No. of Recommendations: 196
1989 vs 2005: Lessons from the past

This is a long post (I hesitate to use the word epic; but it is indeed very long) which incorporates a lot of my current thinking on where we go from here (something I don't usually speculate on) that I have been working on for a couple of weeks and planned to post yesterday to coincide with the Halifax numbers, only it didn't seem appropriate (I also spent much of the day trying to find out where my wife was, fortunately she is okay; then the rest of the day trying to work out how she was going to get home).After some of the O/T comments yesterday I think its appropriate we get back on track. Things are quiet but hopefully a few people will get round to reading it and I will come back to a revitalised PMT board with some healthy discussion.

I do not think it is too much of an exaggeration to confirm that there is some sort of unanimity on this board. I think we are all agreed that prices are probably above (at least in aggregate UK terms) where economic fundamentals would have them, and therefore that they must come down in relation to wages. Where there is disagreement is on how exactly this adjustment will happen. On one side we have the GSD camp who expect (depending on who you ask) to see prices stay about the same in real or nominal terms, and hence gradually adjust as wages "catch up". On the other there is the "bear camp" who expect significant price cuts in real terms and probably in nominal terms; a sharper but shorter period of adjustment. I hope there is nobody left in this country, never mind this board, who expects double digit growth to resume. In trying to decide which scenario is most likely I think it is worth looking at history; although things may indeed be "different this time" I think there are bound to be points of similarity; hopefully we can think about which differences really matter and which are irrelevant.

The most recent reference point we have for a housing crash is the infamous crash of '89; a year that should live in the memory of housing market participants as long as the Crash of '29 does for those who make their living in the stock market. Unfortunately it does not; with most of those buying houses for the first time seemingly ignorant of the events of the first half of the 1990's, and those who should know better sitting comfortably on a large wedge of equity that has wiped out their long term memories leaving them with only a vague memory of the bad times.

1989-1995 was indeed unique because it was the first housing crash where prices fell in nominal terms, on a national scale. Previous crashes had seen real falls, but with much higher inflation. As we know have inflation lower than at any point since the mass expansion of the owner occupied market that begun after the second world war 1989 should be the best reference point to use when we try and work out what is going to happen next. Of course there were a number of important differences, so I shall try and determine what is or is not relevant to our current plight and what might happen over the next few years.

I will talk about three phases around the top of the cycle:

While prices are going up in nominal terms, faster than the rate of wage inflation.

The next period is the standoff when volumes fall but prices are still going up slowly or are static; this is the period we are in currently.

Once prices start falling in real terms (and thus definitely in wage adjusted terms).

Firstly some graphs that you might find interesting. What I did was try and “match” the peaks of 1989 and 2005 (roughly March 1989 and August 2004). I drew both London and the UK since the former is normally a leading indicator – I will talk about London a lot although clearly the whole of the South East would do almost as well. In all of these graphs pink is the entire UK price for 89-2005, yellow the London price rebased to the same level as the UK price at the start of the period. Dark blue is whole UK for 2000-2005; light blue the London price again rebased. 1989-95 prices are on the Left hand axis; 2000-2005 prices on the right hand axis. (these graphs and indeed this analyais are on Halifax figures done before the last release; although one month or quarter isn't much to go on the pendulum does seem to be swinging away from GSD to my own preferred scenario). The axis are adjusted so that prices are at the same level at the start of the period; and to show roughly the same nominal rise for the whole of the UK figures to the peak. Obviously the current figures stop where we run out data, whilst the 1989 figures show the course of the crash. There are also self explanatory top and bottom X axis; these are at the same scale.

In nominal terms then;picture_id=931

and in real terms;picture_id=932

Now for the predictions, I considered 3 scenarios:

- Both London and UK prices would now fall at the same REAL rate as before (dotted lines)

- Both London and UK prices would fall at the same NOMINAL rate as before (lines with circles)

- Both London and UK prices would remain FLAT in nominal terms (Solid line)

The results are shown on three graphs:

Nominal terms (excluded the flat scenario as it is obvious what this would look like);picture_id=937

Real terms;picture_id=938

House Price to earnings ratio (with both Y axis the same);picture_id=939

Some comments: London has gone up less than last time relatively (the rest of the UK has gone up more) and seems to have flattened more than last time when it dropped quickly after a sharp peak. The same real falls will be the fastest and quickest; bringing both London and UK prices back to 2000 levels in real terms; giving a real graph that looks strikingly like the graph last time (with the difference that London goes much lower as it has peaked lower). Because inflation is lower the nominal falls would be larger than last time; but would still only bring us back to the nominal levels of the end of 2002.

