RDP 9602: Consumption and Liquidity Constraints in Australia and East Asia: Does Financial Integration Matter? 4. Domestic and International Effects

The proxies for liquidity constraints in Table 2 are defined in terms of domestic rather than international financial variables for two reasons. The first is that a sufficient run of relevant time series, such as the covered interest differential, which is a measure of exchange controls, is not available. The second is more fundamental. It is difficult to sharply delineate between domestic and international effects, both in the mechanics and the economic and political dynamics of markets.

Consider some mechanics. A secular expansion in the money/GDP ratio is a proxy for financial development but this can reflect not just domestic but also a range of international influences, such as an expansion of domestic deposits due to the entry of competitive foreign banks, the monetisation of capital inflows under fixed exchange rates, or the balance sheet counterpart of the acquisition of foreign assets by domestic banks under an open capital account regime with floating exchange rates. The money/GDP ratio is a domestic variable and its expansion implies greater financial depth and scope for smoothing income, but the factors cited are international. Similarly, a domestic bank funds an expansion of its loan book by taking deposits or borrowing offshore, among other means, and so an expansion of domestic credit may have an international dimension. Domestic and international factors can both be relevant to the most simple of domestic financial transactions.

But more than mechanics, the economic and political dynamics of domestic and international markets are intertwined, as a few examples show. In Australia, there was a major consolidation and reform of banking services (for example, the introduction of ATMs and changes in hours) by domestic banks in the early 1980s, well ahead of the entry of foreign banks in 1985, but in anticipation of it some time in the future, and similar rationalisation has occurred in East Asian banking sectors. In Indonesia and Malaysia, easy access to offshore, mainly Singaporean, financial markets made control of deposit interest rates less effective and stimulated more rapid or effective liberalisation of deposit rates. Liberalisation of both domestic and international financial markets in Japan, Korea and Taiwan has been driven by, among other things, US pressure. There is no clear separation of domestic and foreign forces for reform.

The way that the domestic and international dichotomy has been drawn in this paper is to account for changes in the bind of liquidity constraints by the program of liberalisation and deregulation followed. For example, the capital account was liberalised in Singapore in 1978 and in Japan in 1980 and 1984, but their domestic financial systems were liberalised in 1975 and 1985–1994 respectively. The evidence of declining constraints is stronger for Singapore than for Japan, which suggests that it is the confluence of domestic liberalisation and international openness that is important in facilitating households in intertemporal transfers on income. Domestic markets and the capital account were liberalised in Australia in the early 1980s, and liquidity constraints on non-durables expenditure cannot be identified from at least that time.

Bayoumi and McDonald (1994) have presented a model to distinguish domestic from international effects. Their model is based on an earlier insight that the single-country permanent income conditions for smoothing consumption over time can be expressed in terms of insurance in the form of smoothing consumption between countries at a point in time (Cochrane 1991; Obstfeld 1993, 1994; and Canova and Ravn 1994). They define international financial integration as the equalisation of real interest rates and construct a model in which they interpret declining excess sensitivity of domestic consumption on domestic income as evidence of the development of local financial markets and a rising correlation with foreign consumption as evidence of the development of international financial markets. This model seems flawed on two accounts. Firstly, the derivation relies on real interest parity but this parity condition is stringent (Frankel 1993), such that the real interest differential only equals the expected depreciation of the real exchange rate when there are no barriers to trade in financial assets, financial assets are perfect substitutes, the Fisher effect holds at all points in time and expectations are formed rationally. Even if all these conditions hold, real interest rates are only identical, in general, in the steady state when the real exchange rate is constant. Hence, using real interest parity as the criterion for openness in empirical work is fraught with error. Secondly, consumption becomes less sensitive to domestic income as the set of income-smoothing instruments expands, but this set includes both domestic and international instruments. It is not necessarily correct to classify falling excess sensitivity of consumption to income as evidence of the development of domestic markets. If the cause is due to international factors, then falling excess sensitivity is in fact tantamount to greater consumption insurance.