Freight Transportation & Logistics

Freight Transportation & Logistics ecommerce Trends Prepared for: TD Securities Inc. March 2, 2017

Agenda Market Size and Characteristics Key ecommerce Trends Potential Implications for Transportation 1

US ecommerce is an estimated $340 B market, forecasted to grow 12% per year through 2018 $500 $400 $300 $200 $100 $0 US ecommerce Sales by Year ($B) 2012 2013 2014 2015 E2016 E2017 E2018 Estimated US 2015 ecommerce sales of $340 B 2016 Q1-Q3 grew ~15.5% year over year US ecommerce sales forecasted to grow 12% per year through 2018 ecommerce sales accounted for about 7.3% of total 2015 US retail sales 9.4% of retail sales excluding autos, fuel Source: Internet Retailer; US Census Bureau; Statista 2

Canadian ecommerce is about a $34 B market Canadian online sales estimated at $34 B, about 6.4% of retail sales 45% of online retail sales by Canadians are with non-canadian retailers 62% of Canadians prefer buying from a Canadian business (vs. 2% that prefer to buy from a US company) 60% 45% 30% 15% 0% Percent of People Buying Category Online Canadian Shoppers Source: CIRA The State of E-Commerce in Canada, 2016; Statista; Statistics Canada; TandLA estimates 3

In the US consumer electronics and apparel account for nearly 40% of ecommerce sales E2015 US ecommerce Sales by Product Category ($B) $70 $74 ecommerce sales are along a wide range of product categories $8 $9 $14 $19 $23 $30 $35 $58 Consumer electronics is the largest category at an estimated $75 B, followed by apparel at an estimated $58 B Computer, Consumer Electronics Auto & Parts Furniture & Home Furnishings Toys & Hobby Food & Beverage Apparel & Accessories Books/Music/Video Health & Personal Care Office Equip & Supplies Other Source: emarketer, April 2014, Factiva, our estimates 4

Web only retailers and traditional retail chains account for 75% of ecommerce sales 35% Est. US ecommerce Sales By Type of Seller ($B) 10% 14% 41% Estimated companies in Top 500 ecommerce retailers by segment Web only: 200 Retail Chain: 160 Catalog: 70 Consumer Brand/ Manufacturer: 70 Consumer Brand Retail Chain Web Only Retail Catalog *Based on Consumer Brand share of Top 500 and total estimated 2014 ecommerce sales Source: 2013 Internet Retailer Top 500 5

Amazon.com is by far the largest ecommerce retailer and a leading influencer on the industry 2014 ecommerce Market by Retailer 1% 2% 27% 4% 4% 7% 2% 2% Amazon Apple Walmart Staples Sears Netflix Macy's Office Depot CDW Home Depot Costco All Other Source: WSJ/Internet Retailer Top 500 Guide, AMZN financial statements and website; our estimates Amazon.com accounts for about 27% of the US market Amazon Marketplace allows other companies to sell through Amazon.com website Amazon + Marketplace estimated to account for 38% of US market Can include fulfillment by Amazon Amazon has led many online trends Free shipping Free returns Same day delivery Amazon adding transport capability 195 facilities, more than 78 million ft2 6

Agenda Market Size and Characteristics Key ecommerce Trends Retailer/ e-tailer Capabilities Transportation and Fulfillment Models Relative Economics Potential Implications for Transportation 7

ecommerce is still in a state of flux making it difficult to plan and manage the supply chain Changes How companies go to market Customer expectations on product delivery times Customer expectations on ability to return products Key Questions and Issues Manufacturers/distributors Sell through traditional retailers? Sell through Amazon and other etailers? Sell direct through brand website? Channel conflicts? Retailers Omni-channel approach Marketplaces Slow and cheap, or fast and expensive? Free shipping expectations? What will consumers demand and pay for? How to compete with Amazon? Free shipping of returns? Consumer mentality buy 3, return 2 Where and how to process returns? Fast growth What volumes should be planned for How to maintain flexibility 8

