Saturday, January 19, 2008
Looking now at the latest construction output data released by Eurostat this week, we can see that, despite some volatility the trend is definitely down on a year on year basis - dropping in November by 6.2% when compared with November 2006.
The month by month chart reveals a similar picture, since the months with positive growth have been followed by stronger downward movements. We should also bear in mind that actual construction follows new sales only with a lag, as what has previously been bought still needs to be built. Plus we should anticipate a strong increase in civil engineering activity as the government tries to some extent to offset the impact of the fall in private residential activity, and especially prior to March's elections.
Perhaps the clearest indication of the presence and extent of the economic slowdown is offered by the December data for the EU Commission Eurozone Economic Sentiment Index.
As reported earlier in the month the Commission’s economic sentiment indicator for the entire eurozone is now on a clear downward path. At 104.7 in December, down from 104.8 in November, the index was at its lowest since March 2006.
But perhaps more important than the steady downward drift in the general indicator are the individual country differences, and in particular the sudden movement in Spain and Ireland since the summer.
Wednesday, January 09, 2008
Only last September, José Luis Rodríguez Zapatero, the Spanish prime minister, announced that Spain had joined the “Champions’ League” of world economies. Europe’s fifth largest economy was growing so robustly, and creating so many jobs, it would soon be richer than Germany in per capita terms, Mr Zapatero predicted. That euphoria was short-lived. December saw a spike in inflation, a rise in unemployment and a slowdown in the economy, as the international credit squeeze gripped Spain. The government recently lowered its estimate for economic growth in 2008 from 3.3 per cent to 3.1 per cent, a figure many economists consider is still too optimistic. Inflation last month of 4.3 per cent was at the highest level in more than a decade.
So here is some of the data the FT refer to and some they don't mention at this point. Interestingly they quote the branch manager of a bank in Igualada, a small town not too far from where I live in Barcelona:
"I’ve been a bank manager for 28 years and I have never lived through a situation as dramatic as this....House prices in this town have fallen by 20 per cent, there is no demand, and no mortgage finance. Savings banks have cut off funding. Before the credit crunch, I used to do 12 mortgages a month. Since August, my branch has approved only one new loan."
The Problem of the Cedulas
The picture presented by our regional branch manager more or less sums it all up. For an explanation of why there are no morgages available at this point, and how the credit crunch actually works out in practice see this post here. But also the mention of the regional cajas is significant because of the previous role of AYT Cedulas in the Spanish market. Spains rapid housing expansion was made possible by a process of pooling mortgages from a range of smaller savings banks into one arrangement, creating in the process the euro market’s biggest covered bonds player in the form of Ayt Cedulas. How it was that a small group of previously obscure local savings banks rose meteorically to become the European leader is a story which will be fascinating when someone finally gets to tell it, but for now the important point to note is the danger that this dramatic rise can become an equally dramatic decline. Basically these types of entity are having growing difficulty rolling over their funding, and selling their packages of cedulas hipotecarias, and undoubtedly this difficulty was one of the main factors which lead Jean Claude Trichet to make unlimited finance available to the Spanish banks at 4.21%. Basically, the squeeze on 3 month euribor rates has hit the banks (who have to continually refinance part of their needs and now have to pay nearly 50 base points more - a classic liquidity problem) while the clients - who often pay a mere 0.5% over 1 year euribor - are nothing like so affected. So it is big squeeze time on bank margins. And as my colleague Claus Vistesen points out in this post here, when its big problem started Northern Rock's balance sheet looked solid enough. It was just that a large part of its rolling cash flow requirement was financed through the interbank market, "otherwise there was nothing wrong". Sound familiar to anyone?
Some background information on the system of Cedulas Hipotecarias and how this works under Spanish Law can be found in my working notes on the topic - which are being updated as I learn more. Among the august and venerable names to be found on a March 2001 2.048 Billion Euro Cedula Hipotecaria Bond Offer were (participaing percentages in brackets) Caja de Ahorros Municipal de Burgos (14.67%), Caja España de Inversiones (14.67%), Caja de Ahorros del Mediterráneo (14.62%), Caja de Ahorros de Huelva y Sevilla (8.80%), Caja de Ahorros de las Baleares (7.34%), Caja de Ahorros de Castilla la Mancha (7.34%, Caja General de Ahorros de Granada (7.34%), Caixa D’Estalvis de Sabadell (7.34%),Caja de Ahorros de Córdoba (7.34%), Caixa D’Estalvis Laietana (4.40%), Caja General de Ahorros de Canarias (1.47%), Caixa D’Estalvis del Penedés (1.47%), Caja de Ahorros de La Rioja (1.47%), Caja de Ahorros Provincial de Guadalajara (1.17%), Caja de Ahorros y Préstamos de Carlet (0.59%).
The FT makes quite a fuss about the looming non-performing loans problem - citing real estate developers like Astroc and Llanera who have gone bust, and the rise of 48 per cent in NPLs over the past year (to €1bn) according to data fromthe Bank of Spain, or the fact that banks have €293bn in outstanding loans to property developers. But at the present moment this is not the point.
Fast Deceleration in the Real Economy in Spain
What is alarming about the rate of deceleration in Spain is the rate at which the real economy is slowing, that, and the fact that inflation is spiking when precisely we should expect the deceleration to be producing a downward pressure on prices. My own interpretation of this latter point is that Spain - after nearly 15 years without recession - is unaccustomed to the kind of adjustments which are normal in a society like the UK or the US which see recessions as regular phenomenon. So the reaction and adjustment time is slower in Spain, which might suggest that people are attempting in the short term to compensate for falling sales volumes by raising prices. But this of course won't work, and at some point prices will have to come down if they want to sell (remember that since spain is in the eurozone they cannot devalue to correct price distortions). So the correction process - and particularly if there is a sharp downward correction in house prices (which are reflected in the HICP) at some point, and if there is no further large increase in oil prices which is unlikely - could lead Spain into a downward ternd in price levels, or, if you prefer, price deflation, such as we have seen in Japan. This - if it happens - will produce enormous policy problems over at the ECB, since the eurosystem was never designed to facilitate massive monetary easing a la ZIRP in one individual economy. But then, perhaps we should wait till we get there to talk about this.
As I say, I think the problem posed by defaults still lies some way out in the distance. Only as the real economy sinks deeper, and people begin to realise that this is not temporary but long term, will Spains young people really start to throw the towel in in significant numbers. So I think that we should treat each problem a day at a time here, since I am sure there will be plenty of them.
Now the data. First off two Items the article doesn't mention. Retail slaes and the EU confidence index. European retail sales fell the most in at least 10 years last November declining by 1.4 percent from November 2006 - the biggest drop since at least 1997 when the data series starts - according to eurostat yesterday. More importantly for our present purposes (although of course Europe wide data is important for a Spain which now needs to export) is the fact that Spanish retail sales declined for the third month in succession.
The break that can be seen in both the year on year data and the month on month from the summer is particularly striking.
EU Commission Confidence Index Down All Round.
The general picture of an economic slowdown in the eurozone is once more highlighted by yesterday's realease of the December data for the EU Commission Eurozone Economic Sentiment Index.
The Commission’s economic sentiment indicator is on a clear downward path. At 104.7 in December, down from 104.8 in November, the index was at its lowest since March 2006.
But perhaps more important than the steady downward drift in the general indicator are the individual country differences.
