Home prices France June 2012€¦ · 1 Home Prices in France Over the Long Run Jacques Friggit,...
Transcript of Home prices France June 2012€¦ · 1 Home Prices in France Over the Long Run Jacques Friggit,...
1
Home Prices in France
Over the Long Run
Jacques Friggit, CGEDD, French Ministry in charge o f Housing. Presentation, June, 2012.The analyses and points of view expressed are the a uthor’s, and, in particular, not necessarily CGEDD’s or the government’s.http://www.cgedd.developpement-durable.gouv.fr/rubr ique.php3?id_rubrique=137
2
Preliminary• « CGEDD » = “Conseil général de l’environnement et du
développement durable” = internal audit and prospect ive
department common to the ministries in charge of th e
environment, sustainable development, energy,
transportation, etc. and housing
• Accent on long term perspective
• Various papers, presentations, data series, sources , monthly
updates may be downloaded onhttp://www.cgedd.developpement-durable.gouv.fr/rubr ique.php3?id_rubrique=137
3
PLAN1. Home prices in France, a Historical
perspective
2. Comparison with Other Assets
3. Several Important Properties of Home Prices
4. How can we Explain the 2000-2010 Rise?
5. Home Price Prospective
4
Home price indices in Paris since 1200
Source: CGEDD after d’Avenel, Duon, INSEE, indices Notaires-INSEE and notaries’ databases
Home price indices in Paris since 1200Constant currency
Basis 2000=1
0,001
0,01
0,1
1
10
1200 1300 1400 1500 1600 1700 1800 1900 2000
Centre of Paris 1200-1790
Centre of Paris 1790-1850 before adjustment for obs olescence
Paris 1840-2009 before adjustment for obsolescence
Paris 1840-2009 after adjustment for obsolescence
Divisionby 4
Slope +0,6% per yearon average over 800 years
Little significance
Divisionby 15
1348Great Plague
WW I and IIRent controls
Inflation
100 year war
5
1840-2011: the 1914 -1965 depression
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Duon, Toutain and Villa (CEPII).
Home price indices, France and ParisConstant currency, basis 2000=1
0,01
0,1
1
10
1800 1850 1900 1950 2000
FranceParis
Rent controls+ inflation
Exit from
rent
controls
1914 1948 19651870
defeat
1930-35
« refuge »
out of
stocks 1940-44
« refuge »
against
inflation
1840-1914 1965-20111914-1965
6
What are we talking about?Beware quality/structural effects!
Source: CGEDD after INSEE and notaries’ databases
1.68
3.08
3.12
1.44
1.04
0
1
2
3
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Disposable income per householdExisting-home price index, FranceHousing expense per household (from National accoun ts)Rent index, FranceConstruction cost index
Constant currency, basis 1965=1
(average value, stock)(index, flow)
(average value, stock)(index, stock)
(index, flow)
7
Example of quality effect : surface
Source: CGEDD after housing surveys
Surface par logement, nombre de personnes par logem ent et surface par personne
7073
7983
85 87 89 90 92
2325
2831 33 34 36
3840
3,12,9
2,82,7 2,6 2,5 2,5
2,42,3
0
10
20
30
40
50
60
70
80
90
100
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Surface
0,0
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
4,5
5,0
Nb de personnes par logement
Surface par logement (m²)
Surface par personne (m²)
Nombre de personnes par logement (échelle de droite)
Year of the housing survey
Surface per dwelling, number of persons per dwellingand surface per person
Nbr of personsper dwelling
Surface per dwelling (m²)Surface per person (m²)Number of persons per dwelling (right scale)
8
Stability from 1965 to 2000 then take off ofthe home price index relative to income per
household
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE indices.
Home price index relative to disposable income per household
France, basis 1965=1
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1,9
2
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
Home price index relative to disposable income perhousehold, basis 1965=1 Auxtunnel 0
Tunnel
9
Local differentiation
Source: CGEDD after INSEE, notaires’databases, Notaires-INSEE indices.
Home price index
relative to disposable income per household
Differentiation Paris / Region of Paris / ProvinceBase 1965=1
1.83 (France, Q1 12)
2.54 (Paris, Q1 12)
2.14 (Paris Region, Q1 12)
1.69 (Province, Q1 12)
0,7
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1,9
2
2,1
2,2
2,3
2,4
2,5
2,6
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
FranceParisRegion of ParisProvince
Tunnel
NB: the divider of all four ratios is the disposable income per household over all of France
Exception = « crisis »
10
2000-2010: heterogeneity of home pricegrowth in the various « departments »
The extremes (growth from 2000 to 2010 in the existing-home price indices):•The 3 smallest increases: Territoire de Belfort: +59%; Ha ut-Rhin (deptt ofColmar): +64%, Moselle +70%•The 3 biggest increases: Bouches-du-Rhône (=departt of M arseilles), Paris, Alpes-Maritimes (departt of Nice): (in a draw) +138%•(France: + 107%)(Source: Notaires-INSEE indices and Perval)
Higher 2000-2010 growth if:-More secondary residences-More private rented principal residences-Fewer owner-occupied principal residences-Lower construction rate (elasticity ~ -1 or -2)-Higher population growth (elasticity ~ 1 or 2)-Higher income growth (elasticity ~ 1)-Lower unemployment growth-Results not robust with respect to the period studiedDetails in the paper:http://www.cgedd.developpement-durable.gouv.fr/IMG/pdf/difference-variation-prix-immobilier-par-departement_cle76a2da.pdf
Variation totale
1,056 à 1,381 (30)0,906 à 1,056 (32)0,587 à 0,906 (32)
2010 home price indexBasis 2000=1
11
2000-2010: inversion of apartments / indiv. houses differentiation
•appreciated from 1950 to 1965 (exit from rent controls , which
• had impacted apartments more than individual houses)•depreciated from 1965 to 2000 (while occupants weregetting poorer)•appreciated from 2001(while occupants went on gettingpoorer)
Relative to indiv. houses, apartments have:
Source: CGEDD d’après indices Notaires-INSEE.
