Household Debt, Wealth and Consumption
This issue of The Churchouse Letter is freely available for all to read, if you would like to read more recent issues complete with up-to-date investment ideas, you can purchase your subscription here
Are there Lessons for Asia from the US/Western Experience?
At the core of the global financial crisis is the rapid rise of household leverage in the developed world, much of which has been applied to real estate. Consumers have felt more wealthy as the value of their real estate has risen, and used this as the basis for piling on more and more leverage to support not just acquisition of more real estate but to support even greater levels of consumption. It is true to say that real estate is at the centre of the world’s current financial crisis, but whether it is a cause or an effect is a more difficult distinction to make. Most important is the link between real estate, household leverage and consumption, and how these may come back into some more sustainable balance over time.
We are indebted to a number of very well researched studies that have looked at the issue of household leverage, its role in the current recession, house prices, deleveraging and consumption. (See key references at the end of this section).
We tend to think of the growth of household debt and its impacts on property markets as a distinctly US phenomenon, but it is not. Many western societies have seen household debt ratios (as percentage of GDP, or as percentage of household disposable income) rise considerably in the past decade, some sharply more than has been the case in the US (See Figure 1). In 1960 personal debt to disposable income in the US rose from a modest 55% to a still modest 65% in the mid 1980’s, and then sharply accelerated, particularly in the past decade to more than 130% (See Glick and Lansing). In real terms household debt has increased very much more rapidly than household income, but interestingly real housing wealth has also increased substantially more than household income, despite the recent falls (See Figure 2).
Housing has created a substantial net addition to household wealth in the US and other western societies, but this is a relatively recent phenomenon – until the last decade real disposable income and real housing wealth tracked pretty closely in the US.
This pattern is probably repeated in many western societies that have seen very sharp increases in house prices in the past decade – UK, Ireland, Spain, Canada, Australia, New Zealand are likely contenders. (See Figure 4) Although the US has seen a sharp increase in household debt ratios, in absolute terms it is considerably lower than countries such as Ireland, Denmark and Norway, and the rate of increase of household debt ratios has also been more rapid in these countries than it has in the US. Spain and Portugal have had the greatest rise in household leverage in percentage terms over the past decade. (See Figure 1)
In Germany and Japan, household leverage ratios have remained unchanged over the past decade, and perhaps unsurprisingly these countries have had negative real house price growth over this time frame. (See Figure 4)
We have all watched the rise of household debt and the blowout in house prices and many have postulated the factors that caused or influenced that rise in debt –
- low interest rates,
- advent of securitization of mortgages,
- growth of sub-prime lending to a segment of the population that had hitherto not been able to borrow for housing,
- Government policies that created the impetus to provide such funding to lower income households,
- increasingly lax bank lending standards,
- rise of Fannie Mae and Freddie Mac as essentially government underwritten lending institutions creating moral hazard,
- substantial capital inflows to the US (to a considerable extent from trade surplus nations in Asia).
Pinning the precise contribution of any of these factors to the debt blowout is always going to be difficult, but it is fairly evident that this set of forces increased the total amount of credit available to the housing market, increasing demand for housing and thereby prices.
A new feature of this particular housing cycle has been the re-financing phenomenon, where existing homeowners have refinanced their homes to fund investment elsewhere and/or additional consumption. Remember all those adverts we all saw in north America and UK that claimed you deserved that new car, that holiday, and encouraging you to just do it by refinancing that home which had gone up so much in value.
Mian and Sufi found that existing homeowners borrowed considerably against house price increases to the tune of about 25 – 30 cents of every dollar of house price appreciation in the few years leading up to the start of the decline in 2007.
Much of this debt was used to finance consumption. But a state of affairs in which household debt and consumption keeps rising consistently at a rate faster than income is an unsustainable situation. At some point there must be a cap to the amount of debt that households can service from their income.
In a society such as the US, where consumption accounts for close to 70% of GDP, anything that impacts on the level of consumption, both to the upside and the downside, represents a big impact on GDP.
Recent studies for the US have demonstrated a clear link between rising levels of household debt and consumption through the current recession, with high debt areas seeing highest levels of defaults, highest falls in house prices, sharpest falls in consumption and higher rates of unemployment.