Italy, as I have outlined in greater depth here, has been steadily drifting off towards its next recession since the middle of 2007, but the big news of the moment is what is happening in Spain and Ireland, since as is well known they were the two countries in the eurozone to be most affected by the "housing fever" boom, and if you look at the chart above the correction in these countries since September is striking in its velocity. As the FT also commented yesterday:
Spain is demonstrating that prospects could vary significantly across the eurozone. Since September, confidence in the Spanish construction sector has tumbled, and the Commission’s survey results showed sentiment deteriorated in December in Spanish industry, retailing and among consumers as well.
As the FT point out, unemployment is now trending up, and the underlying situation is reflected in the shifting pattern of employment and unemployment. In this context a comparison between 2006 and 2007 is pretty revealing. If we look at the chart below we can see that in the early months of this year the employment situation was generally up over 2006. Then the situation turned (around July), and since then it is "down hill all the way" unfortunately, with unemplyment rising (as it might well do for seasonal reasons anyway, but in every case the increase is significantly more pronounced than in 2006. Spain's ever productive labour market has, unfortunately, now turned.
As the FT also points out, inflation is on the way up againin Spain,and accelerated in December to the fastest pace since the euro was introduced in 1999. Consumer prices gained 4.3 percent from a year earlier when measured using the EU harmonised index, compared with a 4.1 percent increase in November, according to data from the National Statistics Institute.
Not surprisingly, in this environment consumer confidence is plummeting, touching historic lows for the third month in succession in December, according to the index compiled by the Instituto de Credito Oficial.
OK, since I have already repeated myself enough, and time is of the essence at this point in time, if you want a fuller explanation of some of the points being argued here, please go over to this post.
Tuesday, January 08, 2008
Perhaps the best thing to do is to start off with a nice simple chart showing the essential structure of the Spanish cedulas market. Basically I want to start off right from the outset by explaining that I am in no way suggesting that anything improper has taken place in the expansion of the cedulas market in Spain (ill advised, possibly, but not improper). A need for cheap, accesible and well secured mortgages was felt - this was in the interest of all the participants, from those who bought the flats, to those who lent the money, to the builders who got to do the building, and the town halls who saw a boom in income and local employment, to the large banks who got to sell many of the financial services - without, it must be said, having to assume too much risk - to the regional cajas, who saw a throughput of income and activity that they could never have imagined even in there wildest dreams. But this whole process was posited on one straightforward, simple, but as it turns out probably erroneous, idea, namely that property prices in Spain would rise in a stable and sustainable fashion, as they had done in Germany since 1995, which was the secret of the success of the German Pfandbriefe on which the cedulas model was based.
There were, however, only two underlying yet unnoticed problems associated with this situation: the first was that Germany had had its maximum (demographically speaking) construction boom in the years up to 1995, while the Pfandebriefe really took off following the "correction" in Germany and not before it, and secondly Germany never had negative real interest rates over a sustained period of time in the way which Spain did in the years from 2002 to 2006.
Over the past decade covered bonds, or securities issued by financial institutions that are secured by dedicated collateral, have become one of the largest asset classes in the European bond market and an important source of finance for mortgage lending. The collateral, or “cover pool”, is usually put together so as to obtain the highest possible triple-A credit rating. As a consequence, covered bonds offer an alternative to developed country government securities for bond investors interested in only the most highly rated securities.
The defining feature of covered bonds is the dual nature of protection offered to investors. Covered bonds are issued by financial institutions, mostly banks, which are liable for their repayment. They are also backed by a special pool of collateral – mostly high-grade mortgages or loans to the public sector – on which investors have a priority claim.
The dual nature of protection offered by covered bonds sets them apart from both senior unsecured debt and asset-backed securities (ABSs). The fact that they are secured by a collateral pool in addition to the issuer’s creditworthiness results in a higher rating than “plain vanilla” bank bonds. In contrast to ABSs, the cover pool serves mainly as credit enhancement and not as a means to obtain exposure to the underlying assets. Cover pools tend to be dynamic in the sense that issuers are allowed to replace assets that have either lost some quality or have been repaid early.
The key question when valuing covered bonds is whether or not the cover pool will retain its value in the event of the bankruptcy of the originator. In principle, the insolvency of the originator could endanger the creditworthiness of covered bonds through two channels. First, the credit quality of the assets in the cover pool could deteriorate. Second, even if the cover assets retain their value, creditors of the originator could attempt to seize these assets in order to satisfy their claims. The covered bond legislation and contractual arrangements in place attempt to deal with both threats to the viability of the cover pool by imposing minimum standards for asset quality and by ensuring the bankruptcy remoteness of the cover pool.
Legislative frameworks tend to apply limits on the loan-to-value ratio (LTV) of mortgage loans as well as geographical and, in some cases, rating restrictions for public entities to ensure a high quality of the cover assets. These are sometimes complemented by mandatory stress tests. Such tests are also used by rating agencies to ensure the creditworthiness of the cover pool of bonds issued both inside and outside a legislative framework.
To date, covered bond systems are in place in 25 European countries. In most European countries, the issuance of covered bonds is regulated by specific covered bond legislation setting the standards for bondholder protection, the criteria for eligible assets and a number of other specific features. In some countries, like the UK and the Netherlands, the features of covered bonds are defined on the basis of contractual arrangements.
Spanish cedulas hipotecarias are a specific form of covered bond. In fact covered bonds are less complicated than they might seem to be at first sight and have more common characteristics than might be expected. The European Covered Bond Council (ECBC) has established a Technical Issues working group, where experts from a wide variety of countries work together to carry out a continuous comparative analysis of the different covered bond systems as they evolve. Achieving a common basis of understanding of covered bonds can help identify areas where regulators, issuers, intermediaries and investors can work together to enhance and advance the various covered bond systems already in place.
There seem to be three classes of common characteristics for covered bonds:
The status of the issuer: In all participating countries covered bond issuers are regulated institutions. These can be universal credit institutions (with or without a special covered bond issuance licence), specialised credit institutions or specialised financial institutions. The fact that issuers are regulated allows public supervisory authorities to monitor them closely and to steer the business activities of covered bond issuers. Furthermore, this kind of regulation should normally give investors some sense of security regarding the safety of covered bonds.
Exemption from general insolvency law: The full repayment of covered bonds should not be affected by insolvency procedures on the issuer. Moreover, covered bondholders should be protected against claims from other creditors in case of insolvency of the issuer. Those features establish the preferential claim of the covered bondholders and constitute the basis for the bankruptcy remoteness of covered bonds. These features can be found in all countries where covered bonds are issued.
Adequate cover: The covered bond system should be based on the notion of a cover pool and be operated in a way which ensures that the cover pool generates sufficient proceeds to repay covered bond interest and principal. This requirement is essentially an economic question and involves a variety of features such as the type of eligible assets, valuation of cover assets and Loan to Value (LTV) criteria, asset-liability guidelines, the role and position of derivatives and, last but not least, cover pool monitoring and supervision.
The need to find some cheap and efficient solution for the problem of mortgage funding has been specially important in Spain since large quantities of mortgage loans have been sought and issued. The problem has been that due to the presence of a negative real interest rate environment over an extended period of time bank deposits were not sufficient to meet ongoing demand. As a consequence Spain needed to find an effective supply of mortgage funding, in order to have sufficient cash available to lend.
The Spanish mortgage market law (Ley del Mercado Hipotecario 1981, LMH) served as the legal framework which made the creation of this market possible. According to article 47 of the Spanish Constitution (1978), a constitutional obligation was established of facilitating access to a dwelling for all citizens. The only real and evident downside to all of this was that the more effective and abundant the supply of mortgage funding became, and the cheaper the mortgage loans were the mortgage debtors. Logically this meant that for any given purchaser the quantity they could raise to pay for a house or flat on any given monthly quota went up and up, and with this the underlying value of the asset (or flat) they were trying to buy.
