Prices to fall gradually over the longer term

Quarterly no. 36 April 213 Prices to fall gradually over the longer term Prices are starting to fall. The scale of the adjustment is modest, on average, at -1.7 in the 12 months to end-212 in existing housing, but more marked falls have been observed on particular segments. Sales volumes have seen quite significant falls. As was the case in 28, as soon as the housing market falls, the same old controversies resurface. Are we in a speculative housing bubble? Will the fall-off in sales and prices be violent? Are we going to see the same downward spiral as that seen in the US or Spain with a time lag of five years? Our own diagnosis is unchanged. The ongoing adjustment is cyclical and can be explained by the economic environment and fiscal tightening measures. But the French housing market differentiates itself from most other European markets due to its favourable fundamentals. The economic climate is extremely degraded, with zero growth, falling purchasing power, and steadily rising unemployment. Fiscal rules have been tightened overall, especially for buy-to-let investors. In the first place, there has been a sharp reduction in housing capital gains allowance (excluding the main residence). But there has also been an increase in tax and social security levies on unearned income, and a capping of tax loopholes. The Duflot buy-to-let scheme for the new-build market, which has taken over from the Scellier scheme, is likely to have a mixed success, because although it offers an 18 tax cut on the value of the home, there are stronger conditions attached and higher taxes at the time of resale. The market fundamentals are still favourable. Demand is structurally strong due to population growth, the family breakup, preparations for retirement, the safe haven effect, and so on. There is no oversupply of homes, no lending bubble, few non-performing loans, no marked credit squeeze, and no signs of a speculative bubble. What s more, mortgage-lending rates have been falling for several months past. Relatively weak supply is slowing the fall in house prices. The most probable scenario is, therefore, not for a short, sharp correction, but a gradual adjustment over the longer term. In 213, sales volumes look set to stabilise at a low level in new-build, and could fall by 1 in the existing market. Prices are predicted to fall by -6 y/y in the existing market, for a cumulative drop of 7-8 between end-211 and end-213. Prices are forecast to fall in 214. Contents Recent trends in existing housing...2 Significant fall-off in sales in 212...2 Prices start to fall...2 Recent trends in new-build homes...3 Sharp drop in sales in 212...3 Upturn in the housing stock...4 Fairly sharp drop in housing starts...4 Virtually zero price increases...4 213-214: drawn-out correction... A gradual process... Negative cyclical factors... Four positive structural factors...7 Forecasts...8 Germany: counter-cyclical...9 Untouched by the 1998-27 boom...9 Especially favourable status for tenants...9 Market now picking up... 1 French property market indicators... 11 Olivier ELUERE

Significant fall-off in sales in 212 Demand for pre-owned homes fell markedly in 212. Sales, which were still sustained in 211, fell 12 over 12 months. The number of transactions on pre-owned homes was still very high in 211, at 8, (source: Conseil Général de l Environnement et du Développement Durable, based on fiscal and notary databases), a level close to the record highs of 24-26, of around 82, a year. In 212, only 79, transactions were recorded, 12 fewer than in 211. In addition, this annual total was inflated by an exceptional January, which saw approximately 1, sales, some of which had been brought forward to beat the February cut in the real-estate capital gains allowance. After February, sales fell back, totalling, per month, compared with 67, a month in 211. If we factor out the January 212 bounce, the rate of fall in sales was close to 2 over 12 months. In January 213, the 12-month total fell to 66,. In Greater Paris (Ile-de-France), sales saw the start of a fall in 211. Sales of pre-owned apartments came to 12, units, a drop of 9 over a year. In 212, sales fell 11 to 9, units, their lowest level since 29. The same scale, -2, can be seen over two years in the fall-off in sales. Prices start to fall Prices