In my first post asking whether Hong Kong property is a bubble (see Part 1 here), I focused on historical price trends and ratios. I concluded that residential property prices are currently around 50-60% above historic trend levels; and suggested that, for this not to be a bubble, there had to be something different this time. The aim of this post is to look in detail at some of the possible reasons for the recent run-up in prices to try to determine what if anything is different this time.
Is there a shortage of new housing?
Hong Kong’s high property prices are frequently blamed on a shortage of housing. The Hong Kong government has admitted that a lack of supply is the “core” problem; it is trying to increase the supply of new housing from both the private and public sectors in the next few years.
The limited amount of useable land in Hong Kong means that property will always be expensive. However, there is no suggestion that suitable sites are not available to meet demand from population growth for the foreseeable future. Demand in recent decades has been met by expansion in the New Territories. The Hong Kong government controls the release of new land for building and critics argue that it has artificially restricted supply to maintain high prices; pointing out that a substantial proportion of government revenue is from property-related sources. Concerns are also raised about the role of property developers; the market is dominated by a relatively small number of large companies, fuelling the popular belief that the market is run for the developers’ benefit rather than buyers.
Chart 1 shows the annual number of private housing completions for each class of flat and the vacancy rate since the 1980s, as well forecast completions for the next 2 years. (Flats are classed from A to E, where class A is the smallest at less than 40m2 and class E is the largest at 160m2 or larger.) Completions have declined sharply in recent years; averaging less than 10,000 units per year during the last five years, compared with nearly double that number in the prior five-year period. The figure is also a fraction of the average number of new properties completed annually in the 20 years up to 2005 of nearly 30,000. The annual supply of new housing is a relatively small percentage of the overall private housing stock of around 1.1 million properties; however, it is a much larger percentage of the annual number of housing transactions, typically around 100,000. This is an important distinction.
Overall, a shortage of accommodation as such has not been the main reason for the rapid increase in prices. There almost certainly is a shortage of accommodation in Hong Kong, but in the private housing sector at least, it has not worsened substantially in the last few years: the supply of new private housing has broadly kept pace with growth in the population and the number of households. The housing stock grew at an average rate of 0.8% per year between 2006 and 2011. During this period, the population grew by around 0.6% annually, a substantial slowdown in the rate of growth since the mid-1990s, whilst the number of households increased at a slightly faster rate of around 1.2% per year, probably due to demographic factors such as a higher divorce rate and an ageing population. This has led to a reduction in the vacancy rate in recent years, particularly for smaller flats. There are various reasons why a property might be classed as vacant, but the current level is not out of line with the longer-term average. It also appears that there was an oversupply of new properties coming on to the market during the last downturn in the early 2000s, which contributed to falling prices. The fact that prices have increased faster than rents also supports the view that a shortage of actual accommodation is not the main reason for the rapid price increases.
If property prices were solely determined by supply and demand for accommodation, and people were indifferent about whether they bought or rented, a shortfall in new supply should only affect pricing over the longer term. However, clearly other factors such as interest rates also play a major role in determining property prices. Whilst at certain times, people will prefer to buy rather than rent: renting is not a perfect substitute to buying. Therefore, the supply of properties available for sale and the number of potential buyers can affect prices, particularly in the short-term. In this context, the number of new properties coming on to the market can have a material impact on prices.
One point worth noting from Chart 1 is that the supply of smaller properties has declined even further than the overall number. Historically, classes A and B together accounted for the vast majority of new flats, around 75% on average. In the last 5 years, however, there have been far fewer class A and B flats coming on to the market, just over 5,000 per year, compared with around 15,000 per year in the prior 5-year period, whilst the share of these smaller flats has dropped to less than 60%. In particular, the supply of the smallest class A flats has almost dried up. At the same time, the number of larger flats, classes D and E, completed has increased. This no doubt reflected demand at the time these flats were planned, but the market appears to have changed in the intervening period, with increased demand for smaller flats. This has had important consequences for the market, which are discussed in more detail below.
What action has the government taken?
The Hong Kong government’s policy is to try to ensure a “healthy and stable development” of the property market, although the history of property prices over recent decades suggests it has had limited success in achieving this. The government is trying to achieve a fine balance: it has introduced a number of measures, both supply side and demand side, to try to rein in what it calls “exuberance” in the market, but it does not want to precipitate a market crash. It has been warning of the risks of a “bubble” for several years, although even now, perhaps not surprisingly, it is reluctant to actually admit that one exists. The task is made more difficult by not having control over interest rates, because of the US dollar peg. It has therefore been forced to rely on indirect methods to try to influence the market.