The process of mortgage funding differs from one country to another. Traditionally bank retail deposits have been used for this purpose (as they had been in Spain), but if what the credit institutions want (or need) is a medium or a long term source of sure cash, then one way to achieve this is to structure some kind of passive lending operation, based on a mortgage securities market. There are two great systems of mortgage securities markets today: the German model (Pfandbrief) and the Anglo-Saxon model (mortgage-backed securities), each of these has special characteristics, and its own advantages and disadvantages.
Both mortgage bonds and mortgage-backed securities have special features in Spain that make them unique in the European mortgage securities market (although please note that some at least of the longer term singularities of the Spanish situation have been resolved in the new mortgage legistlation of the autumn of 2007). The structure of the Spanish cédulas hipotecarias is rooted in the original LMH (Ley del Mercado Hipotecario, 1981) and follows basically the structure of the German model of the Pfandbrief, as can be seen in the chart below (which is a simpler form of the general chart I presented above).
Basically a land credit institution (in the Spanish case this was very often a consortium of regional cajas) concedes a number of mortgage loans that create a pool of mortgages with specific features. This pool of mortgages may be funded by issuing mortgage bonds whose value cannot exceed the 90% of the value of all the mortgages of the pool (and here of course we meet our first problem, since if property values drop in Spain, then naturally the value of any given pool will drop, and it will need "topping up"). The mortgage bonds represent debt to the issuing credit institution. The owners of the mortgage bonds (not only institutional but also retail) receive interest periodically until the security becomes amortised. The global amount of outstanding mortgage bonds is backed by all the mortgages owned by the issuer.
In general, there are two great Anglo-Saxon models to structure the securitisation of mortgages:
a) The pass-through securitisation model (USA): in this model, the MBS holders have a property right over the pool of mortgages, because their capital and interests pass through the mortgage debtor directly to the securities holders. In this model the trustee of the issue is at the same time the SPV.
b) The pay through securitisation model (USA and UK with the UK-mortgage-backed securities): in this model, the MBS holders have only a credit right against the pool of mortgages and the security of this credit (either a charge (USA) or a sub-mortgage (UK))is held in their benefit by the trustee of the issue.
The Spanish model is a hybridization of both models and it adapts their structure to the Spanish civil and commercial law.
The Spanish mortgage-backed securities do not securitise the mortgages directly as they are issued on their basis. Before the issue of MBS it is necessary to “participate” the mortgages, which means basically issuing another kind of mortgage security, called “mortgage participations” (participaciones hipotecarias), and these represent a percentage of each participated mortgage. This percentage determines the amount of each mortgage that may be securitised afterwards. The mortgage participations are also securities, and the holder of these becomes a co-mortgage creditor (co-mortgagee) along with the originator of mortgage. The mortgage participation holder receives the principal and interest corresponding to the percentage of the “participated” mortgage from the issuer (originator of mortgages), who has received them from the mortgage debtor (borrower of the mortgage loan). It is a true cession of a mortgage loan, which means that the mortgage participation holder may claim not only against the issuer of the mortgage participation (the land credit institution) but also against the mortgagor -against whom the mortgage participation holder may enforce the mortgage- depending on who has defaulted.
With the issue of the mortgage participations, the mortgage securitisation process has only begun and it does not end until the formation of a pool of mortgage participations (in fact, a pool of parts of mortgages) takes place. The mortgage participations may be issued not only to professional investors but also to the public in general.
Covered bonds are debt instruments secured against a pool of mortgages to which the investor has a preferred claim in the event of an issuer default. In EU countries, the issuance of mortgage covered bonds is regulated by laws that define the criteria for eligible assets as well as various other specific requirements. In most cases, assets are earmarked as collateral for the outstanding covered bond and are kept in separate cover pools. In some countries (such as Spain), all mortgages on the balance sheet of the issuer are acting as collateral for the bonds. Following the ‘cover principle’, the outstanding amount and interest claims on covered bonds must be covered by the amount of eligible cover assets.
The main transformation in the mortgage covered bond market which has taken place has been the issuance of jumbo or benchmark covered bonds. The jumbo model has become the European standard for the issuance of new bonds. It has also been the main driver for a very liquid secondary market, especially through bond standardization and listing on widely used electronic platforms. The jumbo model was first introduced by a syndicate of banks in Germany in 1995. Several features were added to increase liquidity and improve security in order to attract foreign institutional investors.
The main features of the jumbo model are: (i) the minimum size is Euro 1 billion; (ii) jumbos need to be plain vanilla bonds (fixed coupon, paid annually in arrears); (iii) buybacks are allowed; (iv) the bond must be officially listed on an organized market (typically an electronic platform); and (v) there must be at least 3 market makers that quote bid/ask prices simultaneously to maintain a liquid market. The total value of all issues in the jumbo covered bond market in Europe has grown rapidly to over Euro 500 billion by end-2004, about a half of which is accounted for by German and Spanish mortgage covered bonds.
For banks, mortgage loans have always been a mass-market product used to attract customers for subsequent cross selling of higher-margin products. Mortgage margins are fairly thin, so the refinancing of the mortgage book has a major part to play in ensuring profitability and in boosting a bank’s market share.
The arrangements in Spain are the biggest departure internationally from the German model. Any Spanish bank or savings institution may issue Cedulas secured solely by mortgage loans on domestic properties, but it is less demonstrable under Spanish law than under Pfandbriefe that investors would be able to lay their hands on their assets in the event of issuer insolvency.
The Spanish market has enjoyed meteoric growth, with around 240 billion euro worth of outstanding bonds (or 33% of the European total) at the end of 2007. This expansion has been made possible by pooling mortgages from a range of smaller savings banks into one arrangement, creating the euro market’s biggest player Ayt Cedulas at €14,300m. In 2005 the largest primary market contribution in the European covered bond market came from the Spanish Cédulas Hipotecarias segment where issuers launched benchmark bonds in an amount of nearly €54 billion (+48p.c. yoy), overtaking the German Jumbo Pfandbrief segment (€48 billion) for the first time ever. The very dynamic Spanish mortgage lending market was again the driving force behind the strong growth of Cédulas issuance, whereas new lending business of German Pfandbrief issuers hardly compensated maturing loans. As a consequence, most issuers launched Jumbo Pfandbriefe primarily to maintain their presence in the benchmark segment.
Over the past few years Spain has become the largest European issuer of Covered Bonds with the issuance of its “Cédulas Hipotecarias”. These debt instruments are based on Law 2/1981 of the Mortgage Market Regulations. The Mortgage Market Reform Law aims at modernizing the Spaniard legislation with respect to these Cédulas and at promoting mortgage bonds to continue competing in the European capital markets, where a number of changes and new laws have strengthened other similar instruments. Credit rating agencies are normally bound before investors to perform credit risk analysis of this type of debt and it's comparability against other instruments in the European market. This commitment becomes more significant since the buyers of Spanish Cédulas are mostly foreign institutions. In particular Moody's has analyzed the implications of the legal reform on the Cédulas Hipotecarias from a credit risk stand point and also the potential challenges of this type of mortgage bond. Moody's is the leading covered bonds rating agency, having rated all of the new European issuers in 2006, and all of the public issuers of Cédulas Hipotecarias in Spain.
