The Amazon Effect

Confirmation that Amazon is bringing Amazon Marketplaces to Australia has caused a stir, and is one reason Ord Minnett downgraded retail stocks under our coverage to Lighten last week. There are still few details about the US company’s plans, but our analysts expect that Amazon’s entry will have a multi-layered effect on retailers, including; firstly, lowering the price of retail goods; secondly, reducing earnings from FY19 as sales growth slows; and thirdly, restricting the expansion of valuation multiples.

 

The proliferation of Amazon

Despite announcing that it will come to Australia, there are still few details surrounding Amazon’s plans (including timing, pricing and product range), but looking at its rollout in the US can provide a few guidelines.

 

In the US, e-commerce sales growth rates are in the mid-teens, and e-commerce now represents around 8% of retail sales, up from 6% in 2013. Amazon is a behemoth in the online space. Our global research counterpart’s US consumer analyst estimates Amazon accounts for nearly 40% of e-commerce sales in the US (up from 25% in 2013), and we note the following effects over this time:

 

  • Category dominance – Initially launching with books in 1995, Amazon has now expanded into a number of other categories. Based on US experience, most of its sales are from computers, software, electronics and accessories presently (Figure 1), but it also has a large and growing slice of the online market in consumer packaged goods, books, music, toys and apparel. On the other hand, categories where Amazon currently has lower penetration include furniture, appliances and equipment, jewellery and office supplies, although these categories are beginning to exhibit strong online sales growth and Amazon has been picking up market share.

Figure 1: Amazon key top and bottom sales categories, and share of online market 

 

Top 5 Categories   

  Proportion of

Amazon's GMV     

 Amazon's share

of online sales 

Computers, Software, Electronics, Accessories       

 46%

 70%

Apparel & Accessories 

11%

 28%

Consumer Packaged Goods

10%

53%

Books & Magazines

5%

70%

Music, Movies, Videos

3%

70%

Toys & Hobbies 

3%

53%

 

Bottom 5 Categories   

  Proportion of

Amazon's GMV     

 Amazon's share

of online sales 

Home & Garden                                                              

2%

23%

Office Supplies

2%

23%

Furniture, Appliances, Equipment

1%

13%

Jewelry & Watches 

1%

18%

Flowers & Misc Gifts

1%

28%

Sports & Fitness

1%

53%

Source: J.P. Morgan estimate and comScore. GMV - Gross Merchandise Value 

 

  • Pricing competition – It is difficult to separate Amazon’s impact on prices from other factors such as currency movements, lower input costs, and technological advances, but broadly, since 2013, prices of key retailing goods categories have been flat-to-lower across the US. Personal computer prices are down 26%, apparel is down 0.4%, and furniture is down 7% over this period. A case study done by our global research counterparts noted that at one point, the combined price advantage Amazon had over US electronics retailer Best Buy was as high as 20–25% on a Sony or Samsung TV. Furthermore, Amazon’s EBIT margins are low, around 2–3% in recent years (Figure 2), and its international operations are currently loss-making, indicative of a company that is willing to invest for the longer term.

Figure 2: Amazon low historical EBIT margins reflect its investment in the business

 

Segment

   2013    

   2014     

   2015     

   2016    

   2017e    

   2018e    

   2019e    

   2020e   

Total (ex AWS)            

  1.9%

  1.4% 

  2.7% 

  2.9% 

  2.3% 

  4.3% 

  5.0% 

  5.7% 

International

  0.3%

  -0.4%

  -0.3%

  -1.1%

  -2.1%

  1.5%

  2.5%

  3.3%

Source: J.P. Morgan. AWS - Amazon Web Services

 

  • Shape up or shift out – Amazon’s increasing reach and success with consumers means that its competitors are being forced to respond to changing conditions, or otherwise close up shop. Our global research counterpart estimates, for example, that in the year to date there have been 11 bankruptcies in US retail. Those who can narrow the price gap, increase their online presence and distribution capabilities, while reducing costs to sustain margins, will be best-placed to survive. Initially, we expect this will mean increased investment is needed by retailers, and eventually the likely outcome is a fewer number of category leaders in the market. Walmart is one company that has had some success in competing with Amazon, and it is now the third-largest online retailer in the US behind Amazon and Apple.

Multi-level impact of Amazon’s entry into Australia

In some ways, Australian retailers should be better prepared for Amazon’s entry given the shift to online has been progressively taking place for a few years now and Australian retailers are aware of how their global peers have responded to Amazon. Local retailers have already been making efforts to reduce store count, improve their online presence and narrow price gaps. However, online is still at a more nascent stage here. According to the Australian Bureau of Statistics, the proportion of turnover done online in Australia is estimated to be 4%, up from 2% in 2013. As such, we still see an ongoing effect on listed Australian discretionary retailers from Amazon’s entry:

 

  • Price deflation – Our consumer analyst expects retailers here will continue to have to lower prices to compete with Amazon. He forecasts that Amazon could cause like-for-like sales growth to fall by a further 100 basis points for FY19 and FY20.
  • Earnings growth – Investment in lower prices, not all of which can be recouped from the supplier or through cost reductions, will weigh on margins and hence earnings growth. Compared to Amazon, retailers here currently have higher EBIT margins, e.g. JB Hi-Fi at 6.9%, Harvey Norman at 6.1% and Myer at 5.3%. Our analyst assumes Amazon reduces EBIT margins by a further 25 basis points in FY19 and FY20.
  • Price-earnings multiples – Most listed discretionary retailers have already seen a PE de-rating this year. Even so, PE multiples remain above the previous lows seen in 2011 and 2012, and PE multiples of comparable US retailers suggest limited upside. Furthermore, the earnings of US peers on which the multiple is calculated have arguably already been depressed by Amazon’s entry, something which has yet to be reflected in the earnings of Australian retail corporates.

Figure 3: 12 month forward consensus PE multiples of listed discretionary retailers

 

 

 

Description

 Current 12MF PE(x)

 

Sample US Peer Comp

US Peer Comp

  current 12MF PE (x)

Harvey Norman       

     

Electrical & Furniture

13.3

    

 Best Buy

14.9 

JB Hi-Fi

 

Electronics

14.2

 

 Best Buy

14.9

Myer

 

Dept. Store

11.1

 

Macy's

 7.4

Super Retail

 

Auto/Leisure/Sports

13.5

 

Dick's Sporting Goods

11.1

Source: Datastream, IBES, Ord Minnett Research. Comp - comparable

 

 As a result of these assumptions, as well as the weaker consumer environment in Australia, our new forecasts now show little growth in earnings for the discretionary retailers from FY18, and similarly our price targets suggest some further downside to current share prices.

 

Figure 4: Earnings growth estimates FY17-FY19

 

 

 EPS ($)

FY17e

 

EPS ($)

FY18e

 

EPS ($)

FY19e

 

Rating

 

Price Target ($)

 

Harvey Norman       

0.346

       

 0.352

       

0.327

       

Lighten

       

3.50

     

1.6%

 

-7.1%

         

JB Hi-Fi

1.850

 

 1.811

 

1.795

 

 Lighten 

 

20.50

     

-2.1%

 

-0.9%

         

Myer

0.087

 

11.1

 

0.081

 

Lighten

 

0.80

 

     

-0.3%

 

-6.6%

         

Super Retail

 0.642

 

13.5

 

0.700

 

Lighten

 

7.25

     

10.6%

 

-1.4%

         

Source: Ord Minnett Research estimate

 

 

 

Ord Minnett Research Trial

We invite you to sign up for a three month trial of our Ords Monthly investment newsletter. This report includes our latest opinions, research and share market insights that may enhance your current portfolio structure.