Earlier this year, the government sought to take back control over the supply of new land by abolishing the application list system, which relied on developers to initiate land sales by applying to buy land on a government sales list. In future, all land will be sold through a programme of regular tenders. The application list system was introduced after the last property crash, as a way to allow market forces to determine the supply of new housing, in an attempt to avoid problems of oversupply. However, it has been criticised for allowing the large developers to control the market to the detriment of buyers. The supply of new private flats is expected to increase over the next few years to over 15,000 in 2014, which would be the highest level since 2006. In total, an estimated 67,000 new flats are expected to reach the market over the next 3-4 years. The government has stated a target of an average of 20,000 new private sector flats per year.
The obvious danger, given the long time lag between land sales and new housing reaching the market, is that the increased supply reaches the market after it has entered a downturn, resulting in oversupply and further price falls, which is what happened during the last downturn. Nonetheless, increasing the long-term supply of new housing to make it more widely affordable seems a very sensible policy, particularly if there are suitable sites available. However, as an instrument to stabilise prices, it is almost certainly doomed to failure.
Blame the speculators
In addition to the supply side reforms, the government has sought to restrict demand from certain categories of buyers as early as 2010. In the absence of any control of interest rates, it has introduced a range of additional transaction charges targeting so-called speculators and non-local buyers from, with the primary intention of favouring local buyers.
In November 2010, the government introduced the Special Stamp Duty (or SSD); a charge of up to 15% on any property re-sold within 2 years of purchase. The maximum 15% charge was levied on re-sales within 6 months, whilst for re-sales between 1 year and 2 years the charge was 5%. The explicit aim was “to curb speculation”. As evidence of such behaviour, it pointed to the fact that re-sales within a year of purchase had more than doubled in the first nine months of 2010 compared to the same period in 2009. However, the idea that there was widespread speculation at this time seems tenuous. The pattern of sales during this period meant that re-sales within a year would have almost inevitably jumped significantly in 2010. The number of transactions in the midst of the financial crisis, in the second half of 2008 and early 2009, was much lower than average. Therefore, the number of properties available for re-sale within a year would have been low in 2009; and anyone who did sell would probably have done so at a loss. The number of properties available for re-sale within a year would have increased substantially in 2010, following the recovery in transaction volumes in 2009. It is also not surprising that some people, who bought near the bottom of the market in 2009 decided to take profits in 2010, given the rapid bounce back in prices and continuing uncertainty regarding the economy at that time. Furthermore, re-sales within a year only represented 12% of transactions in 2010 (compared with 7% in 2009), which is in-line with historical figures and well below the peak of nearly 25% reached in 1997. The introduction of the SSD virtually eliminated re-sales within a year and significantly reduced re-sales between 1 year and 2 years. Perhaps the best justification for introducing the SSD is that it helped prevent the emergence of widespread speculative activity, although it will potentially catch a substantial number of people who have legitimate reasons for selling a property only a short time after buying it (e.g. divorce, unemployment, change of job).
In October 2012, the minimum re-sale period to avoid the charge was extended from 2 to 3 years with re-sales between 1 and 3 years incurring a charge of 10%; previously the charge on re-sales between 1 year and 2 years had been 5%. For sales within 6 months, the SSD rate has been increased to 20%. One of the reasons given for the changes was a near threefold increase in the number of re-sales between 1 year and 2 years in September compared with March that year and a charge of 5% was not seen as sufficient discouragement. However, the evidence given of speculative behaviour seems weak. The 218 cases in September only accounted for 3% of transactions that month. Even if all of these could be regarded as “speculative”, it hardly seems have been a major problem.
An easy target: Mainland Chinese buyers
An influx of overseas buyers after the financial crisis, in particular from mainland China, is also widely thought to have been major contributor to the rapid rise in property prices.
At the same time as it extended the re-sale period for the SSD in 2012, the government also introduced the Buyer’s Stamp Duty (or BSD): an additional stamp duty charge of 15%, on top of the existing stamp duty charge, paid by all non-local buyers of residential property. Non-locals include anyone who is not a Hong Kong permanent resident. As with the SSD, the aim of the BSD is to exclude certain buyers from the market. In a similar vein, the government has restricted the purchase of flats in certain new developments, including subsequent re-sales, to just permanent residents under the Hong Kong Property for Hong Kong People scheme.