Apartment price indexrelative to
detached house price indexBasis Q4 2000=1
0,95
1
1,05
1,1
1,15
1,2
1,25
1,3
1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015
Region of Paris minus ParisProvince
12
International comparison (1) France, USA and UK:
similar long run trends over 1965 -2000
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Freddie Mac, FHFA, R.Shiller, US Bureau of Economic Analysis, Census Bureau, UK DCLG, UK National Statistics, Halifax
International comparison:home price index relative to
disposable income per householdBasis 2000=1
0,9
1,1
0,5
1
1,5
2
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
FranceUSA (FHFA)USA (S&P/C-S)UK (DCLG)UK (Halifax)Aux 0,9
13
International comparison (2)Diversity since 1995
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Freddie Mac, FHFA, R.Shiller, US Bureau of Economic Analysis, Census Bureau, UK DCLG, UK National Statistics, Destatis, Gewos, Central Bureau voor de Statistiek, Instituto Nacional de Estadística, R. Vergès.
Home price indicesrelative to disposable income per household
International comparaisonBasis 2000=1
0,5
1
1,5
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
FranceUSAUKGermanySpainNetherlands
14
International comparaison (3) France -Germany :
Home sale price Rents(existing-home price index) ( « rent » component of the consumer price index)
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Destatis, Gewos.
Relative to income per household, from 2000 to 2010,• home price indices diverge• but rent indices remain flat in both countries
(impact of German households’ very high long term debt in 2000)
Home price indexrelative to income per household
Comparison France-GermanyBasis 2000=1
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1990 1995 2000 2005 2010
France
Germany
Rent indicesrelative to income per household
Comparison France-GermanyBasis 2000=1
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1990 1995 2000 2005 2010
France
Germany
15
2000-2010: rental return (=rent / price)collapses
Source: CGEDD after INSEE, Notaires-INSEE indices and notaries’ databases
Home price index and rent indexrelative to disposable income per household
France, basis 2000=1
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
Home price index relative to disposableincome per household, basis 2000=1Rent index relative to disposable income per household, basis 2000=1Auxtunnel 0
NB1: the rent index and the priceindex have different perimeters => bias => dividing the former by thelatter does not provide an index ofgross rental income.NB2: the «income per household» used as divider relates is that of allhouseholds, whether tenants or owner-occupier. Tenants’ incomegrows slower than owner-occupiers’ (by ~1% per year).NB3: depends on locationNB4: many structure effects
16
Resilience of the 6% gross rental return• In 1900-1910:
– inflation +0,3%/year, – income per household +1,6%/year, – Paris home price indexs +1,1%/year, – Gov’t debt interest rate 3,1%/year
• The gross rent of the (residential) properties purchased b y La Fourmi Immobilière (a property company) (from 1899 to 1913) was worth 6 to 7% of their price (quoted by F. Simmonet, « La Fourmi Immobilière »).
• « The average gross rental return in Paris would be 6,36% (from 5,13% in the 16th arrondissement to 7,76% in the 20th arrondissement) », of which 33% expenses must besubstracted (P. Leroy-Beaulieu, « L’art de placer et gérer sa fortune », 1906).
• 6% gross rental return ���� « a residential building is worth200 monthly rents »
17
Interest rates are at a historical low
…but their decrease hadbegun much earlier than 2000
Long Term Interest Rate and Inflation
-5%
0%
5%
10%
15%
20%
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
Interest rateInflationInterest rate minus inflation
Source: CGEDD after Ixis, Banque de France and INSEE
•Which rates (nominal, or net of past
inflation, or net of expected inflation) ?•Low with respect to which reference?
1,2%1,5%3,8%3,6%Interest rate minus inflation
1,6%2,8%1,7%5,6%Inflation
2,8%4,3%5,5%9,2%Interest rate
Fin 201020082000Average 1965-2000
18
2000-2010: increase of mortgage initial length
Source: CGEDD after housing surveys before 2005 then OFL
Mortgage initial lengthPurchase of a principal home by an owner-occupier
(years)
0
5
10
15
20
25
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Ancien
Neuf
19
« Affordability », « property purchasing power », etc. indices: what are we talking about?
• Which period? (ex: 2000-2010 or 1990-2010 or 1965-2010)
• Which area? (ex: Paris or France)
• Which price? (« constant quality » index or average price? new or existing?)
• Which income? (buyers’? borrowers’? all households’? « disposable » or net taxable or gross taxable income?)
• Which financing conditions?– Which interest rate? (several series, more or less consistent and continuou s
and reliable; what with adjustable rates, capped adjusta ble rates?)
– From 1973 to 1985, how do we factor in inflation, progressive monthlypayments?
– How do we take into account changes in mortgage length? – How do we take into account changes in downpayment?
• Other variables: transaction costs, interaction with the gov’t (subsidie s and taxes)
• Other points of view: purchasing power in rent years??
20
Example : impact on « affordability » of nominal / ‘real’ (=net of inflation) interest rate, and of mortgage length
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Banque de France, Ixis, housing surveys, OFL.
Affordability index: comparison constant (15 years) and actual mortgage length
nominal or 'real' (=net of inflation) interest rateFrance, basis 2000=1
0,7
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1,6
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Affordability at nominal interest rate, constant (15 years) mortgage length
Affordability at 'real' interest rate, constant (15 years) mortgage length
Affordability at nominal interest rate, actual mortgage length
Affordability at 'real' interest rate, actual mortgage length
Constant mortgage length
Actual mortgage length (= "forgetting" households will have to pay for more years)
High inflation
21
2000-2010: decrease in theaffordability index
at nominal and ‘real’ interest rate
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Banque de France.
Affordability indexBasis 1965=1
0,6
0,7
0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
Affordability index at nominal interest rateAffordability index at 'real' interest rateAuxtunnel 1
22
2000-2010: increase in the mortgagelength necessary to purchase the same
dwelling everything else being equal
Source: CGEDD after INSEE, notaries’ databases, Notaires-INSEE SA indices, Banque de France.
