FRANCE Housing Market: Recent developments and outlook for

Transcription

1 Quarterly No./12 23 April FRANCE Housing Market: Recent developments and outlook for -16 Household solvency improving The correction on the French housing market continued through 14. It was a measured correction in the pre-owned sector, and more pronounced in new-build. In the pre-owned sector, some 7, homes were sold. The market is not collapsing and is down only 14 relative to the 6-7 peaks. Prices have continued to edge down, by 2.2 over twelve months, and by a cumulative 7 since end-11. In new-build, the correction is significantly more marked, with a 4 drop in sales over the year and a cumulative 32 drop relative to 6-7. Over the same periods, new housing starts were down 11 and 27 respectively. Some still-fragile signs of a recovery were perceptible in new-build in early as buy-to-let investment picked up, driven by purchases under the Pinel scheme, and as sales of houses rose and developer morale improved slightly. Conversely, there could be a slightly more marked adjustment in the pre-owned sector in. It s worth noting, however, that the ongoing French market correction is gradual and modest far more so than in other European countries. This exception is due to a combination of negative cyclical factors and positive structural fundamentals. Three negative factors continue to erode demand: higher unemployment, fiscal measures, and excessive price levels. These negative factors are partly offset by the very low level of lending rates, structural support for demand, and the advantages of the French mortgage lending model (cautious offer, no rise in non-performing mortgages). In -16, the factors operating will be more or less the same, with two possible differences in that mortgage lending rates could stop falling and stabilise, while the support measures for new-build housing could drive a slight recovery in the sector. Volumes look set to ease back in the pre-owned sector, and prices are forecast to fall by 2. per year, while developer new-build sales could rebound by a year. Prices are still high, but are less and less overvalued. In view of our indicator, (household ability to purchase homes), presented in the Focus on pages 7-8, the falls in prices and lending rates in recent quarters have brought homes closer to their fair value. If this trend were to continue over the next two or three years, prices would be clearly under-valued and attractive, and household solvency levels would improve. This, combined with a drop in the unemployment rate, would trigger a new upswing in real estate prices from 18. Beware, however, of rising lending rates, which will be a decisive factor. Contents Recent trends in the pre-owned and new-build sectors2 Stable, relatively high sales in pre-owned in Modest falls in pre-owned prices in Further falls in new-build sales in New housing stock levels high, prices down slightly... 3 Record lows in housing starts, building permits : market to continue edging down... 4 The gentle correction continued in Negative factors... 4 Favourable factors... 4 Why the bigger correction in new-build than in pre-owned? : correction to continue gradually... 6 Two potential problems persist... 6 FOCUS Housing prices less and less overvalued... 7 How do you manage price over-evaluation?... 7 Developments in Developments in Towards a newly solvent market around Études Économiques Groupe

2 Recent developments and outlook for -16 Recent trends in the pre-owned and new-build sectors Stable, relatively high sales in pre-owned in 14 For the past three years, the number of transactions in the preowned housing sector has been more or less stable at a high level. After two years of sustained activity, with close to 8, sales a year in -11, sales fell 12 in 12, with 74, transactions recorded, and recovered slightly in 13, with 717, unit sales, a rise of 1.8 over the year. In 14, unit sales totalled 7,, down 2.4 over the year. There were slightly fewer sales in the second half, at, a month, compared with 62, in the first. It is too early to talk of a lasting downtrend, however. In January, the 12-month total was unchanged, at 7, units (source: Conseil Général de l Environnement et du Développement Durable based on fiscal and notary databases). The pre-owned housing sector is thus proving quite resilient, despite the lacklustre economy and tax measures. In comparison with the high points of the recent upswing (8, sales a year in 6-7), the current level is not sharply down, but is 14 off. In the Greater Paris (Ile-de-France) area, sales of pre-owned homes, which were down in 11 and 12, also picked up slightly in 13, with a rise of 3 for pre-owned apartments, to 91, units. In 14, sales of pre-owned apartments eased back at little, dropping 2 over twelve months, to 89, units. Modest falls in pre-owned prices in 14 Prices for pre-owned