The Retail Sales Index (RSI) measures the value and volume of retail sales in Great Britain on a monthly basis. The RSI is a key economic indicator and one of the earliest short-term measures of economic activity. The main output measures include value and volume estimates in both seasonally adjusted and non-seasonally adjusted forms. Further information on estimated prices and the value of retail sales by commodity are published as well as an experimental Internet estimate. It covers the sale of goods made by retailers over a standard period which covers four or five weeks, it does not include the sale of services such as insurance or travel tickets. . The standard period for November is four weeks and covers the dates 31 October 2011 to 27 November 2011. Annexe A gives details of the standard reporting periods for 2010, 2011 and 2012. This article describes how these retail sales estimates are produced, how they can be interpreted and detailed stories from the November 2011 estimates.
The Retail Sales Index (RSI) measures the value and volume of retail sales in Great Britain on a monthly basis. The RSI is a key economic indicator and one of the earliest short-term measures of economic activity.
The main output measures include value and volume estimates in both seasonally adjusted and non-seasonally adjusted forms. Further information on estimated prices and the value of retail sales by commodity are published as well as an experimental Internet estimate.
It covers the sale of goods made by retailers over a standard period which covers four or five weeks, it does not include the sale of services such as insurance or travel tickets. The standard period for November is four weeks and covers the dates 31 October 2011 to 27 November 2011. Annexe A (31 Kb Excel sheet) gives details of the standard reporting periods for 2010, 2011 and 2012.
This article describes how these retail sales estimates are produced, how they can be interpreted and detailed stories from the November 2011 estimates.
The indices are compiled using data from the Monthly Business Survey- Retail Sales Inquiry, a sample of 5,000 is selected from a population of 200,000 retailers in Great Britain. The sample is designed to be reflective of the retail industry and is estimated to cover approximately 95 per cent of all retail sales by turnover within Great Britain.
To achieve this retailers are first allocated to a retail classification group as defined by the Standard Industrial Classification (SIC) 2007 and then by size of business determined by the number of employees. The sample includes 900 large retailers that are continuously sampled, where large is defined as a retailer with greater than 100 employment or turnover of greater than £60 million and employment of 10 - 99, and a random sample of 4,100 small to medium retailers. The sample is supplemented by 30 of the largest retail companies in Great Britain which form the Monthly Commodity Survey (MCS).
All retailers are asked to provide total retail turnover inclusive of VAT, including sales from stores via the Internet, mail order, stalls and markets, door to door and telephone sales. A separate question is asked for total retail sales generated via the Internet only.
Supplementary information on commodity breakdowns sold by retailers is collected through the MCS, where retailers are asked to provide the same information as those in the main sample but also the turnover generated from sales of food; alcohol, tobacco and confectionery; clothing and footwear; household goods; other; and automotive fuel.
Retail sales estimates are published as both value and volume indices rather than in monetary form and are calculated at the lowest aggregation level shown in the SIC structure in Annexe B (31.5 Kb Excel sheet) .
The value estimates reflect the total turnover retailers have collected over a standard period and record the growth since the base period (currently 2008) of the value of sales ‘through the till’ before any adjustment for the effects of price changes. They can be interpreted as how much was spent in the retail sector.
The volume or constant price series are constructed by removing the effect of price changes from the value series. The Consumer Price Index (CPI) is the main source of the information required on price changes. In brief, a deflator for each type of store as shown in annexe B (31.5 Kb Excel sheet) is derived by weighting together the CPI’s for the appropriate commodities, the weights being based on the pattern of sales in the base year. These deflators are then applied to the value data to produce volume series.
Implied price deflators are calculated by dividing the non-seasonally adjusted value indices by the non-seasonally adjusted volume indices. The resulting indices and growth rates provide estimates of how much the prices of goods sold within stores have changed.
Seasonally adjusted estimates are derived by estimating and removing calendar effects (for example Easter moving between March and April), and seasonal effects (for example increased spending in December as a result of Christmas) and trading day effects (for example number of trading days in the period) from the non-seasonally adjusted estimates.
Seasonal adjustment is performed each month and reviewed each year and before adjusting for seasonality, prior adjustments are made for calendar effects (where statistically significant) such as: returns that do not comply with the standard trading period, bank holidays, Easter and the day of the week Christmas occurs.
One off events such as the Royal wedding are not treated as a trading day effect and care is taken to ensure that it does not effect seasonal adjustment in the later periods therefore distorting seasonal adjustments.
There are four growth rates published alongside the retail sales value and volume indices. The month on month growth rate tells the user how much retail sales have grown from the previous month and can be volatile. The three month on three month growth rate provides an estimate of the underlying pattern or trend within the series, in November the three month on three month growth compares retail sales in November, October and September to August, July and June.