Mortgage length to purchase the same existing dwell ing with the same deposit to income and initial payment to income ratios, basis 1965=15 years
34,9, Q1 2012
0 years
5 years
10 years
15 years
20 years
25 years
30 years
35 years
40 years
45 years
1/1 1965 1/1 1970 1/1 1975 1/1 1980 1/1 1985 1/1 1990 1/1 1995 1/1 2000 1/1 2005 1/1 2010 1/1 2015 1/1 2020
Mortgage length 'at nominal interest rate'
Mortgage length 'at real interest rate'
23
2000-2010: New versus existing -home price
•The average price has grown faster for existing homes than for new homes•We have no price index (i.e. at constant quality ) for new homes for now (maybe it’s going to change: INSEE is working on it) sowe don’t know whether a new-home index would growfaster or slower than the existing-home price index
24
2000-2010: the number of existing -home sales hasremained quite constant
Source: CGEDD after CGDD/SOeS(Existan), DGFiP (MEDOC and Fidji), notaries’ databases
Number of property salestaxed at the regular transaction tax rate
France
1 178 000 (Apr. 12)
818 000 (Mar. 12)
0
200 000
400 000
600 000
800 000
1 000 000
1 200 000
12/ 1990 12/ 1995 12/ 2000 12/ 2005 12/ 2010 12/ 2015
Cumul. last 12 months
0
50 000
100 000
150 000
200 000
250 000
300 000
Cumul. last 3 months
Cumul. last 12 months Cumul last 12 months
Cumul. last 3 months (deseasonalized) Cumul. last 3 months (deseasonalized)
Number of property sales taxed at the regular transaction tax rate, all types of property: existing-homes, existing commercial property and VAT-exempt land
Of which existing-homes
All types of property
Of which existing-homes
(Left scale)
(Right scale)
25
2000-2010: construction of new dwellings :no excesses
Source: CGEDD after CGDD/SOeS and INSEE
Construction of dwellingsand change in the number of households
Apr. 2012376 731
0
100 000
200 000
300 000
400 000
500 000
1/1 1950 1/1 1960 1/1 1970 1/1 1980 1/1 1990 1/1 2000 1/1 2010 1/1 2020
Number of ordinary dwellings started, cumulated on last 12 months
12 month change in the number of households
Series changein January 2011
Metropolitan France
26
Sales of new dwellings by developers (NB: only 1/3 of new dwellings)
Source: CGEDD after CGDD/SOeS (ECLN)
Quarterly number of dwellings sold by developers
0
10 000
20 000
30 000
40 000
déc-80 déc-85 déc-90 déc-95 déc-00 déc-05 déc-10 déc-15
Quarterly # of dwellings
Put on sale
Sold
27
Developers’ inventory: reasonable but picking up
Source: CGEDD after CGDD/SOeS (ECLN)
Developers' inventory in months of sales
0
5
10
15
20
25
déc-80 déc-85 déc-90 déc-95 déc-00 déc-05 déc-10 déc-15
# of months of sales
Total inventoryOf which under construction or completedOf which completed
28
2000-2010: increase in the amount ofproperty sales relative to GDP
Source: CGEDD after DGFiP, INSEE and Toutain
Total amount of property sales
as a % of gross domestic productsales of any type of property
(residential and commercial, old and new, incl. lan d) subject to transaction tax, cumulated on 12 months
1995
1980
1990
1984
1853 1881
1844
1831
1848
1871
1948
Apr. 2012: 14.2%
1915
Apr. 2012: 10.2%
Apr. 2012: 8.1%
1/1 2050
1/11800
1/11850
1/11900
1/11950
1/12000
0%
2%
4%
6%
8%
10%
12%
14%
TotalOf w hich taxed at the current regular rate and antecedent ratesOf w hich existing homesAuxabscisse
29
2000-2010: households ’ outstanding mortgagedebt doubles with respect to their income
Source: CGEDD after Banque de France and INSEE
Households' new mortgages and mortgage debt outstanding
as a % of households' gross disposable income
Apr. 1264%
0%
2%
4%
6%
8%
10%
12%
14%
1/11965
1/11970
1/11975
1/11980
1/11985
1/11990
1/11995
1/12000
1/12005
1/12010
1/12015
1/12020
New
m
ortg
ages
0%
10%
20%
30%
40%
50%
60%
70%Mor
tgag
e de
bt o
utst
andi
ng
New mortgages (12-month cum.)
Mortgage debt outstanding (right scale)
30
2000-2010: households ’ outstandingmortgage debt – international comparison
Source: CGEDD after natioanl institutes of statistics and central banks
Households' outstanding mortgage debtas a % of households'disposable income
0%
20%
40%
60%
80%
100%
120%
140%
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
FranceUSAUKGermany
31
Consequences of the increase ofhome prices 2000-2010 (1)
•Increase in incomes indexed on home prices (realtors, n otaries,
etc.) and in departments’ receipts in transactions ta xes
•Wealth creation for owners ( but almost no « equity withdrawals » in France)
•Compared to 2000, cash transfer•From net buyers to net sellers
•From (not too) poor to rich
•From younger than 56 years to older than 56 years
32
Net buyers and net sellers : the 56 year threshold
Source: CGEDD after DGFiP, SOeS, notaries’ databases, EPTB, Filocom
Number of dwellings bought or built, or sold,as a % of the number of households
as a function of age of the householderYear 2006
-5%
0%
5%
10%
15%
20 30 40 50 60 70 80 90 100Age of the householder
% of the numberof households
-200 000
-100 000
0
100 000
200 000
300 000
400 000
500 000
600 000
Number of households
Dwellings bought or builtDwellings soldBalance bought or built minus soldNumber of households (right-hand scale)
56 years
Effect of the baby-boom
Average age of age brackets which are net buyers: 34 years
Average age of age brackets which are net sellers: 74 years
33
Consequences of the increase ofhome prices 2000-2010 (2)
•Injection of the cash provided by increased mortgages
(cumulated over 2000-2010: 15 to 20% of GDP) into•(a little) construction
•(a little) financial savings
•(mostly) consumption => GDP growth, households’ income, increase in tax
receipts, increase in trade deficit – no gain in competitivity
•Future increase in cash disbursements by borrowers (because of
increased mortgage length)
•Buyers increased their debt to purchase an asset •which does not provide any additional income (non productive asset)
•the price of which is higher today but will (as we argue thereafter) revert to its past
trend level with respect to income per household
34
PLAN1. Home prices in France, a Historical
perspective
2. Comparison with Other Assets
3. Several Important Properties of Home Prices
4. How can we Explain the 2000-2010 Rise?
5. Home Price Prospective
35
Price of gold net of inflation: constant in the long run
Source: CGEDD after INSEE, Banque de France, World Gold Council, (Officer, 2002).