Update Saturday 12 February 2007
I am still digging. When I reach some more definitive conclusions I will tidy all this mess up. Basically, reading through Bloomberg this morning I discovered that the Cajas Regionales are now keeping their own bonds on their balance sheets, and financing with temporary funds from the ECB.
Bloomberg 19 October 2007
Caja de Ahorro del Mediterraneo is leading a group of eight Spanish savings banks planning to retain a 1.57 billion-euro ($2.2 billion) sale of bonds on their balance sheets.
The 10-year securities are backed by covered bonds, according to a spokeswoman at Ahorro y Titulizacion, the Madrid- based investment unit used by the banks to sell the debt. The notes are part of a 200 billion-euro program, according to Moody's Investors Service.
Spanish lenders have kept at least 18.8 billion euros of asset-backed debt since September as investor concern that record U.S. home foreclosures will damage bank profits sent interbank borrowing rates to the highest in more than six years. Banks can use the bonds they keep as collateral for short-term borrowing from the European Central Bank.
The transaction is being carried out on behalf of Unicaja, Caja de Ahorros de Murcia, Caja General de Granada, Caja Duero and Caja de Ahorros de la Inmaculada de Aragon, Caja de Ahorros de Navarra and Caja General de Ahorros de Canarias, said the Ahorro y Titulizacion spokeswoman, who declined to be quoted by name.
A group of 17 savings banks last week bought 2.95 billion euros of their own notes backed by covered bonds, she said.
Covered bonds, known in Spain as cedulas, typically get the highest investment-grade ratings by requiring borrowers to set aside assets that can be sold to ensure repayment. The collateral backing the debt remains on the borrower's balance sheet.
September 12 2007
Ahorro Corporacion Financiera SV, the investment group owned by 43 Spanish savings banks, postponed its first sale of covered bonds in dollars.
Ahorro Corporacion planned to sell at least $2 billion of notes backed by home loans to U.S. investors, said Luis Sanchez- Guerra, head of capital markets in Madrid. It would be the biggest dollar-denominated offering of covered bonds from Spain.
The lender, based in Madrid, is the latest company to postpone or restructure debt sales totaling more than $100 billion as investors shun risky assets because of losses linked to the U.S. subprime mortgage market. Banco Popular Espanol SA, Spain's third-biggest bank, last week called off a sale of covered bonds
From Reuters Bonds, yesterday:
Investors took a conservative approach to covered bonds as the 1.7 trillion euro market reopened this week, with short-dated paper from the established German market finding most favour, analysts said.
The covered bond market has moved into sharp focus in recent months as it has faced unprecedented turmoil while also being expected to become an ever more important funding tool for banks eager to raise cash, particularly with the securitisation market effectively shut.
European banks agreed to suspend trading in the $2.8 trillion market for mortgage debt known as covered bonds on November 21 2007, to halt a slump that has closed the region's main source of financing for home lenders. The European Covered Bond Council, an industry group that represents securities firms and borrowers, recommended banks withdraw from trades for the first time in its three-year history. Banks are still obliged to provide prices to investors, according to the formal statement.
Covered bonds are backed by mortgage or public sector loans which remain on the borrower's balance sheet. They have historically been highly liquid and typically rated triple-A by ratings agencies thanks to the quality of the assets and legal support, making them appear a surrogate for government bonds.
UniCredit (HVB) credit analyst Florian Hillenbrand said there had been clear demand for instance for bonds maturing in 2009 and 2010 from German issuers, the mainstay of the covered bond market.
But the market in general remained caught up in the turmoil that started last year.
"The world did not change just because Christmas is over," Hillenbrand said.
That was acknowledged by the European Covered Bond Council (ECBC) on Friday when it eased trading conditions set at the start of the year, allowing for transactions trading at wider spreads to be quoted at triple the normal bid-offer spreads. [ID:nL11231767]
Only the previous week, the ECBC had recommended that covered bond market-makers should commit to trade between themselves at double the normal bid-offer spread on a deal of 15 million euros minimum for all bonds, an improvement from the triple bid-offer spread requirement set across the market late in 2007.
Moody's Investors Service said approved amendments to Spain's Mortgage Market Law, which are due to come into force shortly, will strengthen and clarify the credit position of holders of Spanish mortgage covered bonds (known as 'Cedulas Hipotecarias' or CHs) as well as the timely payment of these instruments following an issuer insolvency.
The amended law will improve the over-collateralisation of the CHs by limiting CH issuance to 80 pct of the bank's eligible mortgages, against 90 pct currently.
The law does not change one of the key strengths of the framework, that is that the whole pool of (non-securitised) mortgages supports the CHs in the event of issuer insolvency. This is in contrast to other European jurisdictions where covered bonds are backed by an earmarked portfolio, Moody's said.
The amended law will improve the asset eligibility criteria in a number of respects, including lowering the loan-to-value ratio for eligible non-residential mortgages to 60 pct from 70 pct, extending the geographical scope to European Union properties, permitting substitute assets up to 5 pct of the outstanding CHs and allowing financial derivatives to form part of the cover pool.
It will also enhance the timely payment of the CHs following issuer default and will oblige the issuer to maintain an internal cover register identifying eligible and non-eligible assets, thus improving transparency, Moody's said.
The amended law will also remove an administrative requirement that has to date impeded the issuance of another type of Spanish covered bond: the 'Bonos Hipotecarios', the rating agency said.
Moody's currently rates 15 mortgage covered bonds in Spain, corresponding to 11 to single issuers and four Spanish funds that pool CHs of multiple issuers.
Mortgage-backed Securities and Covered Bonds
Since Spanish bank deposits did not provide anything like the liquidity needed to fund mortgages during the key years of the boom, lenders increasingly resorted to securitisation. Such securitisation basically takes two forms in Spain: mortgage-backed securities (MBS) and mortgage bonds (cedulas hipotecarias). The difference between these two is basically as follows:
An investor in a “standard” MBS holds a dircet claim on the issuer of the security, and not on the mortgagee, even if he buys securities directly from the originator of the mortgages. However in the Spanish case at least one link in the investor chain has a stake in the original loans. This is because, in order to securitize mortgages, Spanish banks first issue participations, which are shares in each of the mortgages included in a pool. A holder of participations receives a percentage of the interest and principal of the mortgages from the originator, who in turn receives the payments from the mortgagees who obtained the mortgage loans. The participation holder has thus a claim on both the originator and the mortgagee. More importantly, the originator retains a certain fraction — 100 minus the percentage of the aggregate participation — of each and every mortgage created, turning both originator and participation holder into co-creditors. In a second step, the holders of the participations may form pools and then issue securities that represent stakes in those pools of participations.
Such a structure constitutes a form of risk-sharing between originators and participation holders, and this risk-sharing is normally thought to be conducive to higher lending standards (especially when contrasted with off balance sheet SIVs). Originators do not simply get themselves off the hook when they sell the participations, because they keep a share of each mortgage. This may well make them more careful when assessing the creditworthiness of mortgage applicants, and may reduce the chance of overly loose credit standards. And participation holders have a stronger incentive to keep an eye on individual mortgagees than if their sole claim was on the originating institution.
The second source of mortgage funding for originators is the so-called covered bond (in the Spanish variant of the cedula hipotecaria). These bonds are debt instrument issued by a credit institution and secured by a pool of mortgage loans or public debt or even MBS themselves. Such bonds pay coupons and principal, just like any other bond. The investor has a claim on the issuer of the bond and, if the latter defaults, on the pool of loans (not on individual loans). In most European countries, where covered bonds are popular, a set of cover assets is set aside for each bond issue. In Spain, however, all mortgages of the issuer constitute collateral for the bond.
