Economics 15/06/2010: Negative equity 3

Here is the third and last post in the series on negative equity based on my TCD speech (see here).

Rising negative equity has implications for financial stability:

Domestic mortgage lending by the major banks represents over x5 times their core Tier 1 capital in the UK and roughly 10 times in Ireland. Even post disposal of its assets (assuming rosy valuations), AIB’s multiple will be over x11 of its risk-weighted assets. BofI – x7-8. And these are the better ones of the Irish banking lot. In addition, around 40% of all outstanding UK mortgage debt has been used to back securities.

Large losses on these mortgage loans and associated securities can erode banks’ capital positions, affecting both lenders’ willingness and ability to lend and, in extreme cases, their solvency.

Both of the above effects can have implications for aggregate demand and the supply capacity of the economy, highlighting the interdependency of financial stability and monetary policy. Again, in the case of Ireland, the two effects are reinforced by the large exposure of the Exchequer to banks balancesheets and to the property markets.

The defining feature of the materialised losses, and their associated economic effects, is the value of debt at risk (loss given default) and the coincidence of that with the probability of default.

Rising negative equity has implications for loans default probabilities:

In economic literature, negative equity is a necessary condition for default to occur since borrowers with positive equity can sell their house and use part of the proceeds to pay off their mortgage. Transaction costs (high in Ireland) and transaction lags (also extremely long in Ireland) act to further increase this effect. For example, a household with positive equity of ca 10% will still trigger a partial default on the mortgage if it takes a year to close the sale (8% funding opportunity cost per annum) and if closing costs add up to 2% of the sale price.

However, negative equity is not a sufficient condition for default to occur, as discussed in “Negative Equity and Foreclosure: Theory and Evidence” by C. Foote, K. Gerardi and P. Willen (June 05, 2008, FRB Boston Policy Discussion Paper No 08-03). Similarly, in the UK, May and Tudela (2005) find no evidence that negative equity increased the likelihood of a household experiencing mortgage payments problems between 1994 and 2002, although their sample does extend over a lengthy period of robust economic growth and rising incomes. The latter two aspects of the sample are not present in today’s Ireland.

But, as Heldebrandt, Kawar and Waldron (2009) highlight, “if a household is experiencing difficulties meeting their mortgage payments, negative equity can increase the probability of default by reducing the household’s ability to make payments by preventing equity withdrawals.” Benito (2007) found that households are more likely to withdraw equity from their homes if they have experienced a financial shock. Negative equity can affect a household’s ability to do that because of credit constraints.

Furthermore, negative equity can increase the probability of default by reducing household’s willingness to make mortgage payments, since defaulting can reduce the debt burden of the household. In Ireland, despite our anachronistic bankruptcy laws, this option is still available for anyone willing to emigrate. You might as well call this www.book-your-one-way-ticket.ie effect, as households leaving Ireland fleeing bankruptcy will have:

1. a very strong incentive to emigrate; and
2. a very strong incentive never to return for the fear of debt jail.

Negative equity may significantly increase the probability of default of buy-to-let mortgages over and above that of owner-occupiers as costs of defaulting on a buy-to-let mortgage may be lower because defaulting does not lead to loss of residence. In addition, buy-to-let mortgages are, at least in some cases, registered via businesses, implying no recourse on family homes and wealth.

Overall, available economic evidence does suggest that negative equity plays a significant role in mortgage defaults:

• Coles (1992) presents results from a 1991 survey of lenders in which a high LTV ratio was frequently noted as an important characteristic of borrowers falling behind in meeting their mortgage payments.
• Brookes, Dicks & Pradhan (1994) and Whitley, Windram & Cox (2004) find that a reduction in the aggregate housing equity in the UK was associated with an increase in arrears.

Overall, Heldebrandt, Kawar and Waldron (2009) conclude that “evidence suggests that the level of household defaults, and the impact of negative equity on financial stability, is likely to depend on conditions in the broader macroeconomic environment”. And Ireland is at a clear disadvantage since the combined macroeconomic shocks (decline in GDP/GNP, rising unemployment and contraction in private sector credit) are much more severe here.

Rising negative equity has implications for the size of the expected banks losses:

When bank borrowers face negative equity, as probability of default and the value-at-risk in default rise, banks have an incentive to stave off the actual mortgage default. To do this, banks engage in:

• Renegotiations of covenants (loans extension, provision of a grace period, interest only repayments etc)
• In extreme cases – joint equity ownership in asset (though banks will usually engage in this type of transactions under regulatory duress)
  

All of these measures aim to put the borrower into a position to eventually repay the loan in full.

