The Power of Hedging Against Inflation with Real Estate The Hong Kong Experience

The Power of Hedging Against Inflation with Real Estate: The Hong Kong ExperienceSammi Ka Wai HO, and Rose Neng LAI * Sammi Ka Wai Ho is graduate student of Faculty of Business Administration, University of Macau, Macau, China.Rose Neng Lai (Contact Author) is Assistant Professor, Faculty of Business Administration, University of Macau, Taipa, Macao, China. e-mail: RoseLaiumac.mo. Tel. No.: (853)-397-4744, Fax. No.: (853)-838-320.Submitted to the 10th Asian Real Estate Society International Conference July18 - 21, 2005 at The Swiss Grand Resort & Spa - Bondi Beach, Sydney, AustraliaAbstract: Previous literature often concludes that real estate investment is an effective hedging instrument against inflation. Some more recent research finds that the balance of supply and demand of property in the market has a direct impact on real estate hedging ability; and property values are good protection against inflation only when the vacancy rate is low. This study expands existing research to study the ability of real estate in hedging actual, expected, and unexpected inflation in the short term and long term for the Hong Kong economy. We also consider property vacancies in explaining the hedging effectiveness. In particular, we focus on the hedging power of the four private property sectors for the period of 1983 to 2004. Our results provide up-to-date suggestions to which property types are better hedging tools in various inflationary environments. 1.IntroductionReal estate has historically been viewed as powerful hedging tool against inflation. This is because property values tend to increase most of the time, often at a rate that is higher than inflation. As such, the purchasing power can be preserve if real estate properties are included in the investment portfolio. Whether this perception is realistic has been widely studied over the past thirty years in many major markets. This paper serves to supplement the existing literature by studying the hedging ability of real estate against inflation in Hong Kong.There are numbers of reasons for choosing the Hong Kong market for our study. First, the very limited land supply has led Hong Kong to be one of cities in the world with high property prices. Second, property prices tend to be more volatile in Hong Kong than in other places. These two reasons may lead to the prediction that real estate returns have been historically high enough to preserve the purchasing power. However, Hong Kong has experienced a prolonged deflation since the Asian financial crisis in July 1997. Property prices have dropped down significantly. Hence, if the first prediction that real estate returns are good hedge against inflation is correct, what will be the hedging effectiveness after the Crisis? This is the third reason for our study.Fama and Schwert (1977) are among the first to touch the issue. They examine the relationship between inflation hedging ability of properties, equities, government bonds, and human capital in the US between 1953 and 1971. In particular, they separate inflation into two components: expected inflation and unexpected inflation. They find that only private residential properties can provide complete hedge against both expected and unexpected inflation. Among other studies after Fama and Schwert (1977), Rubens, Bond and Webb (1989) show that farmland and residential real estate provide complete positive hedge, while Wurtzebac, Mueller and Machi (1991) find that real estate returns are effective hedge against actual and unexpected inflation in high inflation period, and against actual and expected inflation in low inflation period. In other context, Bond and Seiler (1998) find that both expected and unexpected inflation are significant components of residential real estate returns.In terms of the UK markets, Barber, Robertson and Scott (1997) suggest that commercial property in the UK is a partial hedge, and that property can hedge unexpected inflation shocks better than anticipated inflation trends. Miles (1996) concentrates on commercial properties and concludes that the returns are too sensitive to economic conditions to be good hedge against inflation. Unlike all these previous studies, Stevenson (2000) examines regional housing markets and finds that the hedging power varies among regions, and that the relationships are neither consistent nor stable. He nevertheless suggests that housing and inflation are cointegrated, and therefore exhibit a long-term relationship. Stevenson and Murray (1999) also find that the Irish market is a long-term hedge against inflation, and property returns leads inflation.In other markets, Newell (1996) studies the hedging abilities of various property markets in Australia while taking vacancy into consideration. Vitor (2001) performs the study on the Canadian market. His findings reflect that real estate is strongly positively correlated with inflation in high inflation period but not in low inflation period.Other researches claim that real estate is an ineffective hedge against inflation. For instance, Hamelink and Hoesli (1996) who analyze the Swiss market claim that stocks and bonds are better hedging instruments than real estate; although Hoesli (1994) addresses that the Swiss real estate market seems to provide a better hedge against inflation than common stocks in the long term. nder (2000) also finds that real estate investment in turkey does not provide hedge under high inflationary environment, and that there is no reverse causality between real estate returns and changes in expected inflation. For Asian markets, Sim and Choe (2002) reveal that housing in Korea provides attractive hedge against both actual and expected inflation. Sing and Low (2000 and 2001) claim that the property market provides better hedge against inflation than other financial assets in Singapore. Chu (2003) is one of the first few to study the Chinese property market. Because the Chinese property market is still experiencing fast growth at its early stage, while inflation has been kept low, Chu (2003) concludes that none of the various types of properties is a good hedge against inflation, both in the long and short run. He also finds that the vacancy rates have risen dramatically, which may be a reason why the reduced returns due to vacancy cannot act as a hedge, regardless of the inflation level. In their study on the Hong Kong real estate market, Ganesan and Chiang (1998) conclude that