The growth this month on same month a year earlier provides an estimate of how retail sales have grown year on year and the three month on the same three months a year earlier compares the months November, October and September in 2011 to the same months in 2010.
Revisions to retail sales estimates occur for a number of reasons, the most common being the addition of late data and classification changes both within retail sectors and out of retail. Revisions can also occur from methodological changes such as re-weighting, referencing or seasonal adjustment reviews.
The latest months Retail Sales estimates are published on approximately 65 per cent response in terms of questionnaires which is approximately 90 per cent coverage in terms of turnover. Turnover and internet values are imputed for businesses that do not return. In order to maximise response and thus minimise imputations a response chasing strategy is followed which chases influential businesses first. The impact on published results is fewer revisions as a result of late data.
Each month the retail sales statistical bulletin and accompanying tables provide headline retail sales estimates based on the growth rates for both value and volume seasonally adjusted indices as well as data which provides explanation for these movements.
Seasonal adjustment allows us to make month on month comparisons as seasonal factors such as a increases in sales at Christmas are accounted for but care needs to be taken when using monthly growth rates due to their volatility. Figure 1 shows the month on month growth rates for all retailing sales volume over the last three years, as can be seen it fluctuates from growth to contraction and vice versa.
For example, in September 2011 retail sales volumes increased by 0.6 per cent compared to August 2011. In October 2011 the same series grew by 1.0 per cent compared to September 2011 and in November 2011 decreased by 0.4 per cent. The month on month growth rates in September and October suggest that there was moderate growth in retail sales but in November there was moderate contraction. Adding these growth rates over the three months we find that retail sales have grown by 1.2 per cent suggesting that over these three months there is sales volume growth in retail sales.
A better indicator of the underlying pattern in retail sales is the three month on three month growth rate. Comparing the three months November, October and September to the previous three months August, July and June all retailing sales volumes have increased by 0.7 per cent confirming that there is moderate growth in retail sales volumes.
Figure 1: Month on month and three month on three month growth rate for all retailing
Year on year growth rates can be used to provide a longer term view of sales growth. When comparing non-seasonally adjusted data as shown in tables 7 and 8 of the main RSI reference tables (1.71 Mb Excel sheet) only year on year growth rates are used due to the seasonal pattern in retail sales running on a yearly cycle.
The four different sectors within retail are weighted together to provide the all retail sales estimates. The weighting structure used is based on the amount spent in the retailing sector each year and is shown in annexe B. The predominantly food sector and predominantly non-food sector account for approximately 85 per cent of all retail sales with the food sector accounting for approximately 42 per cent and non-food 43 per cent. Non-store retailing accounts for 5 per cent and automotive fuel 10 per cent.
The retail sector can often provide mixed results and when looking at what drives or influences movements within the retail sales sector it is important to look at the contribution to growth rates that each sector has. For example, table 1 shows the year on year growth rates in June 2011 for predominantly food stores and non-store retailing. It can be seen that the large increase in non-store retailing does not provide as much upward pressure in the form of index points as the much smaller decrease in food stores provides downwards pressure.
|Year on Year Growth||Weight in RSI||Index Points|
|Predominantly Food stores||-3.9||41.7||-1.6|
The size of store is also a factor that should be considered when determining the reasons for movements in retail sales growth rates; and the size and nature of the RSI sample permits comparisons between small and large stores to be made.
In November 2011, non-seasonally adjusted data shows sale values increased by 4.5 per cent compared to October 2010. Driving the growth in sales values was an increase in small stores of 7.8 per cent with large stores increasing by 3.6 per cent.
In the predominantly food sector we have seen the value of sales steadily increase over the last few years as demonstrated in figure 2, however sales volumes have remained flattish over the same period.
Figure 2: Sales volumes and value in predominantly food stores
From this we can assume the increase in sales values within predominantly foods stores is mainly from an increase in the prices of goods sold within these stores. This is further demonstrated by analysing the implied price deflator which over the period November 20010 to November 2011 increased by 4.6 per cent with sales values increasing by 4.1 per cent and sales volumes decreasing by 0.6 per cent over the period November 2010 to November 2011.
Figure 3, shows the indices for each November trading period, 2006 to 2011. The chart shows that over this period, sales volumes are flattish and the value series shows moderate growth. As sales volumes are flat it suggests that the growth in the value series is mainly driven by price.