Mian and Sufi’s work demonstrates clearly the link between rising levels of household debt and consumption through this recession in the US. Their study breaks down household debt increases by counties across America, and sorts them according to the level of increase in household leverage in the 2002-2006 period. The study plots default rates, level of recession and patterns of consumption according to high and low leverage counties. The findings are clear:
- Recession started earlier in high leverage areas, and became deeper than in low leverage counties.
- High leverage counties experienced much higher housing loan default rates than low leverage counties – defaults rising by 12 percentage points (See Figure 8).
- House price declines were much greater in high leverage counties – 40% declines in house prices in high leverage areas against a 10% INCREASE in average house prices in low leverage counties (See Figure 9).
- Auto sales and housing permits also fell much more sharply in high leverage counties than in low leverage areas (See Figure 10-11).
- High leverage counties experienced higher levels of unemployment than low leverage counties (See Figure 12).
- The study finds that while low leverage counties largely avoided the recession in the early stages, durables spending and unemployment both got hit hard in high and low leverage counties at the tail end of the recession, largely due to a sharp drop in short term credit availability in the fourth quarter of 2008/ first quarter 2009 (i.e. post Lehman failure).
- Defaults and the beginnings of household deleveraging started in mid 2006, well before the sub-prime crisis came into full view in the third quarter of 2007 and the major financial meltdown in the wake of the Lehman failure in the latter part of 2008.
Low Leverage Growth Countries
The work seems to confirm what many may have suspected – that household balance sheets are a critical factor in determining broader economic trends and outcomes. And real estate is the primary cause of the blow-out of household debt in this instance.
Given the lesser role of consumption in more developing markets such as Asia, the importance of household balance sheets in macro-economic trends may be expected to be less significant. This work also now begs the question of “what next”.
Stock markets around the world have rallied strongly in 2009, fueled by massive liquidity injections pumped into the global economy by central banks, and central governments. These pumps of adrenalin have pushed asset markets, but have they really had the impacts on the real economy that policy makers expected?
In the west the evidence suggests that a great deal of the liquidity injections has NOT gone into underpinning consumption and credit for small businesses, but has been salted away as reserves in the banking system. Lending to “main street”, despite exhortations from policy makers has been anemic. In the US, bank lending to households and businesses has completely collapsed, with latest data showing lending contracting at an annualized rate of around 12%, the highest contraction since records are available.
Bank lending as a percentage of total assets are also at all time lows. The data for the UK is probably not too dissimilar. So where has all this money gone? Largely into cash on the balance sheets of the banks, where cash assets of banks in the US have grown by roughly fourfold since the Lehman failure.
But at the same time there has been some small growth in consumer spending in the US, despite record unemployment and no sign of new job creation, massive loss of household wealth through housing value declines and surging defaults and repossessions. David Rosenburg of Gluskin Sheff argues that the pattern of home-owners walking away from their homes and the debts they were servicing (or at least trying to service), has meant that this cash is now available for other uses. The taxpayer is of course footing the bill for this.
The prospects for a return to “normal” levels of consumption in many western countries looks remote in the short to medium term. And it is highly unlikely that we will be led out of recession by a surge in capital spending – excess capacity is here to stay for awhile with huge output gaps in many western countries. Labour force participation rates are at record lows, as are employment to labour force ratios.
Low Leverage Growth Countries
Households are deleveraging, and savings rates are rising. That all means less consumption, the driver of recovery in most recessions. The Glick and Lansing work shows quite clearly the relationship between the levels of household leverage and change in domestic consumption for many developed countries. The link between high leverage and consumption may well dictate, or at least strongly influence the pace and strength of economic recovery in the west. They conclude:
“Going forward, the efforts of households in many countries to reduce their elevated debt loads via increased saving could result in sluggish recoveries of consumer spending. Higher savings rates and correspondingly lower rates of domestic consumption growth would mean that a larger share of GDP growth would need to come from business investment, net exports, or government spending” – we are certainly seeing the latter! “Debt reduction might also be accomplished via various forms of default, such as real estate short sales, foreclosures and bankruptcies. But such deleveraging involves significant costs for consumers including tax liabilities on forgiven debt, legal fees and lower credit scores.”