However, in cases where default in unavoidable, the loss to be realised by the bank on any given loan depends on the recovery that can be achieved if the borrower defaults. Negative equity, or positive equity that does not exceed the sum of the cost of carrying the loan during the sale, plus the cost of sale, will imply a net loss on the default. From the bank point of view, the problem is not in the actual level of individual defaults, but in the combined level (aggregate) of negative equity net of costs and recovery values.

Such net losses impact not only the actual mortgage book, but also securitised pools that can be held off balance sheet, as current negative equity puts at risk future revenue against which the book is securitised. The end result on mortgage backed securities side is to reduce the value of MBS asset and, as Heldebrandt, Kawar and Waldron (2009) point out this can induce a second order effect of changing risk perceptions and investors’ sentiment “regardless of the actual performance of any given portfolio of loans”.

Both types of losses lead the banks to record write downs on their mortgage books and securities held. If these are large enough, banks’ capital ratios will be reduced.

In the case of Ireland the problem is compounded as the banks are actively delaying recognition of losses on negative equity. These delays mean that banks are likely to pay elevated costs of external funding over longer period of time. In addition, these delays lead to losses compression – the situation where banks recognize significantly larger volumes of impaired assets later in the crisis cycle. Bunching together losses creates a much more dramatic investors’ loss of confidence in the bank.

In addition, negative equity-driven impairments, plus delayed recognition of such impairments lead to suboptimally high demand for capital from the banks. If this coincides with the period of severe credit crunch, monetary policy aimed at increasing economy-wide liquidity flows can become ineffective, as banks park added liquidity on their balance sheets, creating a liquidity trap. This is evident throughout the crisis, but by all possible monetary policy metrics, it is much more prevalent in Ireland, where even today credit available to the private sector continues to contract. Irish banks are hoarding liquidity and are raising lending margins to offset expected, but undisclosed writedowns. This problem is compounded by Nama which induces greater uncertainty onto banks balance sheets through its hardly transparent or timely operations.

Negative equity and generational asset gap:

In Ireland, the problem of negative equity is further compounded by the generational spread of negative equity to predominantly younger, more productive and more mobile (absent negative equity) households. These households today face higher probability of unemployment (thanks to our unions-instituted and supported ‘last in – first out’ labour market policies). They also much deeper extent of the negative equity because of higher cost of financing their original mortgages and entering the housing market.

The generational effects of negative equity are compounded by geographic distribution of the phenomena – with younger households being more likely to reside in the areas of excess supply of new housing, with poor access to alternate jobs, should they experience unemployment.

Finally, it is the younger households that are subject to twin effects of higher probability (and deeper extent) of negative equity and depleted savings (due to high cost of entry into the housing market). This implies that it is the very same households which have the greatest incentive to engage in precautionary savings motive.


So traditionally, economies grow by:

• encouraging the young to acquire new assets (invest and save); and
• encouraging the old to consume (divest out of savings).

With negative equity, Nama and pressured banks margins, Ireland is:

• forcing the better (more productive today and in the future) young to emigrate;
• keeping the remaining young deep in the negative equity (neither capable of investing in the future, nor of moving to find a lost job today);
• underwriting - at the expense of younger, tax paying generations - continued excessive provision of pensions to retired public sector workers;
  
• forcing younger families to cut deeper and deeper into their children education budgets and own training and education funds in order to assure they continue paying on the asset that will never have net positive return on investment; and
• incentivising the old to remain in their highly priced (if only rapidly losing value) homes backed by slacked consumption due to inability to monetize their pensions savings.

In what economics book is this scenario better than any moral hazard problem that can be incurred in the short run by reforming our bankruptcy system to an American-styled 'restart' button?

Economics 15/06/2010: Negative equity 2

This is the second post of three consecutive posts on the effects of negative equity in Ireland.

