Abstract

The role of Buy-to-Let (BTL) in the contemporary UK housing market is disputed.  Promoters see it stimulating private rental and promoting the consumption of riskier sites for regeneration activity.  Detractors cite damage to communities where BTL stock remains empty, bringing environmental decline, or inflating prices and excluding First Time Buyers from ownership.  But these may not be the most significant effects.  The investment market in housing, including BTL, has skewed new supply to its own needs.  This paper uses nationally available statistics to identify the extent of BTL activity and to propose that the impact of this is to disconnect the current UK housing market from household consumption, a connection that is assumed in many discussions and models informing policy.  The main indicators of this are the volume of BTL lending and investment activity, the cost of housing in relation to household income, and the consequently changed nature of new housing supply.

Introduction

On 11 May 2006 Mervyn King, Governor of the Bank of England, said:

House prices are high relative to the measures that help to put them into context – average earnings and incomes.  By some measures they are remarkably high.  (Reported in the Guardian 11 May 2006)

This reasoning led him to predict no further boom in house prices.  But what actually happened after his statement was 18 months continued growth in house prices.  Whatever his assumptions about the connection between house price rises and household incomes, he seems to have been wrong, at least in the medium term.

This paper presents evidence from nationally available data sources that illustrates this disconnection of prices from incomes and demonstrates that evidence of this has been available for some time despite contrary claims from supporters of the investment market in residential property (Ball 2006, Bradford and Bingley 2007).  The paper also indicates some wider effects that this has had on housing supply.  The latter is important as the evidence suggests that the power of non-consumption purchasers has not only reduced the supply of housing that consumers want to purchase but is probably of a scale sufficient to erode levels of owner occupation, the exact opposite of central policy objectives.

Additionally the standard housing market indicators may mislead investors as they indicate price rises in markets where certain stock may have little or no second hand value (i.e. they pay a premium price for a new product that may hold little realizable equity; if this is the case then buying certain new property is more like buying a new car in that the value drops as soon as you buy it).  Economic models, including Mervyn King’s, assume a reasonable level of knowledge in an efficiently functioning market.  The housing market may have been sustained for several years by levels of ‘confidence’ based on simple price data that does not reflect the market’s actual condition.  Market actor decisions may have been based on ignorance rather than knowledge and this has been sustained by the irresponsible lending that is now being revealed following the Northern Rock crash.  The rapid withdrawal of BTL mortgages (Financial Times 15/12/07) as this crisis emerged indicates a belated realization by lenders that they have been supporting a bubble that has distorted the market.

The remainder of this paper is structured in the following way: Market assumptions will be outlined and two key objections raised to the current applicability of these assumptions.  The evidence will then be presented to support the objections through consideration of housing market activity and the scale of BTL activity using nationally available data.  Finally evidence will be presented on the impact of this investment activity on new housing supply.

Housing demand assumptions

Economic theory about housing market demand contains a number of assumptions.  In summary these are that demand for housing is critically dependent upon household aspirations, tastes, income and broader demographic change interacting with available supply.  Modelling thus assumes demand created at the household level for living space, like Mervyn King’s implicit assumption above, and the linkage is taken to be direct and exclusive.  Household formation and income is thus the main driver of aggregated housing market activity and price fluctuations that in its turn, drives supply.  This assumption also underpins the Barker Review (2004) and the ODPM response (ODPM, 2005).

I am suggesting that investors/landlords may now be the most significant demand driver at some levels of the market.  Evidence for this claim forms the core of this paper concluding with data linking investment activity with changes in new supply being one of the clearer indicators of this break in the links assumed by models.  The traditional assumptions may no longer apply and there are two key assumptions I wish to challenge here:

1. That the main current driver of housing market activity (including new supply) is household level housing consumption decisions.
2. That house prices in the ownership market are consequently driven or sustained by available household income in the market locality.  Dominating this area of discussion is the ratio of prices to average earnings (see Hamnett 1999) that is assumed to limit the size of periodic inflationary bubbles.

These assumptions seem increasingly unrepresentative of actual market behaviour.  I maintain that demand for housing is now more influenced by investment decisions determined by the rate of short term capital return housing produces in relation to other financial assets.  This leads directly to the following points that I will try to demonstrate:-

1. That residential property markets have been penetrated to significant levels by investment money unrelated to occupancy of the target stock.  This has occurred to such an extent that it is no longer primarily household consumer activity that drives the market.
2. This shift in demand has significantly shifted new housing supply towards flats, a favoured target for investors.  This shift has been supported unintentionally by planning policy promoting the use of brownfield sites and higher density building.
3. That the traditional relationship between house prices and average income is now broken causing unprecedented (and possibly unsustainable) increases in residential property values.