One important reason why many observers felt the Spanish financial system was unlikely to suffer the fate of their US equivalents was the virtual absence of Special Investment Vehicles (SIV) and conduits in Spain. These entities allow banks to move mortgage-backed securities off their balance sheets, thus obscuring the exposure of individual institutions and escaping the normal capital requirements. But this did not happen in Spain. The Bank of Spain is right to say that as a result of the banking and financial crisis of the 1980s a condition was imposed that lenders post an 8% capital charge against SIV assets. This has resulted in a relative absence of off-balance mortgage risk. However, as we have been seeing above while the banking system is not exposed on one front it may well be on one or more other fronts.
Securitization in the broader sense (i.e. including covered bonds) has played, and continues to play, a fundamental role in the financing of Spanish mortgages. The proportion of resources which derive from issues on the mortgage market amounted to 37.3% of outstanding mortgage loans (residential or non) at the end of 2006 as against 35.3% at end 2005. At the end of 2006, total funding to the Spanish mortgage market reached 201.3 billion euros, of which 88.3 billion took the form of covered bonds (representing 43.9% of the total of mortgage securities market) and 113 billion was in mortgage-backed securities (56.1%).
Bank liabilities represent an important underlying factor in Spanish securitizations: 27.7 billion euros in 2006, principally in the form of Cédulas hipotecarias (or covered bonds) (34.8% of new issues on the mortgage securities market). The direct securitization flow of mortgage debts (MBS) reached 48.5 billion euros in 2006 of which 38.9 billion was in the form of residential mortgagebacked securities (RMBS). In 2006, MBS’s represented 65.2% of new issues on the mortgage securities market. Outstanding amounts of residential mortgage loans constituted one of the principle supports of securitization in 2006, close to 41% of new issues taking the form of RMBS (38% of outstanding amounts). The growth of these latter has, nevertheless, significantly diminished (+31% of annual growth in 2006 compared to +57% in 2005), reflecting the slowdown in home loans and also the fact that an increasing number of loans no longer satisfyied the rule that loans should not exceed 80% of the estimated value of the property. Combining these two securitization techniques (covered bonds and MBS), mortgage loans (to households and other agents) represented, in 2006, the counterpart of close to 80% of new issues by securitization vehicles (and 83% of outstandings).
The Use of Mortgage Covered Bonds
Renzo G. Avesani, Antonio García Pascual, and Elina Ribakova, IMF Working Paper WP/07/20
The Funding of Mortgage Loans in Spain by the Issue of Mortgage Securities. Their Legal structure.
Dr. Sergio Nasarre-Aznar
Friday, January 04, 2008
Firstly, Eddy makes a very valid point about the 3 month Euribor level and the impact of variable interest rates on Spanish mortgage holders:
"most spanish mortgages (>90%?) are variable, and referenced to 1 year Euribor, so the 3-month euribor/libor "crunch" is inmaterial for setting interests rates. (1 year Euribor has barely moved through the "troubles")."
Now what Eddy says is certainly the case (although we are still unsure about the exact percentage of mortgage holders who have variable rates, the number is undoubtedly very high), but Eddy's key point here is that the recent incrtease in the 3 month Euribor (and Libor euro) rates is not being passed directly on to the banks' consumers, since they are tied to the one year rate, and that hasn't really moved up much. This is a factor I hadn't previously thought about. Basically I hadn't picked up the one-year part of the story, I suppose because I hadn't been paying attention sufficiently to the details here. (this is just one good example of how useful blogs are, since they enable you to contrast opinions quickly in real time).
So the squeeze on the inter-bank rates won't hit the mortgage holder directly, but it will hit someone, somewhere, and the most likely candidtaes are the Spanish banks, since it is the banks themselves who are paying these new, higher, rates to finance their day to day operations. BBV is now offering 4.5% on 12 month fixed term deposits, and 3.8% on two year fixed term deposits, so the short term need for cash is to some extent inverting the yield curve already.
This I suppose helps explain the volume of money Trichet and Co at the ECB recently poured into the banking system system. The ECB - you will recall - supplied €348.6bn to banks in its two-week operation on 18 December - at a rate of 4.21 per cent - after scrapping the normal upper limit on the amount it would lend.This was in part to try and ease the pressure which must be mounting on the Spanish banks (and others, of course, but the Spanish banks must be among the most exposed at this point, see below). While the British Chancellor of the Exchequer Alistair Darling announced a four point plan this week to protect UK citizens' deposits in threatened banks, the Spanish authorities remain strangely silent at this point. Their failure to be able to reassure in the form of concrete guarantees is perhaps more worrying than were they to talk openly about the problems.
I don't have any real data on the banking sector in Spain, but it is this sector which seems to be the key area in the whole Spanish story (as I try to argue in this post, the real economy in Spain is now giving a lot of signs of slowing at a very rapid rate). Most authorities (eg the Bank of Spain) are busy denying there are any issues, but I don't really see how there can't be. Essentially if they are charging "x" base points over one year Euribor, while they are financing with three month rollovers, and now they have to pay more for the three months finance (once the ECB's 3.21 "xmas present" comes to an end, then their operating margins have just received a big squeeze.
Essentially it would be interesting to know the proportion of mortgage lending in the 2004 - 2007 period which was taking place on the basis of "own bank deposits", and how much of the financing was being outsourced by the banks. This measure would give us some sort of vulnerability indicator for the Spanish banking sector. Basically, taking Eddy's point about euribor - this net effect of this is to insulate the borrowers (which was not really the original intention, I am sure), and pass all the additional costs onto the banks, since they need to keep rolling-over borrowing on a three month basis (or to some significant extent they do). So their margins and liquidity must be being squeezed at this point. We will see how long they are able to accept the pain before someone says "ouch".
My guess is that Spain must have one of the highest quotas of home grown sub-prime problems (ie not derivatives from the US situation) on the globe. These principal sub-prime issues are likely to come in 2 areas: a) second-home owners from outside Spain - who have been very numerous in areas like Almeria and Valencia. Basically if property prices tank at some point in Spain, and the debt of many of these "investors" exceeds the evident long term value of the underlying asset then they amy well decide that the rational thing to do is accept the capital loss of their initial payment, and simply walk away from underneath, leaving the keys with the bank as they go, as we ahve seen in the United States; b) and secondly, we have the immigrants. As I show in the chart below, inward migration to Spain in recent years has been very substantial (perhaps proportionately the highest level of any country on the globe).
Here in Barcelona where I live we have whole "barrios" - like, for example, Santa Coloma de Grammanet - where the locals basically sold their homes in what were traditionally poor areas and moved to larger and relatively cheaper properties in housing estates outside Barcelona (Olesa de Montserrat, for example, seems to have been popular in the case of Santa Coloma, Spanish consumer behaviour is very much of this nature, with people tending to look at what others are doing and folloing trends). So basically places like Santa Coloma now have a very significant immigrant population, and many of these family groups have bought property and have outstanding mortgages . Immigrants tend to be badly paid, but they do work long hours, and they do save, and at the height of the boom Spanish banks were accepting four or five "nominas" (or paychecks) as backing for a single mortgage, for a small interest premium, of course (but isn't this the heart of "sub-prime"), and a lot of migrants have bought-in. It is precisely this group who will be the first to feel the impact on the real economy, and then it is difficult to see how they will be able to maintain payments. And again, in extremis, many migrants can simply walk (as Eddy suggests in his comment) either by going to another EU country, or moving back home.