for pre-owned homes are currently falling modestly, but the rate of fall is picking up. Prices were still rising in 21-211, at an annual average of and 6, whereas other real estate markets across Europe were falling. The start of a fall-off became apparent in 212. Prices fell by.3 year-on-year in the second quarter, by 1.6 in the third, and 1.7 in the fourth. Over the full year, the fall was close to zero, at -.4. In Q4, the fall was more marked in the French provinces, at -2.1 y/y, and more moderate in Greater Paris, at -.7 y/y. Prices are starting to fall in Paris, at -1 y/y (source: Insee, Notaries). Quarter-on-quarter, prices are falling very modestly: -.3 q/q in Q4, after -.6 in Q3 and -.4 in Q2 (seasonally adjusted). In view of the data available for Q3 in some regions, a modest and fairly consistent trend can be observed across France: -. y/y in Greater Paris, -1.9 in the Nord-Pas de Calais region, -3.2 in Provence-Alpes-Côte d Azur, -1.1 in Rhône Alpes. In some segments, the trends are, however, far more negative, with falls of at least 1 over the 12 months to end- 212 for pre-owned apartments in cities like Tours, Caen, and Mulhouse. Notaries in Paris and Greater Paris say that according to preliminary sales agreements, prices in Paris would fall by 4. between August 212 and May 213. In Q4 212, the average price per square metre for pre-owned apartments stood at 8,27 in Paris, 4,42 in the Petite Couronne (inner suburbs) and 3,13 in the Grande Couronne (outer suburbs). The average per sq.m. house price in France is estimated at 2,88. Recent trends in existing housing 2 2 1 France: Existing housing transactions Est. number per quarter France: Existing housing prices Chart 3 France: Unoccupied existing houses sold in Greater Paris Chart 1 4 6 7 8 9 1 11 12 Sources: CGEDD, Notaries, CA 2 1 - Est. number of transactions Chart 2-1 4 97 98 99 1 2 3 4 6 7 8 9 1 11 12 change over 12 months Q1 21=1 Sources: INSEE, Notaries 4 3 2 y/y index 1 4 6 7 8 9 1 11 12 Source: Chamber of Notaries Number of transactions 12 1 8 6 No. 36 April 213 2

Sharp drop in sales in 212 In 211, sales of new-build homes (developers, programmes of at least five homes) totalled, units a drop of 8 on 21. This level of sales was still quite high, but fell short of the record average of 122, a year in 24-27. Following a spike in late 211 (3,8 units in Q4) due to investors seeking to profit at the last minute from the Scellier buy-to-let scheme and its 22 tax break, sales fell sharply in 212, to 86,2 units over the full year, an 18 drop in one year. There was no sudden collapse over the year, however, but a stabilisation at around 21, sales a quarter. In Greater Paris, sales had fallen more markedly in 211, by 17 over the full year. In 212, sales stabilised (down.2 over the year) and even picked up in Q4 212 for a total annual increase of 9. Sales of new homes thus saw their lowest level since 28, but demonstrate a degree of resilience, even so. Moreover, according to Insee surveys of developers, demand for new-build homes picked up slightly in early 213, although still at a very low level. Trends are highly differentiated between investors (who buy an asset to let it) and home buyers (who buy a home to live in) so that one can justifiably talk of a two-speed market. The investor segment forms a large proportion in the new-build developer market, running at close to an average 4- of the market, compared with 1- in the pre-owned sector. This proportion does, however, fluctuate considerably depending on tax schemes. It rose to 64 in 29-21 due to the success of the Scellier scheme. In 211, the share fell to 7, with a 17 drop in the number of sales to buy-to-let investors. And it fell very sharply in 212, by 37 relative to 211, to 38, units, and around 44 of the total. The fall can largely be put down to the fiscal tightening measures implemented in 212 and announced for 213, which mainly target high earners and unearned income. Several measures concern buy-to-let new-build investment: the Scellier scheme was severely cut back in 212 and replaced in 213 by the Duflot scheme; higher taxation on real estate capital gains (excluding the principal residence); and higher tax and social security contributions (measures set out in detail on page 6). As a result, investors were very hesitant and wait-and-see in 212. The home buyer segment