The same nominal falls would of course have less effect in real terms since inflation is lower; flat prices would have even less effect. To show how long it would take to reach pre boom price earnings levels the third graph covers a longer time span. You can see with the same real falls we would reach UK P:E's of 3.5 in about 5 years with the same real falls; about the same as last time, with the same nominal falls it would take about 7 years, and with flat prices about 10 years.

The boom

One of the main differences between then and now is the path we took to get here. The housing market got going in 1985. Although prices had picked up in the four previous years of the decade much of that was due to rises due to earnings. The house price to earnings ratio barely budged from just over 3. Over the next few years aggregate prices went up by 15, 17, 25 and 22%. This was well ahead of earnings and indeed prices; with inflation relatively low by the standards of the previous 2 booms. It seems that the causes of the boom were many; real incomes were rising from 83 to 87 by 3% annually and nominal mortgage rates fell by a third. Much of the growth in lending came from the entry of the banks into the mortgage market. It seems likely that this financial liberalisation was a key cause of the boom.

The current boom got going in a big way in late 2001; about the time when most analysts think that prices reached “fair value” having been at less than that for a number of years. However double digit price increases started to appear in 1999; there was a pause to a degree in 2000 and 2001 but when interest rates were savagely cut after September 11th (although they had been heading gradually down for a year) the market took off in 2002 (22% rise), 2003 (18% rise) and only began to stall in July 2004 a few months after I began frequenting these boards – at this point the Halifax YoY showed a 22% rise. The savage rate cuts of 2001 were probably a key trigger; as I discuss later there are no obvious demographic reasons this time wheras last time there were. Real earnings growth was a little higher in normal in 2001 though prior to this it had been very low; but the growth (2.5% at best) was well short of the real growth of 4.5% in the late 80's. It is also worth bearing in mind that with inflation so much lower in 2000 than 1989 the comparision of the real variables is more important – the key variable for me (interest rates adjusted for wage inflation) were negative (after MIRAS) in the late 80's only turning positive at the beginning of 1989 and never reaching 2% in the subsequent downturn. The same variable was over 2% in 2000; real wage growth and lower rates brought it down to around 1% for much of 2002 – so the conditions for a boom were there, but not to the same degree as before. When rates started rising (and to a lesser extent wage inflation slowed down) the same indicator has risen to nearly 3% - as I have said before comparing top of boom with top of boom the economic conditions are much less benign than in 1989.

Financial liberalisation of another kind probably played a key part in the current boom. The extension of BTL mortgages on relatively easy terms, in itself a product of the introduction of a new kind of landlord-tenancy relationship; is definitely one of the reasons why we are here now; of this more later.

Worth noting though is how credit constraints were loosened as the boom proceeded; with banks asking for lower and lower deposits from first time buyers (the 100% mortgage became possible; wheras 15 years before a hefty deposit would have been nedded) and to a lesser extent lending more against a given salary; in general the level of deposits increased in the boom (as tends to happen as poorer buyers are shut out of them market) though not to the same degree as the current boom (24% vs 17%). Another credit / earnings related story was the gradual movement of more women into the workforce, particularly in areas where full time male employment had been badly hurt by the slowdown in manufacturing.

However I am getting ahead of myself. Back to 1989 where the country was in the grip of the "Lawson boom". After abandoning monetarism in the early 80's; and shifting the power base of the country from unions to employers, as well as moving public sector workers (who negotiating en bloc had always been responsible for a fair chunk of wage inflation) into the private sector the government believed it had done enough to ensure economic stability and felt free to pursue a very loose fiscal policy with relatively low interest rates. Combined with the "big bang" in the city the result was a feeling of prosperity, centreing on the city of london (this should sound familiar). With the manufacturing sector in decline (a long term trend though, not one that began in 1980) London was no longer the powerhouse of the economy - it practically was the economy. IT is of no surprise then that this was a London led boom with prices in London starting to pick up in 1984, growing QoQ by 5% or more until the end of 1988 wheras in the north ignoring a brief burst in 1986 the growth did not really hit until 1988 and it continued into 1990. The ratio of London to Northern prices went from 1.6 in 1986, to 2.9 in 1988; at the peak of UK prices it was around 2.3; it then fell to a low of 1.5 in 1992; and it remained below 2 until 1998 (in retrospect a great buy signal for London).