Young Spanish people represent less of a risk to the banks, since - as former Govneror of the Bank of Spain Jaime Caruana cynically noted when accepting that property prices in Spain may be over-valued by as much as 35% -they are basically linked in to the banking system through the DNI cards and paycheck link-ups, and it it is thus very hard for them to just walk away. The only thing such young people could realistically do is emigrate, and again as Caruana noted there is very little tradition of this in Spain (even Spanish migrants to Latin America in the first part of the 20th century had a very high return rate) due to family ties etc. So this group of people will be faced with significant distress, and their consumption patterns will be affected, but they are relatively unlikely to simply default.
Essentially what we can see at the moment is that the whole property sector is seized up, and it is likely to remain this way for many years to come. This seize up is being produced by what is known as a change in the "lending conditions", and this change has much greater importance than any simple and straight forward interest rate effect. What do I mean by a change in the lending conditions? Well banks were previously lending to people in Spain over 50 year terms, with offers of interest-only payments for the first 3 years, and, most importantly, with loans of 100% (or even 110%) of the "banks own" property valuation.
Such days are now well over, and they are gone permanently. Basically it may have made sense to someone to lend to the tune of 110% of an underlying asset when the value of this asset was rising by 20% a year, but now that no one knows for certain the future path of the asset value, and when most rational people accept that the direction of any movement is far more likely to be down than up, then noone wants to know anything about such forms of lending. This type of "financial efficiency" is now well and truly dead.
Banks now won't shift above 80% of the (much more strictly measured) estimated value, so this means that young people need to find a minimum of 20% of the asking price, plus notary fees and furniture. So if we take a bottom level entry flat for a first time buyer in a major urban area to be, say, 300,000 euros, they these young people now need to find 60,000 euros plus expenses. Maybe 100,000 euros in total. And they have very little savings, since in the past these weren't considered to be necessary (the behavioural shift).
Worse, mummy and daddy don't have savings either, since the low interest rate environemnt encouraged them to take their savings out of deposit accounts and sink them in second homes. So for the children to buy the parents now need to sell one of their properties, and this is one of the first points were we will surely see downward pressure on prices, since someone somewhere has to unlock something in the was of frozen liquidity for the current gridlock to "unseize" itself. Also the children will need to start to save that 60,000 plus, so immediate consumption will also be affected, as possibly this impact is already being felt.
A second factor which is going to seize up the works is the behavioural shift which is taking place from buying first and selling second, to selling first and then buying. Most people who bought new properties in recent years in order to move "up market" from their existing home bought on the basis of architect's plans (which is why construction is still ticking over, since all the property that was bought up to August still has to be built), and only selling their current home at the last minute. This made sense while prices were rising, but now they are falling.....
I think it is important to realise just how much in the avant-guard the Spanish banking sector were in introducing the so-called "financial efficiencies" which made the boom possible. Now some of the "efficiencies" achieved are real and permanent (like some of the benefits of financial globalisation, you can borrow cheap in yen, eg), but some - like raising the % of the loan, lending to risky individuals, lending on the basis of issuing paper rather than deposits - were most definitely not efficiencies, but "irrational exuberance". As I say, the Spanish banks - especially BBV, Santander Hispano, Sabadell Atlantico - seem to have been in the forefront of all this. My feeling is that the caixes have been rather more conservative (at least here in Catalonia) but that may vary from region to region (and remember many regional cajas are involved with local political institutions who may now be seeing property-boom-related income drying up, and hence some of the institutional indebtedness they have on their books may now become unsustainable. I would also be thinking about corporate balance sheets, if the value of the urban office "park" starts to fall, some of the major companies may be widly over-leveraged in this context (think of all the euphoria at the time of E-On/Endesa, and all those warning from the ECB about excessive leverage).
One measure of the issues raised by financial globalisation can be obtained by looking at the contributions to a recent Bank for International settlements conference on the topic (2006) Basically the whole conference is interesting, but in particular the Panel on "Review of recent trends and issues in financial sector globalisation" session and in especially the presentations from Christine Cumming (Federal Reserve Bank of New York), José Luis de Mora (Banco Santander Central Hispano) and David Llewellyn (Loughborough University) are of interest. De Mora is fascinating, since he gives us an indication of just how the Spanish banks were taking their new "methodology" to the UK (think Northern Rock).
Basically one part of my work as a macroeconomist is related to the role of demographic processes and construction booms. Especially I have been following Germany, Japan and Italy, since these are the oldest societies that have been produced to date, and they seem to have quite distinct characteristics when it comes to domestic consumption dynamics and property sector activity in recent years as compared to their younger counterparts. My feeling is the nearest comparison we have to what just happened in Spain is Japan 1992. Especially given that the move into property in Japan post the 1989 stock market bubble crash and the post internet boom collapse move into property in Spain. The most worrying aspect about what is happening at the moment is undoubtedly the recent sharp spike in inflation. With the real economy near to entering constraction, and demand falling all over the place this inflation simply cannot be sustained, so Spain does have, IMHO and especially given the lack of a sensitive interest rate policy from the ECB, a real risk of turning itself completely upside down, and falling into Japan style deflation in the mid term. This is something which I never would have imagined possible only a year or so ago, but you need to look at the continuous above-trend growth in prices we have seen in Spain over the last six or seven years, and think about how an economy can correct for these in a common currency zone, and especially with global factors (the fall in the dollar, the disintegration of Bretton Woods II) maintaining the euro at possibly very high levels. Structurally the eurosystem was never designed with the need to operate a country specific ZIRP ( zero interest, massive liquidity easing) policy, but that is exactly what may be needed in Spain if the worst case scenario is fulfilled.
Demographic factors are important here. Basically the 25 to 50 age group just peaked in Spain (see chart below), as it did in Germany in 1995. This means that natural demand for housing, and home loans, is likely to trend down from here on in. The only unknown is immigration. Spain just sent its estimates for 2007 migration in to Eurostat (and they must have a pretty good idea from the Padron Municipal) and we have yet again had a record year it seems, with the estimate pushing the 700,000 mark (again see chart). But the big question is what is going to happen to all these people during the downturn.
In conclusion I would say I have three large questions in my head:
1/ What happens to the immigrants and inward migration?
2/ What level of distress will we see in the banking sector going forward?
3/ What will happen to all these young people who are mortgaged up to their eyeballs on property whose value may now enter a protracted period of secular decline. Will there be political pressure in Spain - as in the US - for the situation to be eased (possibly by adjusting downwards the book value of the debt). If any such political measure is taken, who will pay the bill? Basically Spain's "broken plates" will need to be paid for by some apportionment between the young people themselves, the banks, and the taxpayer, but the big question is in what proportions?
I don't anticipate that this is going to be a short term affair at all. Everything got a very heavy push start back on 9 August, we are now in January, and things have hardly started to happen. If we go back to Spain's last real recession - foolowing the 1992 "Olympic" boom, see chart below - then property prices needed till at least 1995 to recover their earlier values. This time I think the situation is much, much more serious. Japan really still isn't out of the problem that started in 1992 even now. So I would say a conservative estimate on this one is a five to ten year window of problems, and that is imagining we get a "soft" landing, which may well not be the case.
So basically, while I don't have time to comment on Spain's issues on a day by day basis, I will try a more or less monthly report on this blog as new data comes in.
Well, if you've got this far, the above is a summary of my original comment, and now here is Eddie's most revealing reply:
1/ What happens to the immigrants and inward migration?