homes continued to fall very slowly. They had continued to rise in and 11, respectively by.1 and 6 annually, but began to fall in 12, by 2.1 over the 12 months to Q4, and in 13, by 1.9 over the 12 months to Q4. We saw a similar drop in 14, at 2.2 in the 12 months to Q4. As with sales, the drop was slightly more marked in the second half than in the first, at 2.2 y/y in Q4, and 1.3 in Q2. The cumulative fall in prices in the preowned sector since the start of the correction in late 11 has, therefore, been very small, at 7, compared with a cumulative rise in prices between 1998 and 11. The trends are fairly comparable across the large regions. In Q4 14, prices fell by 2.3 y/y outside greater Paris; by 2 y/y in Greater Paris; and by 2.1 y/y in Paris itself (source: Insee, Notaries). Given the data available for certain regions, in Q3 14, prices fell by 2 y/y in Nord-Pas de Calais and Provence-Alpes-Côte d Azur regions. Prices fell by.8 in Rhône-Alpes. On a city-by-city basis, the trends are more contrasted: in Q4 14, prices fell by 13 in Caen, 11 in Reims and Toulon, and in Saint Etienne. Conversely, prices rose by 3.3 in Toulouse, 2. in Strasbourg, and 1.3 in Bordeaux. According to the Notaries of Paris/Greater Paris, prices are expected to have fallen more markedly in Q1, by 3.8 y/y for apartments in Greater Paris, and by 3. in Paris, partly due to a base effect. Further falls in new-build sales in 14 Sales of new homes continue to fall. Developer sales (developers, programmes of at least five homes) experienced a fairly marked correction in 11 and 12, respectively of 8 and over 12 months. Full-year unit sales stabilised in 13, at 89,3, up.3 compared with 12. Sales were again down in 14, to 8,8 units, a drop of 4. Relative to the highs of 6-7, at 127, sales a year, the current level is sharply down, by 32. We must also mention an increase in on-plan reservation cancellations totalling 2 France: Existing housing transactions Est. number of transactions per quarter Sources: CGEDD, Notaires, CA y/y - - France: Existing housing prices change over 12 months Q1 = index Sources: INSEE, Notaries, Crédit Agricole S.A France: Sales of new-build homes s per quarter On market Source: Ministry of Ecology Sales index No./12 23 April 2

3 Recent developments and outlook for -16 8,7 in 14 a increase over 12 months (withdrawals relative to sales reservations made in previous quarters). However, in early, there are signs of a recovery in buy-to-let sales. The trend is even more pronounced when it comes to the sale of individual, non-developer homes. These stood at around 99, units in 14, a drop of 4 over 12 months and 48 down relative to the highs of 6-7 (189, unit sales a year). There was a slight improvement in January, with a 12 y/y increase to 7, sales. Obviously, this improvement is fragile and needs to be confirmed. In Greater Paris, new-build (developer) sales picked up smartly, by., in 13, then stabilised in 14, with a rise of 1.2 y/y. This was partly offset by the level of sale cancellations. According to the INSEE survey of developers, the trend in demand for new-build homes was still highly degraded in January, and well below its trend level, but has improved slightly relative to October 14, at -49, compared with -4. New housing stock levels high, prices down slightly The number of houses up for sale which was still quite high in 12, fell sharply in 13, by 13 y/y, to 4, units. There was a further correction in 14, with a drop of 13.3 y/y to 89,9. In addition, a number of construction programmes were cancelled. Developers have reduced their supply due to the fall-off in sales and the fairly high level of housing stock up for sale. This number rose to 6,7 units in mid-14, then fell back slightly to 3,3 units in Q4 14. The ratio of stock to sales, which measures the average time it takes to sell a home, stood at 14.4 months on sale, slightly less than in the previous quarter s.6 months, but significantly higher than the average level of 8.3 months in Inventories per se, ie, the stock of homes up for sale that are finished or under construction, are far lower, however, at 4, units: 7 of the stock is completed, 37 under construction, and 6 at the planning stage. In Greater Paris, the stock of homes continues to rise. By months on sale, the ratio amounted to 12.9 months on the market in Q4 14 compared with 11.4 months at end-13 and 9.6 months at end-12. Prices for new-build homes have reacted only very slightly to the drop in sales and to the high stock of homes for sale. Prices for apartments were down a little in Q4 14, by. y/y, compared with 1.3 in Q4 13. Prices for houses fell by 2 y/y in Q4 14 compared with a.9 rise in Q4 13. Record lows in housing starts, building permits Home building, which was still sustained in 11, with 423, new housing starts, fell in 12, by.7 over 12 months, and stabilised in 13, (+.7 over 12 