This time of course we have had Mr No More Boom and Bust Brown with his Golden Rule. In theory Mr Brown has used contrcyclical fiscal policy to ensure that the public finances remain balanced over the economic cycle. In practice he seems to have let rip with spending to the extent that at the time of writing, when a recession seems likelier than ever, he will be forced to reduce spending or increase taxes to keep to his golden rule. From that point of view I do not see much difference between then and now; although he does not have the monetary levers any longer.

London has led the boom again, although there seems to have been more of a lag with London slowing down earlier around 2 years ago. As a result although the ratio reached similar levels (3.0 in 2001 and 2002) it has since fallen, and at about the approximate peak (somewhere between last summer and early this year) it is around 1.9. Hence on the graphs we see more of a rise in the whole UK figures compared to 1989.

Another key factor (and one that is almost always used to justify any price rise) was demography and the demand for housing stock. The 80's saw an acceleration in the expansion of the owner occupied sector with a higher rate of council house sales (it did not, contrary to popular myth, only happen after 1979). This expansion results in a higher proportion of FTB'ers than in a stable situation where the owner occupied stock is roughly stable. There were certainly a great deal
more FTB'ers in 1989 than now - 50% of buyers rather than 30% now. A key point is also that the expansion of owner occupation meant that first time home owners were much poorer, relatively speaking, than before. Another demographic effect was that the mid 80's were probably when the last of the baby boomer generation bought their first houses. Both of these effects (the expansion of housing and the baby boom) led to a higher number of FTB'ers than would otherwise be seen than in a market where the flow of owner occupying households was roughly constant. Although we might have expected to see a fall in FTB's going into the last boom in fact the demographics overwhelmed the price rises and numbers were pretty stable although the proportion did pick up into the bust reaching a peak of nearly 60%. It is only in the last 5 years that there has been a definite fall from below the 50% level – the BTL brigade may be to blame if they have been responsible for many of the recent buys; however the underlying demographics are likely to be key as well since the fall seems to have begun well before prices began to accelerate. Looking at quarterly numbers of FTB'ers (given that the change in transaction volumes can make the proportion move around) these ran at around 130K from 1984 onwards, growing into the boom and then dipping significantly from about 1989 to 1992. From 1992 numbers then picked up, although never reaching the boom levels were usually over 100k. The current picture is clearly different; with FTB numbers staying level until 2002 before falling off a cliff (sustained dropping below 100K for the first time since 1992); prices continued to rise strongly for 18 months after January 2003 with relatively few FTB'ers buying though overall transaction volumes also held up - BTL'ers may have filled some of the gap but increasing chain lengths have occurred as well (a consequence of losing volume at the bottom of the chain).

As well as the regional differences different types of houses did better than others in the boom. Flats, 1 and 2 bed houses and new build all did much better. The primary reason of course was that as affordability got stretched the younger and poorer buyers had no choice except to pay more for entry level housing. This boom has been no different in this respect.

Trigger happy

In retrospect the end of MIRAS double taxation relief on is usually cited as the trigger that caused the boom to end. Everybody likes a trigger, it seem unacceptable that a boom can merely collapse under its own weight although this is usually what happens - there are no more buyers left; so we have to find a reason for it. For example the reason quoted for the '87 stock market crash - some bad trade figures - was particularly unconvincing. God only knows what trigger will be found in retrospect for any forthcoming crash; perhaps mortgage "A day" last yeas when the FSA took over mortgage regulation. What certainly were not triggers were a rise in interest rates when we tried to stay in ERM (several years too late) or unemployment / recession. These factors probably made the housing crash worse, and the housing crash in itself made the recession worse; and indeed may have caused it. But the time sequence is wrong for them to be the cause of the crash.

