E-Commerce

Ahead of the Curve ™

Assessing Amazon’s U.S. Demand & Delivery Capacity

May 4 2018

Report by John Blackledge, Helane Becker, Jason Seidl, Nick Yako, Conor Cunningham, James Kopelman, William Kerr and Tyler Seidman

 

We expect AMZN’s continued shift to faster, costlier delivery speeds to drive the Prime value prop. Other carriers (Ind. 3PLs, AMZN, DHL) will gain significant share of AMZN’s US last mile delivery volume (2BN packages in ’18 rising to 4.3BN in ’23), while USPS, FedEx, and UPS share declines, in our view. Underpinning our work is expected strong sustained growth of AMZN’s US eCommerce biz.


 AMZN US Delivery Analysis (Blackledge, Becker)

Cowen’s Internet, Airlines and Air Freight, and Washington Research teams collaborated on what we believe to be the most comprehensive analysis of Amazon’s US delivery and infrastructure to date. In the report, we assessed AMZN’s (Outperform, $1,572.08) US last mile delivery capacity requirements over the next five years and expect AMZN’s last mile delivery package mix to shift to an Other bucket of carriers comprised of Independent 3PLs, Amazon, and DHL, with USPS, UPS (Market Perform, $109.81), and FedEx (Outperform, $242.15) losing share either by choice or AMZN’s intentions. As part of our work, we held discussions with several companies and/or gov’t entities directly involved in AMZN’s last mile delivery as well as delivery consultants.

Framing the Debate

One of the most active Amazon investment debates centers on the company’s delivery costs and fulfillment infrastructure. Recent comments by President Trump have raised questions over whether Amazon is “taking advantage” of the U.S. Postal Service, and, as such, if USPS shipping rates will increase to a point that will affect Amazon’s margins. We address this debate. We also delve into the broader economics of providing US shipping and logistics services to Amazon. We estimate Amazon will ship ~2.0BN US packages this year, at a cost for last mile delivery of ~$15 billion. As US shipping volumes rise to 4.3BN packages in five year’s time, Amazon will need to ensure sufficient capacity and capabilities.

We analyze the company’s existing relationships with UPS and FedEx. Given AMZN’s increasing choice of speed of delivery driving the Prime value prop despite the costs, we expect AMZN’s last mile delivery package mix to increasingly shift to the Other bucket of carriers. We forecast Other carriers including Independent 3PLs, Amazon and DHL deliver via last mile 17% of AMZN’s US packages in ’18, rising to 34% by ’23. AMZN will in part facilitate this mix shift to Other LMD carriers (including AMZN) via our estimated $42BN capital investment (31% of aggregate WW capex) in US fulfillment network over the next five years to support rising demand and enable closer points to customers.

What Is New for Investors (Blackledge, Becker)

(1) We provide our AMZN global GMV and Unit forecast ’15-’23 by (i) US GMV, (ii) Other NA GMV, (iii) Int’l GMV and, (iv) Digital GMV; (2) We forecast AMZN’s US package volume delivered via last mile ’15-’23 across USPS, UPS, FedEx, and Other bucket; (3) We forecast AMZN’s last mile delivery costs by carriers ’15-’23; (4) US Postal Service deep dive; (5) Assess AMZN’s changing last mile delivery mix toward the Other bucket; and (6) AMZN fulfillment investment analysis.

AMZN US eCommerce Demand to Remain Strong

Underpinning the analysis is our expectation for AMZN eCommerce demand to remain strong. We estimate US eCommerce rises from $481BN in ’18 to $860BN in ’23, with AMZN’s share of US eCommerce rising from 36% in ’18 to ~45% in ’23, driven by strong customer value prop, rising Prime subs (~60MM estimate) and positive secular trends.

What Is Good for AMZN Is Good for USPS

We estimate the US Postal Service delivered ~59% of total AMZN US packages in ’17. Following discussions with persons in the US Postal Service ecosystem and a review of USPS financials, it’s clear that AMZN has been a key driver of USPS for years. AMZN has driven USPS’ fastest growing business, Shipping & Packages, as revenue rose from 20% of total USPS revenue in FY14 to 28% in FY17. We forecast Shipping & Packages to rise to ~50% of USPS revenue and 8% of volume by ’23.

AMZN & USPS are Co-Dependent, but USPS Doesn’t Have ALL the Leverage

While we forecast USPS will nearly double the number of US packages delivered for AMZN from 1.1BN in ’18 to ~1.9BN in ’23, its packages delivered of total should drop from 55% in ’18 to 45% in ’23. Based on our discussions, USPS certainly wants to grow in line with AMZN, and we expect AMZN’s total US packages delivered to rise from 2.0BN in ’18 to 4.3BN in ‘23, a ~16% CAGR.