Gold price index, constant currencybasis 2000=1
0,1
1
10
100
1800 1850 1900 1950 2000
in constant French currency
In constant $
In constant £
36
Fixed income: long term interest rate = inflation + 3% + wide fluctuations
Source: CGEDD after Vaslin, Loutchitch, Ixis, Chabert, Lévy-Leboyer, Homer & Sylla, national institutes of statistics and central banks.
Long term interest rates net of inflation
-60%
-50%
-40%
-30%
-20%
-10%
0%
10%
20%
30%
1800 1850 1900 1950 2000
FranceUSAUK
37
Stocks have been providing a 6.6% trend total return above inflation over two centuries (except catastrophic wars)
Source: CGEDD after (Arbulu 1998), SGF, Euronext, (Chabert, 1949), (Lévy-Leboyer & Bourguignon, 1985), INSEE, (Schwert 1990), (Shiller 2000), S&P, STAT-USA, US Bureau of Labor, (Dimson, Marsh & Staunton, 2001), UK Office of National Statistics
Value of an investment in US, French and British stocks,
constant local currency, dividends reinvested, basis 2000=1
y = 9E-58e 0,0654x
R2 = 0,9944
0,000001
0,00001
0,0001
0,001
0,01
0,1
1
10
1800 1850 1900 1950 2000
US stocksFrench stocksBritish stocksTrend of US stocks, constant $
TunnelSlope =+6,6% /an (« Siegel’s constant »)
TunnelSlope =+6,6% /an
(« Siegel’s tunnel »)
38
Value of an investment in stocks (dividends reinvested)
relative to long term trend
Source: CGEDD after (Arbulu 1998), SGF, Euronext, (Chabert, 1949), (Lévy-Leboyer & Bourguignon, 1985), INSEE, (Schwert 1990), (Shiller 2000), S&P, STAT-USA, US Bureau of Labor, (Dimson, Marsh & Staunton, 2001), UK Office of National Statistics
Value of investments in US, French and British stocks
relative to US stocks long term trend
0,1
1
10
100
1000
1800 1850 1900 1950 2000
US stocks
French stocks
British stocks
Tunnel
0,5
2
Tunnel
(« mean reversion »)
Distribution:-bimodal (?)-Standar deviation slightly (and less and less) growing- platikurtic
39
Value of an investment (total return: dividends reinve sted),French stocks relative to US stocks (both in constant local currencies)
Source: CGEDD after (Arbulu 1998), SGF, Euronext, (Chabert, 1949), (Lévy-Leboyer & Bourguignon, 1985), INSEE, (Schwert 1990), (Shiller 2000), S&P, STAT-USA, US Bureau of Labor
Value of an investment in French stocks, div. reinv 'd, constant French currency_____________________________________________________________
Value of an investment in US stocks, dividends rein vested, constant $
0,1
1
10
100
1000
01/3800
01/3810
01/3820
01/3830
01/3840
01/3850
01/3860
01/3870
01/3880
01/3890
01/3900
01/3910
01/3920
01/3930
01/3940
01/3950
01/3960
01/3970
01/3980
01/3990
01/4000
01/4010
01/4020
Date + 2000 years
Basis: average 1965-2005=1
40
Other yardsticks for stocks
• Price/earnings ratios [which earnings: past (overwhich duration)? future (over which duration)?]
• Price / dividends ratio [which dividends? past (overwhich duration)? future (over which duration)?]
• If the ratio departs from its long terme average, willit revert to it by the numerator, the denominator or both?
41
Which yardstick : « Siegel’s tunnel » or PER (earnings smoothedover 10 years)? They have been diverging since the 1990’s
Source: CGEDD after (Schwert 1990), (Shiller 2000), S&P, STAT-USA, US Bureau of Labor
Comparison Total return LT trend
versus adjusted PER
End-of-month values, USA
Feb. 2009
Dec. 1999
Aug. 1987
Apr. 1942
Feb. 1937
Nov. 1916
Jun. 1877
Mar. 1842
Jun. 1835
Nov.1802
Dec. 1814
Aug. 1982
Jun. 1949
Jun. 1932
Sep. 1929
Dec. 1920
Jun. 1881
Oct. 1857
Dec. 1852
0,5
0,4
0,6
0,7
0,8
0,9
1
2
3
1/11800
1/11850
1/11900
1/11950
1/12000
(1) Total return index / Siegel's long term trend(2) Shiller's 10 year adjusted PER / 15(2) / (1)0,5
42
Value of various investments (yearly returns reinvested)
Source: Source: CGEDD after Arbulu, Euronext, Vaslin, Loutchitch, Ixis, Banque de France, ECB, notaries’ databases, Notaires-INSEE indices, Duon, INSEE, Schwert, Shiller, S&P, World Gold Council, Officer.
Value of various investments, French constant curre ncy, basis 2000=1
1E-06
1E-05
1E-04
1E-03
1E-02
1E-01
1E+00
1E+01
1E+02
1800 1850 1900 1950 2000
GoldMoney marketBondsFrench stocksUS stocksRented residential property in Paris
1914-1965
43
Trend total return of an investment in housing on the basis of 1965- 2000
•(a) Capital gain:GDP growth inflation + 2,5% Households’ disposable income growth idem : inflation + 2,5%Minus growth in the number of households - 1,2%Equals: growth in income per household inflation + 1,3%
Capital gain idem : inflation + 1,3%
•(b) Net rental income :Gross rental income 6,0%Minus expenses 37% (incl. heavy repairs) - 2,2%Minus purchase expenses (11%) depreciated over 20 years - 0,5%Equals: net rental income 3,3%
•Total return = (a)+(b) inflat ion + 4,6%
Somewhat too high?