All of the above is significant for housing supply decisions at local level and for housing policy predicated upon the desirability of and potential for the continued expansion of owner occupation.

What Drives the Housing Market: The Role of Investors

The combination of legislative change promoting private rental and financial deregulation leading to new lending products from Banks and Building Societies have all contributed to an overall growth in what is broadly categorized as ‘private rental’ in the UK.  The mass media have also been a stimulus by illustrating how easy it is to make money from the property market.  UK terrestrial television transmits many programmes focused on the purchase and renovation of residential property and the print media carries adverts and promotional material.  For example ‘Mum of two, Jo Seward [who] has bought a buy-to-let property to invest money for her children’s future.’  Jo, who lives in Worsley, has bought a ‘one bed apartment in a new development in Wolverhampton” that isn’t built yet and she hasn’t even been to Wolverhampton.  She has relied on the ‘research’ done for her by the ‘Armchair Property Company’ who negotiated with developers for her to pay £101k for a property listed at £116k “but already worth £122k’.  (M.E.N. Lite Wed 12 April 2006).

Entrepreneurs create ‘investment clubs’ and other vehicles to help members of the public access wealth through property, and estate agents have adopted marketing strategies reflecting this new demand.  “Suit first time buyer or investor” is a common tag for property adverts.

As a result of all this activity ‘approximately £80bn’ had been invested in the sector since the mid 1990s “much of it by small players owning one or two rental homes” (Times May 6 2005).  The role of small investors in the PRS has been reported by Crook and Hughes (2001), Rhodes and Bevan (2003) and Scanlon and Whitehead (2005) and is supported by other recent research (Stoke-on-Trent CC, 2006).  The capital figure quoted in the Times is likely to be an under-estimate although investment growth is hard to assess.  An obvious support to this growth has been the Buy-to-Let mortgage scheme (BTL) the simple expansion of which is shown in Table 1.

Table 1: The Growth of By-to-Let Mortgages 2004-2006

Mortgages Advanced during the year	2004	2005	2006
Total mortgages (number) Table ML1	2,613,000	2,170,000	2,260,000
BTL Mortgages (number) Table MM6	217,700	223,800	330,300
BTL mortgages as a % of total mortgages	8.3%	10.3%	14.6%

Source: CML Tables, CML website

The value of this lending is huge.  The Council of Mortgage Lenders (CML) reports BTL lending rising from £6.9bn in 2001 to £45.3bn in 2007 (CML Table MM17).

In his report undertaken for the Association of Residential Letting Agents (ARLA) and others, Michael Ball (2006) maintains that while BTL is a phenomenon that has helped to stimulate the private rental sector (PRS) into renewed life, it has not had a major impact on the house sales market overall.  He is also sceptical about the reported damaging effects of BTL and sees detractors as taking an unbalanced view of the sector.  He summarises the opposing positions well, commenting that:

Buy-to-Let has been variously blamed for: creating house price inflation; being full of wild property speculation and reckless borrowing; a threat to the stability of the rest of the housing market; being a danger to mortgage lenders; pricing people out of owner occupation by removing the bottom rungs of the property ladder; destroying the character of city centres and encouraging binge lifestyles; hindering the revival of rundown neighbourhoods because of unsavoury tenants and landlords; creating ghost neighbourhoods of new blocks of flats; and to cap it all being run by inept rank amateurs.  It is not the purpose of this report to provide commentary on all of the criticisms thrown at Buy-to-Let.  Even so, a more considered view of Buy-to-Let inevitably discounts as wrong or highly unlikely many of the wilder claims made about it.  (Ball 2006 P3)

Overall his report is written as if the price growth trend in the UK would have happened regardless of BTL.  In fact given the rapid rise of prices, BTL has actually been a benefit to would be homemakers or prospective First Time Buyers (FTBs) in giving them access to housing they can afford at a time when purchase is impossible for them.  Wilcox (2007) writing later indicates the possibility of BTL impact on prices but from an apparent under-estimate of the volume of activity (Wilcox 2007).

That FTBs are struggling to gain access to the market has been confirmed by the CML which has acknowledged that “Half of all young first time buyers may be getting help from parents” (CML) which is a clear recognition that even lower priced homes are now beyond the means of lower income earners.