Figure 3: Monthly indices for volume and value in November 2006 to 2011
In November 2011, sales volumes decreased by 0.4 per cent and sales values decreased by 0.1 per cent. Table 2 shows the contribution of each sector to the sales volume growth rate.
|Volume SA month on month growth (%)||Weight in RSI||Contribution to all retailing (% points)|
|All retailing including fuel||-0.4||100||-0.4|
|Predominantly food stores||-0.8||41.7||-0.3|
|Predominantly non-food stores||-1.0||43.2||-0.4|
|Textile, clothing and footwear stores||1.1||12.2||0.1|
|Household good stores||-1.2||9.7||-0.1|
|Predominantly automative fuel||1.9||10.2||0.2|
The greatest downward pressure to the all retailing index came from the other stores sector in particular watches and jewellery, computers and telecomms and carpets and rugs. The non-store retailing and predominantly automotive fuel sectors provided upward pressure. In the non-store retailing sector feedback from retailers suggests this increase is due to an increase in sales in the run up to Christmas and from promotions.
Concentrating now on the three month three month growth rates all retailing
Sales volumes increased by 0.7 per cent.
Sales values increased by 1.8 per cent.
Table 3 shows the contribution of each sector to the three month on three month sales volume growth rate.
|Volume SA month on month growth (%)||Weight in RSI||Contribution to all retailing (% points)|
|All retailing including fuel||0.7||100||0.7|
|Predominantly food stores||0.7||41.7||0.3|
|Predominantly non-food stores||0.8||43.2||0.3|
|Textile, clothing and footwear stores||-0.6||12.2||-0.1|
|Household good stores||0.9||9.7||0.1|
|Predominantly automative fuel||0.6||10.2||0.1|
The predominantly food and predominantly non-food sectors provide the greatest upward pressure to the all retailing three month on three month growth rate.
Compared to November 2010, all retailing sales volumes increased by 0.7 per cent and sales values increased by 0.7 per cent and slaes values increased by 4.6 per cent. The non-store retailing sector and predominantly automotive fuel sector provided upwards pressure increasing by 18.9 and 2.7 per cent respectively and the predominantly food and predominantly non-food sectors provided downward pressure decreasing by 0.6 and 0.7 per cent respectively.
Figure 4 shows the month on month growth rates for value and volume for the November trading periods from 2006 to 2011 that is it compares the growth between October and November in the years 2006 to 2011.
Figure 4 Month on month growth rates in Novembers 2006 to 2011
The graph shows the sales volume and value decrease in November 2011 compared to October 2011 is not unusual as we saw decreases in 2008 and 2009 and no increase in 2010.
Comparing this to the three month on three month growth rates in figure 5 we see in the November trading periods 2006 to 2011, the underlying pattern in sales volumes is flattish and there is more growth or contraction within the value series again these movements are likely to have been driven by changes in price.
Figure 5: Three month on three month growth rates in Novembers 2006 to 2011
In November 2011 non-seasonally adjusted sales values data shows that small stores fared better than large stores, increasing by 7.8 per cent with large stores increasing by 3.6 per cent with all stores increasing by 4.5 per cent.
Table 4 shows the year on year growth in sales values, non-seasonally adjusted by type and size of store. The tables shows that the non-store retailing sector saw the largest sales value growth and that all stores with the exception of other stores saw sales values in small stores fare better than in large stores.
|All retailing including fuel||4.5||3.6||7.8|
|Predominantly food stores||4.4||3.5||8.3|
|Predominantly non-food stores||0.3||-0.1||1.5|
|Textile, clothing and footwear stores||2.3||2.1||3.6|
|Household good stores||-0.4||-6.0||14.1|
|Predominantly automative fuel||15.7||N/A||N/A|
Looking now at the value non-seasonally adjusted index by size of store as shown in figure 5, prior to 2010 the underlying pattern in sales values in small stores is relatively flat but also it is at a higher level than that of large stores and subsequently small stores, the growth rates that we see in small stores in recent months could be in part attributed to small stores returning to the level they were at prior to 2010.
Figure 6: Value non-seasonally adjusted index by size of store
Each month retail sales estimates include experimental statistics on Internet Sales. These statistics cover sales made over the internet by all retailers that is they include on-line sales from supermarkets, department stores, clothing stores and predominantly non-store retailers.
In November 2011 the non-seasonally adjusted value of Internet retail sales was estimated at £787.9 million which was approximately 12.2 per cent of all retail sales (excluding automotive fuel), compared with November 2010 which was £593.4 million which was approximately 9.5 per cent of retail sales (excluding automotive fuel).
Figure 7 shows the percentage of internet sales of all retailing from 2006 and we can see from this chart there has been an increasing amount of retail bought online over the past 5 years. In November 2006 around £1 in every £33 that was spent on retail excluding fuel was actually spent online.
Now looking 5 years later this has increased to an extent that now around £1 in every £8 spent on retail is over the internet.
Figure 7: Internet sales as a percentage of all retailing
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: firstname.lastname@example.org
These National Statistics are produced to high professional standards and released according to the arrangements approved by the UK Statistics Authority.