has been more resilient. In 211, its share in total sales had risen to 43, compared with 37 in 21, with a 6 rise in transactions. In 212, sales in this segment rose by 6 over the full year, accounting for around 6 of the total. This rebound may seem surprising. It is probably linked to the usual support factors, especially demographics, and to the fact that the PTZ+ interest-free loan had been refocused on new-build in early 211. See page 8 for the outlook for 213-214. Recent trends in new-build homes 49 41 33 2 France: Sales of new-build homes s per quarter France: Inventories of new-build homes Chart 4 18 On market Sales 1 91 94 97 3 6 9 12 13 11 9 7 3 91 94 97 3 6 9 12 Stock for sale 4 4 3 3 2 2 s months Chart Stock by months for sale (rt.sc.) Chart 6 France: Sales and inventories of new-build homes Sales s months for sale 1 91 94 97 3 6 9 12 Sales Stock by months for sale 23 19 11 7 3 24 2 16 12 8 4 No. 36 April 213 3

Upturn in the housing stock Some 123, homes came on to the market in 211, a 13 rise over one year. The figure was lower than the record, homes a year for sale in 26-27, but picked up as a result of the relative resilience of sales and a near-normal inventory level. In 212, the number of homes coming on to the market fell quite sharply, by 11 over the year, to 19, units. The trend was quite marked in Q4, with just 2,2 homes coming on the market, for a drop of 34 over the year. The drop can be explained by the fall-off in sales and the increase in the stock of new homes available for sale. This stood at 89,9 units in December, 17, more than a year earlier. The figure was, however, down slightly over the quarter (94, in September). The ratio of stock to sales, which measures the average time it takes to sell a home, was a high 12.4 months on the market in December 212, compared with 7. at end-211 and 6.1 months at end-21. The December 212 figure was far higher than the average 7.7 months for 2-21. The inventory as such (ie, the number of homes for sale that have been completed or are under construction) is far smaller, at 4, units: of the stock is completed, 39 is under construction, and 6 in the planning stage. In Greater Paris, the stock of homes for sale is also rising, with a ratio of stock to sales of 9.6 months on the market in December 212, compared with.4 months at end-211 and 3.8 months at end-21. It is well known that housing needs are very high in Greater Paris and that supply is structurally insufficient, largely due to the lack of building land. Despite all this, demand is falling on a slight drop in sales and the cancellation of some reservations, and the unsold stock is growing. Fairly sharp drop in housing starts In 211, the pace of home building was still fairly sustained on relatively resilient demand and a fairly low level of unsold stock. Housing starts totalled 421,3 units, a 22 increase over the year, with building permits up 18 over the year to 34,8 units. This recovery was, above all in the collective housing sector, and in the social housing sector in particular. In 212, the trend went more clearly into reverse due to a fall-off in sales and an increase in unsold housing stock. Over the full year, housing starts totalled 346, units, a drop of 18 over the year, while building permits were down 7 over the year to 49, units. In the last three months for which data are available, (December-February), housing starts were down 22 over 12 months, and building permits down by 13. This trend is likely to continue in 213. Virtually zero price increases Prices for new-build apartments are not falling, but hardly rose at all due to the fall-off in sales and the increase in unsold housing stock. Prices rose just.8 y/y in Q4 212, compared with a 3. y/y increase in Q4 211 and.1 at end-21. In Greater Paris, prices have fallen quite sharply, by 3.4 over 12 months, compared with a rise of 1. at end-211. Prices are not falling in France due in particular to the need to offset increases in production costs, especially in connection with environmental standards ( low energy consumption building ). These concern reservation prices. 2 17 14 11 8 France: New-build inventory (Greater Paris) 91 94 97 3 6 9 12 Stock for sale 6 4 3 s France: Housing starts and building permits s, cumulative 12 months France: Home prices months Chart 7 Stock by months for sale (rt.sc.) Housing starts Building permits 3 2 2 1 Chart 8 2 1998 2 22 24 26 28 21 212 2 1 - annual change Nnew appartments Existing dwellings Chart 9-1 1 2 3 4 6 7 8 Sources: Ministry of Ecology, INSEE-Notaries 9 1 11 12 No. 36 April 213 4