London Development Research (2006) report that only 30% of new supply in London goes to owner occupiers.  LDR also claims that investment has stimulated more supply but had no detrimental market effect.  However there is contrary evidence even within that report as “…investors generally seek smaller units…”  (LDR 2006 p4) and CLG Live Table 252 shows that in 2000/01 66 per cent of London’s new supply was 2 bed or smaller, by 2004/05 this had risen to 84%.

Data from national sources indicates that BTL and investor activity is of a scale that has the potential to drive up prices in a comparatively fixed supply market.  Ball claims that this is not the case as BTL accounts for only 9 per cent of mortgage lending (Ball 2006 p 4).  Wilcox (2007) also indicates BTL accounting for 1 in 10 sales.

The basis for this claim seems to be the Council for Mortgage Lenders (CML) statistical reports.  However this simple assessment could be misleading as total mortgage lending includes re-mortgages.  Re-mortgaging is a significant volume of lending but does not increase house sale transactions (unless the equity release is reinvested in purchasing another home).  A simple summary of the Mortgage Lending is shown above in Table 1.  The complete figures for 2006 were not available when Ball produced his report but his figure of less than 9 per cent of mortgages in 2005 is supported by CML figures.

The figures for 2004 to 2006 show a slight growth trend over the period and reinforce Ball’s point that BTL seems well established.  But using this comparison BTL is perhaps not significant enough to impact on the market as a whole.

CML data also shows that almost half of new BTL mortgages are re-mortgages (CML Table MM17).  This clearly complicates calculations of volume of activity in the market as a whole however Ball also says in this report (ibid. p1; main text and footnote) that only 54 per cent of BTL landlords use a mortgage.  BTL mortgages are therefore not all funding new acquisition activity but seem reasonably representative of the overall acquisition activity in the market in any one year by PRS landlords.  Recent interviews with landlords have confirmed that BTL mortgages are used by around 50 per cent of landlords but are not the only funding mechanism (PRS Study Stoke-on-Trent CC 2006, unpublished research for Gateway and West Yorks Housing Partnership).

While Ball tends to write of the sector in terms of overall supply and total housing stock this paper will discuss BTL activity in the context of the active market of transactions taking place.  This distinction is important as the active market usually involves less than 5 per cent of total stock and includes the trickle of new supply.

I will progress with the assumption that BTL mortgages are a rough approximation (probably conservative) of the scale of investor acquisition of residential property (or, if they are equity release, they are matched by other purchases as indicated by Ball above).  Thus there is an assumed link between BTL lending volumes and the number of investor transactions in the house sales market.  This concentration on transactions significantly increases the involvement of BTL in the active market.

The set of figures in Table 1 includes re-mortgages.  The more relevant figure for us to consider is the lending specifically for house purchase.  The CML tables are not explicit on this but calculations across tables indicate that BTL mortgages are part of their total “Loans for house purchase” figures in table ML1 as presented in Table 2 below.

There are two fascinating features of this table.  Firstly the total number of mortgages for house purchase remains a fairly constant, with fluctuations usually less than 10 per cent per year.  Secondly the role of BTL is no longer insignificant and its rate of growth as a proportion of house purchase lending is staggering and accelerated throughout 2006.  The total of BTL mortgages for the first 6 months of 2006 was 152,000 and in the second half of 2006 it was 177,000.  This is a 20 per cent increase in identifiable investor consumption of housing in a very short period.  For the first six months of 2007 the number has fallen slightly to 172,000 but this level sustained through the year would show a further increase to 340,000 BTL related transactions in a year.

Table 2: Buy-to-Let Mortgages and House Purchase Mortgages 1999-2006

	Total number of Mortgages for House Purchase	Total number of BTL mortgages	BTL mortgages as a % of total house purchase mortgages	Growth in proportion of BTL mortgages from previous year
1999	1,253,900	44,400	3.5%
2000	1,123,300	48,400	4.3%	0.8%
2001	1,313,700	72,200	5.5%	1.2%
2002	1,396,900	130,000	9.3%	3.8%
2003	1,251,800	187,600	15%	5.7%
2004	1,242,600	217,700	17.5%	2.5%
2005	1,016,400	223,800	22%	4.5%
2006	1,141,000	330,300	28.9%	6.9%

Source: CML Tables ML1 and MM6: CML website

This analysis seems to produce an intuitively a more accurate figure for those of us involved in recent housing market research that has considered the investment activity in city centres for example.  Far from being insignificant, we can see how any source of funding accounting for nearly 30% of house purchase lending has the potential to have a “price driver” impact.  Of particular importance is the fact that BTL probably operates in specific segments of the market and is not evenly spread across the range of transacted stock.  Flats, terraced houses, and lower value areas seem to have been particular targets for example.