AMZN’s interest in increasing speed of delivery to improve its Prime value prop likely mitigates USPS ability to grow in line with AMZN US package growth over time. AMZN has PrimeNow which delivers goods via independent carriers in 1-2 hours for Prime members in over 30 US markets and Prime Same Day which offers ~1MM different items for same day delivery in 8K US cities and towns. For USPS to deliver same day, goods have to be in USPS facilities by 6am.

AMZN’s Shift to Faster Delivery Driving USPS Shipping & Packaging Volume Deceleration

AMZN started offering faster delivery programs in late ’14; shortly thereafter USPS parcel volumes, where we believe the bulk of AMZN’s USPS package volume sits, have decelerated. USPS parcel volume growth has decelerated from ~+26% y/y in CY15 to +21% y/y in CY16 to +15% y/y in CY17. The deceleration is even more pronounced in CY4Q, the biggest seasonal quarter for AMZN, as USPS parcel volume growth has decelerated from +27% y/y in CY4Q15 to +18% y/y in CY4Q16 to +10% y/y in CY4Q17. USPS clearly states in its SEC filings it has 3 large customers, among hundreds of other customers in its competitive offerings, so the mix shift and deceleration could be driven by other customers.

AMZN’s Contract with USPS Appears Profitable

The Postal Regulatory Commission determines if competitive contracts are legal and whether contract covers the USPS costs, including a minimum 5.5% unallocated contribution to USPS overhead. Based on discussions, the overwhelming majority of USPS contracts covered their costs last year and USPS contributed an additional 23% to overhead, ~4x above the minimum requirement. As such, we do view the AMZN contract as profitable.

USPS Rate Hike Sensitivity on AMZN Delivery

We estimate AMZN paid USPS $2.19 per each package in ’17 and USPS rates increased 4% for Competitive services segment this year. Per our discussions, we don’t view a material rate hike ahead, but our ’19 sensitivity analysis suggests every 5% increase above our estimated mid-single pricing increase would lead to $150MM shipping cost increase or ~1.3% op income impact for AMZN.

USPS Ramp Alongside AMZN Is Achievable Despite Challenges

While we believe USPS has the ability and desire to scale with Amazon over the next 5 years, issues exist. The USPS has faced a plethora of challenges in recent years which have left USPS in a dire financial situation including i) operational mix shift, ii) burdensome legislation, and iii) chronic under investment in its operations. Based on our analysis, USPS will start generating operating cash flow losses in ’21 persisting through ’23 (our forecast period), but will be able to maintain a positive cash balance. However, our analysis assumes USPS does not make all of its required expense payments or pay off its current portion of debt due. Furthermore, we are not accounting for the capex required to update its fleet, which we expect to cost $6.3BN (180K new vehicles at $35K/vehicle) spread over 7 years. In the report, we offer two potential scenarios that could drastically improve USPS’ near term financial situation.

The Big Leap…AMZN’s US Last Mile Delivery Mix Shifting to the Other Bucket

We estimate the Other bucket, a combination of Independent 3PLs, AMZN Flex, and DHL US will pick up the changing last mile delivery mix, accounting for 17% of AMZN’s LMD delivery volume in ’18 and rising to ~34% by ’23 as (i) AMZN wants to control more of its shipping and wants to drive faster delivery options for Prime subs, (ii) independent carriers can handle faster delivery options than USPS, though are much costlier per our analysis, and (iii) FedEx doesn’t intend to scale with AMZN expected volume growth. Additionally, AMZN is leveraging learnings and confidence from the UK, where AMZN contractors deliver >50% of its package volume.

AMZN shipping costs are housed within COGS and since AMZN instituted faster and costlier US shipping methods in late’ 2014, gross margins have risen from 29.5% in ’14 to 37.1% in ’17. While we estimate US shipping costs as % of COGS rises from 11% in ’18 to 15% in ’23, we expect overall AMZN gross margins to rise from 39% to 45% during that time period.

Could a JV with DHL be Possible? (Becker, Blackledge)

DHL allows Amazon to use its Americas Hub at Cincinnati/Northern KY International Airport (CVG) during the day to sort packages. DHL completed a $108 MM expansion at the airport last year that allows for greater automated sort capabilities and 40 new reload positions. In addition, there is a new ramp providing an additional 16 aircraft parking spaces, bringing capacity to 65 aircraft gates. Atlas Air (Outperform, $67.40) and ATSG (Outperform, $20.38), both of whom count Amazon and DHL as customers, also have operations at CVG. Atlas has a crew base at the airport and ATSG operates aircraft into the airport. Amazon intends to invest $1.4 billion to build their operation at CVG.