All of this work paints a picture of a long, drawn out, slow recovery with the consumer unable to form its normal function of driving economies out of recession this time around.
- Atif Mian and Amir Sufi “Household Leverage and the Recession OF 2007-2009” University of Chicago and NBER, October 2009.
- Reuvin Glick and Kevin J. Lansing “Global Household Leverage, House Prices, and Consumption”, FRBSF Economic Letter, January 11, 2010.
- Reuvin Glick and Kevin J.Lansing “U.S. Household Deleveraging and Future Consumption Growth” FRBSF Economic Letter, May 15, 2009.
It’s Different in Asia! Or is it Really?
While the seeds of financial crisis in the west can be traced to a greater or lesser extent to consumer excess and the consumer living beyond his means via a never ending stream of debt, it is common currency to think that Asia is different. The popular perception of Asia as nations of savers rather than consumers is very much the norm. Asian consumers are imagined to be averse to debt, and seen to be, if anything, underleveraged. With respect to the property markets there is very much the view that “it cannot happen here” – Asian property markets are not riddled with piles of sub-prime mortgages, layer upon layer of collateralized mortgage obligations, derivatives on derivatives attached to housing or other property. This is largely true. Most lending to the property sector is very vanilla. Borrowers go to the bank for a property loan, and the bank lends, based on its own deposit base and based on its assessment of the borrower’s ability to pay. There is very little packaging and on-selling of mortgages in Asian markets. Moreover, most banks restrict lending to a 70% or perhaps 80% loan to value ratio. In some markets “top –up” loans may be available from other sources, but tend to be expensive. It appears as a very balanced and traditional old school bank lending model.
But the image of the Asian consumer as being lowly geared is not entirely born out by the facts. We have attempted to cobble together some data on household debt and income to compare with that shown above in the work of Glick and Lansing and that of Mian and Sufi. This is not as easy as it might sound as data is not as readily available as it is in many western countries. However, our numbers are shown in Figures 14-17 along with our sources.
The popular image of Asian households being lowly geared is not entirely born out by the facts. Household debt/income ratios in Korea and Singapore are up there with many western societies but is very much lower in places such as Hong Kong, Japan and Thailand. Australia and New Zealand also stand out in the highly leveraged household stakes. The countries with higher debt/household income ratios have also seen biggest rises in household debt/ GDP ratios.
The first point to note is that household debt to income ratios have shown very different patterns across the main countries in the region over the past decade. Debt/income ratios have risen very sharply in countries like Korea, less so in Malaysia but very sharply in Australia and New Zealand. These latter two countries have more or less mimicked other Anglo Saxon based economies on the other side of the world.
In Korea’s case, this rise in household debt levels I believe was driven largely by government policy in the wake of the Asian financial crisis that actively promoted a culture of domestic consumption to fill the gap in the domestic economy from a decline in exports that came out of Asia’s economic decline at the time. A very central part of that policy was to promote lending to the housing sector, which in turn led to a sharp rise in property prices in the country.
Notable by contrast is the experience of Hong Kong, Japan, Singapore and Thailand where household debt ratios have remained flat over the period.
But the debt ratios in some countries in the region are as high, or even higher than they are in many western nations where there is much hand-wringing over the levels of household debt in relation to income. While I cannot vouch that the data is provided on the same basis as the numbers for western nations, it is nonetheless clear that debt ratios in most of our survey universe are as high, or higher than most western nations. Only Hong Kong, Japan and Thailand have markedly lower household debt ratios.
A slightly more encouraging picture is shown when we look at household debt to GDP ratios. But again Australia and New Zealand stand head and shoulders above the rest in terms of consumer indebtedness.
Although the data is not easily available, it is probable that a very high proportion of household debt is attached to housing. Credit card usage and credit card debt, as well as other forms of consumer debt such as hire purchase, auto loans etc. are much less evident in Asia than in many western societies.
The takeaway here is that we should perhaps be a little more cautious about spouting the “Asia is all right” mantra, something that is very much a common theme in the region. There are pockets of high consumer leverage in the region that could conceivably pose problems for the financial systems in those countries. A major downturn in property values could be expected to produce some downturn in consumer spending but should not be anything like as significant as we are seeing in many western nations.