Supporters of the BTL phenomenon would argue that the PRS brings in new supply to the market.  The difficulty here of course is that neither BTL growth nor the house price rises of the past few years has impacted much on the volume of new supply (most housing transactions are trades of existing stock rather than new stock).  Overall what is happening in the market is that more and more money (significantly supplied through mortgage lending) is chasing a relatively stable volume of available stock offered for sale including a tiny volume of new supply.

Table 3 below shows the total number of completions of new stock by the private sector for 1998/9 to 2005/6 (later figures not available when this table was compiled) again using CLG Live Tables.

Table 3: Housing Completions 1999/2000-2005/06

	Total Housing Completions in the private sector: United Kingdom
1999/00	160,730
2000/01	152,577
2001/02	153,333
2002/03	163,938
2003/04	171,845
2004/05	183,651
2005/06	188,653

Source: CLG Live Table 209

Most obvious here is that the number of completions stays very low compared to the total United Kingdom stock of dwellings. New completions every year are usually less than 1% of stock regardless of price change.

I now consider BTL in relation to transactions rather than total housing stock or total mortgage lending.

Table 4: Buy-to-Let Mortgages as a Proportion of Transactions 1999-2006

	Total Transactions	BTL Mortgages as a % of transactions
1999	1,469,000	3%
2000	1,433,000	3.4%
2001	1,458,000	4.9%
2002	1,586,000	8%
2003	1,344,000	14%
2004	1,787,000	12%
2005	1,531,000	14.6%
2006	1,774,000	18.6%

Source: CLG Live Table 532 and CML Table MM6

In recent years BTL activity has achieved almost 20% of the market transactions annually. At 330,000 acquisitions this is higher than the volume of new supply and would indicate therefore that home ownership is in decline despite policy objectives to extend it. As Peter Malpass (2006) has pointed out one of the reasons for promoting owner occupation is that its equity value in later life relieves the state of the burden of care funding in old age.

This is also significant for specific segments of the housing market and prices in those segments not just in terms of the simple total volume of the BTL activity.  The average BTL mortgage in 2006 was around £116,000 at a time when average house prices involved in transactions approached £200,000.  This means that BTL activity is probably concentrated in the mid to low price ranges so the proportion of stock that BTL takes in the market segments it operates in is likely to be consistently higher than any of the figures given above.  On new estates for example, developers have reported up to 60 per cent of sales going to investors (Stoke-on-Trent CC 2006).  The capacity of BTL activity to drive local level prices is therefore amplified as its acquisitions are concentrated in specific locations.

Impact on First Time Buyers

The impact on areas of traditionally low priced housing of the type accessible to first time buyers (FTB) cannot be calculated but can be inferred.  When residents of lower value areas talk in interviews about landlords buying everything that comes on the market the figures above illustrate exactly how that can work within specific localities.

CML figures also indicate a slump in the volume of lending to FTBs and in order to keep pumping money into this market sector lenders have devised new “financial products”.  It is now possible to secure a mortgage of several multiples of salary (6 times salary is the highest this author has read about) and for longer periods than previously available (again 40 years is the longest period this author has heard about).

This is always presented as the building societies “helping” young people and first time buyers onto the property ladder.  But committing households to higher mortgage payments for much longer periods both secures the building society a higher ‘take’ from the households’ total lifetime earnings but also helps protect the value of the assets they hold through mortgages.  Increasing individuals’ borrowing capacity helps to prop up prices by allowing them to enter the market and compete with BTL investors, individuals for whom the purchase price of the property they bought no longer had to be connected to the income they had.

What is also quite clear is that far from ‘adding to supply’ and helping to expand the housing market, BTL investment is simply switching ownership of a fairly static stock.  Since 2003 BTL purchases have significantly exceeded the level of new supply (measured as new completions above) so they are eroding the stock owned by household owner occupiers.  Noting this fact implies no moral judgement but it is a significant issue for housing policy predicated on promoting home ownership.

The CLG Live tables have interesting figures relating the lowest quartile house price to the lowest quartile income (this would predominantly be first time buyers and other low income households buying into terraced or small flatted property in moderately desirable parts of towns and cities).  Table 5 below includes some city examples to compare with England and Wales averages.

