Wealth, volume and stock market volatility: case of Hong Kong (1993–2001)

Abstract

This article attempts to answer the question of whether the gain and loss in property market speculations and rate of information flow play a significant role in stock market volatility in Hong Kong. To test for our wealth–volume–volatility hypothesis, two different measures of volatility: absolute (absolute value of SD from mean with monthly dimension) and conditional (EGARCH) are used and results are compared. In both measures, we find evidence of a statistical presence of a wealth effect on stock market volatility, particularly in the investment of luxury class of property in Hong Kong. To account for this result, we apply the prospect theory, house money effect and the newly developed conditional confidence theory. Although we fail to establish a volume–volatility relationship in our estimation, we offer additional dimensions to the explanation of our observation.

Acknowledgements

I would like to thank Andrew Somerville for very valuable comments. I am in debt to Mike Harrison for technical advice and suggestions. I also gratefully acknowledge the support of a Trinity College Dublin Ussher Fellowship Award while writing this article. All errors and omissions remain the responsibility of the author. Any opinions expressed are those of the author.

Notes

¹ Unfortunately for some they also become heavily in financial debts later on following the onset of the Asian Financial Crisis.

² Finance Bureau, Government Secretariat, Hong Kong Government.

³ Hong Kong Stock Exchange Fact Book 1999.

⁴ Could be applicable to property owners all over the world.

⁵ Risk aversion for gains and risk seeking for losses at high probability, and risk seeking for gains and risk aversion for losses at low probability.

⁶ Noise traders are defined as irrational and uninformed traders. Experiments on mass psychology have shown that noise traders have a tendency to overreact to news (Tversky and Kehneman, 1982) and become overconfident and take on excessive risk (Shleifer and Summers, 1990).

⁷ Our choice of a 3-period setting is a subjective one that is based on the fact that we are using monthly financial data. It certainly can be extended according to the model and the nature of data that are being studied.

⁸ This article focuses on properties that are solely for residential purposes.

⁹ The general formulation of volatility measure is developed in Schwert (1990b), we take the absolute value of this measure as a modification.

¹⁰Frequently, this shock is also referred as ‘news’, or ‘error’, or ‘unpredictable component’ of a time series.

¹¹ See table of property class categorization in Section IV.

¹² For a discussion of these potential factors see Schwert (1989).

¹³ See Karpoff (1987) for a list of previous studies that fail to find a trading volume and stock price relationship.

¹⁴ Since the amount of original investment in luxury property could easily be many times over that of residential class, therefore, a 2% return for luxury property could give a substantial gain in money terms over other classes of property.

¹⁵ We assume the event of paying up the full amount at one time being an unlikely and unreasonable event, as this will only substantially reduce the amount of capital available for additional property purchases.

¹⁶ Unfortunate in terms of the huge amount of losses sustained from both property and stock markets.

¹⁷ Our suspicion is partly based on empirical results of this article which finds a wealth effect in the luxury class property investors and also the news coverage at that time regarding the intensity of people's preoccupation with the markets.

¹⁸ Either for reasons of panic or liquidity.

¹⁹ This could be particularly applicable for monthly data such as those used in this article.

²⁰ A new class of people emerged as a result of the property market collapse and in Hong Kong they are now called the ‘negative asset class’.

²¹ Since 1998 the Hong Kong economy has suffered a recession and many blame the collapse of the property market as the main cause of the event.