9 Most traditional companies are set up to ship and receive pallets, not packages companies have to add capabilities Outbound Shipment Volume Warehouse Operations Traditional Retail DC to store Low frequency of large shipments (TL, TL multi-stop) Racked pallets, moved with fork lifts Staging of multiple pallets to move into trailers Limited handling or packaging some mixing, shrink wrapping ecommerce Fulfillment DC to consumer High frequency of small shipments (parcel) At individual product level eaches Manual picking Packaging lines Can have pre-sort for packages for parcel carriers Can be highly automated picking systems with large capex investment

10 Technology to support single pick is very different and requires additional capabilities Inventory management at various levels Pallet Case Each Order picking Premium on real-time information in both directions What is in-stock Applying orders immediately to inventory Shipping cost Integration with various systems, often crosscompany Web front end (Demandware, IBM, Oracle, SAP) Various order management systems (Jagged Peak) Warehouse management systems (JDA, Manhattan) Various ERPs (SAP, Netsuite) More complex returns

11 Web-only retailers have significant scale issues ecommerce Challenges Typically begin with single point fulfilment Insufficient scale to stock multiple DC s Leads to high delivery costs, longer delivery times When grow to multi-dc Causes technology issues Legacy DC s in wrong location Requires more inventory Options/Issues DIY vs. Outsource Single vs. multi-user facility Managing returns Returns restock vs. liquidation

Returns and sales spikes are two other ecommerce differences that increase the supply chain challenge Returns are much higher in many product categories in store is typically low single digit percentages 20-30% of apparel purchased online ~10% of hard goods (home goods, toys) purchased online 87% of retailers allow for online returns to their stores encourages additional purchases Free shipping encourages returns Returns frequently are part of the sales model (buy 3, return 2) Sales spikes are greater than in physical stores Flash sales Subscription models Sources: Kurt Salmon; TandLA experience 12

Agenda Market Size and Characteristics Key ecommerce Trends Retailer/ e-tailer Capabilities Transportation and Fulfillment Models Relative Economics Potential Implications for Transportation 13

14 US final mile/delivery - there are a range of delivery options at different service levels and price points Express Single fulfillment center vs. regional centers Service Parcel carrier air or ground Time definite Ground USPS Delivery (DDU Induction) Same Day Parcel Parcel carrier Courier ground does bulk pickup, delivery carrier line haul USPS provides final mile delivery 2-5 days 2-7 days Same day Have brand/manufacturer perform fulfillment (drop ship)? Speed Next Day/2 nd day Cost High Medium Lowest High Position Easy Easy, less Lowest service High expensive Lowest cost quality than Express No DIM Expensive DIM issues <5 Lb. packages Oversized products can come LTL or specialized 3PL Amazon in major metros Difficult for most others to follow

FedEx and UPS US shipments with now include dimensional pricing (DIM) on Ground shipments, but not on their USPS delivery services, Canada had different DIM s 15 Description FedEx and UPS instituted dimensional pricing for ground shipments Parcel pricing now based on both package weight and dimensions Light bulky shipments are assigned a higher weight and billed at that weight Shipments in Canada are also subject to DIM pricing Impact Shippers of relatively large, low weight packages pay higher parcel rates Operational impact - Increased emphasis on better packaging. Smaller boxes Merchandizing impact - Potential to add products to a shipment without adding shipping cost Service impact Shifting to Parcel Select (slower)

DDU Induction (called Parcel Select by USPS) products are growing fast as a low cost US parcel option, without a DIM $4.0 $3.5 $3.0 $2.5 $2.0 $1.5 $1.0 $0.5 $- USPS Parcel Select Annual Revenue $ Billions 2012 2013 2014 2015 USPS Parcel Select product growth has been rapid 2012 2015 Revenue CARG: 27% 2012 2015 Volume CARG: 16% DDU induction providers, such as Newgistics, UPS SurePost, FedEx SmartPost have been growing fast Parcel Select is dependent on USPS pricing Parcel Select 2012 Revenue/Package: $1.32 Parcel Select 2015 Revenue/Package: $1.74 Sources: USPS Annual Reports; FedEx financial filings; Stifel Nicolaus; UPS 2014 Annual Report; TandLA expertise 16