"Many immigrants will lose their jobs (construction, services etc). This is already happening. Many will return home or will seek greener pastures in Europe. "
"See for example here
"Page 22-Foreign unemployed total increased by 24,5% y/y
Foreign unemployed in construction increased by 53.5% y/y "
2/ What level of distress will we see in the banking sector going forward?
Although data is disperse, my hunch is that the problems for the financial sector will come short term from the immigrants (where a lot of shoddy financing has been going on, without adequate income history) AND the R.E. firms, who are clearly overexposed.
But the main threat for the banking sector short term IMO, is not an increase in delinquencies but the inability to roll over their debts in the European Inter bank Market or to sell their mortgages (cédulas) abroad at all, or at least, to do it in profitable terms.
Margins for the Spanish banks in their mortgages are razor thin. A typical mortgage with Euribor 1Year +0,50% is paying now around 5,25%.
(There is a market failure here, cause they are using short term rates to finance long term credits, with a very low margin, anyway, great for borrowers)
A flattening of the curve is causing clearly pain for the banks (and they have convinced a lot of clients to sell their “fondos de inversion” and transfer the money to deposits as a short term remedy)
In the same manner as thecrash in US mortgage providers, Spanish banks could very well be soon unable to sell the mortgages they originate, or roll over the existing debts.
A la Northern Rock, but much, much bigger.
This would be a text book “sudden stop” in external financing …and with a 10% c.a. deficit this year, something to worry A LOT.
3/ What will happen to all these young people who are mortgaged up to their eyeballs on property whose value may now enter a protracted period of secular decline.
For me the main problem now is not “negative equity” , but the effects in general activity of a housing recession. Of course a lot of people will become house-slaves (unable to sell or to move away). But barring any financial accident, the could cope with their debts if Mr Trichet doesn’t increase rates.The problems will come later, with a general recession (given the oversize weight of housing in Spain’s economy), unemployment etc. Of course a general recession would decimate public finances
4) Variable/Fixed Mortgages
Well, I don’t have the total numbers, but my hunch is that 90+% of the mortgages originated in the last 10 years are variably type.
Take the numbers for last month, for example here ( October 2007)-97,6 % of mortgages where variable (incidentally, notice the fall in total volume for residential –7.32% y/y and –5,67% m/m)
Now in response to the following comment from Eddy:
"The problems will come later, with a general recession (given the oversize weight of housing in Spain’s economy), unemployment etc. Of course a general recession would decimate public finances"
Yes, but if you look at the data I am presenting, we may be heading for recession quite - indeed alarmingly - quickly. There was already some indication of a slow-down in the third quarter. The quarter on quarter increase was 0.7% which was down from 0.9% in quarter two. More significantly perhaps was the fact that the slowdown was lead by a deceleration in domestic demand, while the quarterly growth did not fall more due to the fact that exports increased significantly, so external trade became less of a drag.
When we come to the fourth quarter we should expect this downward trend in domestic consumption to be continued, and this is exactly what we are seeing, first and foremost in the case of retail sales,which are now steadily going down month by month.
At the same time construction is weakening steadily, industrial output started to contract in December (going by the PMI) and the services sector is now barely growing. Given that foreign trade is a still a drag rather than a boon, my guess is that the Spanish economy will enter negative growth in the 1st quarter of 2008, and that it will stay there for some time. As we can see from the chart below, private domestic consumption (in terms of year on year growth rates) actually peaked back in 2004/5, and the rates of increase have been slowing all the time since - as the last drops of juice get gradually squeezed out of the lemon - so we should only expect this trend to continue:
And if we come to look at the quarter on quarter changes for the last seven quarters, then I think the trend is very clear - for my taste at least. In Q3 2007 private domestic consumption only grew at 0.37% over Q2, we have to go back to Q1 2003 to get a slower rate, and if we look at 2006 and note the rebound there was in that year, then we should ask ourselves about the difference in global conditions in end 2006/early 2007 and then I think the position should be clear enough.
If we now come to look at the long term charts for Spanish GDP, we can see that there hasn't been anything approaching a recession since in Spain since 1993, and we should be asking ourselves why that is. Oh, I know, Spain is different, but it really can't be so different that it can simply ignore all the laws and rules of marcoeconomics. Basically I think the answer is simple enough. In priciple Spain ought to have had a recession in 2002 along with virtually everyone else, but cheap interest from the ECB kept putting it off and off. I mean I think it is important to note that Spain hasn't even had one single quarter of negative growth (on a seasonally adjusted basis) since the 1992/93 recession. So now we should expect downside underperformance to follow all the upside overperformance. That is what my macro instincts tell me.
As can be seen from the combination of charts (which overlap slightly since they are based on different time series) not only has Spain not had a recession since 1993, it has not had even one quarter of negative GDP growth (on a seasonally corrected basis, remember there is Spain's "other" dependence, apart from construction that is, the dependence on tourism).
Now if we look at the first chart (which shows annual rates of change by quarter, while the second shows quarter on quarter change), we can note the importance of the Nasdaq correction at the end of the internet boom, since there is a marked slowdown in the rate of annual growth, but there is no recession - I think this is the key point I want to make here - and there is no recession since construction activity took over from the stock market, as the money moved over from one "bubble" to the other (just like Japan in 1989 and 1992).
All of this can be seen from looking at a chart for monthly cement construction since the late 1990s which simply goes up and up.
Cement output serves as a nice proxy for the levels of construction activity (which puts a whole new meaning onto the old expression about getting "stitched up in a concrete overcoat") and as we can see the level of this activity has yet to fall in any substantial way - due I suspect to an uptick in government civil engineering activity, as well as the completion of outsanding purchases. Yet Spain's economy is already slowing very noticeably . we can just see it in the way GDP starts to point down in Q3 2007 (at the very end of the upper graph), we should expect to see more of the same in Q4, and then I conjecture real negative territory in Q1 2008. I may be out by a quarter here (ie negative growth may come in Q2, but I doubt the margin of error is greater than that). So the present slowdown is more the result of movement in the financial services and real estate sectors than in construction itself. Which means, of course, that there is a lot more to come as construction activity itself starts to decline.
If you want to know the size of the correction the Spanish construction industry has to make, then just look at the cement production increase since the late 1990s - from roughly 2,500 thousand tonnes monthly to around 5000 thousand tonnes, or the double. That is how far back we need to unwind I think.
Some additional indication that the process has barely started can be gained from the next chart, which shows the progression of buildings approved for construction since 1992 on a monthly basis. This series appears to have spiked at a local peak in the summer of 2006, made a temporary resurgence towards the summer of 2007, and now started to head down. The latest month for which we have data here is September 2007, but all the evidence points to the fact the decline continued across the last quarter.
Another interesting measure of the extent of the problem comes from thinking about the share construction activity has come to assume in Spanish GDP. As we can see from the next chart this share has risen from around 70% in the mid 1990s to 11.5% in the last two quarters of 2006. Logically it is now about to trend down again, leaving us with the question of what can come online to fill the gap. My own view is that ultimately Latin America will be the salvation of Spain, but this will mean transforming Spain from an internal consumption driven economy to an export oriented (possibly services dominated?) economy along the German lines. But the road from one thing to the other is likely to be a long and hard one, and there lying in the middle is the question of what to do about the extraordinary indebtedness of the under 40 generation in Spain.
The next chart gives us some indication of the extent of the house price rise. I say some indication, since it is simply Eurostat's housing cost component for the harmonised CPI, which is not a dircet measure of house prices, but does give us an idea of the relative order of magnitude. By way of comparison I have included the data for Germany over the same period, and this should lead us to ask why - given the same interest rate was applied in Spain as in Germany - Spanish costs rose so much more rapidly. Oh, I know, because inflation expectations were so much stronger in Spain. But this argument is in danger of becoming circular, since why were expectations so much stronger in Spain than they were in Germany?
