A study on the impact of changes in consumer sentiment on the housing market in Korea

ABSTRACT

This study examined the effect of consumer sentiment on housing market upon the trading volume and price in the housing market through the panel Vector Auto-Regression analysis employed on monthly data collected from 15 metropolitan cities and local provinces in Korea. Results of the empirical analysis revealed the positive (+) influence of consumer sentiment upon trading volume and price in the housing market together with the negative (−) influence of interest rate of CDs and actual construction of residential buildings and the positive (+) one of the index of industrial production upon trading price in the housing market. The results imply that the moves of trading volume and price in the housing market would be dependent on the speculation on trading price of houses in the housing market. It also suggests that the consumer sentiment can be employed as a variable to predict moves in the housing market.

KEYWORDS:

• The index of consumer sentiment
• trading volume in housing market
• housing price
• impulse–response analysis
• panel vector auto-regression model

JEL Classification:

• D1
• R1

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 Korean government fixed the level of the LTV and DTI by 70% and 60%, respectively, regardless of the regions and financial institutions on 1 August 2014. In particular, the apartments placed in the capital region priced over 600 million KRW experienced a 20% increase in the LTV from 50% to 70%. The government intended to provide the positive psychological prospect to the housing market by easing the regulations upon the limit of loanable amounts to encourage the real estate market, which had been stagnated since the global financial crises triggered in 2008.

2 Debt-to-income ratio is the rate of principle and interest redemption amount of the creditor in income, and is one of the standards for handling secured loan.

3 Lastrapes (2002), McCarthy and Peach (2002), DiPasquale and Wheaton (1992), Poterba (1984), Jung (2006), Lee (2008)

4 The GDP or GDRP are normally employed as a proxy variable representing the economic growth. However, those values are currently collected quarterly in Korea. Thus, the monthly data of the index of industrial production were used in this study as that of the proxy variable representing the domestic economic growth with references made to prior studies conducted by Kim and Kim (2009) and Chun and Park (2012).

5 The Korea Research Institute for Human Settlements has been announcing the index of consumer sentiment on the housing market since July 2011.

6 According to the 2010 housing census of ‘National Statistical Office’, the portion of apartment houses occupied 58% of the total housing in Korea, which is now recognized as the representative housing style. Apartment houses in Korea are similar to condominiums in US.

7 In Korea, the data associated with the Index of consumer sentiment on the housing market are announced by the ‘Korea Research Institute for Human Settlements', ‘The Bank of Korea’, ‘Korea Housing Institute’, and other real estate trading information websites, such as the ‘Doctor Apartment’ etc. ‘The Bank of Korea’ announced the survey on the outlook of asset value since August 2008 till December 2012, whereas the ‘Korea Housing Institute’ has been releasing the quarterly Housing Business Survey Index for brokers and construction companies, and real estate information websites generally collect the data from the real estate brokers. The indices of consumer sentiment on the housing market have been released by the Korea Research Institute for Human Settlements recently since July 2011, but the data prepared for the indices were collected from 2240 brokers placed in 150 local unit regions and 6400 ordinary households in Korea. The value of the index over 100 indicated the number of respondents with positive prospects on the real estate market and exceeded the number of respondents with negative prospects compared to the previous month. Trends of selling or purchasing land and houses, including rentals, trends of trading prices of house and land, and trends of transactions of house and land are surveyed through 29 questionnaires every month from brokers and by every quarter from ordinary households. The index of consumer sentiment on the housing market is represented by the numbers between ∼0 and 200. The brokers and general households placed in the enumeration districts of cities or counties would have equally assigned weights, and the final weights to be assigned for the prediction of consumer sentiment will be corrected by a unit of analysis or post information. In the case of an analysis by a unit of each district of individual cities or counties, the general households are distinguished into the types of ownership (one's own or on lease) and of houses (apartment or other types of residence) to assign the post-corrected weights because the samples were equally assigned to the 150 domestic districts. Otherwise, the finalized weights calculated from a subject population of entire households included in respective analytic units of combined districts of more than two cities or counties are assigned thereto. The index of consumer sentiment on the housing market is classified into three categories of stationary, descending, and ascending. The values of the index placed in the interval of ∼95–115 indicate that the real estate market is in a stationary phase; the values <95 indicate the decreasing phase of the real estate market, and the values >115 imply the increasing phase of the real estate market. The data collected from each broker in local counties and cities and from each household have the identical weight in the prediction of the index of consumer sentiment. The weight for the final application for the prediction of the index is corrected according to the unit of analysis or post information.