213-214: drawn-out correction A gradual process The correction on the housing market began in 212. Transaction volumes did not collapse, but fell quite markedly, by 12 over 12 months in pre-owned, and by 18 in the developer new-build sector, including a fall of close to 4 in new-build buy-to-let. Construction fell sharply. Conversely, prices have so far fallen only slightly, by eg, 1.7 y/y in the pre-owned sector. The fall-off is extremely marked as regards housing lending. The (12-month) rate of increase in lending slowed to 3 in the 12 months to December 212 and to 2.4 in February 213 (compared with increases of 6.1 at end-211 and 8.2 at end-21). The production of new loans fell sharply, by 31. in the 12 months to December 212 (cumulative, 12 months). Monthly new lending bottomed between March and May 212, at 6.4 billion euros a month, which can be explained not only by the drop in the housing market but also by a fall-off in remortgaging levels. This is because lending rates rose in 211 and early 212 (4.7 in March 212 for a fixed-rate of more than a year, compared with 3.41 at end-21). Monthly new lending picked up in the second half, to an average 9 billion euros a month, and in early 213 (1.2 billion euros in January and 9.3 billion in February). It is probably that the drop in lending rates since summer 212 is stimulating remortgaging and slowing the decline in new lending. This is still well below the monthly levels achieved in 21 and 211, however, of close to 13 billion euros a month on average. New lending is likely to see a drop of 1 this year and the rate of increase in total lending to slow to 1.. The gentle adjustment process observed in 212 is likely to continue through 213-214. A market correction is normal, and even desirable, following a cyclical upturn of record scope and duration: prices for pre-owned homes have risen by since 1998 and are clearly over-valued. The very poor economic climate and fiscal tightening measures have stopped the upside cycle in in its tracks. But France differs from the United States and most European countries, especially Spain, the United Kingdom and Ireland, due to its favourable structural fundamentals. It is, therefore, quite unlikely that we will see the marked downward spiral observed in these countries replicating itself on the French market. Rather, it is moving towards a gradual, drawn-out adjustment process. 8 6 4 2 France: Mortgage lending to households Bn 98 2 4 6 8 1 12 Amounts Sources: BdF, Crédit Agricole S.A. 16 12 8 4 million Growth over 12m (rt.sc.) France: New mortgage lending Cumulative 12 mths y/y 3 4 6 7 8 9 1 11 12 13 Amounts Growth over 12m (rt.sc.) Sources: BdF, Crédit Agricole S.A. y/y Chart 1 2 1 Chart 11 8 6 4 2-2 -4 Chart 12 Negative cyclical factors The ongoing market correction is mainly due to the combined effect of a steadily increasing unemployment rate and fiscal tightening measures. Near-zero growth and a marked rise in unemployment. GDP growth stagnated in 212 (an annual + by volume) and should be close to zero in 213 (+.1). The unemployment rate is rising month by month and is forecast to stand at an annual 1. in 213 in continental France, compared with 9.9 in 212. In 214, the recovery will be modest, at 1.2 and the unemployment France: Existing apartments sold unoccupied in Paris 4 2 2 4 3 1 3 2-2 -1 83 8 87 89 91 93 9 97 99 1 3 7 9 11 Sales 12-month price variation (rt.sc.) Source: Chamber of Notaries No. 36 April 213

rate unchanged at 1.. Purchasing power is falling, and household confidence is very low. This background makes buyers cautious and apt to wait and see, while some potential buyers have to abandon the idea of buying. Higher taxation on housing. Several fiscal tightening measures implemented in 212-213 are impacting the housing market, namely a sharp reduction in the capital gains allowance on housing (excluding the principal residence); an increase from 13. to. in tax and social security contributions on unearned income; a refocusing of the PTZ+ interest-free loan on new-build housing; a higher wealth tax; and a capping of tax loopholes at 1, a year. The new capital gains