Retail Trade, Australia, Preliminary

This release provides a preliminary estimate for Australian retail turnover for April 2020. This estimate is compiled from the monthly Retail Business Survey and is based on preliminary data provided by businesses that make-up approximately 80% of total retail turnover and is therefore subject to revision. The final monthly estimate will be published in Retail Trade, Australia (cat. no. 8501.0) on 4 June 2020.

• The seasonally adjusted estimate fell 17.9% (-$5,383.3m) from March 2020 to April 2020. This result is the strongest seasonally adjusted month-on-month fall in the history of the series.
• In seasonally adjusted terms, Australian turnover fell 9.4% in April 2020 compared with April 2019.
   

• The fall in seasonally adjusted terms in April 2020 was driven by the Food retailing industry, which fell 17.1% (-$2,444.3m) in April, following a 24.1% rise in March 2020. Sales in Food retailing are 5% above the level of April 2019.
• Analysis of supermarket and grocery store scanner data shows that monthly retail turnover fell in original terms for Non-Perishable Goods, Perishable Goods and All Other Products by 23.7%, 15.3% and 24.5% respectively in April 2020 compared to March 2020. These falls follow significant unprecedented demand in March 2020 where Non-Perishable Goods rose 39.0%, Perishable Goods rose 21.6% and All Other Products rose 30.5%.
• The April month saw further strong falls in Cafes, restaurants and takeaway food services and Clothing, footwear and personal accessories retailing. Businesses reported that regulations regarding social distancing measures limited their ability to trade as normal for the entire month. Turnover in Clothing, footwear and personal accessories, and Cafes, restaurants and takeaways in April 2020 is around half the level of April 2019.
• April also saw a strong result for Online retailing, with 10% of total retail turnover purchased online.
   

Caution should be exercised in interpreting preliminary estimates as they may be significantly different to the final published estimates. This is due to several factors:

• Estimates are based on preliminary data provided by businesses that make-up approximately 80% of total retail turnover.
• Where respondents have not yet provided their data, it is estimated (or 'imputed') based on previous responses or averages from similar responding units. The level of imputation in preliminary estimates is significantly higher than for final estimates.
• The quality of imputation for preliminary releases may also be poorer than for final estimates, due to the higher level of non-response. Furthermore, historical imputes which are based on data from previous months, may not accurately reflect changes in the economy due to recent events.
• Changes to imputation methods were been made from the March monthly release to ensure non-respondents are more accurately reflected by the responding units in the current COVID-19 environment.
• Until February 2020 Retail Trade used the concurrent seasonal adjustment method, meaning that seasonal factors were re-estimated each time a new data point becomes available. If not appropriately accounted for, unusual real-world events, such as COVID-19, can distort estimates calculated using this method. From March 2020, seasonal factors are calculated using data up to and including February 2020, then projected from March 2020 onwards. This approach, known as the forward factor method, ensures that the seasonal factors are not distorted by COVID-19 impacts.