Negative equity can lead to a reduction in consumer spending, collateral & credit

This can take place via four main pathways:

1. Housing equity can be used as collateral to obtain a secured loan on more favourable terms than a loan which is unsecured. This channel for lower cost financing is cancelled out by the negative equity.
2. Falling collateral values may also affect the cost of servicing existing mortgages if borrowers have to refinance at higher interest rates when their existing deals expire (eg when exiting temporarily fixed-rate or tracker mortgages). That would reduce income available for consumption, which may further reduce demand.
3. Households on adjustable rate mortgages are facing additional pressure of higher banks margins. Since vintages of many ARMs are coincident with 2005-2007 period, negative equity has direct and significant impact on them. Nama exacerbates this impact by forcing banks to up their margins on performing loans, pushing more and more households into not just negative equity, but virtual insolvency.
4. Fourth, falling values of housing equity also reduce the resources that homeowners have available to draw on to sustain their spending in the event of an unexpected loss of income (eg due to redundancy, illness or a birth of a child). By reducing the value of housing equity, falling house prices may lead some homeowners to seek to rebuild their balances of precautionary saving at the expense of consumer spending and investment.
   

Note that precautionary savings are held in highly liquid demand deposits – a fact that I will use below. In general, households with high amounts of housing equity may not respond much to falling house prices, because their demand for precautionary savings balances may already be satisfied through their positive net worth balances on the house. Households with low or negative equity have an asymmetrically stronger incentive to save in a form of short-term deposits.

Rising negative equity can also result in a reduced supply of credit to the economy as a whole:

Negative equity can raise the loss that lenders would incur in the event of default (loss given default) and the probability of a loss. That can make banks less willing or able to supply credit to households and firms.

Per Benford and Nier (2007) Basel II regulations, which require banks to hold more capital against existing loans when their anticipated loss given default rises, can reinforce that.

If credit is more costly or difficult to obtain, households and firms are likely to borrow less, leading to lower demand through lower consumer spending and investment. This, in turn, can lead to reduced business investment.

Expectation hypothesis suggests that negative equity effects on willingness to borrow and lend can extend beyond those immediately impacted as other households anticipate their own asset value decline toward negative equity.

Blanchflower and Oswald (1998) paper showed that a reduction in credit availability may also have some effect on the supply capacity of the economy by reducing working capital for smaller businesses and the capital available for small business start-ups. In addition, a recent (June 2009) paper “Reduced entrepreneurship: Household wealth and entrepreneurship: is there a link?” by Silvia Magri, Banca d’Italia, published by the Bank of Italy (Working paper Number 719 - June 2009) shows that negative house equity can result in reduced entrepreneurship, as many new businesses are launched on the back of borrowing secured against primary residencies or other real estate assets.

Rising negative equity can also result in a reduced household mobility:

Negative equity can affect household mobility by discouraging or restricting households from moving house. Two fathers of behavioural economics, Tversky and Kahneman (1991) argued that households may be reluctant to move because they would not wish to realise a loss on their house. Notice, that our so-called ‘smart’ politicians often claim that negative equity is never a problem unless someone wants to move. Actually, it is a problem even if someone does not want to move, but has to move because of their changed employment or family circumstances.

Tatch (2009) shows that a household in negative equity would be unable to move if they were unable to repay their existing mortgage and meet any down payment requirements for a new mortgage on a different house. This is even more pronounced in Ireland due to the nature of Irish bankruptcy laws.

Hanley (1998) shows that the effect of negative equity on mobility were quantitatively significant during the early 1990s in the UK. The paper estimates that of those in negative equity in the early 1990s, twice as many would have moved had they not been in negative equity. The paper argues that reduced household mobility leads to a reduction in the supply capacity of the economy by increasing structural unemployment and reducing productivity.

Reduced household mobility implies a reduction in the number of households moving home. This can have adverse implications for tax receipts, spending on housing market services and certain types of durable goods as highlighted in Benito and Wood (2005). So as negative equity increases, tax revenues and economic activity in the housing sector and associated white goods sectors falls.

15/06/2010: Negative equity 1

Yesterday, I gave a speech at the Infinity Conference in TCD on the issue of negative equity (see newspaper report here). The following three posts (for the reasons of readers' sanity) reproduce the full speech.

What effects can negative equity have in the case of Ireland?

I did a troll of the literature on negative equity and below I summarize the main findings, relating some to the case of Ireland.

Broadly-speaking there are three dimensions through which negative equity can have an effect on Irish economy:

1. Macroeconomic channels via negative equity impact on aggregate supply and demand;
2. Monetary channels which lead to negative equity impacting adversely banks balance sheets and increasing the cost of default and probability of default for mortgage holders; and
3. Growth channels, which relate to the adverse effects of current negative equity on future demand and investment, and directly on growth.