regime, especially, is likely to be fairly dissuasive for buyto-let investment in pre-owned housing. Duflot scheme destined for mixed success. The scheme is designed to encourage buy-to-let investment in new-build homes, and replaces the Scellier scheme from 213. It presents a mix of positive and negative aspects. It is attractive from a tax viewpoint as it offers a tax reduction of 18 of the value of the home, compared with 13 under the 212 Scellier scheme. This is a notable benefit at a time when most financial investments are being hit with higher tax and social security levies. But other conditions have been tightened. The scheme only applies to areas suffering housing shortages, rents are capped at around 2 below market prices, tenants must be meanstested, and so on. In addition, when the home is sold on, which is possible after nine years, higher taxation will apply (see above). More generally, frequent changes to tax rules create a sense of instability. Overall, investors are likely to be hesitant and selective. Sales on this segment (buy-tolet developer new-build) could stabilise at 212 levels in 213, but a rebound seems unlikely. Sales prices over-valued and acquisition costs very high. So far, prices have hardly fallen and homes are quite significantly over-valued, the cumulative increase in prices since 1998 having been only partly offset by a drop in lending rates, the increase in mortgage duration and higher incomes. This overvaluation may be estimated at around -2 in the pre-owned segment in view of the levels reached by theoretical affordability ratios 1 of 37 in Q3 212 compared with 3 in 23 (see chart 13). An approach based on risk premiums 2 gives a similar outcome of an over-valuation of around in 212. Mortgage rates have been falling very gently for several months (see below), but the duration of credits is relatively stable and overall, acquisition costs are very high. They have even risen for many first-time buyers due to the abolition of the PTZ+ interest-free loan for pre-owned homes. All this has discouraged some buyers or requires them to make marked sacrifices as regards the size or quality of the home. Note also that as prices start to fall, some potential buyers are expecting lower prices and deferring their purchases, further accentuating the fall-off in demand. 39 37 3 33 31 29 27 France: Theoretical affordability ratio 2 87 89 91 93 9 97 99 1 3 7 9 11 Annual mortgage repayments/income Sources: BdF, Notaries, CA calculations 3 2 1-1 -2 France: Price of new-build homes annual change 91 94 97 3 6 9 12 France 2 2 1 Greater Paris France: Existing housing sales and prices s Chart 13 Chart 14 Chart -1-3 4 6 7 8 9 1 11 12 Sales Change in price over 1 year (rt.sc.) Sources: CGEDD, Notaries 2 1-1 The affordability ratio is the annual repayment cost of a mortgage (principal and interest) divided by the annual income of the purchasing household. The theoretical affordability ratio is calculated using market conditions (prices, interest rates, length of mortgage), but assumes that buyers always have the same income profile and always buy the same type of home in terms of surface area and quality). 2 The risk premium compares the internal rate of return of a property investment with a fixed income rate. In 21-212, we use a smoothed fixed income rate in view of the record low rates during this period. No. 36 April 213 6

Conversely, mortgage lending rates are falling once more. They stood at 3.36 in February 213 for a fixedrate mortgage with a maturity of over one year, compared with 4.7 in March 212, and are likely to stay low, at around 3. over the coming months. Short-term rates will remain stable at low levels:.2 for 3-month Euribor out to end-213. Rates on French OAT 1-year Treasuries are likely to rise only modestly, towards 2.4 by end- 213, compared with 2.2 at the start of the year. Lending standards are likely to remain unchanged or tightened very slightly due to a drop in demand for homes, resulting in more or less stable lending rates at record low levels, which will have a marginally positive impact relative to the negative factors described above. Four positive structural factors 1) Demand for homes is underpinned by solid, lasting drivers: the desire for home ownership (only 8 of French households are owner-occupiers, compared with the EU average of 6), a favourable