ATSG & Atlas Air Provide 767Fs to Amazon

ATSG provides Amazon with a combination of 20 B767-200Fs and B767-300Fs. Atlas Air also has a contract to provide Amazon with 20 B767-300Fs by year-end 2018. We believe there are at least 33 aircraft in service now for Amazon Air, 20 from ATSG and 13 from Atlas Air. Amazon has warrants to buy 19.9% of each company, and has the right to buy up to 29.9% of Atlas Air if Atlas provides Amazon with more aircraft. We believe there is currently an RFP in the market for additional aircraft. There are currently a limited number of 767s available for passenger to freighter conversion, and FedEx, UPS and the US Air Force have aircraft on order with Boeing that take up all of Boeing’s capacity to produce B767Fs for the next five years. As a result, we believe Amazon could look for other aircraft, like the 777F for international service or a converted A321P2F for additional domestic US service.

Does a Merger with FedEx or UPS Make Sense?

We believe the answer is no. We are skeptical as to whether or not a merger would receive the necessary regulatory approvals. In addition, we believe Amazon is no more than 3% of FedEx’s revenue, so an acquisition of FedEx would be disruptive to the supply chain as we believe a large number of the company’s customers would leave. (Although where would they go?) We believe Amazon is close to 10% of UPS’s revenues, so an acquisition might make more sense, but again, we believe existing customers would leave. UPS is more leveraged to retail than is FedEx. Furthermore, we do not believe Amazon has any interest in dealing with unions. UPS is heavily unionized while FedEx has only the pilots’ union on the property. We believe a JV or deeper relationship with DHL makes the most sense for Amazon.

Transport Stock Implications (Becker)

In our view, the implications of Amazon expanding their transportation and logistics network on transportation stocks is clear. FedEx has consistently stated they will not scale with Amazon, while UPS has said they will. We believe Amazon’s contribution to FedEx and UPS margins are below their system averages, which implies any incremental volume from Amazon will further dilute margins. FedEx prefers to focus on B2B, while UPS has embraced the growth in B2C. UPS needs to significantly invest in their network to improve the profitability of e-commerce packages if they plan to scale with the market. We continue to rate FedEx common shares Outperform while we rate UPS shares Market Perform, given the difference in philosophies towards Amazon as well as the difference in their views on the B2B and B2C markets. We also rate the common shares of Atlas Air and ATSG Outperform as they continue to scale their business with Amazon (both currently account for all Amazon’s aircraft) and would likely be beneficiaries if Amazon continues to expand their fleet.

Other Near- and Longer-Terms Questions

US Delivery Capacity Appears Fine Next 4-5 years…(Blackledge, Becker)

If the retail world were flipped on its head today and 90% of retail were eCommerce, while 10% were brick and mortar, the current infrastructure could not come close to supporting and fulfilling consumer’s purchasing. We estimate US eCommerce penetration rises from 10% of total retail sales in ’18 to 15% by ’23. Based on our AMZN delivery and fulfillment work, over the next five years, we think current delivery infrastructure can support the demand.

By ’23, it will still be early for US eCommerce penetration in our view, with several of the largest retail categories still under-penetrated. For instance, we estimate Food & Bev grocery will be a $950BN US retail market, but only 8% eCommerce penetration and the ~$600BN Consumable US retail market will only have 19% eCommerce penetration. As such, it appears that material investments will need to take place from a range of players to support the long-term changes in consumption trends.

…We Expect AMZN to Continue to Invest in Its US Network

In order to support the level of US eCommerce growth, we estimate AMZN will annually invest 2-3% of US GMV (x-Digital) from ’18-’23, in line with historical investment per our analysis. In total, from ’13-’23, we estimate AMZN will invest roughly ~$42BN via capital investment in its US fulfillment operations. Our analysis assumes i) no improvements in efficiency via automation/robotics and ii) its US network is operating at 100% capacity, both of which are highly unlikely and would suggest we are overestimating the future spend needed.

Amazon Policy Risk More Bark Than Bite (Chris Krueger, Cowen WRG Policy Analyst)

The United States Postal Services (USPS) is a government monopoly with a captive audience that is governed by numerous regulations set by legislation. With almost 32,000 post offices and locations scattered among every Congressional district in the country and a heavily unionized work force with numerous military veterans, they have a lot of political juice. This background and political reality underscores why the policy risks to Amazon on contracts and pricing strike us as a lot more bark than bite. The structure of the USPS – and the lack of ANY appointees to the Board of Governors – should give investors relative pause on overreacting to any Presidential Statements delivered via Twitter on the ability to restructure postal contracts. The only wild card here – other than more Tweets – is the USPS Task Force that was created by President Trump in mid-April “to evaluate the operations and finances of the USPS” that is chaired by Treasury Secretary Steve Mnuchin. The Task Force is set to unveil recommendations within 120 days (late August), though they will still be bound by the USPS and current law. The ticking clock – and the glaring absence of ANY Board Members – suggests to us that a big rewrite of postal laws is pretty low on the policy totem pole. Without new legislation, the Task Force is likely more headline risk than actual policy risk.

 

TO READ THE FULL REPORT, CONTACT YOUR COWEN SALES REPRESENTATIVE