Here are more detailed explanations of these channels.

Why the problem of negative equity is likely to be greater in Ireland than in the UK

A forthcoming paper “House Price Shocks and Household Indebtedness in the United Kingdom” by Richard F. Disney (University of Nottingham), Sarah Bridges (University of Nottingham) and John Gathergood (affiliation unknown), to be published in Economica, Vol. 77, Issue 307, pp. 472-496, July 2010, used UK household panel data to explore the link between changes in house prices and household indebtedness. The study showed that borrowing-constrained by a lack of housing equity households make greater use of higher cost, higher risk unsecured debt (e.g. credit cards or personal loans). Crucially, when house prices revert to growth, “such households are more likely to refinance and to increase their indebtedness relative to unconstrained households”.

These effects – present in the case of the UK – are likely to be more pronounced in the case of Ireland, because Irish households which find themselves in negative equity today experience much severe deterioration in their net worth base due to the following factors:

• Majority of Irish households have been forced to front-load property taxes into their purchase costs and often mortgages. Thus average LTVs are more likely to be higher here in Ireland, for more recent mortgages vintages, than they were in the UK.
• Ireland has experienced a much more severe contraction in house values than the UK to date.
• Because of significantly higher entry-level taxes, younger buyers in Ireland had to be subsidized more heavily by their parents than their UK counterparts, implying that once true levels of indebtedness are factored in, real mortgages and debts held against a given property of more recent purchase vintage might be higher than those recorded on the official mortgage books.
  

In many cases – we do not know how many, but anecdotal evidence suggests quite a few – credit unions and building societies, as well as non-mortgage banks were engaged as sources of top up loans to younger buyers, implying that once again the true extent of house purchase-related debt in Ireland, for younger households, might be higher than official records on mortgages suggest.

Another recent study, titled “The Economics and Estimation of Negative Equity” by Tomas Hellebrandt, Sandhya Kawar and Matt Waldron (all Bank of England) published in Bank of England Quarterly Bulletin 2009 Q2 looked at the effects and extent of negative equity between Autumn of 2007 and the Spring of 2009. Over that period of time, nominal house prices fell by around 20% in the UK, suggesting that negative equity impacted around 7%-11% of UK owner-occupier mortgage holders by the Spring of 2009.

By now, in Ireland:

• house prices fell down ca50% already (accounting for the swings in terms of premium to discount on asking prices – by closer to 55%),
• vintage of many purchases was much closer to the peak valuations, so
  

for Ireland, estimated negative equity impact is now around 35-40% of the mortgage holders.

Extent of negative equity here is compounded by:

1. High entry costs into the homeownership (100+% mortgages due to stamp duty costs and poor quality of real estate stock);
2. Lax lending – cross-lending by banks and credit unions and building societies;
3. Hidden nature of some of borrowing – parents’ top ups etc;
4. Coincident borrowing – with younger households being more likely to engage in borrowing for a mortgage, while borrowing for car purchase etc.
   

BofI, which holds ca 25% of all mortgages in the country (about 190,000) has reported that of these, more than 20% were already in negative equity (over 40,000) around the beginning of 2010.

The aforementioned Bank of England paper provides a good starting point for outlining the set of adverse impacts that negative equity can have on the Irish economy.

Negative equity can have implications for monetary policy:

A rising incidence of negative equity is often associated with weak aggregate demand as households in negative equity are more likely to cut their expenditures across two channels:

• due to reduced marginal propensity to consume out of wealth; and
• due to increased marginal propensity to save.

The direction of causation is not always obvious, implying a possibility of feedback loops – as households experience (or even anticipate) negative equity, they start reducing their borrowing against depreciating assets, the effect of which is amplified by the banks reduced willingness to lend against such assets. In addition, households rationally interpret these declines in today’s wealth as declines in future wealth, implying greater exposure to pensions under-provision in the future, plus greater exposure to the risk of sudden collapse in earnings (due to, say, unemployment or long term illness). As the result, these households tend to reduce their consumption today and in the future.

The reduced consumption leads to a loss of revenue to the exchequer and thus to additional pressures on future public pensions and benefits provision. This, in turn, leads households to further tighten their belts and attempt to compensate for the risk of reduced future benefits by lowering consumption exposures today.