This crackdown on purchases by non-locals is aimed primarily at mainland Chinese buyers. There have been plenty of reports of cash buyers from mainland China buying new flats, pricing out locals. It seems to play to Hongkongers more general fears of being overwhelmed by mainland Chinese. The number of non-local buyers has increased in recent years. The government estimates that, overall, they made up around 6.5% of buyers in 2011, compared with 3.1% in 2007; and a higher proportion of sales of new property, around 20% in 2011. However, even before the new measures were introduced, the proportion of non-local buyers appears to have started to fall; according to the government’s estimates, in the first nine months of 2012, they accounted for just 4.5% of total sales. It is difficult to assess the precise impact of non-local buyers on prices. Whilst it is unlikely that they are the main cause of the rapid increase, they have almost certainly contributed to it.
In February this year, the government doubled basic stamp duty rates for any buyer who is not a Hong Kong permanent resident or who already owns a property. On the new scale, the lowest rate is now 1.5% for properties up to HK$2m, rising to a maximum of 8.5% for properties over HK$20m. The previous, lower rates still apply to Hong Kong permanent residents, who don’t already own a property. It is not clear how many people will actually pay the higher rates. Non-locals have effectively been shut out of the market already by the BSD, whilst it would seem relatively easy for most permanent residents to avoid paying the higher rates. Note that the old rates don’t just apply to first-time buyers as was widely reported in the media at the time of the announcement. If a Hong Kong permanent resident, who already owns a property, buys another property, but sells the first property within 6 months, they can claim a refund of the extra stamp duty paid. In practice, it is intended to discourage permanent residents from buying more than one property, although it may lead them to buy cheaper flats, which incur a lower stamp duty rate.
Lending tightened for some borrowers
In addition to the measures introduced by the government, the Hong Kong Monetary Authority (HKMA) as the banking regulator has tightened lending criteria on several occasions in the last few years. As early as October 2009, it reduced the maximum loan-to-value (LTV) ratio for the most expensive properties from 70% to 60%, citing rapid price increases and the fact that luxury flat prices were above their 1997 peak levels. The maximum ratios have been reduced further since then, so that for properties over HK$10m it is now just 50%. Only on properties costing HK$6m or less is the maximum LTV still 70% (without mortgage insurance). However, the vast majority, around 80%, of sales are for less than HK$6m and are unaffected.
The HKMA has also limited the amount that can be borrowed by certain other buyers, for example, those who already have mortgaged properties. The amount that can be borrowed is also limited by the maximum debt servicing ratio (total mortgage costs as a percentage of income) of 50% in most cases; previously the maximum was 60% for higher earners. Nonetheless, with interest rates so low, it is still possible for buyers to borrow around 9x income. In February this year, the HKMA also increased the risk-weighted capital that banks are required to hold against residential mortgages from 10% to 15%, leading the main lenders increased mortgage rates by 25bp.
The government-owned Hong Kong Mortgage Corporation, which provides insurance for certain mortgages with a LTV ratio of more than 70%, reduced the maximum loan size it would insure up to 90% from HK$6m to HK$4m. However, around half of all property transactions are still potentially eligible for insurance up to this threshold.
Whilst there has been a considerable amount of tinkering with the rules, none of the changes are aimed at the lower end of the market: the majority of buyers will be largely, or completely, unaffected by them.
It is important to recognise that the Hong Kong property market is not a single market: as Chart 2 shows, in recent years, prices for the different classes of flat have increased at substantially different rates. Price rises for different locations have also varied considerably. From the end of 2008 up to the middle of 2010, all classes increased at roughly the same rate. However, from the middle of 2010, there has been a clear divergence between the classes in order of size, particularly in 2012. Since the end of 2008, the smallest, class A flats have increased in price by 145% compared with 78% for the largest, class E flats, and an increase of 127% for the market as a whole, which is heavily weighted towards smaller flats. Since the middle of 2011, the largest flats have increased by only 5% compared with 32% for the smallest flats. What is clear is that price increases for the mass-market, class A and B flats have largely driven the market over the last couple of years. A similar pattern can be seen in the rental indices since the middle of 2011: rents have been flat for the classes D and E, whilst those for class A flats have increased by 16%.