I think it is hard to point to a single trigger for the crash that seems to be about to happen. The key moment appears to have been somewhere last summer. Since then prices have wobbled sideways and unlike previous wobbles in the past few years do not seem to be ready to continue their increase. Personally I cannot think of a single strong factor; however when interest rates started to rise over a year before the peak that was probably the beginning of the end.

Stand off

As we know London led the fall and other regions were slower to follow. Ultimately though there were nominal price falls of greater than 20% in most “southern” regions of the UK. As I noted before the regions that went up the most subsequently fell the most and fell first. The subsequent correction probably took prices well below their fundamental values. It was a long hard trough at the bottom with the relatively quick nominal falls followed by several years of flat prices and real falls.

The nominal falls came out because there were a lot fewer buyers; many had already bought in the boom, others had become priced out and as for the few left that could buy initially they saw a flat market and no reason to overstretch any longer to get on the property ladder (a situation we are beginning to enter now); and then once the nominal falls began they were scared off as the collective pysche of the nation turned against housing for nearly 10 years. However in the initial stand-off situation where the buyers stopped buying there was no reason for prices to fall if all the sellers stood firm (a classic "prisoners dillemma"). Looking at the nominal pattern of the top of the cycle there were a few months (from Q4 89 to Q3 '90 or so) when prices held steady. Once sellers began to crack however nominal prices could start coming down.

The flat / standoff period in the current market appears to have started early last year in London and a few months later in most parts of the UK. A key difference; which I will examine at length, is the stand off period in London appears to be much longer than in the previous boom. Could this be the mythical beast GSD? Of this more in a minute.

I think it is worth examining in detail the process by which the nominal prices probably began to slip. Of course in earlier real price falls inflation had been so high there was no need for this to happen.During the "stand off" period buyers wages would have been rising significantly. Although volumes fell drastically (the sign of the stand off period) sellers knew they only had to hang for a while and let inflation catch up with their prices. The real falls were
terrible (particularly after the mid 70's); but nominal falls were avoided - a classic case of "money illusion" on the part of the sellers. Thus the country had at least 35 years of rising nominal prices and with home ownership in many families less than 35 years old effectively no collective memory of falling nominal prices at all.

Thus the stand off period had to be longer than before to allow for correction without nominal reductions. If prices are overvalued by 20% then with 25% inflation (as in the 1970's) it takes less than a year to correct the problem. But with 5% inflation in 1989 it would have taken 4 years. As the level of inflation is common knowledge sellers in the late 80's must have realised they did not have the luxury of waiting less than a year - barring economic crisis many sellers can stay put for that long - before they could sell. Also lower inflation means that the real value of debt is eroded slower. So if a buyer had overstretched themselves in the 1970's they could sit tight knowing that they only had to make the payments for a few months; inflation would be eroding them anyway and there was a good chance a buyer would turn up without nominal cuts being needed. In the late 80's when people had overstretched themselves (which clearly they had with deposits much lower than now, and P:E ratios of FTB'ers higher than normal) they did not have the same get out of jail free card.

There are many reasons why people might be forced to sell. At the beginning of the crash the reasons would have been different to the end. Some of the reasons at the beginning of the crash might have been cashflow overload (an initially overstretched buyer, after running up debt for a few months to pay the mortgage and the bills; throws in the towel - without any exogenous / external shock), job relocation, death of homeowner. With the exception of the first the latter are happening all the time even when the market is strong and these sales are easily absorbed except when there aren't enough buyers. There are also people wanting to trade up or trade down; but in a slow market they can usually delay their decision more easily than those in the previous group who are forced sellers to a degree. It may also be that the longer the standoff, the larger the stock of delayed sales that builds up; and the more savage the nominal cuts that follow have to be to. This may be particularly relevant for anyone expecting a smooth GSD all the way down to the bottom of the cycle – the longer the stand off (which looks like a GSD) the worse the crash might be.