Negative equity tends to become more prevalent when house prices fall, which usually reflects weak demand for housing, since housing supply is fixed in the short term. In the case of Ireland, this is compounded by the fact that we have severe oversupply of properties in the market. Demand effect, therefore, reinforces supply effect. Once again, in Ireland there is one more additional channel of induced market uncertainty due to Nama operations.

Weak housing demand often coincides with weak consumer demand in general, due to

1. reduced availability of credit to consumers and potential home buyers; and
2. precautionary savings as households respond to decline in their nominal wealth.

If negative equity leads to a further contraction in the availability of credit to both households and firms, as in Ireland – exacerbated in the case of Ireland by Nama – second order effects reinforce first order effects.

Lastly, as negative equity in Ireland is coincident with construction sector bust, we have twin effects of decreased households’ mobility and increased unemployment. This once more reinforces the uncertainty levels in the markets for housing, implying that risk-adjusted negative equity becomes even more pronounced here.

Economics 13/06/2010: Mapping Dublin's weight in economy

Here is an interesting set of data for some of the world's leading cities in terms of their contributions to country GDP and their share of total population. Sizes of the bubbles reflect the ratio of contribution to GDP to share of population. Greater Dublin is taken as per CSO: Dublin plus Mid-East region.
To me, this really does put into perspective the necessity for continued investment in the Greater Dublin region and the futility of our serial National Spatial Development Strategies.

We hear so much about the massive urbanization in the emerging economies, especially in the East. This process, in fact, is very much a reality. But what we do not hear about is urbanization of our own economic activity. The fact that Dublin stands out amongst the most urbanized zones, relative to the rest of the country, in the world is telling me that Ireland should focus more attention on developing the Greater Dublin region to reflect the reality of demand of the firms' and workers' for its location.

Updated: see the same chart with Dublin only:
Note: per CSO "The Mid East region (Kildare, Meath and Wicklow) and the Dublin region are affected by a substantial proportion of their workforce living in one region and commuting to work in another." That's as much of a 'definition' as we get to reflect the most likely fact that vast majority of commuters are to Dublin...

And here are the inputs that went into the above charts:
And the source for Dublin figures: National Accounts, regional incomes:
Notice that Dublin City's overall relative weight in the economy (ratio of share of GVA to share of population) rises relative to Greater Dublin, as expected, in line with the evident 'bedroom communities' nature of Mid-East region. Just look at the CSO figure for Mid-East share of GVA.

Economics 11/06/2010: What's going up might be also going down

Irish retail sales have surprised on the positive side, posting a 0.3% increase yoy for sales ex-motors in April 2010. Sounds impressive, especially considering this was the first yoy increase since March 2008, or over some 25 months now. But hold on to that thought of a recovery signal. Check out the charts:
Things are still very much up in the air as to whether retail sales are actually on a mend or not. The figures above plot seasonally adjusted series ex-motors. More importantly, sales in the categories that are correlated with overall household investment activities - household equipment (down 3.1% in value mom, and down 1.2% in volume mom), electrical goods (-3.2% in value mom and 2.3% in volume mom) and Furniture & Lighting (down 5% in value and 6.2% in volume) - all signal no growth in the core leading indicator of a recovery - improved domestic investment. Only Hardware, Paint & Glass category related to investment showed increases of 2% and 4.2% in value and volume in mom terms.

Interestingly, Ireland bucked the EU-wide trend in April:

Economics 11/06/2010: Private Sector Credit data

Central Bank data released yesterday show Private Sector (non-financials) credit fell 9.3% in Q1 2010, to €355 bn. Total outstanding mortgages volume fell €1.2 bn to €146.4 bn. Two thirds of the total amount of credit decline came from writedowns on existing loans, which means that there is continued pressure on loans (keep in mind that Nama transfers are not yet in the data). Charts below update, as usual:
Aggregates first
Notice the rise of securitizations - banks shifting stuff off their balance sheets at an aggressive rate.
Chart above shows monthly flows. There is some improvement here, but absent seasonality corrections it is hard to say what exactly is going on. However, it does appear that the latest monthly transactions uptick is not in line with pre-crisis dynamics for Non-Financial Corps, but in line for Households:Year on year changes:
Clearly, yoy things remain bleak, although the rate of contraction is getting reversed for households. This is a seasonally consistent result, so I would not be reading too deeply into it.