First, there has been little evidence of refinancing of housing as a prop for consumer spending in Asia. Second, given the low loan to value ratios, there is a greater cushion of equity in the housing markets, meaning less risk of default and foreclosure. Third, consumption, in any event, is a smaller proportion of the economy than is the case in most western societies.
It is probably fair to conclude that Asia IS different, but it should be sobering to note that household debt levels in many Asian countries are high and have risen, just as in the west in the past decade or so. However, the link between household debt, consumption and GDP is probably not as tight as it is proving in many western countries. But the experience of the past couple of years should provide a valuable lesson for Asian central banks and policy makers to encourage restraint on the growth of consumer lending, and prudent lending practices on the part of banks. Let’s hope those lessons are well learned.
Most importantly it is noteworthy that many Asian countries that have seen rapid rebounds in their property markets in the past year, following global injections of liquidity post Lehman’s failure, have sounded warnings of property bubbles. Several countries have reacted to the sharp recoveries in property markets by tightening liquidity, and tightening rules surrounding property lending and property markets generally.
It is a pity the likes of Greenspan, Bernanke, King and Brown (and others) did not take a leaf out of the Asian policy makers’ books!
A Note to Stansberry – Quick Thought on China
A couple of days ago I received and email from my friends at Stansberry Research asking my view of Jim Chanos’s by now famous prognostications on the China economy and property market.
As an aside, I was not able to open the attachment contained in the email as the website proclaimed it is accessible only to people inside the USA. I found this a little curious given the loud huffing and puffing we hear constantly from Americans about censorship of the internet by China’s authorities (and others). This example of apparent censorship is perhaps yet another example of the double standards that plague western thinking!
Jim Chanos has been widely noted recently for his view that China is in the midst of a massive bubble, particularly in its property markets, and it is all about to end very badly. A huge property crash, financial market crash and economic decline is going to be the result of the excesses that he sees today.
Anyway, here is the brief note I penned back to the folks at Stansberry Research on the China property market.
Dear Porter, Steve and team,
Yes, China property is entering a bubble phase, once again, fueled by massive state directed lending that was force-fed into the economy last year in the wake of the global financial crisis. Some US$1.3 trillion of NEW lending was poured into the economy last year, and of course massive amounts of that went into real estate in one form or other. Some estimates suggests as much as 20% – 25% went into real estate in various forms either on the supply side or demand side. A huge surge of new building has emerged with all major developers loading up afresh on land banks and beginning construction on new buildings at an even more rapid rate than before. Most China property companies listed in HK have increased construction levels by 50% – 100%, with sales (and profits) surging around 40% for 2010.
New Lending continued at a very rapid clip again in Q1 of this year with central government still showing anxiety about a possibly slowing economy, and demonstrating a desire to maintain loose monetary policy. Although policy is to slow lending down from last year’s massive levels, the central authorities still envisage total lending very much higher than in the years leading up to the financial crisis. Officials at the People’s Bank of China have been sounding warnings signals for some months now about the potential fallout from the flood of new lending. They are right to do so.
Some 8000+ new companies have emerged from the country’s municipalities and entered the property market. Municipal organisations are not allowed to borrow from banks so they have set up these companies to do property development – and that is where huge amounts of the debt has gone.
FUNDAMENTAL FLAW: There is a fundamental flaw in China property investment right now. For major cities such as Beijing and Shanghai, Guangzhou, Shenzhen and even many tier 2 cities, commercial floorspace vacancies are rising rapidly, and rentals are either flat or falling. But prices in many locations are continuing to go up at double digit rates. (See Figure 18-19) THIS FALLING YIELD TREND IS TOTALLY UNSUSTAINABLE – either rentals have to rise (which is highly unlikely given the massive 30% vacancy rates in some cities and very extensive new supply still coming on stream) OR PRICES NEED TO COME DOWN – the more likely scenario in my view.
