5625.0 - Private New Capital Expenditure and Expected Expenditure, Australia, Sep 2009 Quality Declaration 
Previous ISSUE Released at 11:30 AM (CANBERRA TIME) 26/11/2009   
   Page tools: Print Print Page Print all pages in this productPrint All RSS Feed RSS Bookmark and Share Search this Product



This feature article examines the use of realisation ratios in relation to the Survey of New Capital Expenditure. There is a particular focus in this article on the use of a five-year average realisation ratio as a basis to project future capital expenditure. This method was used to calculate the Experimental Projected Capital Expenditure Series which has been published in recent years in Private New Capital and Expected Expenditure, Australia (cat. no. 5625.0) and has now been discontinued.


Realisation ratios for the Survey of New Capital Expenditure are related to the seven financial year estimates of capital expenditure compiled from a combination of actual and expected expenditure. As such, for the purposes of this survey there are six realisation ratios (with the seventh always equal to one as it represents actual data for the full financial year).

The realisation ratios are calculated once actual expenditure for a financial year is known (Estimate 7). A comparison between actual expenditure for the financial year and the expected expenditure is made and the resultant realisation ratios (subsequent actual expenditure divided by expected expenditure) then indicate how much expenditure was actually incurred against the amount expected to be incurred at the various times of reporting.

Realisation ratios can be used as a tool in understanding and interpreting expectation statistics for future periods. The application of realisation ratios enables the adjustment of expectation data for known under (or over) realisation patterns in the past and hence provides a basis for comparison with other expectation data and actual expenditure estimates.

The nature of realisation ratios is that they move closer to 1 with each progressive financial year estimate. Therefore, the realisation ratio for Estimate 1 has been used in the following examples. This is because this estimate most clearly demonstrates the concepts being examined. These same conclusions can be drawn from subsequent realisation ratios, although the effect observed is less pronounced for each ratio as it approaches Estimate 7.


The pattern of realisation ratios generally align closely with changes in overall economic conditions. That is, when economic conditions decline realisation ratios will fall as businesses, at least in the short term, underestimate the reduction in their Capex plans. Similarly when economic conditions improve realisation ratios will increase as businesses underestimate the growth in their spending plans. The outcome of this pattern is that realisation ratios can be a used as tool to help track the movements of the economic cycle, particularly in relation to identifying turning points in an economy undergoing rapid change.

The five-year average realisation ratio is a smoothed estimate for the realisation ratio of 2.5 years ago rather the current year. Therefore, it is not a timely estimate for the current year. When examining the behaviour of the five-year average realisation ratio over the life of the economic cycle it is evident that at both ends of the economic cycle, that is during rapid growth or decline, the suitability of these estimates to provide reliable projections diminishes. An example of this can be seen in the following graphs which look at each asset type, equipment, plant and machinery and buildings and structures, separately.

Note in these graphs the Estimate 7 (financial year actual capital expenditure) has been used as a form of measuring actual economic activity at the time.

Five Year Average Projected Time Series, Equipment
Graph: Graph Five year average in Equipment

Five Year Average Projected Time Series, Building
Graph: Graph Five year average in Building

Firstly it is noticeable that the equipment series is generally more volatile than the building series. This may be due to the ability of firms to adjust investment levels for equipment at a more rapid pace than those of building, which are generally more long-term plans. The result is that building expectations are less responsive to short term changes in economic conditions and that even if actual investment in building changes, the change in expectations may take longer to flow through.


Looking closer at the behaviour of the realisation ratios, and more specifically the five year average, it is evident that there is a significant issue that arises when the economy undergoes significant change. This is noticeable with a specific examination of rapid decline and rapid growth periods of an economy as seen in the graphs below, which are excerpts from the graphs seen previously.

Equipment - Five Year Average Projected Time Series, Period of Decline
Graph: Graph Five year average decline in Equipment

Building - Five Year Average Projected Time Series, Period of Growth
Graph: Graph Five year average growth in Building

Firstly it is clear that there is an identifiable lag when realisation ratios from both the previous year or a five-year average are applied. This highlights the problem of predicting future outcomes using historical data which may no longer be relevant or appropriate. However it is noted that, as expected, the single year series does react quicker to the change. Furthermore, at least with the rapid economic decline seen in the equipment series, when the effects of the economic downturn began to flow through, as seen in the decline in Estimate 7, the forecast for both the single and five year average series' actually increased at an increasing rate.

The outcomes seen during these periods of rapid change in the economy suggest that by applying a five-year average realisation ratio, the turning points of an economy will be distorted and smoothed over. Therefore the usefulness of this method for forecasting future capital expenditure is severely limited during a period of change in the economy.

This graph also reveals another factor which impacts the effectiveness of the five-year average. The five-year average is often under the influence of significant patterns of activity which do not necessarily represent an accurate picture of the current overall conditions of the economy. For example, this current five-year average is heavily influenced by the rapid growth and relatively above average realisation ratios of the Mining industry during the period of 2004/05 to 2006/07, particularly evident in the building series.

This period of very high realisation ratios would be strongly contributing to the five-year realisation ratios predicting continued growth in investment in the next couple of financial years. This example of the Mining Industry shows the impact of a 2-3 year period of extreme growth, or decline, in realisation ratios on five-year averaged estimates. This again reflects on the underlying issue that using potentially unrelated and misleading historical data, which may provide little or no insight into future plans, will not provide reliable results in times of economic change.

Realisation Ratio Industry Comparison - Estimate 1, Building
Graph: Graph Realisation Ratio industry comparison

An examination of these two end points in an economic cycle would certainly seem to reveal the inherent problem with five year average realisation ratios. That is during periods of significant economic change, they are unreliable and often provide results which are significantly wrong.

Furthermore this problem is cyclical in that it is prevalent at both the start and finish of an economic cycle. A five-year average is sufficient, and even reliable during periods of moderate growth in investment. However, when under going rapid change in investment they are actually misleading.


In summary, an average of the previous 5 years’ pattern of capital expenditure realisation of expected expenditure is not considered a sound approach to projecting future capital expenditure when the economy has experienced a significant shift in the economic cycle, with a specific emphasis in an event such as a turning point.

It is considered a better approach to provide the user community with a full history of available capital expenditure realisation ratios and for users to determine which realisation ratios or multiple term average realisation ratio may be suitable for projection of future actual capital expenditure, in a given economic context.

This step has been taken with the inclusion from July 2009 of two additional tables on the ABS web site which feature time series of these statistics back to the first available data. These tables are titled:
  • TABLE 12A. Financial Year Estimates Combining Actual and Expected Expenditure by Type of Asset and Industry - Current Prices 
  • TABLE 12B. Realisation Ratios Comparing Actual to Expected Expenditure by Type of Asset and Industry - Current Prices


Data for all realisation ratios and seven estimates is in Current Prices, Original.