Table 5: Lowest Quartile Prices and Income to Price Ratio 2000-2006

	2000	2001	2002	2003	2004	2005	2006 (prov)
Eng & Wales Lower Quartile Price	£54k	£60k	£70k	£85k	£105k	£115k	£122k
Ratio lower quartile price to lower quartile income	4	4.2	4.7	5.2	6.3	6.8	7.1
Manchester	1.8	2	2.2	2.1	3	3.8	5.2
Birmingham	3.2	3.4	4	5	5.8	6.2	6.4
London	5.6	6.3	7.3	7.7	8.3	8.5	8.6

Source: CLG Live Table576 Lowest Quartile Price, Ratio Lower Quartile Price to Lower Quartile income.

The problems faced by young first time buyers in this market are very clear as prices accelerate away from their low inflation incomes.  It is also of interest to note that the average lowest quartile house price in England is only slightly higher than the average BTL mortgage.  The use of BTL mortgages in this segment of the market is confirmed by recent landlord interviews in Stoke, Hull, Leeds, Bradford, Calderdale, Wakefield, and Kirklees (Stoke-on-Trent, 2006; recently completed research for Gateway and West Yorks Housing Partnership).  This is certainly not a localized phenomenon exclusive to one or two areas.

Lending strategies and Price Rises

Total mortgage lending (including sales and equity release) by all major lenders had a total annual value of £33.3 billion in 1990.  There was of course a price crash around that time and this annual figure dropped to £15.1 billion by 1995 and then began to rise steadily.  By 2003 the annual figure was £101 billion (CLG, 2007).

In the whole 11 year period from 1990 to 2001 total mortgage lending amounted to £296.3 billion.  In the 5 years since it has probably totalled £470 billion (all figures from CLG live tables 2007).  Willingness to lend these huge amounts is totally dependent upon the putative value of the entire housing stock (and the mortgaged stock lenders hold title to) as calculated against the sale of only 5 per cent of the total stock each year.  The value of those sales in a constrained supply market is, in turn, entirely dependent upon the willingness of mortgage lenders to prop up and inflate the prices of that 5 per cent of sale stock.  If the lenders get cold feet and begin to insist on more conservative property valuations then the market could shudder to a halt and fall.  At the time of writing this appears to be what is happening.  While the bubble was expanding because of irresponsible lending the lenders claimed they were responding to demand, but once the dangers in the bubble start to hit home they exercise responsible restraint to limit that demand.

In a report on the Northern Rock crisis the Financial Times quote one authority (Richard Sexton of E.surv) saying that the reaction of the Building Societies was likely to be just this.  In a Money Supplement the FT reports that over the past few years they had not been too concerned about the ‘accuracy’ of valuations as the market was rising so fast that an over generous valuation would soon be caught up to.

Sexton said it was crucial that lenders had accurate valuations now that property prices have started to cool.  In a rising market this is less of an issue, as even if a property was over-valued at purchase, a few months down the line its value should have caught up.  (FT Money Supplement)

This behaviour is both inflationary and high risk in ways that we do not associate with Building Societies.  The modern market mantra is that where there is high risk there is high profit.  However, in the US, when the risk stopped being a dog with a bark and became a dog with a bite the market has collapsed affecting share values and also home-owners.  The effect has spread in the “credit crunch” and as Building Societies now start to seek resale valuations on city centre flats before lending, as opposed to accepting surveyor’s assessments, they may find that the resale values do not exist.

While this is not exclusively a BTL mortgage issue there several detailed stories about BTL acquisition and inflated valuations.  The main headline in the Financial Times of February 9 2008 was “Mortgage fraud spirals out of control”.  The BBC (Panorama 04/02/08) has reported on valuation and lending practice problems in Nottingham, London and Leeds and their Inside Out and You and Yours programmes (BBC) have covered a specific series of mis-valuations relating to property in Leeds.

A valuation report was carried out on behalf of a well-known mortgage lender for a flat in Leeds.  It was valued at £238,000.  But less than a year later [when the owner wanted to sell] another surveyor valued the property at £133,000, a fall in value of £100,000.

