4.2 Stock measures of saving (continued)

To test whether or not the stock measure of saving shifted from a high saving state to a low saving state, we fitted the following Hamilton (1989) Markov switching model by maximum likelihood

(4.4)    

where denotes the saving rate, is a coefficient, is the mean of the high and low saving states and .[16] The Hamilton model estimates the probability of a regime shift from a high mean saving rate to a low mean saving rate or vice versa.[17]

The results show two regime shifts. Household saving switched from a low saving state to a high saving state in 1981 and back to a low saving state in 1988. Over the high saving period 1982-87, households, on average, saved almost four times more than in the low saving state, as measured by changes in households’ net wealth as a percent of disposable income. However, despite the fall in household saving in 1988, the average saving rate from household balance sheets, at about 16.5 percent, remains significantly larger than the average saving rate from the national accounts over the period 1987-2001 of -1 percent.

Figure 4.7 - Stock and flow measures of business saving (as a percent of GDP)

Source: Statistics New Zealand and authors’ estimates.

A stock measure of business saving can be derived from the Annual Enterprise Survey as the change in firms’ net wealth (assets minus liabilities). Business saving in terms of stocks is plotted in Figure 4.7 together with the sectoral and NZISA flow measures, all as a percent of GDP. The stock and flow measures of business saving show even larger level differences than household saving (note that they are plotted on a scale of 20:1), with the discrepancy mainly arising because of asset price movements. Figure 4.7 also shows that the stock and NZISA flow measures of business saving move in the same direction for the three years that data are available for both series. This is probably because the Annual Enterprise Survey provides the main source of information for the corporate sector in the Institutional Sector Accounts.

Figure 4.8 plots the stock and flow measures of government saving as a percent of GDP. Unlike for the household and business sectors, the stock and flow measures of government saving are very similar. The main reason for this is because the crown’s balance sheet is less subject to asset price movements than firms’ and households’ balance sheets.

Movements in the market value of households’ and firms’ assets and liabilities can be caused by a whole range of factors. These include changes in economic conditions or monetary policy, investor sentiment, exogenous shocks affecting specific sectors. As a result, it is often difficult to identify what are the main factors leading to the swings and hence changes in private net wealth.

Figure 4.8 - Stock and flow measures of government saving (as a percent of GDP)

Source: The Treasury and Statistics New Zealand

Many of the crown’s assets and liabilities (e.g. schools and state owned enterprises) are not priced in an actively trading market and hence are less affected by factors driving market valuation. Moreover, the value of some of the crown’s assets and liabilities has remained fairly stable in recent years, like notes and coins issued by the Reserve Bank of New Zealand. Consequently, movements in the crown’s balance sheet can often be related to the specific factors driving them, re-evaluation and measurement changes. For example, in 2001 crown net wealth increased largely because of the inclusion of the state highway network.