New softwood forecasts show modest rises before downturn hits in JQ’17
New short-term forecasts of sawn softwood sales volumes indicate prospects for continued growth through to the March Quarter of 2017. However, the short-term forecasts then anticipate a decline in the June Quarter of 2017.
This is one of many projects which form part a three year (2015-16 to 2017-18) arrangement with ABARES funded by industry. This project has capitalized on ABARES experience in econometric modelling and the data generated through the FWPA industry data aggregation program. These forecasts apply to four products in the FWPA Softwood Data Series which includes monthly sales data from 14 softwood processors and covers some 83% of Australia’s sawnwood production.
At a headline level, the forecasts, supplied by ABARES as part of its relationship with FWPA, address four wood products contained within FWPA’s monthly Softwood data series. Combined, sales of the four products (they are listed below) totaled 1,618,408 m3 for the year-ended June 2016. A year on, they are forecast to rise a modest 4.3% to 1,687,538 m3 for the year-ended June 2017.
However, the individual product volume forecasts are not so clear for the June Quarter of 2017, as the table below shows.
Forecast Sales Volumes for Selected Sawn Softwood Products: JQ’16 & JQ’17 (m3 & %)
|Sawn Softwood Grade||JQ’16 (Actual)||JQ’17 (Forecast)||Forecast % Change|
|Termite treated (H2F)||178,946||176,898||-1.1|
Comparing the June Quarter of 2016 with the forecast for the June Quarter of 2017 shows that despite the forecast full-year rise in sales volumes for the selected products, by the end of the June Quarter 2017, a downturn is expected for every product other than the landscape products which has a strong seasonal/cyclical element.
In the charts below, displaying the forecasts for each of the products, there are three forecast lines. The centre line is the forecast and the other lines represent the upper and lower limits of the forecast model. All of the commentary here is based on the centre/forecast line.
As expected from ABARES, the forecasts are delivered using very detailed and tested methods and assumptions. The methodologies report, for those with a head for such things – the Statistics Count team consumed it with pleasure – provides more than just a statement about the integrity of the forecasts.
The econometric models estimated in the report are not based on formal structural economic models of the relevant wood product markets. Instead, they are simple, statistically robust, predictive models that are based on historical relationships and patterns observed over the sample period. Its relevant to understand the broad data examined by ABARES in constructing the forecasts.
They start with the FWPA Sawn Softwood data series. Forecasts were then developed by assessing:
- Historic sales, taking into account factors such as seasonality and relative sales (between the four products) and identifying ‘outliers’ or inconsistent data;
- Short-term sales drivers, including production and demand
- Extent of correlation between the different products and the factors impacting them
The forecasts that arose are in large part based on the evidence that “….implies that sales in the current quarter are a strong predictor of sales in the next quarter.”
Even at this early stage in the forecasting journey, it is worth industry asking itself if the forecast outputs, taken as a whole and compared one with the other, as well as with what you expect in your business and what you see in the general economy, pass the ‘sniff test’.
The summary you might consider for such a personal analysis is that Treated softwood <120 mm sales will continue to grow until the June Quarter next year, while sales of the Untreated <120mm product will rise only until the end of 2016, after which they will decline, as will be the case for Untreated >120 mm. The declines coming as housing commencements begin to drop away due to the slowing of new housing approvals. As houses are concluded, they need gardens and thus, the forecast shows relatively strong growth for Landscape wood products through the entire forecast period.
Does that pass your ‘sniff test? We think it’s a very reasonable start!
FWPA’s mission to support improved economic literacy and data driven decision-making is supported by forecasts delivered with this type of transparency, because they aid comprehension and point to opportunities for further use of the same methods and similar tools.
In interpreting or using the ABARES forecasts it should be noted that the FWPA Softwood Data series do not include all producers in the industry and the number of participating producers has increased by about 14% of volume in the past 3 years. As such, the series are not entirely representative of national production or sales and the effects of new entrants in the data series has the potential to affect the accuracy of the estimated models. The models estimated in this study, and approaches used to generate forecasts, will be periodically reviewed by ABARES in order to improve the accuracy of forecasts over time.
This forecasting is now a quarterly exercise, allowing for assumptions for each period to be assessed by industry, as well as by ABARES, as we continue progress towards more accurate analysis of the future.
The detailed report of the methodologies and assumptions deployed by ABARES can be found by clicking here.