months). The rate of fall accelerated in 14, with new housing starts falling by 11.3 over 12 months to 36,, while 376, building permits were issued fewer. There has been a very sharp, 27, fall in new housing starts since the 7 peak of 489,. Readers will notice that the statistical series has been revised upwards, in particular for -14, due to the use of a new methodology. This includes estimates (using statistical and econometric methods) for building permits and housing starts in order to offset the time it takes (several months) to extract data for building permits, along with the lack of exhaustive data for housing starts Stock for sale Source : Ministry of Ecology 4 3 s France: Inventories of new-build homes months Stock by months for sale (rhs) France : housing starts and permits thousand, aggregate12m housing starts Source : Ministry of Ecology - - a/a France : housing prices permits Newhousing units Existing housing units Source : Ministry of 'Ecology, Insee-Notaries 7 3 No./12 23 April 3

4 Recent developments and outlook for : market to continue edging down The gentle correction continued in 14 The French housing market has been undergoing a correction since 11. Housing is a highly cyclical sector and it is logical that there should be a downswing after the boom years of The trend is fairly measured in the pre-owned sector, and more pronounced in new-build. In 14, the pre-owned market was again fairly resilient, with a high level of transactions (7,) and a measured, 2.2 fall in prices over the full year, for a cumulative 7 drop since end-11. The correction has been far more marked in new-build, with a 4 fall in sales over the year, and a 32 fall relative to the highs of 6-7. Given the proportional importance of transactions in the preowned sector (three-quarters of all sales), overall, the market is only moderately down. France is still a case apart in Europe, where in most countries (Spain, Italy, the Netherlands, the UK and Ireland), there has been a substantial correction in sales and prices since 8. The French soft landing can be explained by a combination of generally negative economic environment factors and positive structural fundamentals. 3 Housing price Indices Index base =1997Q UK Spain France Ireland Sources : Halifax, Ministerio de Fomento,Insee, DS Negative factors Low growth and rising unemployment GDP growth was very modest in 14 at an annual.4 by volume, and is forecast to pick up, but only modestly, by 1.1, in, and by 1.3 in 16. The unemployment rate stood at 9.8 in 14, is forecast to continue rising, to.2 in, and to ease back very slightly, to.1 in 16. Expectations around incomes are extremely low, and some first-time buyers are abandoning their plans to buy. France : housing prices and unemployment rate 6 8 Fiscal measures Fiscal measures had a negative impact between 12 and 14 due to capital gains on real estate and the PTZ interest-free loans, and the very mixed success of the Duflot scheme: despite the tax break it offered, it came with fairly tight restrictions on rent ceilings and tenant solvency requirements. A raft of support measures was announced in 14, including a widening of access to the PTZ interest-free loan, an easing of the Duflot scheme (since renamed the Pinel scheme) with 6, 9, and 12-year packages, a simplification of standards, limits on rent controls, and a 3 reduction in capital gains on land sales until the end of. Overvalued selling prices Selling prices are still overvalued, but less and less so. They were overvalued by close to in early 12. As a result of falling prices and lower mortgage lending rates, they were only overvalued by - at end-14 (see Focus, page 7). Prices are still high, however. They need to fall still further and become clearly undervalued to be seen as genuinely attractive and trigger a new uptrend and a return of first-time buyers. For the time being, high prices are a brake on demand. Favourable factors Mortgage lending rates at record lows Rates fell throughout 13 to an annual average of 3.2 and fell again in 14, to as low as 2.6 in December. This decline in the cost of borrowing is driven by the steady fall in OAT rates. Very low mortgage rates, coupled with the risk that sooner or later they will yearly price increase unemployment rate (inverse right scale) Sources : Notaries, Insee Source : 11, Eurostat Ownership ratio in Europe (as a of total households) 12 No./12 23 April 4