The smoking gun and the role of the banks

In the latter stages of a crash, particularly the 1989-93 crash we can identify other reasons that become more prevalant. As the crash "feeds" on itself creating waves in the wider economy (reduced consumption, unemployment...) and once nominal falls enter into the public mindset there will be very few buyers and even more sellers. Some reasons for sales once the crash was well underway would have been people responding to interest rate shocks (rates went....) and those having to sell due to unemployment (unemployment went....). Here we come to a key point. These forced sellers were often in arrears of their mortgage payments. In many cases they simply handed in the keys and let the banks reposses. Many of them mistakenly believed they were exercising their "free option" and that they only had to share in the upside of house price ownership and the downside was somebody elses problem (astonishingly I speak to plenty of people today with the same
mistaken belief). The banks were definitely culpable though; in many cases they encouraged people to hand back the keys (see, but more importantly they were very aggressive in pursuing repayment with the result that many "marginal" borrowers defaulted when with sympathetic handling (a years payment holiday perhaps) they might have kept their houses and eventually resumed payment of both interest and principal. Recall that many homeowners were relatively new and from poorer echolons of society, and hence more likely to fall into debt.

Additionally one of the results of the de-industrialisation of Britain and government encouragement had been a rise in self employment; self employed people formed the bulk of the bankruptcies at any given time as they lack the steady income that would allow them to weather downturns and are often highly geared. The entry of the banks into he mortgage market had also encouraged aggressive lending to gain market share; in retrospect credit constraints were loosened too much. As a result there were many more marginal borrowers than in previous market booms and the banks, new to the mortgage game (not strictly true, but before the 1980's they had generally dealt with richer borrowers) had dealt with them in such a way that provided short term gain (they got the security back) for medium and long term loss (the housing market crashed, so the security was worth much less). By taking houses off marginal borrowers, and then of course selling them at auction - often for any price (frequently below even the depressed market value), the banks in themselves created many of the nominal falls that occurred; certainly once the crash got going their debt management policies probably accelerated falls with a substantial portion of the houses being sold after a repo and below even the declining market value just to get the non performing asset off the books (contrary to what the borrowers believed the banks could, and did, pursue them later for the difference plus several years in interest).

This loosening of credit constraints, then very tight debt management; is an example of procyclical behaviour by the banks. This is the counterpart to the excessive optimism shown by governments, borrowers and consumers in booms; followed by excessive pessimism in busts. Such behaviour makes market, business and credit cycles more pronounced. And as I have already stated, this behaviour has not gone away; if anything it has become even more pronounced.

Even if borrowers did not get repo'd the market was stifled further due to the difficulty of getting loans with negative equity; in retrospect another countercyclical move as allowing such loans would have given the market more liquidity. The solution that many borrowers took was to rent out their old house (often without informing the lender) and buy another if they had to move, or really had to trade up. In retrospect we see here the germs of the late 90's BTL market; at the time it was a survival strategy. The lenders (and the underlying legislation) going into the current boom have made it much easier, and cheaper, to legally BTL or Let to Buy; but the key point will be wether they act as before and tighten up on this lending just when borrowers need the escape route the most.

Thus although one key component required for nominal falls (low inflation) is in place it will be up to the behaviour of sellers, buyers and banks as to wether they occur, how quickly and for how long. The initial nominal falls will probably come from forced sellers as before; we have already had the interest rate rises (though as before they will probably come down again) and there are signs that the wider economy may be beginning to suffer. These effects will be magnified and enhanced by the relatively high level of non secured debt and higher price earnings ratios compared to last time - more "throwing in the towel" sellers whose finances have collapsed under their own weight. The higher deposit ratio means they will probably not be in negative equity for some time, but I suspect that these average deposits mask a range between 130% mortgages and those with rich parents; the deposits may have been garnered directly or indirectly from unsecured lending - and in any case it will be a cash flow rather than a balance sheet problem that will trigger the forced sale.

However most sellers will be in a standoff situation, so in the inital standoff period will it be they or the buyers that will crack? In the standoff economic conditions are still benign (as last time), but with low inflation unless buyers change their minds over time more and more sellers will have to sell for one reason or another (and with lower inflation the standoff period will be even longer). So the role of the buyers will be crucial. AS before the role of momentum is probably the key here. Buyers buying through fear of getting left off the ladder or greed for the profits to be made when the market ews rapidly rising will stop buying once the market is flat or falling and appears to be staying that way.

Perhaps the key statistic for me is the difference between the bell-weather London markets in each boom. There does indeed appears to have been a genuinely flat market in London for at least a year wheras in 1989 we can see clear sharp top. Is it then possible that the UK could follow, with the first Gradual Slow Down in the UK housing market (at least in nominal terms), although it should more accurately be called a gradual fall, and in real terms it is already a fall.