44
Return X volatility : 1840-1914
Source: CGEDD after Arbulu, Le Bris & Hautcoeur, Euronext, Vaslin, Loutchitch, Ixis, Banque de France, ECB, notaries’ databases, Notaires-INSEE indices, Duon, INSEE, Schwert, Shiller, S&P, World Gold Council, Officer. NB : ASIS = Arbulu-SGF-INSEE-SBF250 series. LB = Le Bris series
1840-1914
US stocks
Fr. stocks (LB, dep. 1854)
French stocks (ASIS)
Housing ParisFrench bonds
French money marketGold
Infla
tion
0%
20%
40%
60%
80%
100%
0% 5% 10% 15% 20% 25%
Yearly return
5 year return volatility
45
Return X volatility : 1914-1965
Source: CGEDD after Arbulu, Le Bris & Hautcoeur, Euronext, Vaslin, Loutchitch, Ixis, Banque de France, ECB, notaries’ databases, Notaires-INSEE indices, Duon, INSEE, Schwert, Shiller, S&P, World Gold Council, Officer. NB : ASIS = Arbulu-SGF-INSEE-SBF250 series. LB = Le Bris series
1914-1965
US stocks
French stocks (LB) French stocks (ASIS)
Housing Paris
French bonds
French money market
Gold
Infla
tion
0%
20%
40%
60%
80%
100%
0% 5% 10% 15% 20% 25%
Yearly return
5 year return volatility
46
Return X volatility : 1965-2011
Source: CGEDD after Arbulu, Le Bris & Hautcoeur, Euronext, Vaslin, Loutchitch, Ixis, Banque de France, ECB, notaries’ databases, Notaires-INSEE indices, Duon, INSEE, Schwert, Shiller, S&P, World Gold Council, Officer. NB : ASIS = Arbulu-SGF-INSEE-SBF250 series. LB = Le Bris series
1965-2011
Gold
French money market
French bonds
Housing Paris
French stocks (ASIS)French stocks (LB)
US stocks
Infla
tion
0%
20%
40%
60%
80%
100%
0% 5% 10% 15% 20% 25%
Yearly return
5 year return volatility
47
Trend return X volatility
Source: CGEDD after Arbulu, Euronext, Vaslin, Loutchitch, Ixis, Banque de France, ECB, notaries’ databases, Notaires-INSEE indices, Duon, INSEE, Schwert, Shiller, S&P, World Gold Council, Officer.
Trend(assumption: inflation 2%)
Gold??
French money market
French Bonds
Housing Paris
French stocks
US stocks
Infla
tion
0%
20%
40%
60%
80%
100%
0% 5% 10% 15% 20% 25%
Yearly return
5 year return volatility
(Inflation +4,6% , 28%) (NB: without leverage )
(Inflation +6,6% , 42%)
(Inflation +6,6% , 50%)
(Inflation +3,0% , 20%) (NB: not protected against unexpected inflation
(Inflation +2,0% , 8%)
(return , volatility)
(Inflatn +0,0%, ?)
48
Other aspects of investing
•Leverage
•Management costs and transaction costs
•Taxes
•Risks other than price volatility(valued differently depending on the investor)
•Diversifying power
49
Now
Relative to their respective long termtrend ,- stock prices are low (*)- gold, bonds(**) and housing prices are high
(*) although a decrease is possible in the short term(**) except sustained very low inflation
50
PLAN1. Home prices in France, a Historical
perspective
2. Comparison with Other Assets
3. Several Important Properties of Home Prices
4. How can we Explain the 2000-2010 Price Rise?
5. Home Price Prospective
51
Several properties of house prices (1)
•Series are short => limits the significance of
results (incl. problem of lack of robustness of results)
•High autocorrelation of 1 year price changes
=>strong cyclicity (in the sense of high autocorrelation)
Deciders’ « myopia » (= expectations based on the recent p ast = « short
memory »), self-reinforcing phenomenon
Conversely, no short term autocorrelation for stocks ( ~ random walk)
•No periodicity other than seasonality
52
Seasonality
Source: CGEDD after Notaires-INSEE indices.
Seasonal variation coefficientsfor existing-home prices
-3%
-2%
-1%
0%
1%
2%
3%
Q1 Q3 Q3 Q4
Quarter of the sale
France - apartments
Paris - apartments
Paris region except Paris - apartments
France minus Paris region - apartments
France - individual houses
Paris region - individual houses
France minus Paris region - individual houses
53
Several properties of house prices (2)
•Link with households ’ income :
•in time : intuitive in appearence , empiric in
reality
•and in space
54
Link home price X income per household :by city in the Paris region
Source: CGEDD after notaries’ databases and Filocom (DGFiP)
Average home price (apartments and individual house s)as a function of gross taxable income per household
2006, 1000 biggest cities in Paris region(averages exclude the 10% extreme values)
y = 6,63x - 38958,29
R2 = 0,81
0
100 000
200 000
300 000
400 000
500 000
600 000
700 000
800 000
0 10 000 20 000 30 000 40 000 50 000 60 000 70 000 80 000 90 000 100 000
Gross taxable income per household (€)
Home price (€)
The surface of the circle is proportional to the number of households in the city
55
Idem by urban area (impact of secondary homes)
Source: CGEDD after notaries( databases and Filocom (DGFiP)
Average home price (apartments and individual house s)as a function of gross taxable income per household
by urban area, year 2006
Lunel
Thonon
Annemasse
Annecy
ColmarMulhouse
Orléans
Saint-Malo
La RochelleMartigues
Avignon
Arles
DijonNancy
Cluses
Antibes
Paris
Chantilly
Menton
BayonneLa TesteFréjus
Rambouillet
Maubeuge
Longw y
Vierzon
Saint-Louis
Fontainebleau
Cavaillon
Nice La Ciotat
Dinard
ToulonCarpentras
Istres
Saint-NazaireDraguignan
Montceau-les-Mines
Epinal
Saint-Avold
RodezLens
Haguenau
Alès
Miramas
y = 8,24x - 97272
R2 = 0,55
50 000
100 000
150 000
200 000
250 000
300 000
350 000
20 000 25 000 30 000 35 000 40 000 45 000 50 000 55 000
Gross taxable income per household (€)
Home price (€)
The surface of the circle is proportional to the number of households in the urban area
56
Interpretation of the link price X incomeHouseholds ’ asset (1)-100% of users are households (housing expense = 1/5 of their income)•29 million households, 34 millions dwellings of which 1/10 secondary homes (differences = vacancy)
-95% of buyers are households (a household’ biggest purchase duringits existence)
-8 dwellings out of 10 are owned by households•difference =8/10 « social housing » (« HLM ») + 2/10 owned by other non-individuals
- ¾ of households own a dwelling at least once in theirexistence, 63% of households own at least one dwelling, 59% of households own their principal residence
Source: estimates by CGEDD after Housing surveys and Filocom, SOeS and various sources
57
Households’ asset (2)-Out of 10 households:
-4 own no dwelling-4 own at least 1 dwelling-2 own more than 1 dwelling (on average 2.8 dwellings, incl. their principal residence)
By comparison only 2 households out of 10 own stocks (and only 5% own a significant amountof stocks)
-Out of 10 households:-4 are tenants (of which 2 get a housing benefit)
-2 are owner-occupier and reimburse a mortgage (« accédants à la propriété »)-4 are owner-occupier and don’t reimburse a mortgage (« propriétaires non accédants »)
-Out of 3 dwellings purchased–1 is the first principal residence of the buyer–1 is a principal residence of rank >1 (the 2nd one, the 3rd one, etc.) of the buyer–1 is a rental investment or a secondary home (of which: 2/3 rental investmentsand et 1/3 secondary residence)
-A household purchases on average 2,5 dwellings during its existence
Source: estimates by CGEDD after Housing surveys, Filocom, SOeS and various sources
58
Households ’ asset (3): number of dwellingsgetting into or out of an individual’s ownership during his lifetime
Source: CGEDD estimates after various sources
1,1Dwellings owned at death
0,4Dwellings inherited
0,0Net balance of dwellings given or received as donationsWithout
payment
0,7Difference purchased or built minus sold
1,8Dwellings sold
(0,7)(of which purchased new or built)
(1,8)(of which purchased existing)
2,5Dwellings purchased or built
Withpayment
Entry intoor exit fromownership
Number ofdwellings
During the lifetime of an individual as part of a household
59
Several properties of house prices (3)•Low univariate correlation of house price changes and interest rate changes
•counter-intuitive but is the basis of the diversifying power of housing investment withrespect to bonds (maybe different for whole buildings owned by large investors?)