Of course the second valuation is still only ‘true’ if the property sells at that price; prices in these pure investor markets with low residential demand may actually be lower than valuations, even conservative ones.  Investment clubs also seem to pose a particular risk (see Practical Property Portfolio example in Sprigings 2007)

Impact on new supply

Finally, a tangible result of speculative investment is the impact on the type of new supply.  The change in the supply of new dwellings is very obvious both nationally and locally.  This may be simply coincidental with other demands on construction and planning such as the requirement for brownfield and city developments at higher density but I believe the city centre markets would not have taken off to the extent they have without BTL and other investment related finance (as opposed to home ownership acquisition).

Table 6: Flats as a percentage of private sector completions for sale

	1997/ 8	1998/ 9	1999/ 00	2000/ 01	2001/ 2	2002/ 3	2003/ 4	2004/ 5
Flats as % of new: England / Wales	12%	14%	15%	17%	22%	26%	33%	40%
North West	10	13	15	17	24	19	27	39
Yorkshire & Humber	6	5	6	15	11	19	29	31
East Midlands	4	6	9	4	8	12	14	20
South West	8	8	9	11	19	24	26	31
London	58	64	57	58	66	70	81	85

Source: CLG Live Table 251 and 253

Table 6 above shows that flats are now almost half of new dwellings constructed nationally. This is presented in the table as both a national average with some regional comparisons.

The growth of flatted dwellings is a consistent trend across regions but is not of a consistent rate.  There is an obvious time lag between the rapid growth of investor interest in the housing market in the late 90s and the development of popular investment properties (flats are popular with investors for security reasons and for letting to households without children, often young professionals, new property is popular because of low maintenance costs etc).

CLG figures (Table 547) also show that there is a trend for new dwellings to be smaller than they have been for 15 years with property offering less than 5 habitable rooms being 40% of supply (up from 21 per cent in 1997).  The last time the figure was as high as 40% was around the time of the last price crash in 1990.

The local impact on supply is interesting.  Table 6 above shows the rapid growth, even from an already high base, of the flats market in London.  More flats equates to smaller space standards.  In 1996/7 in London 1&2 bed stock accounted for 68 per cent of new supply.  By 2004/5 this had risen to 84 per cent.  Three bedroom new supply had fallen from 23 per cent to 11 per cent in the same period.  Three and four bedroom HOUSES have fallen from 30 per cent of new supply to 13 per cent of new supply.  These market shifts are significant and as developers still have a ready market for their products it is likely that the development pipeline is similarly if not more committed to smaller dwellings dominated by flats.

At more local level the change in the nature of new housing supply has been dramatic.  DTZ report that

In 2000, 86 per cent of all new build properties in the Leeds District were houses and 14 per cent flats.  In 2006, this has changed to 74 per cent flats and 26 per cent houses.

One effect of this is that the type of dwellings that most household consumers want to buy (namely houses), are now a dwindling proportion of new provision.  Not only does BTL acquisition come into direct competition with aspiring owners for existing stock but it seems to have had the effect of reducing the annual new supply of available houses thus increasing the competition for an increasingly scarce resource.

Conclusions

This paper has drawn on readily available data to show that BTL activity, to which other investment purchases could be added, has been demonstrably of a level to impact on housing markets locally and nationally in terms of influencing both price and supply.

Firstly the scale and rate of growth of BTL is far higher than is commonly assumed.  When related to the volume of actual transactions BTL can be seen to have been acquiring at least 20% of the transactions in any one year for several years.  This may still be negligible in relation to total stock, but most stock is not active in the market in any one year and therefore does not have any influence on price or new supply.

Secondly the impact of disconnecting house prices from household incomes, particularly obvious for lower income households and lower value stock, affects policy makers’ ability to predict market trends.  Significantly the Bank of England may have responded slowly to the expanding housing bubble.  Failing to identify its cause, they assumed it would run out of steam long before the collapse of a bank triggered a halt in lending.

Thirdly, the dominance of BTL (and other) investors in the market has been attractive to developers.  Proposed stock is often sold at the design stage giving developers guaranteed cash flow.  Developers have probably also worked out that the prices they can secure easily are inflated well beyond their normal expectations.  This has meant that developers have responded to immediate demand by increasing the supply of stock to investors.  More flats and smaller properties is the consequence that actually puts further pressure on the housing that consumers want to buy.  Combined with new financial products extending the borrowing capacity of household consumers this effect has added further to the speculative bubble.

Finally another impact of this has been to force would be purchasers into rental as the total home ownership stock declines for the first time in a century or more.  This is happening while public sector policy is increasingly assuming growth in home ownership that will fund welfare in the future (Malpass, 2006).