demographic, family break-up phenomena, preparations for retirement, saturation of the rental market, and, above all, a safe-haven effect. The economic and financial environment is uncertain still, and households are faced with multiple concerns: the poor economic climate, rising unemployment, fiscal tightening. French households are still riskadverse vis-à-vis the financial markets, and the bond market in particular, which are likely to see increased volatility in 213 against a backdrop of political instability in Southern Europe. All this contributes to strengthening the safe-haven image of real estate. Bricks and mortar are reassuring due to their tangible nature, fairly steady trends, and favourable valuations over the long term. 2) Lenders are cautious and lending criteria fairly stringent. Non-performing loan ratios are low, at 3.44 in Q3 212 for consumer lending, and 1.26 in 211 for housing loans. These loans are granted on the basis of the borrower s ability to repay. The ratio of repayments to income must not exceed one third of the borrower s income. Most mortgages are fixed rate and maturities are reasonable (an average 18. years in 211). Unlike some other countries, France has not developed an array of nonstandard loans. Lending supply has accordingly been prudent, and no marked credit squeeze has been necessary. According to recent surveys, lending criteria were unchanged in the second half of 212 and were very slightly tightened in early 213. 3) Housing supply is not excessive and there are even shortages in some sectors, especially in Greater Paris. Developers have been cautious and have adjusted supply to changing demand. Nor should we forget that developer new-build housing stocks are increasing, although only have been completed, 39 are under construction, and 6 are at the planning stage. Supply should not, therefore, be over-abundant in relation to falling demand, unlike the situation in the United States or Spain, for example. 2 1 France: Housing prices and unemployment rate - Housing price rise over 1 year Unemployment rate (inv. rt.sc.) -1 87 89 91 93 9 97 Source: Notaries, INSEE 99 1 3 7 9 11 8 7 6 4 3 France: Mortgage lending rates 2 2 4 6 8 1 12 Sources: BdF, Crédit Agricole SA 27 2 22 2 17 12 1 quarterly average Av. rate for fixed-rate mortgage (APR) 1-year OAT rate France: Real estate and the stock market Deflated series, 1=Q1 1988 7 88 9 92 94 96 98 2 4 6 8 1 12 Chart 16 6 8 1 12 Chart 17 Chart 18 4 3 2 1 France, pre-owned prices Paris, pre-owned prices CAC 4 (rt.sc.) Sources: Notaries, Crédit Agricole S.A. No. 36 April 213 7

4) The factors which led to the emergence of a speculative bubble in 198-1991 are not at work in the current cycle. That bubble was concentrated in Greater Paris. The price of homes in Paris saw a rapid, sharp increase totalling an aggregate 167 and an annual average 19. Prices fell back by 3 between 1991 and 1997. The proportion of property dealers and speculators involved was important. For classic buyers, expectations as regards rents, inflation, and incomes were too optimistic, where they were not simply wrong. Risk management by lending institutions was less stringent. The non-performing loan ratio for household credits saw a significant increase, of up to 7 in 1994. None of these factors was in operation in the 1999-212 boom years. Forecasts These favourable fundamentals mean that demand is fairly resilient and make a market collapse scenario unlikely. One can see that sales have fallen significantly, but have not collapsed. They are likely to fall by around 2 over two years (212-213). In Spain, when the correction began, sales fell by close to over two years (28-29). Housing supply is tending to increase (ie, the stock is growing), but remains relatively moderate. In the pre-owned sector, some potential sellers are loath to sell and to accept marked price falls in the face of higher capital gains taxation. They prefer to take their asset off the market. This is all resulting in a gradual adjustment in prices. If the erosion in sales continues, stocks of unsold homes will increase and prices should logically fall more sharply. But that will depend on seller behaviour. In 213, sales should stabilise in the new-build sector and fall by around 1 in the pre-owned segment. Pre-owned prices are forecast to fall by -6 y/y, for a total fall of 8 