A few things bother me about this theory. The first is that London itself is many micro markets, many of which have fallen sharply (i.e. Tower hamlets down 6% June 04-05 Hometrack) and others that have continued to rise (up 4% in the same period).It may just be that the correlation / lag between intra London price movements is less than last time; creating an aggregate flat market where non exists; it is also worth pointing out that this flat market has been against low volumes – a sign of a standoff and a build up of unsold stock. On the upside we have seen lower correlation between London and the rest of the UK. As a result when London stalled the regions carried on going and as a result (clear from the graph) the regions have risen more than London than in the last boom as the London / North ratio illustrates. Thus the experience of the regions this time is likely to be worse (since they have probably risen far beyond their fundamental value compared to London), and so even if London does managed a flat market the regions will not; and neither will the UK in aggregate. I think the 2 speed economy of Britain has developed further in the last 15 years and the prosperity in London may be enough to keep the housing market struggling on and if recession hits the regions worst then they will do very badly both absolutely and relatively. Docklands and other marginal areas of London will be the key indicators. Although Canary Wharf is now complete and unlikely to go bankrupt there is a great deal of vacant commercial office space in London (with the exception of the West End that has done very well from hedge funds); and a very large number of new build flats coming on stream. So although prime London will probably weather the storm the more marginal areas could yet drag it into a proper nominal fall.

One thing that disturbs me when comparing the booms and that could be a key factor when deciding if the buyers break the standoff is the makeup of the buyers. Firstly chain lengths have got very long in recent years - implying fewer people coming in at the bottom of the market with "new money" and secondly the number of first time buyers is extremely low, with some of the difference made up by BTL'ers. Partly this can be explained by the one off effect that massive expansion of home ownership had; effectively resulting in many more first time buyers than in a stable market. Demography also played a
part; with the demographics now at an all time low in terms of people reaching typical FTB age compared to 1989 when younger and poorer baby boomers were still entering the market. Because of these two factors the natural effect on the market would have been a reduction in FTB'ers even if the market had not boomed. This would have meant longer chain lengths in any case, but in a non bubble market should have kept prices down. The arrival of BTL'ers saved the day to an extent, but perhaps more importantly the few FTB'ers left were then allowed to overstretch themselves like never before - so though few in number they bought in enough money to create a boom where the fundamental demand was not really in place. As a result as the boom went on the key indicators of relative potential FTB affordability - FTB age and salary relative to the median; are much higher than before (Age as we discovered recently has partly increased due to the number of previous cohabitees divorcing- without this group; who carry much higher deposits than true FTB's the numbers would be even more miserable). The previous boom had its roots in demography and relatively low levels of home ownership; credit allowed it fly. This boom has no such solid foundations of first time buyers; lengthening chains and the number of BTL'ers show that.

So the buyers that have come in have been relatively richer, older and with much higher price to earnings ratios than before. This is a very limited pool of buyers to build a sustainable housing market from. The pool of money that has formed the relatively large deposit base of these recent buyers shows signs of coming to an end. Whatever its source; MEW'd from relatives (or alternativley downsizing parents) or principal residences, borrowed from the future by increasing unsecured credit or reducing savings and pension contributions - none of these sources are sustainable, relying on a ponzi pyramid of money. Wheras buyers in '89 stretched themselves on low deposits and relatively high income and relatively low unsecured debts, todays recent buyer has a substantial deposit - but significant unsecured debts to pay; and an even higher multiple. It is hard to see, with banks already beginning to tighten up on easy credit self cert and its ilk; and with the pool of deposit money run dry (only to be replenished by the old fashioned method of saving), where the buyers are going to come from. Surely any FTB'er that could beg, borrow or steal a deposit has already bought?