17 Same day was considered too costly, but Amazon is making an aggressive push, leveraging its local fulfillment network Advantages Highest speed to customer Best alternative to physical shopping - need it now Challenges Inventory must be close to customers Very difficult to have density to make low cost delivery Amazon Prime Same Day Free Shipping Cities - 2015 Amazon offerings 1 and 2 hour delivery (Prime Now) Sunday delivery Retailers offering same day through 3PLs such as Deliv Without a local network Same day is cost prohibitive Amazon Prime Same Day Free Shipping Cities - 2016 Sources: Amazon.com, Best Buy website; TandLA expertise

Companies fulfill orders with various models Single Fulfillment Center Multiple Fulfilment Centers Advantages Fits in existing technology/network as a store Can start small, easy to start and manage Lower final mile costs Better delivery transit Stores Close to customers Existing space and inventory Facilitates same day Dis-Advantages High cost final mile delivery Long and variable delivery transit Omni-channel technology issues/complexity Inventory levels Scale to make centers efficient Greatest omni-channel issues High cost DC labor, per sq ft Interrupt merchandizing and store operations 18

Moving to multiple DC s dramatically lowers final mile costs, but adds significant complexity Significantly lowers average Zones Lower cost Faster delivery Reduced variability However Creates a lot of issues and challenges Omni-channel technology and operations issues Increased inventory Required scale to support multiple operations Note: optimal network requires moving the initial DC Example UPS Zones 1 DC 8 4 3 2 8 6 4 8 4 7 4 5 5 Example UPS Zones 2 DCs 5 2 4 2 4 3 5 4 2 5 4 5 6 5 Sources: UPS website; TandLA expertise 19

20 Many companies outsource fulfillment to a 3PL to gain more sophisticated capabilities Outsourcing Benefits 3PL invests in technology 3PL expert in handling eaches Can use space in a shared facility to gain scale benefits Faster way to build capability Can have lower cost labor Can have lower cost transport Leverages scale of 3rd party provider Provider may have volume to pre-sort for carrier Fulfillment is non-core for many companies Leading Players Amazon Fulfillment Radial PFSWeb Speed Commerce Newgistics Saddle Creek

Agenda Market Size and Characteristics Key ecommerce Trends Retailer/ e-tailer Capabilities Transportation and Fulfillment Models Relative Economics Potential Implications for Transportation 21

22 There are many drivers to lower ecommerce supply chain costs companies must balance across them Inbound Cost Facility Cost Handling Cost Inventory Cost Delivery Cost Volume to support full loads Multiple facilities to limit distance, re-shipping Drivers to Lower ecommerce Supply Chain Costs Scale benefits, dedicated or shared Low cost locations Amount of inventory Increased automation Managing staffing to volume Intelligent picking logic Less complex product profile Multiple products per shipment Fewer facilities = less inventory Tech to manage inventory as pool Faster inventory turns Cost of capital Service level Distance from customer Volume to support density Number of products per shipment

23 Supply chain costs are about 21%-23% of total store-based retail, or about three times the cost of the stores themselves Est. US Store Retail Sales Cost Buildup Total $4.33 T Supply Chain 25%-30% 21%-23% Stores 6%-10% Est. Retail Store Supply Chain Cost Category % Sales Spend ($B) Inbound Fgt. 7% $304 Fulfillment 4% $173 Inventory 5% $217 Outbound 3%-5% $130- $217 Returns 2% $87 TOTAL Supply Chain 21%- 23% $911 - $998 Source: US Census Bureau; TandLA experience and model estimates Store costs can be 6-10% (or more) of sales, $260 B - $433 B