between end-211 and end-213. In 214, the trend is likely to continue, with sales stabilising at a low level and prices again falling by around. One of the main risks in relation to this central projection would be that of a sharp rise in French bond yields and hence in mortgage rates, which would trigger a more pronounced falloff in sales and prices. This scenario is fairly unlikely. For one thing the eurozone has implemented effective shielding mechanisms, namely the ECB s debt purchase programme and the permanent rescue fund (EMS), to ward off panic reactions on the markets. France is also likely to continue to be seen by investors as one of the most reliable sovereign names in the eurozone, especially in view of the large-scale economic difficulties and risks of political instability in Southern European countries. 6 4 2 France: Risk premium on real estate risk -2 199 1994 1998 22 26 21 Sources: INSEE, BdF, Crédit Agricole S.A. 14 1 6 2-2 -6 France: Existing home prices and stock of new-build homes -1 87 89 91 93 9 97 99 1 3 7 9 11 12-month price increase New-build stocks by months for sale (inv. rt.sc) Sources: Notaries, BdF 9 8 7 6 4 months France: Existing housing sales and prices s forecast Chart 19 Chart 2 4 8 12 16 2 Chart 21 2 1 - -1 3-92 94 96 98 2 4 6 8 1 12 Sales 12-month price increase (rt.sc.) Sources: CGEDD, Notaries, Crédit Agricole S.A. No. 36 April 213 8

Germany: counter-cyclical Untouched by the 1998-27 boom Between 1998 and 27 there was a sustained housing boom in almost all European countries, with, in some cases, the formation of twin housing and credit bubbles. Germany escaped this housing boom. The cumulative increase in housing prices was very low between 1997 and 27, at just, compared with in France and around 2 in Spain and the United Kingdom over the same period. Similarly, mortgage lending was virtually unchanged during that time, rising at just 1 a year between 24 and 27, compared with a 14 annual increase in France. There are several reasons behind the sluggish housing market: Low population growth, limited increases in incomes, and fairly gloomy household confidence levels on average throughout the period. During Germany s reunification process, and as a result of the tax incentives implemented at that time, a property boom developed in the 199s. Price rises were relatively limited, but home building surged spectacularly and created over-supply. Building permits, which totalled 37, units in 199, rose to 7, in 1994 and ran at 8, a year in 1993-1998. German households preference for tenancy as opposed to ownership (see following paragraph) and a fairly weak desire for home ownership: the owner-occupier rate is only 46 in Germany, the lowest level in the European Union (an average 6 in the EU 27). A limited number of second-home investors (sellers/buyers) and a fairly low number of transactions (, in 211 in the preowned sector, compared with 8, in France). In general, German households prefer to wait to buy the home corresponding to their aspirations and live in it for much of their lives, unlike French households, many of whom buy and sell their principal residence several times over during their lifetimes, depending on their income and family size. 14 13 12 11 1 Germany: Housing prices Germany: Building permits Chart 22 9 9 92 94 96 98 2 4 6 8 1 12 Source: BulwienGesa AG 8 6 4 2 s 12 mth. cum. Housing prices, 1= 199 Chart 23 9 92 94 Source: OECD 96 98 2 4 6 8 1 12 Especially favourable status for tenants The proportion of tenants is very high, at 4 of households, compared with an EU average of 3). This is because tenants enjoy an especially favourable status: The model lease is unique, and not time-limited (since 1971). The only possibility for a landlord to reclaim possession of the property is if he wants to occupy it himself. In practice, this is little-used. Tenants can terminate the rental agreement with three months notice. Rents are fixed freely, but rent increases are in general fairly limited when leases are renewed, since the housing supply is relatively abundant and the tenant can apply to a court if he feels the rent is excessive. Chart 24 Owner-occupiers in Europe as of total households 8 6 4 2 Source: 27 data, Eurostat No. 36 April 213 9