•=> one has to factor out many other phenomena to see the sensitivity of home prices withrespect to interest rates•Link by the return to the hierarchy of trend return-risk couples but this return is notimmediate
• No univariate correlation of house price changes with stock investment [nevertheless stock krachs have often (1929, 1987,
2000, 2008) but not always (1882) been followed by an increase in house prices (particularlyrented house prices)]
•Time-series analyses (with autoregression) methods don’t yield better results•Multivariate analyses (incl. changes in offer and suppl y, income) impaired by the series brevity (at most 46 years)]
=> Diversifying power of housing investment with respect to financial investme nts
60
Several properties of house prices (4)
A fundamental property : the elasticity ofhousing price with respect to housingsupply seems in the -1 or -2 range, maybe(?) slightly more (-3?) in the Paris region
Details in the paper:http://www.cgedd.developpement-durable.gouv.fr/IMG/ pdf/elasticite-prix-immobilier-nombre_cle093f5d.pdf
61
Elasticity of housing price with respect to housing supply (complement 1)
• Quite few works• Complicated because
– Many variables must be taken into account– Reverse effects– Time lags– Time series are short => results not robust with respect to the
period studied– Analyses in space may compensate the lack of memory… but few
local data
• Many more works on a reverse problem: sensitivity ofconstruction to housing price changes
62
Elasticity of housing price with respect to housing supply (complement 2)
• Barker report + Oxford team: e=-2 in the UK• Murphy, Duca & al. (Oxford + Fed Reserve):e=-1 in the
USA• Other references: widely dispersed results (by a factor 8 for
the USA!)• Often, economists’ assumptions may be seriously
debatable (incl.: arbitrage is not instantaneous) • Result robustness is rarely mentioned• Economists may be wrong: cf. (McQuinn, 2004) about Ireland,
(OECD: Girouard, Kennedy & al., 2006) about the USA
• But results form a cloud centered around an order of sizeof -1or -2
63
Elasticity of housing price with respect to housing supply (complement 3)
References
Source: CGEDD after table 3 of (OCDE: Girouard, Kennedy et al., 2006), as well as (Duca, Muellbauer & Murphy, 2009) , (Cameron, Muellbauer & Murphy, 2006).
Country, period Elasticity Reference
Ireland, 1977-2004 -0,007 (ancien), -2,0 (neuf) (OCDE, 2006)
Ireland, 1980-2002 -0,5 (M cQuinn, 2004)
Netherlands, 1970-2002 -0,5 (OCDE, 2004)
USA, 1979-2007 -1 (Duca, M uellbauer & M urphy, 2009)
Netherlands, 1980-2003 -1,4 (Verbruggen & al., 2006)
Norway, 1990-2004 -1,7 (Jacobsen & Naug, 2005)
UK, 1969-1996 -1,9 (M een, 2002)
UK, 1972-2003 -2 (Cameron, M uellbauer & M urphy, 2006)
Idem Idem (Barker, 2004)
Danmark, 1984-2005 -2,9 (Wagner, 2005)
USA, 1981-2003 -3,2 (M cCarthy & Peach, 2004)
Paris, 1986-2004 -3,6 (Bessone, Heitz & Boissinot, 2005)
Australia, 1975-2003 -3,6 (Abelson & al., 2005)
Spain, 1989-2003 -6,9 à -8,1 (OCDE, 2005)
USA, 1981-1998 -7,9 (M een, 2002)
64
Elasticity of housing price with respect to housing supply (complement 4)
• Our contribution– Comparison France / UK over 1970-2005: elasticity =
minus a few units
– Comparison of the various French departments (1994-2010) ( multiple regression of housing price change with respect to change in income, population, number of dwellings, etc.)
• Confirms an order of -1 or -2 ,• maybe (?) slightly more (-3?) in the Paris region• Results are sensitive to the subperiod studied (problem of
robustness)
65
PLAN1. Home prices in France, a Historical
perspective
2. Comparison with Other Assets
3. Several Important Properties of Home Prices
4. How Can We Explain the 2000-2010 Price Rise?
5. Home Price Prospective
66
How can we explain the 2000-2010 price rise ?
•Supply -demand of housing service? •Inflationary impact of housing subsidies?•Other explanations except financialenvironment ?•Financial environnement?
•for the investor•for the buyer of his own principal residence
67
How can we explain the 2000-2010 price rise ?
•Supply -demand of housing service?No because:
•Elasticity price /supply too low (-1 or -2)•No rent rise beyond historical trend•Qualitative effects?
•Decrease in household size?•Ageing?•Foreigners?
no or not at the scale of the problem
68
How can we explain the 2000-2010 price rise ?