24 With ecommerce, supply chain costs are about 25% to 30% of the total cost Est. 2015 US ecommerce Sales Cost Buildup Total $340 B Supply Chain 25%-30% Est. ecommerce Supply Chain Cost Category % Sales Spend ($B) Inbound 4-6% $14-$21 Fulfillment 5% $17 Inventory 6% $21 Outbound 6%-9% $21-$31 Returns 4.5% $15 TOTAL Supply Chain 25.5% - 30.5% $88 - $104 Source: US Census Bureau; TandLA experience and model estimates Assumes a well run supply chain If sub-optimal, costs escalate fast!

Outbound costs are critical - ecommerce turns a 40,000 pound truckload to a store, into 10,000 four pounds shipments to homes $1.75 $1.50 $1.25 $1.00 $0.75 $0.50 $0.25 $- Example Relative Shipping Economics Per Pound (by shipment weight) $1.74 2 to 5 lbs. $0.39 20 to 50 lbs. $0.18 500 to 600 lbs. Parcel LTL Truckload $0.11 2000 to 2500 lbs. $0.03 40,000 lbs. Outbound transportation cost Up significantly in ecommerce environment Parcel cost/pound can be over 10x the cost of LTL and TL to store Dimensional pricing further increases costs for light products 25

Based on some example moves TandLA has modeled, economics change significantly based upon number of DCs, products per shipment, and delivery service Inbound transportation cost benefits from local move from port in 2 DC scenarios Handling cost benefits from multiple products per customer shipment Delivery cost benefits from multiple products per shipment DDU induction rates are much lower than standard ground rates Estimated Sample Cost/Unit High Volume Shippers Illustrative 1 Unit; 1 DC; Grnd Delivery 2 Units; 1 DC; Grnd Delivery 2 Units; 2 DCs; Grnd Delivery 2 Units; 2 DCs; DDU Delivery $2.73 $1.92 $1.92 $3.63 $3.45 $1.92 $1.86 $4.38 $6.56 $6.39 $5.97 $0 $3 $6 $9 $12 Cost Per Unit Inbound Storage/ Handling Delivery $10.13 Sources: UPS and USPS websites, FedEx financial filings, client fulfillment economics, Stern School, Port of LA/LB; TandLA estimates 26

Agenda Market Size and Characteristics Key ecommerce Trends Potential Implications for Transportation 27

Full truckload is likely to lose volume to LTL as products flow to smaller DCs in/near market, or into parcel Positive for LTL carriers, minimal impact for TL carriers 28 Market Truckload Built around old retailing model Likely to lose some volume More fragmented inbound results in smaller shipment sizes Less dedicated store delivery Truckload is huge, so doesn t significantly impact market LTL Gains from ecommerce Smaller shipments to more DC s/ fulfillment centers benefits LTL Home delivery of larger items (> parcel), a high growth market Increase in drop ship could be plus or minus (TL to LTL, LTL to parcel) LTL much smaller than TL, so gains could be meaningful to overall LTL industry volumes Carriers Highly fragmented, share gains can more than make up difference Performance remains based on execution, not market Shift to LTL to more locations will be positive for most carriers Some LTL carriers struggling to develop home delivery model who will be successful at B2C?

29 Rail carload likely unaffected by ecommerce, while rail intermodal could be marginally hurt by shift to ecommerce Market Carriers Rail Carload Little to no impact on carload Most commodities not B2C Limited differences across railroads Rail Intermodal Likely to lose some volume due to ecommerce Shipments likely to be more fragmented outside of core intermodal lanes Increased transloading of ISO to domestic containers near ports allows for leakage to transload to truck In Canada, intermodal less impacted due to concentration of population in intermodal served cities and lanes Canadian roads may be minimally impacted US may be impacted more given dispersion of population