The rental market is also favoured by special tax measures (interest on buy-to-let investments is tax deductible). This results in modest rent increases and a fairly active rental market. Note also that the proportion of social housing is low, involving only 6 of households, compared with 17 in France. Conversely, the desire for home ownership is limited by several factors: Very low price rises over the past years, which means that the outlook for capital gains is poor and wealth effects are very limited; The benefits of tenant status, as described above; The personal deposit required to obtain a mortgage is quite high; Taxation is not particularly favourable for buyers: mortgage interest is not tax deductible, and transfer taxes are high. Chart 2 Mortgage debt (mortgage debt/household income) 12 9 6 3 26 27 28 29 21 211 212 France Germany Spain UK Source: Central Banks Market now picking up The German market is counter-cyclical. It was sluggish during the global real estate boom. The recession of 28-29 brought the boom to an end, and housing markets almost everywhere are undergoing a correction, with cumulative price falls of 2 in the United Kingdom, of 27 in Spain and in Ireland. Conversely, in Germany, the market has been picking up since 21. House prices are rising, with increases of 2 in 21, in 211, and 4 in 212, for a total 12 over three years. The upturn in demand can be ascribed to the recovery in Germany s economy, low price levels (and which are probably fairly significantly undervalued) and attractive financing conditions, due in particular to low interest rates. In addition, the over-supply has gradually been absorbed. The construction industry is gradually picking up after shrinking significantly in recent years, to, housing starts a year, compared with over, a year in the 199s. In 211, housing starts rose to 183, units, a increase over 12 months, and are seeing further growth in 212. What we are seeing, however, is more a normalisation than the emergence of a marked boom. The recent increase in prices has been modest in comparison with the changes seen in Europe during the first decade of this century. Total mortgage lending has been growing very modestly, at 1.2 per year in 211-212. What s more, the mid-term outlook is fairly unsettled. The economic climate is poor in 213 and GDP growth is unlikely to top 1. a year in 214-216. Germany s population should also contract slightly in the years ahead, limiting the corresponding need for housing. The desire to own one s own home seems likely to remain quite limited. And, lastly, there is no major mismatch between housing supply and demand. The German housing market should thus see a modest uptick in sales, prices, and new construction in the coming years. The threat of a bubble seems very unlikely. Chart 26 Germany: Mortgage lending to households y/y 4 3 2 1-1 -2 24 2 26 27 28 29 21 211 212 Source: ECB Rise in lending over 12 months No. 36 April 213 1

French property market indicators Residential Non-residential Existing New-build Sales Demand Prices Supply Construction Vacancy rate Rental values Completed on 24 April, 213 Crédit Agricole S.A. Group Economic Research 12 place des Etats Unis 92127 Montrouge Cedex Chief Editor: Isabelle Job Editorial board: Olivier Eluère Sub-editor: Fabienne Pesty Contact: publication.eco@credit-agricole-sa.com Website: http://economic-research.credit-agricole.com Subscribe to our free online publications This publication reflects the opinion of Crédit Agricole S.A. on the date of publication, unless otherwise specified (in the case of outside contributors). Such opinion is subject to change without notice. This publication is provided for informational purposes only. The information and analyses contained herein are not to be construed as an offer to sell or as a solicitation whatsoever. Crédit Agricole S.A. and its affiliates shall not be responsible in any manner for direct, indirect, special or consequential damages, however caused, arising therefrom. Crédit Agricole does not warrant the accuracy or completeness of such opinions, nor of the sources of information upon which they are based, although such sources of information are considered reliable. Crédit Agricole S.A. or its affiliates therefore shall not be responsible in any manner for direct, indirect, special or consequential damages, however caused, arising from the disclosure or use of the information contained in this publication. No. 36 April 213 11