This is not entirely a question of selecting a particular starting date. Over the very long-term, since the early 1990s, the prices of larger flats have increased more than those of smaller ones. But this is entirely due to faster growth for large flats in the period up to the 1997 peak, when the small flats lagged. On a peak-to-peak base, from 1997 to now, there is not a big difference between the categories.
What has caused this divergence? One possibility is the measures introduced by the government and HKMA. As discussed earlier, these have primarily affected the upper end of the market, although how many buyers in this market are reliant on large mortgages is unclear. Despite the fact that class A flats are now 50% above their 1997 peak level, the HKMA has not reduced the maximum LTV for less expensive flats. Together, small and medium-sized flats (classes A-C) make up the majority of the market; three-quarters of all sales are of properties for less than HK$5m. If the main rationale for the HKMA’s lending restrictions is banking stability, then logically they should also be applied to these flats as well. The measures introduced by the government have probably also had a greater impact on sales of larger properties. Anecdotal evidence suggests non-local buyers from mainland China (and no doubt most expatriates) affected by the BSD are predominately interested in the luxury end of the market. The recent doubling of stamp duty may not affect most local owner-occupiers, but investors (or speculators) may decide to buy cheaper flats, given the lower stamp duty charges.
As discussed earlier, the supply of new smaller flats has declined substantially in recent years. Whilst it is quite difficult to reach any firm conclusions, I suspect that as prices have increased, increasingly buyers will have been “squeezed” downmarket: incomes have risen but not as fast as prices. Therefore, since mortgage rates have not changed significantly since the middle of 2009, the maximum amount, which can be borrowed, will only have increased in line incomes. This is ignores the problem of trying to save for an increasingly large deposit in a rapidly rising market. Therefore, increasing numbers of people are likely to be chasing a fixed number of smaller, cheaper flats, at a time when the new supply is very limited and government actions are potentially pushing buyers (unintentionally) towards these flats. It is probably not surprising therefore, that prices of these flats have increased at a faster rate.
Limited speculative activity
One of the common features of an asset bubble is the existence of widespread speculative behaviour (some would argue it is a prerequisite). Despite the concerns of the government, there does not appear to be any evidence of widespread speculative behaviour in the current market, although almost inevitably some issues will come to light, if (when?) the market undergoes a sharp correction.
In one sense, it makes no sense to talk about the “market view”, as the market clearly doesn’t have a view. Nonetheless, the majority of commentary on the Hong Kong property market has for some time focused on the increasing lack of affordability and the risk of prices falling. It is hard to discern any sense of market euphoria in the prices rises, although no doubt people who bought a few years ago are feeling considerably wealthier. Whilst this is purely conjecture, if anything the market appears to be driven as much by the fear of being priced out of a rising market. The authorities would no doubt attribute the lack of speculative activity to their early pre-emptive action. But, it might also be due to caution on the part of buyers and lenders, given the recent history of the financial crisis and the collapse in the US housing market; it would also be surprising if the last Hong Kong property crash has been entirely forgotten even if memories do sometimes fade quickly.
Mortgage lending has increased substantially since the financial crisis (see Chart 3), increasing by around 50% since the end of 2008. As a proportion of GDP, it is now back to prior peak levels. There are some concerns about the build-up of debt, notably the possibility that large amounts of non-property lending has found its way into the property market. However, growth in mortgage debt at around 8% per year has not been very high. LTV ratios have remained relatively stable. After rising to around 66% in the middle of 2009, the average LTV ratio for new loans has since fallen to around 54%. This is probably due to the HKMA’s tightening of LTVs, but there is a danger that the average figures masks substantial underlying variations: some purchasers of lower valued properties may still be highly leveraged. Nonetheless, from a systemic perspective, the measures adopted, as well as the relatively low number of transactions, should make the fallout from a major market downturn less of a problem.
Declining market activity
The lack of speculative behaviour is also reflected in transaction volumes. In this regard, the current market is very different from that in 1997. Historically price changes and sales volumes have tracked each other closely: rising prices generally help to lubricate the market. However, as Chart 4 shows, the two have decoupled in the last couple of years. The number of transactions has declined markedly since 2010 and is now well below average.