The next chart shows movement in quarterly house prices themselves (for both new and second hand properties) since 1995 using the data prepared by the Spanish housing ministry.
As can be clearly seen, the most important part of the boom was between the end of 2001 and the middle of 2006. An what was happening during that time? Well just to remind us I am reproducing the chart for spanish inflation and ECB interest rates that I prepared from my last post. As is pretty evident, the height of the boom coincides pretty closely with the period during which Spain has negative interest rates.
The next graph shows the net migration annual figure for Spain in recent years using data provided by the INE to Eurostat. Of particular interest I feel is the number of migrants still ariving in Spain in 2007, despite the first signs of slowdown. This is definitely a key data number to watch going forward.
Finally here are two additional demographic charts. Firstly the fertility data for Spain, since this gives us an idea of the longer term native domestic demand for housing.
And lastly, the chart for the 25 to 49 age group as a proportion of the total population. As we can see this group peaked in 2006, at around what ar far as I can see the highest proportion for any society yet to date of 40%. Is this just a coincidence, or does it have more significance in all this?
Thursday, January 03, 2008
A 70 percent surge in oil prices over the past year - the price of crude oil touched $100 a barrel for the first time in New York yesterday - together with a global surge in food prices is driving up inflation across Europe even as the euro's gains against the dollar make imports cheaper. The "strong, short- term'' increase in inflation is a source of concern to the European Central Bank, the Bank of Spain said yesterday. It certainly is, since
Adverse weather conditions in food-producing countries and increasing demand from rapidly growing emerging economies, where food consumption composes a significant proportion of household expeniditure, are also pushing up food prices, increasing the inflationary pressure. Droughts in Australia, Canada and Ukraine pushed the price of wheat to a record in December while soybeans touched a 34-year high.
Inflation across the euro region accelerated to 3.1 percent in November, the fastest pace since the currency was introduced. Meanwhile consumer confidence is plummeting, touching historic lows for three months in succession in December according to the index compiled by the Instituto de Credito Oficial.
The cost of borrowing - as measured by the British Banking Association's three-month Libor rates has been falling during the last couple of weeks as central bank measures to relieve a year-end logjam in money markets continue to show signs of success, but they are still way above where most people would like to see them.
The three-month euro interbank offered rate, or Euribor, dropped almost 2 base points to 4.67 percent, the European Banking Federation said today. It was 4.93 percent Dec. 17, the day before the European Central Bank injected a record $500 billion into the banking system.
Short-term borrowing costs have fallen since policy makers in the U.S., U.K., Switzerland, Canada and the euro region announced plans Dec. 12 to counter a credit shortage threatening to undermine global economic growth. Up to the present time ubprime-mortgage defaults in the U.S. have forced the world's financial institutions to write down about $100 billion in fixed-income securities.
But these measures have yet to achieve their underlying objective of pushing borrowing costs back down to the levels of July, before the collapse of the U.S. subprime-mortgage market caused banks to stop lending to all but the safest borrowers. The three-month euro rate is still 66 base points above the ECB's key financing rate of 4 percent, and this is up considerably on the average difference of 25 base points during the first half of 2007. The comparable dollar rate is 43 basis points more than the Federal Reserve's benchmark rate, up from 11 basis points in the first half.
Basically Libor rates for one-to-three month funding in dollars, euros and sterling are global bellwethers for interbank lending.
Three-month euro Libor was fixed down again yesterday at 4.661313 percent having fallen at eleven out of the last 12 fixings and marking a second consecutive week of declines. Although still well above the 4 percent ECB base rate, this was the lowest fixing for three-month euros since November 22. So what we can say is that while the massive easing operations have reduced the spread, the underlying difficulty is still there, and this is very important in a country like Spain, where an estimated 80% of mortgage holders have variable rate packages which are normally related in some form or other to the Euribor rate.
So really what we should note here is that Spain is facing a very strong tightening on nearly all fronts - interest rates, inflation, credit conditions, and, last but by no means least, exchange rates. The euro simply rose and rose against the dollar throughout 2007 in a way which, I must admit, certainly surprised me.
So what we are faced with is a Spanish economy which is entering its most important "correction" in many years - certainly since 1992, but quite possibly the most important one in its entire history, with all the policy indicators set on tightening. This, it seems to me, is very serious indeed.
As I have pointed out on numerous occassions in the past, this is the downside for Spain on euro membership, since it means that Spain will not have anything like the policy mix that is appropriate to its current needs (the upside was, of course, the property boom, which was in part produced by having negative real interest rates - ie interest rates which were below the actual inflation level in Spain - over a considerable number of years.
The underlying situation in Spain is also highlighted by the shifting pattern of employment and unemployment, and a comparison between 2006 and 2007 is pretty revealing. If we look at the chart below we can see that in the early months of this year the employment situation was generally up over 2006. Then the situation turned (around July), and since then it is "down hill all the way" unfortunately, with unemplyment rising (as it might well do for seasonal reasons anyway, but in every case the increase is significantly more pronounced than in 2006. Spain's ever productive labour market has, unfortunately, now turned.
Also if we look at retail sales we get the same picture. Retail sales across the entire 13-nation euro region fell in November according to the most recent eurostat data, dropping by 0.7 per cent from October to a level which is just 0.2 per cent higher than the Novemeber 2006 one. Of course this average hides considerable variance, with the weakest performances coming from Germany, Italy and Belgium.
But of particulr of note is the performance of retail sales in Spain (the zones 4th largest economy) since strong growth in Spain has previously offset weaknesses in Germany and Italy at previous critical junctures. But this time it will be different, since Spanish retail sales have fell in both October and November 2007, and while the year on year readings are still in positive territory, they will not remain there for long since the earlier strong readings will eventually drop out of the data.
Also manufaturing industry in Spain may well now be contracting. Euro zone purchasing managers surveys for the manufacturing sector confirm the economy is likely to have slowed down in the fourth quarter despite pockets of resistance in some countries. The Royal Bank of Scotland purchasing managers index for the euro zone manufacturing sector eased to a final 52.6 in December from 52.8 in November, a figure was revised up slightly from the provisional reading of 52.5 issued in mid December. The Spanish manufacturing PMI fell to 49.5 from 50.7, dropping below the critical 50 dividing line between expansion and contraction. Since it was also below 50 in October this means that in 2 out of the 3 months in Q4 2007 Spanish industrial output may have been contracting. When we add this to the decline in retail sales and construction, and take into account that with a large trade deficit foreign trade is a drag not a plus for Spanish GDP, it is hard to see where the growth is going to come from in the last quarter, except of course from an expansion in civil engineering projects financed by the government. With elections looming in March, my guess is that they are deficit spending as much as they can at the moment.
The Spanish services sector purchasing managers index (PMI) stood at 51.0 in December, up from 50.7 in November, according to market research group NTC. "Commercial activity continued expanding... even though recent calculations suggest that the growth will not be maintained in 2008," Nathan Carroll at NTC Economics commented. In particular new orders fell for the second month in a row, coupled with moderate additional declines in pending orders, all of which point to a continued slowdown in activity. Basically services - which constitute the lions share of the Sanish economy - continued to expand in December, but - as can be seen from the chart below - at a much slower rate than earlier in the year. Looking at the chart it is not difficult to imagine that we will enter negative growth in services in Q1 2008, and I imagine that at that point the Spanish economy will enter recession.