beliefs about market value causes a lurch into risk aversion by banks and others. Bankers in well-capitalised banks, seeing their less well-capitalised competitors suffer, turn risk-aversion into their brand, making it more entrenched, not less so. This is augmented by the use of regulatory-approved risk models that translate past volatility into future risk (Persaud 2001).

Policy responses have been stymied by the widely held but historically incorrect view that borrowing short to lend long is the unchangeable essence of banking. All that can be done, according to this view, is to tighten bank lending requirements, increase capital, and for central banks to offer near limitless liquidity. More capital is required and the more capital the better it is for the safety of banks and tax payers, but it does not sever the link between the housing cycle and the economy as much as many hope. Capital buffers are not used to hang on to troubled assets. Moreover, the descent into risk aversion is one of the reasons why aggressive central bank lending may keep banks alive but does not quickly restore lending and spending.

Policy remedies

More can be done. The mortgage dominance of and by banks is in direct response to the incentives on offer from bank regulators. It is not the natural result of technological, market, or competitive forces. Today’s banking model follows the regulatory model. Regulators favour, for instance, the securitisation of mortgages with lower capital requirements allowing banks to outcompete traditional mortgage lenders. It was not always so. The proportions of mortgages to household debt; of bank mortgages to all mortgages; and of real estate lending to total bank lending, are each dramatically higher than forty years ago.

There are types of mortgages that could reduce the sensitivity of the economy to housing cycles. One example is a mortgage that automatically reschedules interest and principal payments during an economic recession, forestalling foreclosures, and holding up household expenditure.4 Banks cannot provide these types of mortgages because doing so would add to their already dangerous pro-cyclicality – their liquidity and asset values rise and fall with the economic cycle. Institutions that have long-term liabilities, such as life insurance companies and pension funds are better suited to doing so because what matters to them is that their investments pay off in the long run when life insurance and pensions come due, not that they provide an income beforehand. They can offer mortgages that provide the flexibility that suits the homeowner, offsets the economic cycle and meets their own investment objectives. They don’t because current and oncoming capital requirements for insurers make it costly for them to do so.

Turn to insurance and pension funds

The solution is for the natural capacity of long-term insurers and pension funds to spread liquidity risks over time, and the current inability of banks to do so, to be better reflected in financial regulation. The global capital adequacy requirements for insurance assets recommended by the Financial Stability Board must be adjusted for the maturity of the liabilities they are set against. The Basle Committee’s Long Term Stable Funding Ratio for large international banks, which requires banks to fund illiquid assets with stable funding, must no longer be delayed and the definition of liquid assets needs to be narrower. 

These steps will reduce the danger of housing finance, but unless we address housing inequality there will be political pressure for easy financing that could reintroduce danger. Housing is a politically charged issue. The authorities should have an annual target for the new supply of affordable homes and employ planning laws, fiscal policy, and direct building to help meet it, but they should refrain from subsidised mortgages. Cheaper finance without cheaper homes only deepens housing inequality.

By not reflecting liquidity risks in capital requirements, banking and insurance regulators in the advanced economies have conjured up a dangerous system where those without liquidity take liquidity risks, and those with liquidity fail to do so. Systemic risk committees were created after the crisis to identify just these kinds of systemic flaws and develop integrated responses. Two regulatory steps, one in banking and the other in insurance, would enable the financial system to take long-term risks, like long-term mortgages, more safely. History suggests that a safer housing market would do more to make the financial system safer than all of the other recent initiatives put together.


Canham, C (2008) “UK professor calls for mark-to-funding measurement”, The Accountant, 15 November.

Glick, R and K J Lansing (2010) “Global household leverage, house prices and consumption”, Federal Reserve Bank of San Francisco Economic Letter, 11 January.

Mian, A and A Sufi (2015) House of Debt, University of Chicago Press.

Mollenkamp, C (2007) “In home-lending push, banks misjudged risk,” Wall Street Journal, 8 February.

Persaud, A (2001) “Sending the herd off the cliff edge: The disturbing interaction between herding and market sensitive risk management practices”, Jacques de Larosiere Prize Essay, BIS Papers, 2: 223-240.

Persaud, A (2016) “Breaking the link between housing cycles, banking crises, and recession”, Peterson Institute for International Economics, CIYPERC working paper series, 2016/02.

Quigley, J M (2001) “Real estate and the Asian Crisis”, Journal of Housing Economics, 10: 129-161.


[1] For an excellent survey of the link between housing cycles and recession (Mian and Sufi 2015).

[2] Singapore is of course also an advanced economy.

[3] See my proposal for ‘mark-to-funding’ that would allow for those with long-term liabilities to loosen the procyclicality of market-to-market accounting (see Canham 2008).

[4] Mian and Sufi (2015) have proposed a ‘shared responsibility mortgage’ which has equity like features. This has strong merit, but I believe that a more aggressive interest and principal holiday is required. The essential point though is that banks are not in a position to offer either of these instruments as they draw on bank liquidity when they need it most.



Topics:  Financial regulation and banking Global crisis

Tags:  housing market, housing inequality, real estate, business cycle, banking cycles, capital requirements, liquidity risk, insurance, Pension Funds, regulation

Emeritus Professor of Gresham College; Non-Executive Chairman of Elara Capital PLC


CEPR Policy Research