real estate returns are not good hedge against inflation.From the various studies on different geographical regions, we can safely assume that property market returns may be a good hedge against inflation in some regions and time periods, but not others. Furthermore, as an indicator of market disequilibrium, vacancy can play a crucial role in the real estate return, thus affecting its effectiveness against inflation hedging. We therefore consider it necessary to reexamine the hedging ability of real estate returns against inflation in the Hong Kong market from three aspects. Firstly, we attempt to contribute to the literature by updating the study by Ganesan and Chiang (1998). In particular, we will repeat the three tests that they employ by using quarterly data for the period of 1985 2004: (1) the Fama and Schwert framework for testing short-term hedging, (2) test for co-integration, and (3) long-term relationship between real estate returns and inflation. Notice that our data series covers almost the whole study period of Ganesan and Chiang (1998) (except for 1984). Secondly, we enrich the literature for the Hong Kong market by including vacancy as an explanatory factor into the models for short-term hedging. Finally, since our period of study is made up of high and low inflation periods, and also incorporating the remarkable Asian Financial Crisis (in which we suspect an acute structural change in the economy), we further perform individual studies for each of the periods. This paper is organized as follows. We first briefly introduce the real estate market situation in Hong Kong over the period of study in the next section. We then discuss the methodology in Section 3, and the test results in Section 4. Finally, Section 5 concludes.2. Real Estate Market Performance in Hong KongThere are four major property sectors in Hong Kong: residential, retail, office and industrial. Residential properties are made up of public and private housing. Only private housing prices are relevant to inflation because public housing benefits from the governments supports. Housing is the most important type of savings for many households. In Hong Kong, households have always looked upon ownership not only as a social necessity, but also as the best hedge against inflation. The housing supply constraints also gave rise to high housing costs. On the other hand, retail property exhibits close relationship with consumption. When the economy is doing well, retail properties will also perform well. Therefore, it is less successful as a defensive asset. In this study, we will also include commercial properties and industrial properties. Because the data of other properties such as private specialized factories are limited, industrial properties are regarded as flatted factories, which refer to premises designed for general manufacturing processes and normally intended for sale or letting by developers. Figure 1 compares the levels of the price indices for the four types of private property sectors. The property prices are more volatile than elsewhere, with a number of large structural changes in the past two decades. Hong Kong is well-known as a city with serious scarcity of land and highly dense population, and thus relatively high property prices compared with other parts of the world. Furthermore, as He and Webb (2000) have shown, the commercial and residential real estate markets are both very sensitive to important economic and political news, as indicated by the rapid responses in their prices and rental rates. As a result, Hong Kong real estate prices and returns are volatile. The property prices decreased sharply with high interest rates caused by U.S. Fed policy to beat inflation and political uncertainty caused by Sino-British negotiations during 1982 to 1984. After that, the property market recovered and maintained an upward trend with peaking in 1997 after signing of the Sino-British Agreement to return Hong Kong to Chinese sovereignty in 1984. Hong Kongs property market has experienced significant downward fluctuation after 1997 Asian financial crisis and 2003 SARS outbreaks. The property values of retail, residential and office properties have drastic declined for six years because of persistent deflation.Figure 1Property Price Index of Hong Kong for 1980Q1 to 2004Q4Since the Asian financial crisis in 1997, Hong Kong has witnessed an unprecedented lengthy period of gloomy economy. This is further fueled by the government adopting a housing production target of 85,000 units per year since 1998 that have driven the private property prices to fall sharply (the government eliminated the policy after a couple years). The effect of the Severe Acute Respiratory Syndrome (SARS) epidemic in 2003 (we refer it as the SARS period in the later context) further depressed domestic demand within Hong Kong. Real estate returns from different sectors have remained below or close to zero after these events, as depicted in Figure 2. The inflation rate from Figure 3 also shows similar movement. Inflation rate has experienced a gradual decline for six consecutive years starting from the fourth quarter of the 1998. The cumulative effect from these factors was to technically push the economy into the third recession in six years. As a result, the oversupply of stock has created, and vacancy rate has stood at a high level. Figure 4 plots the vacancy rates of four property sectors over 1982 to 2004. Figure 2Property Nominal Returns of Hong Kong for 1982Q1 to 2004Q4Figure 3 Trend of Inflation Rate in Hong Kong for the Period of 1982 to 2004Unlike other real estate sectors, the price index for flatted factories reached a peak much earlier than the other price indices and started to fall from the early 1990s. It reflected a decline in demand for factories because many investors move their factories to Mainland China where operating costs are lower. There has also been high vacancy level in industrial market since 1994, especially in 1995. Since factories have limit potential for other applications, vacant units will usually be left idle for longer periods. Fortunately, after the SARS period in mid-2003, investors have gained confidence in the Hong Kong economy, and become more aggressive on real estate market. All property sectors are at the stage of recovery.Figure 4 Vacancy Rates of Property in Hong Kong for 1982 to 2004As mentioned earlier, with a glance of Figure 2 and 3, it seems that the inflation rate and the rate of return from real estate are closely