•Decrease in household size: is not new and goes on at the previous pace
Nombre de personnes par ménageFrance métropolitaine
2,3
2,4
2,5
2,6
2,7
2,8
2,9
3,0
3,1
3,2
3,3
1960 1970 1980 1990 2000 2010Source: CGEDD after INSEE
Number of persons perhousehold
69
How can we explain the 2000-2010 price rise ?
•Ageing : impact >0 or rather <0 ?
Households who prepare their retirement would have realizedthat their pensions might be lower than for previousgenerations and as a consequence would save more andaccept lower expected returns? * True they accept lower rental returns, but is it sustainable? * This demographic interpretation may be reversed: increasein the % of households older than 56, threshold beyond whichhouseholds become net sellers.
70
How can we explain the 2000-2010 price rise ?
• (Net) purchases by foreigners ? Too few bar exceptionsPurchases of existing dwellings net of salesby foreigners as a % of the number of sales
1,0%0,9%0,8%1,0%1,1%Of which others
-0,1%-0,1%0,0%0,0%0,1%Of which Germans
0,3%0,2%0,2%0,2%0,3%Of which Portuguese
0,6%0,5%0,5%0,7%0,6%Of which MATT
1,8%1,5%1,4%1,9%2,1%Other
0,4%0,9%1,7%1,3%0,5%Britons
2,1%2,4%3,2%3,2%2,6%2,3%All foreigners
20082006200420022000Moy 94-99 Evol.
(mostly resident)
(mostly resident)
Source: CGEDD after notaries’ databases
71
How can we explain the 2000-2010 price rise ?
•Inflationary impact of subsidies?Not at the scale of the 70% to be explained-Households’ housing expense = 15% of GDP-Amount of dwellings purchased or built by households = 250 Billion € = 13% of GDP in 2007 (at its highest)-Amount of property inflation generated in 2007 by the 70% increase in home prices relative to income: ~100 Billion €
To be compared to- Transfers organized by gov’t in favor of housing = 1 to 2% ofGDP (depends on how one counts)
- Leeway on these transfers = ten times smaller- « PTZ »: around 2,5 Billion € equivalent-subsidy (at its maximum)
72
How can we explain the 2000-2010 price rise ?• Other explanations except financial environmt ? (1)- Resale financed the price rise?
By nature resale feeds rises (and falls: reversible effect) but a) Resale financed the same % (21%) of purchases (of principal residences by owner-
occupiers) in housing surveys 2002 and 2006b) The number of existing-home sales remained constant (800 000/year) from
2000 to 2007 => the « rotation speed » of the housing stock did not increase(rather it decreased)
c) Departments where home price rised most were those where there were thefewest owner-occupied principal residences as a % of all dwellings
- Inheritance and donations finance (and will finance) the price rise?Not that much and not more than previously- One inherits from parents around 55- Donations financed a low (3%) and constant % of purchases (of principal
residences by owner-occupiers) in housing surveys 2002 and 2006- Lagged and reversible effect
73
How can we explain the 2000-2010 price rise ?
• Others explanations except financial envirnt ? (2)- The price rise since 2000 results from land price rise and
scarcity? NoLand price- The market price of constructible land is determined by the price of existing
dwellings in its neighborhood => the rise in home prices caused the rise in land prices, not the reverse.
- The average price of land used for building individual houses did not growfaster than the average price of existing homes, whereas the construction cost index grew much slower.
Land « scarcity »- The elasticity of housing price with respect to housing supply being around -
1 or -2, an increase in the supply of constructible land parcels (by regulatorychanges or by sellers’decision to sell) decreases housing prices only slightly.
- When, from 2004, construction grew from 300 000 to more than 400 000 peryear, finding land was not a problem.
74
How can we explain the 2000-2010 price rise ?
• Other explanations except financial environmt ? (3)- The price rise is just a continuation of the increase in th e
weight of housing expense as in households’ budgets experienced since 1965? No since the increase in housing expenses from1965 to 2000 took place while the house price index was constant relative to households’ income: it resulted from an increase in the quality of housing, withoutany equivalent in the 2000-2010 period.
75
How can we explain the 2000-2010 price rise ?• Financial environment?A. Housing as an investment : arbitrage againstother assets (risk X expected return)
•« Rational » investors value housing as a rent-indexed perp etual bond (Rnet
initial~r-i+k-a where r= interest rate, i = expected inflation, k= risk premium, a= expected growth rate of rent, net ofexpenses and inflation)
•In 2010 interest rates were low (relative to their trend level) => could justifyhousing prices in spite of low rental returns…: housing inve stment wascompetitive with respect to bonds , which probided low expected returns too(but isn’t it risky to finance an indexed perpetual bond by a 25 year bond?)•…but stocks were low (relative to their trend level) and their expected retu rn washigh in the long term
•Certainly many households don’t arbitrage housing agains t stocks in any way• but only « myopia » (after the 2000 stock krach) can explain that the others didn’t move to stocks .
•Parallel with 1930-1935
=> The 2000- 2010 home price rise can be explained by arbi trage only if one assumes deciders’ «myopia »
76
How can we explain the 2000-2010 price rise ?• Financial environment?B. Housing as principal residence (=majority ) : what can one buy for a given monthly payment?
•Lower interest rates impact owner-occupiers less than inve stors (15-20 year mortgage less sensitive than perpetual bond to interes t rate)•Longer mortgage length
•In the short term : to be relativised (increased the amount purchased by 12% to 15% everything else being equal), •In the long term : repayment takes longer ( => from which budget willhouseholds take the cash? )
•Downpayment as a % of price has decreased from 2000 to 2006 (consequenceof increase in indebtness allowed by longer mortgages and lower interest rates)•Other conditions have fluctuated as they have since 1965 – may have contributed moderatelyto the price rise•Conclusion: for owner-occupyers, lower (net) interest ra tes relative to the1965-2000 reference (3.6%) have not compensated the price i ncrease, eventaking into account longer mortgages
77
How can we explain the 2000-2010 price rise ?