Not only are gross rental yields lower than the cost of debt capital in these cities, there is very significant tax leakage of rental income. I see, for those companies that actually do pay their taxes etc, (and many I believe do not do so via various dubious practices), for every $100 of total rent received, only about $45 – $50 actually flows to the operating profit line. And then the company still has to pay its own operating and management costs and any possible profits tax. So while gross yield might be quoted as say 5%, net operating yield is more like 2.5% – maybe 3%. This does not adequately reward investors for their risk, and there is PLENTY of that. Regulatory risk is very high. Governments at various levels change the rules of the game on almost a daily basis, making the whole real estate investment game fraught with uncertainty.
DEBT: Most of the China property companies listed in HK have debt/equity ratios of around 50% – 90% – high but not totally off the charts. The HK based listed property companies have much lower gearing of around 30% or less – the lowest for any property sector anywhere in the world. But most China property companies are developers with a build and sell business model – they have very little investment property. But someone has to own all that commercial real estate – a great deal of it is in the hands of State owned Enterprises (SOE’s), or municipal organisations, state organisations, provincial, military affiliates etc. No one knows the gearing levels of these organisations, but I suspect it is typically 100% – 200%. They cannot help themselves!
This whole property situation is a house of cards where we will see a very sharp correction in commercial property prices in major cities, sharp falls in rentals, and a very substantial new round of NPL’s at the banks. This happened in the mid – late 1990’s. It is just a case of when, and how big the pack of cards actually is.
For example, in Shanghai in 1994, prime office rentals were around US$8- $9/sq ft/month. By 2000, following a massive building boom which produced a 73% vacancy rate in Pudong and about 45% for the entire city, prime rentals fell to around US$1.30 – $1.50. They have only recovered during this decade to around $3.50 – $3.90 – much lower still than in 1994, – and are now heading south again as vacancy rates rise. This must put pressure of balance sheets of the developers/owners of these buildings.
RENTALS ARE THE TRUE MEASURE OF VALUE OF REAL ESTATE as they more truly reflect underlying forces of supply and demand for the asset for genuine occupation. Capital values often do not reflect earnings potential of buildings, but the future price expectations of a buyer who has access to cash – usually someone elses! In other words, speculative forces.
Nothing wrong with that, but when it is done with borrowed money, it can pose big risks to a country’s financial system, and its broader economy. Residential rentals in major cities also are flat to down given the very substantial amounts of new supply.
DEBT LIKELY MUCH GREATER THAN HEADLINE NUMBERS. I suspect that real estate related debt is way higher than one might be told from bank sources. Firstly, a great deal of loans taken out by state enterprises and quasi state organisations I strongly suspect is not recorded accurately. Second, a great amount of loans are probably recorded in the banks as “industrial” loans, or “general working capital” loans – but have actually been applied to land and property. That was certainly the case in other parts of Asia in the late 1990’s financial crisis.
CENTRAL AUTHORITIES clearly see the dangers of the current property lending bubble and the huge construction boom and speculative buying it is fueling. Already a series of measures have been introduced to curb borrowing and construction, as well as curbs on consumers who are borrowing for property investment/speculation purposes. As in the last cycle when the authorities did this, they kept introducing new measures until the market cracked. Problem is that it is very difficult to fine tune the market such that measures simply lead to a slowing of the property price cycle. The danger is that it leads to a sharp correction as happened last time around. Government does not want to inflict a major downturn on the market – it knows what that will do to the banks. But my guess is that a sharp downturn will occur anyway given the huge amounts of new supply, rising vacancies etc.
More Yield Mismatch
RESIDENTIAL MARKET: Prices are up 60% plus certain in parts of some cities since the early part of last year, and nationally, average prices are up about 11.7% according to government’s own figures for 70 cities. (See Figure 20-23) This is the highest rate ever for this data series, and disguises very much higher increases in certain cities, and certain property types. This is very definitely a bubble. I am hearing stories that some 60% of some sales of new projects are being bought by “investors”, not for self occupation. Average flat prices for “mass” housing are 8 – 12 times average income, and rising sharply. Some places stand as high as 20 times average income. This suggests that affordability is well stretched. However, for many sales of new releases, very high proportions of buyers are paying cash.
There is very strong genuine demand for housing, that is for sure, but there is an increasing wave of speculative buying with large numbers of people buying second and third homes for “investment” purposes.