5 Recent developments and outlook for -16 rise, are encouraging some hesitant households to buy despite the high prices and this is propping up housing demand. Structurally high demand for housing This is driven by solid, lasting factors including the desire for home ownership, a dynamic demographic, the family separation phenomenon, preparations for retirement, saturation in the rental market, and the safe haven effect. The very low level of shortand long-term interest rates is making real estate (seen as less risky, more reassuring and more profitable than most financial assets) a more attractive proposition. A cautious lending offer Mortgage lending criteria remained strict during the cyclical upswing of the noughties and remained so thereafter. There was no surge in lending or excessive risk-taking, unlike in other countries. The ratio of non-performing loans was low and fairly stable, at 3.3 in Q4 14 for consumer lending, and 1.4 in 13 for mortgages. It has not been necessary, therefore, to sharply tighten lending. Mortgages are granted mainly on the borrower s ability to repay. The affordability ratio must not exceed onethird of the borrower s income. Most mortgages are fixed rate and maturities are reasonable. The cautious mortgage lending model largely explains why the correction is more gradual and more moderate than in most European countries. Recent surveys suggest that mortgage lending criteria are more or less unchanged. Why the bigger correction in new-build than in pre-owned? Developer new-build sales have fallen by 32 from their 6-7 highs, and sales of new, non-developer houses, by 48. There are several reasons for the trend. In the developer new-build sector, the fall can mostly be laid at the door of buy-to-let investors, who form a substantial proportion of buyers, with sales down 4 since 7 due to fiscal measures, including the limitations of the Duflot scheme (rent ceilings and tenant solvency requirements) and capital gains measures. This made buy-to-let investors more hesitant. The relative weight of this segment fell to 44 in 14 compared with in 7. The home-buyers segment is more resilient, but is still down 26 from its 7 level, driven by high unemployment, restrictions on the PTZ+ interest-free loan, high prices, and exacting building standards governing energy use, environmental factors, accessibility, and so on. These negative factors had an even greater impact on new, non-developer individual houses. Pre-owned sales have been more resilient and have fallen only 14 relative to 6-7. This can partly be explained by the high numbers of second-time buyers people selling one principal residence to buy another. Sales on this segment are proving highly resilient, as structural support factors such as preparations for retirement or the safe-haven effect are vital, while high prices are less of a deterrent. This is a growing segment, involving of new lending at end-13 compared with 4 in early Conversely, the proportion of first-time buyers has fallen, from 32 in early 12 to 18 at end-13. High unemployment, uncertainties about future income, excessively high prices, and the abolition of the zero-interest loan on pre-owned houses have all had a major impact on this segment and in particular on young households on modest incomes. The market has shifted in favour of second-time buyers, and to a lesser degree towards buy-to-let investors. Prices are down only moderately due to the fairly high number of sales and relatively modest supply. 2, 1,8 1,6 1,4 1,2 1,,8,6,4,2 France: non performing loans/total credits for housing loans, Source : ACPR 3 France: Unoccupied existing houses sold in Greater Paris Number of transactions Sources: Chamber of notaries France: Sales and inventories Sales s of new-build homes months for sale Sales On market Source : Ministry of Ecology Data for 14 are not yet available. No./12 23 April

6 Recent developments and outlook for : correction to continue gradually The correction is likely to continue, slowly and steadily, in and 16, as the same positive and negative factors should continue to operate. Two points are worth looking at more closely, however. Mortgage lending rates could stabilise. These were slightly down again in early, at 2. in February. They could stabilise over the coming quarters in connection with the stabilisation or very slight rise in -year OAT rates, which means that lending rates will continue to be highly attractive. However, the very fact that lending rates will have stopped falling could mean that prospective buyers might sit tight, leading to a slight drop in sales in the pre-owned segment, slightly more marked falls in pre-owned sales and in prices. The new housing support measures should halt the decline in sales in the developer new-build sector and lead to a small recovery, especially in the buy-to-let segment, driven by the new provisions of the Pinel scheme. A few encouraging signs were, in fact, perceptible in early, namely a recovery in house sales and a pick-up in sales to buy-to-let investors 2. In, pre-owned transaction volumes are forecast to be down slightly. Pre-owned home prices look set to fall 2. y/y, compared with a drop of 2.2 at end-14. Developer new-build volumes are forecast to recover somewhat, however (by ): + for buy-to-let investors and for home buyers. In 16, the trends