To summarize :• Mortgage conditions (rate + length) favored some priceincrease ,• + deciders ’ « myopia » (mainly investors)These factors impact rental investment more than purchases by owner occupiers=> explains that since 2000, prices have been growing faster•for apartments (3/4 rented) than for individual houses (3/4 owner-occupied)•In departments with lower % of owner-occupied principal residences
The same factors seem to have caused the rebound in home p rices in 2009-2010:- Additional fall in ‘real’ interest rates (not sustainable )-Deciders’ « myopia » , even more so after the 2nd stock crash (2008 after 2000) (reversible )-Higher rebound in larger urban areas (where more rented d wellings) -Difference 2008-2010 France – UK - USA:
-Few housing investors + « subprimes » and repossessions in the USA -Fewer investors + (adjustable!) interest rates fell by more in the UK,
78
PLAN1. Home prices in France, a Historical
perspective
2. Comparison with Other Assets
3. Several Important Properties of Home Prices
4. How can we Explain the 2000-2010 Price Rise?
5. Home Price Prospective
79
Home Price Prospective
Indice du prix des logements rapporté au revenu disponible par ménage
France, base 1965=1
0,1
1
1/11930
1/11940
1/11950
1/11960
1/11970
1/11980
1/11990
1/12000
1/12010
1/12020
1/12030
1/12040
2
0,2
0,3
0,4
0,90,80,70,6
0,5
1,1tunnel
F
D
BA
E
C
Source : CGEDD after notaries’databases, Notaires-INSEE indices and INSEE
Divergence
Level change
Return tothe tunnel
Not useful to anticipate the future(impact of rent controls)
Home price index relative to disposable income per householdFrance, basis 1965=1
80
Scenario F (divergence) looks unlikely
• Rental returns can’t decrease indefinitely• => after a certain while, rents would be disproportiona te
relative to households’ incomes• => Scénario F is unlikely• => the home price index will probably resume a progressi on
parallel to income per household
81
Will the « tunnel » change level? (= scenarios C, D et E)
Does the 2000-2010 price rise signal permanent « levelchange » resulting from irreversible phenomena?
• Will the causes of the 2000-2010 price rise have a permanent impact?– Explanations by « supply / demand »: have been rejected
– Explanations by inflationary impact of subsidies: have been rejected
– Other explanations other than financial environment: have been rejected(or are reversible)
– Remains: financial environment and deciders’ « myopia »
• New phenomenon : level of public debt
82
Financial environment: A. Housing as an investment
• Interest rates will revert to the trend level = 3% plus inflation• Risk - return couples will revert to their trend levels for other investments• Arbitrage => idem for housing investment: total return will revert to its trend• « Level » => housing investment’s capital gains will be at its trend level
=> Rental return L/P (total return minus capital gains) will revert to its trend level (net ~ 3,5%)
Rents L should not grow faster than income per household:• Have grown at the same pace up to now• Low elasticity / supply and demand• Tenants can not pay much more than they do as a % of their income• Tenants’ income grows slower than average households’ income
=> Housing price P should revert to its past trend level r elative to income per household
• In addition deciders’ « myopia » (the impact of which is temporary and reversible by nature)
with respect to stocks will end and invert its impact at some point
• => « Return to the tunnel »
83
Financial environmentB. Housing as principal residence
• Interest rates will revert to the trend level = 3% plus inflation
• Mortgage length– In the long term,
• Reverse effect of additional monthly payments on the amount of furtherpurchases [where will people find the cash? Lower housing purchases? Lower other housing expenses? Lower other expenses (cars, vacations, etc.)?]
• Part of the increase in average price will be an increase in quality (cf. mortgage lengthening of 1965-1975) => further reduces the impact on thehouse price index
• In the USA, mortgage lengthening in the (21 years in 1963, 27 years in 1980) has not coincided with a « level change » (rather the opposite)
• Other: no reason not to assume past constants will change + reversibilityin some cases
=> No cause for a significant level change in the long ru n.
84
Government deleveragingWill impact in some way households’ housing purchases
Households' mortgage debtand Maastricht government debt
Apr.1264%
Dec. 1186%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1/11965
1/11970
1/11975
1/11980
1/11985
1/11990
1/11995
1/12000
1/12005
1/12010
1/12015
1/12020
Households'mortgage debt as a % ofhouseholds'disposable incomeMaastricht general government debt asa % of gross domestic product
Source : CGEDD afterBanque de Franceand INSEE
85
As a conclusion: a « level change » looks unlikely
• Rejection of explanations by « supply-demand »• Rejection of explanations by subsidies and misc.• Reversion to « trend » interest rates (3% plus inflation)• For the investor : stability of the risk/return hierarchy of the various
investments (return to rental returns of years prior to 2000) and end of « myopia »
• For the owner-occupier : reverse impact of mortgage lengthening
• Government deleveraging⇒ a« level change », if any, should be small: we reject
scenarios C, D and E⇒ Remain : scenarios A and B: « return to the tunnel »
86
How fast shall we revert to the tunnel?
Indice du prix des logements rapporté au revenu disponible par ménage
France, base 1965=1
0,1
1
1/11930
1/11940
1/11950
1/11960
1/11970
1/11980
1/11990
1/12000
1/12010
1/12020
1/12030
1/12040
2
0,2
0,3
0,4
0,90,80,70,6
0,5
1,1tunnel
BA
Source : CGEDD after notaries’databases, Notaires-INSEE indices and INSEE
A (fast) = nominal pricesfall by 35 to 40% in 5 to 8 years
B (slow)= nominal pricesconstant for 15 to 20 ans (« Japanesescenario »)
Home price index relative to disposable income per householdFrance, basis 1965=1
87
How fast shall we revert to the tunnel?
•Based on years 1965-2000, scenario A (fast) est likelierthan scenario B (slow) which may not be excludednevertheless•Low interest rates lessen the likelihood of scenario A until deciders’ « myopia » reverses its effect•Local scenarios may differ (depending on the 2000-2010 change in home price and on the past and prospective change in income, unemployment,
population, number of dwellings, % of secondary residences, etc.)
88
Prospective: households ’ mortgage debtprojected to 2030
Source: CGEDD after (up to 2011) Banque de France, INSEE, Federal Reserve, Bureau of Economic AnalysisPaper: http://www.cgedd.developpement-durable.gouv.fr/IMG/pdf/dette-immobiliere-2030-friggit_cle7496ca.pdf
Households' mortgage debtas a % of their disposable income
0%
20%
40%
60%
80%
100%
1970 1980 1990 2000 2010 2020 2030
USA, observedFrance, observedFrance, projected, scenario DFrance, projected, scenario AFrance, projected, scenario B
Macroeconomic consequencesof an increase in households’ mortgage debt?