In the US, shift to market-based DCs likely to result in volume where railroads don t currently offer intermodal service. Shipments also go direct to consumer Store-Based Example ecommerce Examples Traditional Store-Based Intermodal to DFW DC DC 500,000 ft2 Truck from DFW DC to stores in other Texas cities Customers shop at stores ecommerce, High Volume TL or LTL to in-market fulfillment centers DC 10,000 to 30,000 ft2 Local package delivery from inmarket fulfillment centers ecommerce, Low Volume Out of market fulfillment center parcel direct to consumer 30

31 Parcel carriers are experiencing volume surges from ecommerce this is both good and bad Market Carriers Integrators Volume increases are good and bad Significant volume increase High B2C makes delivery economics worse High seasonality and spikes (how much capacity to invest in?) Love/hate Amazon relationship Move in US to grow Parcel Select (USPS delivery) B2C likely to continue to grow for major integrators despite handling lower share of Amazon Margins continue to be an issue Alternative Small Package Small companies focused on ecommerce space designed for B2C Typically regional Big upside growth Often high customer concentration Some have high reliance on USPS for delivery Typically have lower cost structure than integrators Many different models Have to develop long-term sustainability (not subject to Amazon insource)

32 Amazon.com has been developing its own transportation capabilities. It is not clear what its end game is Amazon Transport Capabilities Obtained NVOCC operating license for shipments from China to US Added air cargo capacity through relationship with ATSG and Atlas Air Acquiring a private truck fleet to transport shipments between facilities Developed several programs for final mile delivery Amazon Flex (Uber-like) model Use of local delivery companies Fleet of branded delivery trucks for final mile delivery Amazon Strategy Amazon could follow a number of potential strategies: Support Peak: Build guaranteed capacity to support holiday peak; other carriers will remain a core aspect of delivery network Handle Amazon Business: Take control of shipments where density allows it beyond what is required for holiday peak; Support companies selling through Amazon marketplace For-Hire Parcel Carrier: Leverage its own volume and capabilities to become full for-hire carrier

33 Air Freight carriers may win, but likely only in niches. Barge likely not impacted Market Carriers Air Freight Grounded air freight networks are participating in ecommerce via line haul, sort and delivery for larger items (e.g., seasonal charcoal grills) Extend two-day network Nationwide for B2C Forward Air, other grounded air freight providers? Providers to Amazon Prime Air Barge Little to no impact on Most commodities not B2C No differences across barge lines

34 ecommerce is a strong positive for warehousing and distribution ecommerce typically requires significantly more warehousing space than brick and mortar retail Demand for new large warehouses has skyrocketed 120 built between 2010 and 2016 ~30 more planned or being built Most are near large population centers Many built by Amazon ecommerce now driving demand for small regional facilities Shift in demand from a few large national DCs to many regional/local fulfillment centers (think Amazon Same Day service locations) Needs to be very close to population, speed delivery, cut cost (zone 2) Older, smaller warehouses of all shapes and sizes are in demand Historically cast-off in favor or mega-warehouses too small, ceiling too low, located close to population in congested areas Now ideal for ecommerce fulfillment - don t need high ceilings, or large scale and close to population is a benefit

35 3PLs and technology providers are in a position to manage complexity and provide better solutions Market Providers 3PLs ecommerce adds complexity good for 3PLs Retailers not used to eaches may outsource B2C fulfillment Requirements of systems and processes very different for B2C ecommerce start ups want to scale fast often done more easily through a 3PL Can be economic benefits to shared facilities Requires re-think of inbound, too Traditional 3PLs frequently are not equipped to do ecommerce Targeted efulfillment providers many have struggled to-date Tech Providers Strong positive from having to manage the complexity Inventory and order management at the eaches level, not case or pallet Integration between front end and back end, across partners Management across more locations, including stores (OmniChannel) Heavy returns component Likely to be some big winners Major players and niche providers in the game

36 Lee A. Clair Managing Partner 1910 First Street Suite 300 Highland Park, IL 60035 www.tandla.net