Interest rate cuts are much less effective when credit is tight

Friday 19 December 2008

Professor Gilbert, commented: “While simulation models cannot predict the future, they can show what could happen and can help us to be suitably cautious. The model informing this report is one of the new breed of ‘agent-based models’ that are being used to understand complex social and economic processes from the spread of epidemics to the growth of social networking sites.”

John Hawksworth, UK head of macroeconomics at PricewaterhouseCoopers LLP, commented: “The Bank of England has made large and very welcome cuts in interest rates recently, which will help to ease the debt burden on both individual and corporate borrowers. But our model suggests that cheaper mortgages will only have a limited effect in boosting house prices so long as the availability of credit remains a binding constraint on many potential homebuyers.”

The PwC agent-based model shows how realistic house price behaviour can be simulated where buyers and sellers are assumed to interact following relatively simple rules. It allows users to change key input assumptions and see in ‘real time’ how this affects the evolution of prices and other key variables such as transactions numbers. The model display shows how people move around a schematic ‘town grid’ in which some new houses are built and others demolished over time. Key model parameters have been calibrated to match actual UK data.

Notable results of model simulations include the following:

•Both house prices and the number of housing transactions tend to be volatile in the short term, helping to explain why short term housing market forecasts are often subject to significant margins of error;
•In the longer term, average house prices tend to show a stable relationship to average income levels for given levels of mortgage interest rates and loan-to-value ratios; if interest rates fall or maximum loan-to-value ratios rise, then the equilibrium house price to income ratio rises in response;
•A key insight from the model, however, is that the interaction between these two assumptions is critical: the interest rate effect on house prices is much weaker when the maximum loan-to-value ratio is lower. In particular, the model suggests that moving from a 100% to an 80% maximum loan-to-value ratio - which is similar to what has happened recently in the UK mortgage market due to the credit crunch - can have a major depressing impact on house prices through keeping first-time buyers with limited savings out of the housing market. This can outweigh any positive effects on house prices and transactions volumes from interest rate cuts;
•Increased inward migration puts significant upward pressure on house prices given that the supply of houses takes a long time to respond to such changes (and may be constrained by the availability of land with planning permission to build on). The number of people seeking homes but unable to find affordable properties also tends to increase significantly in such simulations; and
•If there is a significant ‘yuppie invasion’ of migrants to an area with significantly higher income levels than existing residents, the long term effect is to price these residents out of the market and push house prices up to a higher level but with the same ratio to incomes as before; the opposite tendency is evident with a large and persistent inflow of lower income inward migrants to an area;

In policy terms, the model results suggest that:

•building more new houses can reduce prices in the short to medium term, but as empty land is used up there is a strong tendency for house prices to return to previous levels in the longer term; and
•stamp duty changes do not seem to have a significant effect on house prices.

The model also shows how clusters of high house prices can arise over time due to a self-reinforcing tendency of a few high value houses in a particular area influencing valuations of other nearby houses. This occurs even in a model where house prices are initially randomly distributed across locations and there are no obvious locational advantages assumed, such as good schools or access to transport links.

John Hawksworth, UK head of macroeconomics at PricewaterhouseCoopers LLP, added: “Most economic models ignore one of the key aspects of real world housing markets: location. Our model brings this to life by showing how buyers and sellers move around a schematic town grid and how house prices emerge from their interactions. The model provides a ‘laboratory’ for the exploration of the effects on house prices and transactions volumes of changes in factors such interest rates, credit constraints, migration, housing supply and stamp duty.”

Editors' Notes

1. A report describing the results of the model in more detail ‘Agent-based modelling: a new approach to understanding the housing market’ is available on request from john.c.hawksworth@uk.pwc.com   
2. Nigel Gilbert is Professor of Sociology at the University of Surrey. He is Director of the Centre for Research in Social Simulation (CRESS) and the editor of the Journal of Artificial Societies and Social Simulation. He has written or edited several books on research methods, including Agent-Based Modelling (Sage Publications, 2008) and Simulation for the Social Scientist (with Klaus G. Troitzsch, Open University Press, 2nd edition, 2005). For more details of his work in this field see: http://cress.soc.surrey.ac.uk/

Media Enquiries

Peter La, Press Office at the University of Surrey, Tel: +44 (0)1483 689191, or Email mediarelations@surrey.ac.uk