are likely to be similar, which would amount to a cumulative 12 drop in prices between 11 and 16. Two potential problems persist The market should, therefore, continue to adjust gradually in and 16 and the trend is even likely to continue in 17. A scenario where housing demand collapses seems unlikely, but two concerns should nevertheless be raised. For one thing, first-time buyers are partly excluded from the market. The proportion of such buyers has fallen sharply and should remain low in. The market is treading water, demand is not being renewed, and could fall, and sales/purchases by second-time buyers cannot go on indefinitely. A genuine recovery in the market will require first-time buyers to return, and for that to happen they have to improve their solvency level. This process is ongoing, but is unlikely to be complete before (see page 8). Secondly, the market is extremely, if not overly, dependent on lending rates. The steady decline in rates and their low level are stimulating demand, but the cost of credit is historically low and longterm interest rates and lending rates are likely to rise, sooner or later. As detailed above, our scenario is for a stabilisation over the coming quarters followed by a drawn-out, gradual rise. However, we cannot rule out a scenario where French long-term rates rise more rapidly, in the event of a surge in investor concerns about French sovereign risk, for example, even if this scenario seems unlikely to materialise. If it did, the subsequent rise in lending rates would trigger a fairly marked correction to real estate sales prices France: new mortgage lending breakdown Mar-12 Sep-12 Mar-13 Sep-13 First-time buyers Second-time buyers Remortgaging Rental investment Source : ACPR, Crédit Agricole S.A France : existing housing prices and stock of new-build homes months yearly price increase new-built stocks by months for sale (inv. rhs) Source : Notaires, BdF s Sales Sources: CGEDD, Notaries France: Existing housing sales and prices Change in price over 1 year (rt.sc.) Official data are not yet available. No./12 23 April 6

7 Recent developments and outlook for -16 FOCUS Housing prices less and less overvalued Are prices still significantly overvalued, and how to you gradually restore buyer households to solvency, especially first-time buyers? The most recent data have somewhat modified our conclusions. How do you manage price over-evaluation? The overvaluation of a price is the differential between the observed price and the equilibrium price, or fair value, which reflects a balanced market. Price is thought to be determined by fundamental variables such as demographics, income, rents, rates, etc., and not by cyclical variables that lead to wide fluctuations in real estate markets, from growth, to unemployment, false expectations, speculative trends, unsuitable housing supply, and lending supply that is either too accommodative or not accommodative enough, etc. There are a variety of approaches, none of which is perfect. The most commonly used are the ratio of prices to rents, and the ratio of prices to income. The Friggit Tunnel, based on an analysis of the long-term trend in the ratio of home prices to income per household, does not take financial variables into account. However, the vast majority of real estate purchases are made on credit, so lending rates and terms are two essential factors that influence both demand and prices. This is why we think they should be factored in when calculating overvaluation. Econometric approaches pose a variety of problems. It is difficult, for example, to establish an econometric link between prices and several explanatory variables; nor is it certain that these variables take in the fundamental factors. The risk premium (ie, the difference between the internal rate of return on a buy-to-let investment and bond rates) was commonly used in the recent past. A risk premium of around 2 corresponds to correctly evaluated prices. But it is very difficult to use at present due to the extremely low level of bond rates. To calculate the over-valuation of prices, we use an approach based on solvency ratios. This ratio is equal to the annual repayments (capital and interest) for someone taking out a new fixed rate mortgage as a proportion of their annual income. The household in question earns the average French household income. At each period, we calculate the surface area of the home they could buy at market prices, rates, and mortgage terms, with a solvency ratio of 3 (a rule applied fairly strictly by French credit institutions). The calculations are made using pre-owned home prices. The indicator therefore measures a household s ability to purchase a home. Over the period, we observed that the surface area that could be purchased by the average household changed cyclically around an average surface area of about 9sq.m. The approach consists