The vested interests and even several people on this board often think that the standoff can be broken by some bold move on behalf of government. Several of these ideas are just silly (raising the stamp duty threshold to the point where you could nearly buy a 1 bed flat in Birmingham without paying duty, giving grants to key workers, the £50k home, property in SIPP's) in that they affect a very small group of people. However more sensibly the question of cutting interest rates
is worth discussing. It is an appropriate time to question wether it will be BTL'ers that will end the standoff. The reasoning being possibly that BTL'ers are more likely to be fundamentals rather than momentum buyers. The logic being that BTL is much easier, and a known force which is a clear difference between the last boom and the current one. I see a number of problems with this scenario (or indeed with a group of fundamentals based FTB'ers jumping in and saving the

Firstly is the difficulty of judging the fundamental price. Any calculation based on very low current interest rates is flawed since it is likely that rates relative to inflation will rise in the future (this should be a long term investment). Capital appreciation is clearly going to be higher the better the initial yield as well. IT is hard to believe that a sensible landlord will buy property at a P:E of 5 even with rates at 3% when in the last crash it nearly reached 3. Lenders might tighten up their restrictions on BTL - especially if they get stung. There is an argument that there is already enough rental property to satisfy demand; many areas are saturated. And even the most die hard fundamentals buyer will find it difficult to buy when it would clearly be catching a falling knife. Apart from the rental demand argument these all hold true for FTB'ers as well. All in all I do not think there will be early knife catchers in significant enough numbers to make one jot of difference.

One thing is clear - the interest rate has to come at the perfect moment - during the standoff period. Once that is over then it will be too late. Unfortunately cuts seem to take a while to work their way through the system, so it may never be possible to achieve the required aim. The worse scenario (Japan) could see very low interest rates but no recovery. After all at these low nominal rates it is the repayment of principal that dominates; only savage nominal price falls
will bring that down.

Once the standoff period is over and nominal price cuts have begun, and worked their way into the public conciousness; the bust will probably gather momentum. Factors that will speed it up include (like last time) increases in unemployment (which in themselves are more likely since the housing, credit and consumer markets are linked creating more extreme business cycles in all) and short term thinking by panicking banks. It is hard to see a fiscal stimulus saving us; no more boom and bust Brown is in the position (no matter what his sums say) of having to tighten policy just at the wrong moment. In any case the likely increase in the savings ratio and the repayment of secured and unsecured debt (which thanks to low wage inflaton will be much more of a problem than before) will swamp any fiscal loosening. If the banks have learnt from last time they might slow down the rate of nominal falls - fewer repo's and auction sales bringing down prices even quicker. However this may or not be a good thing.

How far and how fast: Short sharp shock or long painful nightmare?

The banks may have done us a favour before. With the relatively low inflation of last time a long nominal standoff with a illiquid housing market would have inflicted serious damage on the economy. With even lower inflation this time the fall would be even longer unless the nominal cuts are even sharper than before. We only have to look at Japan to see the dangers of prolonged pain. You only have to look at the graphs to see how long it would take to reach pre boom P:E levels if nominal prices are static or only fall as far/fast as before. The question is, once prices start falling, will they do so quicker or slower than before; and for how long (or to put it another way when and where is the bottom?).

Let us look a little further back in history: all figures in UK wide real terms

1971 – 74 Boom: 3 years, 53% rise
1974-77 Bust: 3 years, 26% fall

77 – 80 Boom: 3 years, 28% rise
80-83 bust: 3 years, 18% fall

83-88 Boom: 5 years, 65% rise
88 – 95 bust: 7 years, 32% falls

95 – 2005 Boom: 10 years, 108% rise
2005 - ???

A few things are clear here. Obviously the rises are larger than the subsequent falls because in real terms the market has gone up in the long run. The 80 boom /bust was small fry compared to those around it. The cycles seem to be getting longer, and larger in amplitude – larger rises, larger falls. And the fall in real terms is roughly half that of the previous rise; plus a long boom seems to lead into a longer bust (although it is too soon to say). With longer and larger swings we might expect the speed of change to be roughly constant and it is; between 8 and 12% a year in the boom and –6% and 10% a year in the bust; so slower on the way down and particularly slow in the last bust.

There isn't much data here to give us a statistically strong case but we might think about what is underlying these relationships if they are real. The faster boom than bust is easy: nominal prices are very sticky downwards but adjust upwards with remarkable ease. With a higher rise it will take a high fall to “undo” the damage. The increasing amplitude might be a red herring, but the long bust after 1989 probably took the market well below fundamental value in a significant over correction. The relatively low inflationary environment probably contributed to this as well, meaning it would take longer for wages to catch up. This was the slowest slide in real terms as well, probably a result of nominal stickiness in a relatively low wage enviroment. One reason for the increase in amplitude could be the gradual loosening of credit and the growth in unsecured lending which would result in more cycle reinforcing procyclical behaviour by borrowers and lenders. The length of boom and bust might be related to “buyer memory” with the unique nominal falls of the early 90's sticking in the mind for longer than the real falls of earlier years.