The official measures introduced might have successfully reduced speculation and demand from non-locals. The danger is, however, that they have had other, unintended consequences by reducing transaction volumes. It is difficult to say how much of the decline in sales is due to the government’s actions. Nonetheless, higher transaction costs will almost certainly have reduced the number of properties for sale and led to fewer transactions. Sales volumes declined noticeably after the introduction of the SSD in 2010, although this may just be coincidence. The new charge largely eliminated re-sales within 12 months, which had accounted for up to 15% of sales prior to its introduction, so some reduction would be expected. However, as the charge only applied to purchases after the measure was introduced, this shouldn’t result in an immediate fall in the number of properties available for sale, although (as intended) it may put off some potential buyers. The resurgence in volumes in 2012 also came to an end after the measures introduced in November of that year, whilst the tightening of lending criteria by the HKMA will also no doubt have had some impact. The biggest short-term impact may be indirect, by affecting market confidence: if people believe that prices are likely to be affected then this may be self-fulfilling.
Reduced market liquidity is not generally a good thing, particularly when the number of transaction falls below normal levels. Looking at Chart 4, it appears that the number of transactions is several thousand per month below the level, which might be expected. It is not entirely clear to me that measures such as the SSD are a good idea; removing a substantial number of potential buyers and sellers from the market is unlikely to improve market efficiency. I am not convinced it would hold back price rises: a shortage of properties for sale, for whatever reason, may have pushed prices up further. The decline in affordability and “squeeze” on buyers downmarket may also have contributed to a decline in volumes in certain market segments.
Low interest rates: fuel to the fire
The various policy actions worldwide following the financial crisis are widely regarded as having contributed to the rapid rise in Hong Kong property prices. The US Fed’s on-going programme of quantitative easing (QE) and the Chinese government’s post-crisis stimulus package are seen as the main culprits. In 2009, Hong Kong had particularly large inflows of foreign currency, some of which no doubt found its way into the property market. Nonetheless, whilst M3 money has experienced strong growth since the financial crisis, as Chart 5 shows, growth has continued at the pre-crisis trend rate. I am not an economist, but it does not seem to me that the growth in money supply has been a direct cause of the rapid increase in property prices, although it has clearly enabled prices to rise. In such circumstances, a rapid increase in asset prices is not inevitable, as the subdued performance of the stock market in the last few years demonstrates (see Chart 6).
Interest rates fell globally during the financial crisis and have remained low since. Short-term rates in Hong Kong have been close to zero since 2009. The Hong Kong dollar has been pegged to the US dollar at the current rate since 1983. As a result, Hong Kong’s nominal interest rates generally track those in the US reasonably closely, although divergences do occur. Overall, for an economy like Hong Kong’s, the benefits of the currency peg may outweigh the disadvantages, but it is inevitable that interest rates will not always be appropriate for conditions there. This was the case during the period of protracted deflation after the 1997 Asian financial crisis, when real interest rates, including mortgage rates, were extremely high (see Chart 7), peaking at around 12%. Real interest rates have been on a declining trend for more than a decade, as Hong Kong has returned to a more normal level of inflation. Interest rates in the last few years have clearly been too low for the economic conditions in Hong Kong. Real rates have been consistently negative since 2009, which amongst other things has helped to fuel property prices.
Mortgages in Hong Kong are almost all variable-rate, with the majority priced by reference to the individual bank’s best lending rate (BLR). Chart 8 shows the BLR (for HSBC) together with an estimated “typical” actual mortgage rate, based on the figures for new mortgages from the HKMA’s monthly Residential Mortgage Survey. Before 2000, new mortgages were typically priced at BLR, or BLR plus a premium. However, since 2000, lower funding costs and increased competition has meant that most new mortgages have tended to be priced at a substantial discount, typically of around 2.5-3%. Hong Kong lenders are primarily financed through deposits and, to a lesser degree, the interbank market. Since late 2001, funding costs from both sources have for the most part been very low. It is worth noting that, for a 2-year period from the middle of 2009 to the middle of 2011, the effective average cost of new mortgages was well below the level shown in Chart 8. During this period, most new borrowers chose mortgages priced by reference to HIBOR, typically at a rate of only around 1%, well below the cost of BLR-linked mortgages; previously such mortgages had not been widely available. In 2011, the premium charged on HIBOR-linked mortgages increased, bringing the cost into line with BLR-linked mortgages.