in saying that prices are correctly evaluated if the average household is able to purchase this average surface area. Prices are over/undervalued if the surface area is lower/higher than the average surface area. When prices are overvalued, the household has to sacrifice surface area. Overvaluation is measured in terms of the difference between the average surface area and the affordable area France : mortgage loan rates average housing loan rate Sources : BdF, Crédit Agricole Source : Chamber of Notaries y OAT rate France: Existing apartments sold unoccupied in Paris y/y No. of sales per quarter Price rise per sq.m. (rhs) Developments in In the previous real estate cycle ( ), following a period of sharp increase ( ), between 1991 and 1997 the market corrected very rapidly in Paris and Greater Paris (where a form of speculative bubble had developed), but more slowly and gradually in the rest of the French pre-owned housing market. Prices for preowned homes fell cumulatively by around 7 between 1991 and No./12 23 April 7

8 Recent developments and outlook for a modest fall. However, the combination of this slow decline in prices, falling lending rates and rising incomes led to a fairly marked improvement in household solvency in light of the indicator described above. The affordable surface area rose from 73sq.m. in 1991 to 112sq.m. in The market began to recover in 1998 for a number of reasons, but largely due to the fact that prices had become attractive once more and household solvency had improved. First-time buyers on low incomes gradually returned to the market. Existing housing prices sq.m / euros fcast 2 Developments in 8-14 History is not necessarily repeating itself, but a similar scenario seems to be playing out in the current phase of the cycle. The prices of preowned homes were significantly overvalued, by, at the top of the cycle in early 8. They were again markedly overvalued in early 12, by 18 (the decline in prices observed in 8-9 was offset by a rebound in -11). The market correction phase really began in late 11-early 12. Since then, the cumulative fall in prices currently stands at 7. Prices were still overvalued, by around, in 13. But household solvency has been restored more rapidly than expected. With the drop in prices, and above all the decline in lending rates over recent quarters, at end-14, early, prices are hardly overvalued at all at between and ). They therefore seem close to their fair value. This does not mean that first-time buyers will very shortly be piling en masse into the market. As we saw in the previous cycle, prices need to adjust further, and become clearly under-valued before they are felt to be really attractive and trigger a new market uptrend. Towards a newly solvent market around 18 If prices continue to decline gently in -17, at a rate of 2. a year, we will have seen a cumulative fall of 14 since 11. Assuming that lending rates remain more or less stable for three years at the early- level (quite a big assumption, obviously), and there is a steady acceleration in nominal household incomes, household solvency will see a gradual improvement. The affordable surface area could be around sq.m. by end-17, compared with 77sq.m. at end-11 and 83sq.m. in early 14. This factor, combined with an improvement in the economic environment and lower unemployment could kick off a new cyclical real estate upswing from 18. This is only a simulation, of course. The aid used to calculate price overvaluation is an over-simplification, but other methods can also be used. What happens to interest rates will be decisive, and buyer solvency is likely to deteriorate once more in the event of a significant hike in lending rates. The new-build market, which has its own special characteristics, could move differently. However, in the framework of the assumptions used, and allowing the pre-owned market to gradually adjust, we could see buyers becoming more solvent and a pick-up in the housing market getting off the ground within two or three years yearly increase, sq.m/euros Source : Fnaim, Notaires, INSEE, Crédit Agricole SA sq.m Household ability to purchase homes repayment to income ratio at 3 Source : BdF, Notaries, calculations CA fcast No./12 23 April 8

9 Recent developments and outlook for -16 Completed on April Crédit Agricole S.A. Group Economic Research 12 place des États-Unis Montrouge Cedex Publication manager: Isabelle Job-Bazille Chief Editor: Jean-Louis Martin Information centre Gisèle Walter - Statistics: Robin Mourier Sub-editor: Fabienne Pesty Contact: publication.eco@credit-agricole-sa.fr Consult Economic Research website and subscribe to our free online publications: Website: ipad : Etudes ECO application available on App store platform Android : Etudes ECO application available on Google Play platform 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./12 23 April 9