Extraploating then I make the following predictions based on the data above:

- We have had a long boom and seem set for a long drawn out bust; at least as long as 89-95; maybe even 10 years or more.
- We have had a substantial boom (doubling of real prices) and thus I think we will see a substantial bust (halving of real prices)
- With low inflation and slow painful nominal cuts the speed of real cuts is likely to be as slow (averaging 5% a year) as before if not slower

Fortunately these three predictions are self consistent. They are of course predictions above averages. In the last bust the real falls, once the bust got properly going after a year or so, were 8-10% a year. In todays inflationary environment that means nominal cuts of up to 8% a year, for several years. I think there is an upper limit on nominal cuts.

However these are not just estimates based on a scarce few data points. There is some further evidence to support them.

Prices are now, in my opinion, much further above their fundamental value than they were in 1989; with demographics looking worse as well (as the baby boomers reach downsizing age). With the procyclical, self reinforcing, mindset of buyers and banks an overcorrection below fundamental values seems inevitable. In relation to wages I think the P:E ratio will fall from 5.5 to somewhere below its current fundamental value of around 3.5 – so a fall of at least 36% in relation to wages and more likely to be 40% plus. Thus a 50% real fall would be needed as with such low real wage inflation that even over 7-10 years wages would not be catching up fast enough.

Because people have taken on a lot of debt with low wage inflation they will be paying it back as a large chunk of their income for a lot longer than before. They will also probably be saving more (although many have not been saving at all) and putting money into pensions (“my pension is my property” perhaps?) possibly with Government compulsion. They will also be trying to eradicate the memories of several years of nominal price falls. Hence a long time scale seems highly plausible.

These are not just predictions about averages but about aggregates. I have already posted on the likely regional nominal and real price changes. The key difference to last time is that it is the regions that are further above their correct fundamental level; especially given that I see no real improvement in their relative economic prospects which at best have remained the same. Thus the areas of the North that escaped the 20% plus nominal cuts of the South in the early 90's will be nursing very substantial nominal cuts indeed. One possibility that I have also mentioned before is that the long nominal cutting period could result in correlations between regions increasing and hence larger nominal cuts at the aggregate level. If that is the case (and it certainly did not occur in the last crash) then we may be looking at 6 years rather than ten.

One final point would be what would happen if the economic fundamentals worsened. The variable that has the single biggest effect of fundamental value is interest rates discounted by wage inflation. We can decompose this as:

Nominal interest rates – (General inflation + Real wage inflation)
= Real interest rates – Real wage inflation

Hence we would need to see real wage inflation growing; or real interest rates falling to see the economic environment improving (and a higher “fair” p:e). As I have said before the sorts of numbers required to justify the current high p:e are pretty ridiculous; effectively substantially negative real interest rates. I do not think with an independent Bank of England that we are going to get those sorts of numbers. Hence I reiterate that to be in line with the fundamentals property prices relative to wages need to come down considerably; and I see no reason why this process once it begins should stop until the fundamental level – if not lower – is reached.

Returning to the graphs I had at the top of this post then I think the circled line – similar nominal falls is the most likely option at least for the first couple of years. But we will then see them for longer and after 2008 or 2009 will be entering unknown territory; at this point it is even possible the market will get enough of a shock to scar enough generations sufficiently that there is no property boom for at least another 20 years. That is wishful thinking though - I think the next boom will be going again by the end of the next decade. I'll see you again in 2020 then, to find out if Bialy will be boasting about his perfect timing (buying a house in 2014 for tuppence ha'penny and selling it again in time to buy into god knows what bubble)?

Print the post  


Practical Property Investing
We have another board on which to discuss the practical details of investing in property.
Sources of property data
A long useful post with lots of links to useful PMT data.
Investing in Overseas Property
See this board for discussion regarding overseas property investing.
Closure of the UK Discussion Boards
The UK Discussion Boards are now closed to new posts.