Tuesday, 9 May 2017

Differing House Price Linkages Across UK Regions: A Multi-Dimensional Recursive Ripple Model

By Chris Hudson, John Hudson and Bruce Morley
Link to paper

The dynamics of the UK housing market have become increasingly topical following its role in the financial crisis of 2008. This study analyses the ripple effect across the UK’s regional housing market covering different housing types, whilst also incorporating any reverse effects back to the origin in London. This effect refers to the rippling out of changes in house prices in London to the surrounding regions and potentially also then reversing the ripple back to London. This ripple effect is important to the UK housing market as a whole, as movements in house prices usually originate in London, and recently London’s house prices have risen substantially after capital inflows from overseas were directed at the capital’s housing market. This has made the London market more volatile and thus more risky for investment purposes, but it is important to determine how this volatility in London could affect the rest of the UK housing market. 

The 2008 financial crisis initially started in the housing market when assets backed by the sub-prime housing sector collapsed in price, undermining the solvency of the international banking sector. As a result all central banks now monitor the relationship between financial stability and the housing market more closely, especially through macroprudential policies which are specifically targeted at housing, such as the loan to house value ratios. This also requires some idea about how house prices are likely to evolve over the near future, which is why the understanding of house price dynamics and the ripple effect have become so important.

House prices across types and regions

We can see in the Figure that the three types of house prices tend to move together, but there are differences. In particular in the period between 1990 and 2008, old and modern house prices caught up the new house prices. The Figure also shows the trend in old prices as we move north from London. Again there are similarities between the curves, but also substantial differences. For example, as we move northwards, and compared to London, the decline in prices in the early 1990s tended to be delayed and more muted, but the slowdown was then more prolonged. This gives plausibility to the hypothesis that shocks emanate in London and then spread out through the country. But our analysis suggests it is not that simple. Firstly, if there is some shock in another part of the country, we would anticipate it also rippling outwards including towards London and the South East. But secondly for the shock that does begin in London, as it ripples outwards so there will also be a recursive ripple back to London itself. The analogy can be made to a stone which is thrown into a pool of water. Waves will ripple outwards, but once they meet an obstacle they will begin to ripple back again.

This is complex enough, but our analysis, and this is one of its main contributions to the literature, suggests it is more complex still. It is as if there are three interconnected pools of water, with ripples from one impacting on the other two. A regional shock to new house prices will impact mostly on new prices in neighbouring regions and thence their neighbours. But there will also be some interaction between old, and in particular, modern house prices, both in the same region and neighbouring ones. A similar analysis applies to shocks to old and modern house prices.

Thus, a positive shock to new houses in London and Outer Metropolitan, will first impact on the South East and thence on the South West, the West and East Midlands and East Anglia. There will then be further knock on effects on e.g. the North West, all with differing speeds. Initially the largest impacts will be on other house types in London and new houses in adjacent regions. But the impact will eventually spread across all house types and regions. There will then be ripples, or echoes, back, from these regions to new, and other, house types in London. Hence the ripple acts recursively on at least two dimensions, firstly the spatial dimension and secondly the house vintage dimension. In effect there are multi-dimensional recursive ripples between different house types both within the same region and between regions.

There are lessons for policy makers in this. For example, if they introduce a policy which impacts on new houses, either across the UK or in a single region. Then there will be repercussions across all regions and all house types and it will take several years before the full effects have worked their way through the system. Policy makers should be aware of these repercussions and not just focus their attention on the new housing market or one particular region. 

No comments:

Post a Comment