As can be seen from Chart 9, mortgage rates are also currently very low in real terms. And have been for some time: averaging -1.0% since the start of 2009, and just 0.4% since the beginning of 2004. In such an environment, it is almost inevitable that people will look to borrow as much as they can: provided property prices keep pace with inflation, the mortgage, in real terms at least, is cost-free. Note that for the period prior to 1997 I do not have any data on what the typical mortgage rate was; I have therefore assumed that it is the BLR. What Chart 9 also shows is that in the early to mid-1990s real mortgage rates were also low.
Chart 10 shows a clear inverse relationship between changes in property prices and mortgage rates, with the latter inverted. In my view, interest rates are one of the most important drivers of property prices. However, the relationship with contemporaneous nominal rates isn’t perfect: it doesn’t take into account a number of other factors, such as the level of real rates or changes in expectations for future rates.
A number of additional observations are worth making on Chart 10. First, whilst mortgage rates in nominal terms fell rapidly during 2001 to below 3%, the recovery in the property market didn’t start until mid-2003; continuing deflation meant that real mortgage rates did not fall decisively until 2004. Second, the price fall in 2011 can at least partly be attributed to the increase in rates for the then popular HIBOR-linked mortgages, discussed above, although in Chart 10 this only shows as a smaller blip up in BLR-linked mortgage rates. Finally, does the recent increase in mortgage rates and rollover in property prices a pause or the start of a more substantial correction in prices?
Chart 11 is the same as Chart 10 except that it shows real rather than nominal mortgage rates. Deducting inflation from the mortgage rate creates some additional “noise”, particularly as inflation is backward looking, but I also think it provides some useful additional information in understanding the relationship between price changes and interest rates.
The majority of buyers still take out a mortgage to finance the purchase. Cash buyers face the same dilemma as investors worldwide in a period of prolonged negative real interest rates: invest in cash or short-term bonds, and accept the inevitable erosion from inflation, or invest in riskier assets if they want to maintain the real value of their savings. Since the end of 2008, inflation has eroded the value of the Hong Kong dollar by nearly 15%. In such circumstances, it is logical that investors may choose to invest in property, even if they think it is overvalued, rather than accept a guaranteed loss in real terms by holding cash.
Interest rates have been so low for such a long time that there is a real danger that borrowers assume that mortgage rates at the current level are normal. Despite the recent increase in longer dated US Treasuries, short-term US interest rates are virtually unchanged and are likely to remain low for some time yet, probably until at least the end of 2014. Nonetheless, the recent increases in long-term interest rates may focus attention onto the fact that mortgage rates may increase in the not to distant future. When mortgage rates do eventually rise, the expectation is that they will only do so by a few hundred basis points at most. This seems to be the view of the HKMA, whose mortgage stress test, only assumes an increase of 300bp. This may be a reasonable estimate of future mortgage rates; mortgage rates are certainly unlikely to go higher in the next few years. But is it a worst-case outcome? This is particularly important for Hong Kong buyers, given that their mortgages are effectively tied to interest rates in the US and the problems this has caused in the past.
Affordability: low cash cost
In Part 1, I looked at affordability in terms of the price-to-income ratio. This is currently close to peak levels. However, low nominal interest rates mean that in cashflow terms at least, property is still reasonably affordable. With typical mortgage rates remaining low, despite a recent tick-up, and interest on savings close to zero, the actual funding cost, taking into account interest paid on the mortgage and interest foregone on the deposit, is still only around 2%, assuming a buyer puts down a 30% deposit. Chart 12 shows the interest funding cost compared to the rental cost. On this basis, the cost of buying is also well below the cost of renting, and has been for most of the period since 2001. But a buyer does need to be able to down a deposit and the comparison does not take into account the other costs of home ownership.
Chart 13 shows the total buying cost, which includes capital repayments on the mortgage in addition to the interest funding cost, compared to the cost of renting. Even adding in mortgage repayments, the on-going cashflow “burden” of the buying is still not much more than the cost of renting (see Chart 13).
Whilst Charts 12 and 13 show the relative affordability of buying versus renting, Chart 14 shows affordability in terms of the ratio of mortgage costs, including both interest and capital repayments, to household income for a notional typical purchaser of a typical flat, as with the price-to-income ratio in Part 1. The ratio is calculated by dividing the estimated mortgage costs for a 50m2 in the New Territories by the 75% percentile household income. Mortgage costs are for a typical new mortgage, assuming a 25-year repayment term and a 30% deposit.