A raft of measures has been introduced to slow down lending to buyers who are getting into a second or third home. It may slow down the pace of sales, but has had precious little impact yet.
Residential yields for good quality properties have fallen from around 12% – 14% in 2002/03 in some major cities to less than 5% today in good locations in major cities.
Even CEO’s and Chairmen of major companies that I know are predicting a fall in property prices at some point over the coming year or two, but all are still building out their land-banks. Like always, they all think that they will be able to dodge the bullets.
What we do not know is what the reaction of consumers will be to a major fall in property prices – will they continue to service the mortgages? They do in HK even when prices fell 60% – 70% – but in the iron rice bowl culture of China this might prove different.
In any event, banks are going to take a caning, certainly from NPL’s in the developer community, and possibly in the mortgage scene. It could easily be to the tune of US$200 billion. There is not the level of mortgage securitisation that we saw in the US, but there is some. Insurance companies and pension funds have bought some portfolios, but do not think it is a huge amount. Banks are already signaling the need to raise capital – and that is just to support existing loan portfolios and keep desired capital ratios. Those needs will rise sharply again once the NPL’s start coming through. That is probably a 1 – 2 year time frame event I would guess.
Although the scale of impacts for the property market and banks could be quite large, it is still quite tough to accurately pin down a reasonable estimate of the actual size of the property finances hole that is likely over the coming couple of years or so.
The China-blinkers-on school of thinking, of which there are many believers, still assert there is no bubble, and that there is no significant financial risk to the economy or the banks. I simply don’t buy it.
There will be fall-out in HK also, as some of that liquidity has rubbed off on to that market. Banks however, consumers and corporates are much better positioned though with low levels of debt amongst the property companies, and relatively modest consumer debt. There has been nothing of the refi’s, the no income loans, credit card debt etc that we saw in many western societies.
I could go on at great length on the subject…..
As a follow-on from that comment, there has been announcement of yet another round of government measures aimed at cooling the China property market and bank lending to the sector. It is highly likely that volumes of property sales will come off sharply in the coming weeks, and I would not be surprised to see developers offering incentives and cutting prices to move inventory. There is likely to be a substantial new wave of supply hitting the market for sale in the second half of this year, and first half 2011, reflecting the great increase in lending in 2009 to the developer sector, that in turn produced a wave of new construction starting in the second half of 2009.
China’s authorities are demonstrating more prescience in addressing the debt fueled property bubble than their counterparts in the US and other western societies did. But there again, the bubble is directly of their own making, fueled by US$1.3 trillion of new state directed lending that was force fed into the economy during 2009 in the wake of the global financial crisis.
There is no doubt that genuine demand for housing in China’s cities is huge with rural-urban migration of around 20 million people per year. There is also great need for upgrading of poor quality housing of existing urban residents. But price increases of 20% – 60% in some cities and very substantial debt fueled speculative activity should send some very clear warnings signals of overheating. Affordability is also being stretched well beyond “normal” sustainable levels. On top of that, the market is about to see a significant surge of new supply coming up for sale.
At least the Chinese authorities recognize the signs and are committed to doing something about it – lessons for the likes of Greenspan/Bernanke in the US and Gordon Brown in the UK who oversaw the great lending binge and housing bubble in that country as Chancellor of the Exchequer.
China property stocks listed in Hong Kong have come off significantly since the government started signaling its intentions to rein in the property market. (See Figure 28). Stocks are now trading at substantial discounts to NAV, and many are now trading at single digit forward earnings multiples. However, just as last year saw investors and analysts sharply raising earnings forecasts as prices picked up and volumes surged, this year they will likely be doing the opposite, but as yet I am not seeing big downgrades from the analyst community. Those PE’s might not look so cheap as sales and prices get revised downwards – though much of the 2010 completions have already been sold. Net Asset Values (NAV’s) are also likely to be revised downwards as prices correct.
It is still too early to increase weightings in the China property sector. Remain underweight near term. Hong Kong property stocks are also likely to be lack lustre performers near term given the prospects in China and the Hong Kong Governments own efforts to put the brakes on Hong Kong’s buoyant property markets. Stay underweight developers, but property investment companies focused in the office and retail sectors should fare better.