related. To verify this, Table 1 shows the correlation between the property returns and the inflation rates, as well as vacancy for the period of study (1982 2004). Also mentioned before is that the residential properties should moves most closely with inflation among the four sector. However, the correlation figure shown in Table 1 implies that their relationship is more moderate than expected. This, however, does not mean that they are not related because lead-lag relationship may still exist. Furthermore, as high vacancy should imply lower return, there should be strong correlation, which is proven also from Table 1. It should also be noted that the nominal returns of four properties are negatively correlated with vacancy rates. According to Wurtzebach, Mueller & Machi (1991), vacancy rates indicate a strong negative correlation with inflation. The nominal return on office exhibits strongest negative correlation (r = 0.721) with vacancy, while the housing nominal return has lowest negative correlation against vacancy. This is probably because Hong Kong has experienced a shortage of housing supply and sizable population, which is why a small gap between supply and demand cannot exert much pressure on domestic property prices. Table 1 Correlation of Real Estate Returns with Inflation and Vacancy for the period of 1982 2004VariablesCorrelation withInflation RateVacancy RateResidential return0.435-0.489Retail return0.285-0.596Office return0.187-0.721Industrial return 0.280-0.494 3. The Methodology 3.1.Short-term HedgingMost studies on short-term inflation hedge follow the methodology of Fama and Schwert (1977). Fama and Schwert (1977) propose the regression tests based on the work of Fisher, in which the expected nominal rate of return has three segments: the expected real rate of return, the expected inflation, and the unexpected inflation, that is (1)where E() refers to the expected value, “” denotes that the variables are random, represents the nominal rate of return from property j at time t, is the information set available at time t 1, is the real rate of return from the property, and Dt is the inflation rate at time t. The last term in equation (1) represents the unexpected inflation rate, which is the difference between actual inflation rate and the expected inflation rate. This is the term that Fama and Schwert (1977) propose on top of the original model by Fisher (1930).The estimation of equation (1) can be done by running the following regression equations: (2).(3)where ejt and hjt are the stochastic disturbances. In the first regression equation, it is hypothesized that the return from property sector j is related to actual inflation in the same period; while equation (3) is further extended by breaking actual inflation into expected inflation (the first regressor) and unexpected inflation (the second regressor).Furthermore, because vacancy rates have noticeable impact on real estate returns, we follow Wurtzebach, Mueller and Machi (1991) by adding vacancy rates property sector j into the regression models to form(4) (5)where is the vacancy rate of property sector j at time t.Rubens, Bond and Webb (1989), and Wutzebach, Mueller and Machi (1991) evaluate the hedging performance against inflation according to the magnitude of the regression coefficients of the inflation rates in equations (4) and (5). In particular, a complete positive hedge against inflation is said to be achieved when a positively signed coefficient (bj or bj and gj) is not statistically different from positive one; whereas a partial hedge against inflation is obtained when a positively signed coefficient is statistically different from both positive one and zero; an indeterminant hedge is obtained when the coefficient is not statistically different from zero; and finally, an effective hedge occurs when a positively signed coefficient is statistically greater than positive one. We will follow such definitions in our tests.In addition to these commonly applied regression equations, we further extend the study to incorporate a dummy variable to account for the structural change as a result of the 1997 Asian Financial Crisis. Recall that the real estate prices in Hong Kong had suffered significant declines as the aftermath of the Crisis. We therefore define a dummy, , to equal one for the observations after the Crisis, and 0 otherwise. The coefficient of the dummy thus measures the effect of the Crisis on the property returns. The estimate of equation is as follows: (6)3.2.Long-term HedgingMany studies (see, for example, Ganesan and Chiang, 1998) do not support Fama and Schwert (1977) model because the assumption that real rates of return are constant is not applicable to illiquid assets such as real estate. It is therefore suspected that real estate markets respond to long term change of inflation, even if the relationship cannot be reflect in the short run. Recent researchers focus on cointegration test as a measure for assessing the long term hedging ability of properties. Following Engle and Granger (1987), the two-step procedure in the cointegration test include firstly checking the stationarity of the two time series (return rates of property sectors and the inflation rates) using the augmented Dickey Fuller (ADF) unit root tests, and secondly checking if the residual is stationary. If so, the two series are cointegrated which means that inflation and property return have long term relationship.Two variables are cointegrated if they are integrated of order one. In other words, each data series should be stationary in their first differences. Thus, the ADF regression test states (7)where y represent the series, t is the trend component, and e is the white noice. Then the cointegration test for series x and series y is given by . (8)The ADF test is then applied once again to the estimates of ut to test for stationarity using (9)3.3.Causal RelationshipThe long run causal relationship between inflation and nominal property returns can be tested by the Granger causality tests. Based on the test, if X causes Y, then changes in X should precede changes in Y significantly. In other words, the past value of X can help improve the prediction of current values of Y. The test involves estimating two regressions: (10) (11)where the third term in both equations is the error correction term which must be included if the series are cointegrated. The null hypothesis that x does not Granger cause y (that is, all bs are not significantly different from zero) is then t