By this measure, affordability has steadily decreased in the last few years, as prices have risen faster than incomes, but it is not particularly stretched yet. Even an increase in mortgage rates of 300bp would only lift the ratio to 41%; at this level, affordability would be tight, but the ratio would still be well below previous peak levels. Of course, mortgage rates might increase by more than 300bp. This seems unlikely in the next few years at least, but is possible in the medium-term. Although if incomes continue to rise, this would obviously offset any higher mortgage costs in calculating the ratio. Even if affordability remains reasonable for the typical buyer, buyers will still tend to be “squeezed” downmarket, whilst some will be priced out of the market entirely.
The importance of confidence
Other economic factors, such as unemployment, household income and wealth, not just interest rates, will clearly play an important role in people’s confidence and willingness to buy property. Prices are unlikely to go up, if unemployment is rising and incomes are falling. Confidence would have been particularly low in the period in early 2000s. The unemployment rate rose to an unprecedented 8.5% in the middle of 2003 (see Chart 15). Household incomes fell in 2002 and 2003; only in 2004 did they start to rise again. Price falls can further reduce confidence, as equity in property is eroded, resulting in further price declines. At the low point in the market in 2003, 22% of borrowers were in negative equity (31% of the value of mortgages outstanding). It is therefore not surprising that property prices initially remained depressed even after interest rates started to fall.
These economic factors are currently supportive of high prices: unemployment is low; incomes are rising; and past price rises create a feeling of wealth, so that existing owners are likely to feel comfortable upgrading.
In Part 1, I concluded that Hong Kong property prices are currently around 50-60% above prior previous trend levels. But the question remained: what if anything was different this time? A lot has been made of the shortage of supply of new properties coming on to the market and the impact of overseas, particularly mainland Chinese, buyers. Whilst these have almost certainly contributed to the rise in prices in recent years, particularly the shortage of smaller flats, I do not think that they have been the main driving force behind the increases.
In my view, the main reason for the rapid price growth has been the low interest rate environment since the financial crisis. Low interest rates are also the main reason why this cycle is different from the last one: at the peak of the market in 1997, the BLR was nearly 9%. Real rates have been negative and have been too low given the generally supportive economic backdrop. Even at current prices, property remains relatively affordable on a cashflow basis. Some buyers will have been priced out of the market, but the majority of potential purchasers can still afford to buy in the current market. Furthermore, relative to the cost of renting, buying still looks attractive, for the moment at least. In many ways, buying is a logical decision, even in the current market, given the inherent uncertainty on future prices and interest rates. After all, even after detailed analysis it is still difficult to come to any firm conclusions on where the market is heading. However, a painful readjustment may well occur when interest rates do properly start to rise, if not before. Another important difference compared to the last cycle is the apparent lack of speculation this time.
When setting a budget, buyers tend to base it on how much they are allowed borrow. The HKMA has tightened lending criteria, but generally, these have not affected the average buyer. The banks of course may have tightened lending criteria to such borrowers, but there is no evidence that they have done so to any significant extent. The HKMA has acted in its role as the banking regulator to maintain banking stability. However, the suspicion is that for political reasons it has avoided targeting this part of the market. The government’s measures, in the absence of any control over interest rates, have in the short-term sought to ration properties for those considered the most deserving, have also not affected the average buyer. I am not convinced by how effective these measures have been in cooling the markets, beyond perhaps some short-term psychological impact. To a degree they seem have been populist initiatives, aimed at easy targets. They may also have contributed to higher prices by reducing market liquidity.
One of the biggest concerns for me is the divergence in prices in the last couple of years between smaller, mass-market flats and larger, high-end properties. In my view, an increasing number of buyers have been “squeezed” downmarket, chasing a fixed number of available properties (at least in the short-term), leading to rapid price growth. The danger is that a bubble exists in the mass market, even in the absence of widespread speculation, exacerbated by a shortage of supply. What could trigger a fall in prices? Certainly, a significant rise in interest rates would be likely to, but a fall could occur without that. Prices appear to have fallen somewhat recently; whether this is a pause, a minor correction or the start a larger fall is unclear, but the feeling is that market is performing a high-wire act.
In Part 3, I intend to look in more detail at current valuations and actually try to quantify if and to what extent prices are overvalued.