Published Last Week

FTAI Aviation: Short Report Review

Two weeks ago, Muddy Waters released a short report claiming FTAIs reporting and accounting is misleading. While management and the sell-side have both rebutted the claims, the report focuses on a couple key points:

  1. FTAI is selling whole engines, not modules as reported - this overstates its addressable market for ‘module swaps’
  2. FTAI acquires, records, and depreciates engines in its leasing division before shifting engines at subjective residual values to the MRO division. This reduces COGS and drives higher MRO EBITDA margins

Last year, we also spent time trying to understand how a ‘module swap’ works and published various research: 

  1. FTAI Aviation; a curation of all our work on FTAI

This interview with a Former FTAI Executive explores the major claims in the short report. 

Fever-Tree: Molson Coors Partnership & US Off-Premise Distribution

With any software point solution, niche CPG category, or any specialised vertical offering, the question is always how to scale distribution. And not just to scale at any cost, but whilst protecting the economics and brand equity. 

Over the last five years, Fever-tree has increasingly moved to own more of its distribution. It faced challenges launching US bottling and production and recently took back Australian distribution in-house. But this is old news: last Thursday, FEVR reported a strategic partnership with Molson Coors in the US. TAP is now responsible for production, bottling, marketing, and distribution of Fever-tree in the US.  

This feels similar to Monster’s deal with Coca Cola in 2008. Monster was originally distributed by Anheuser Busch before partnering with Coca Cola globally. Monster balanced ABI for on-prem and Coca Cola both off and on-premise globally. Since the Coke deal was signed, Monster’s sales have increased by +5x, gross margin has expanded from ~52% to over 60% and the EBIT margin has increased ~600bps to ~35%.

We believe that on an overall basis CCE has a very extensive network of salesmen, merchandisers, trucks and a very professional selling organization behind them. They have got very deep reach; they reach a large number of additional accounts that have not really been available to us through an alcoholic system, because in many cases they just do not go to the same universe of accounts. There are many areas in the US where there are dry areas or the alcoholic and non-alcoholic systems do not line up." - Rodney Sacks, Monster Beverage CEO, 2008 

There are five US national beverage distribution networks: Coca Cola, Keurig Dr Pepper, Pepsi, Anheuser Busch, and Molson Coors. These networks all have over 500k accounts across off and on-premise from bars, hotels, to convenience and grocery stores plus access to hundreds of distributors.

While Southern’s provided a moat versus other premium mixers, it lacks reach in off-premise mass merchandisers and c-stores. It also doesn’t have the same leverage as ABI or Molson in the grocery channel.

We’ve been asking for years why Fever-tree didn’t follow a similar strategy as Monster. Especially as Fever-tree does not have operational or distribution expertise internally. We compared Monster and FEVR 3 years ago: 

It’s difficult for Fever to partner exclusively with Diageo or Pernod because the company is brand agnostic and co-markets with every spirit brand. But maybe a distribution agreement with a beer company would work? Or Nestle and Unilever? A company with beer or water distribution seem to be the options - Monster, Fever Tree & Non-Alcoholic Beverage Distribution, IP Research 2022

This interview with a Former Molson Coors Executive, who spent over 15 years at the company and experience distributing its non-alc portfolio, explores how Molson may scale Fever-tree. The interview focuses on how Molson grew ZOA, The Rock’s new energy drink, to $100m sales in year 1. One interesting insight was how ZOA faced challenges after Year 1 because Molson’s distributors had exclusive deals to carry Red Bull.  

Launching an energy drink category through the Molson Coors network was challenging because Red Bull has an exclusivity partnership or distribution deal with beer distributors that sell their product. Outside of the markets where they have their own route to market, there's a barrier of entry within their distributor contract stating that you cannot sell another energy drink as long as you sell Red Bull. - Former Senior Executive at Molson Coors Beverage Company

While FEVR won’t have this challenge, selling premium mixers is more complicated than energy drinks. Energy is a clear category; it has a simple, obvious use case. Premium mixers are more complex; Fever-tree is marketed as a mixer but can also be consumed straight. It's harder to drive consumer trial compared to energy drinks.

How Molson uses its leverage to sell FEVR off-premise is critical for Fever-tree achieving anywhere close to Monster’s returns. This may also encourage Fever-tree to shift towards non-mixer occasions such as RTDs and premium soft drinks. The details of the partnership also seem to reduce the volatility in FEVRs future earning power whilst improving its ROIIC. We plan to speak to 1-2 more Molson distributors to understand the nuances in beer vs spirits distribution. 

Sartorius: Upstream Bioprocessing & Chinese Competition

This interview with a Former Sartorius Senior Executive explores how competition has changed since COVID and the threat of Chinese competitors:

Additionally, competition increased. During Covid, everyone wanted to join the market. Distributors turned into integrators and assemblers, and many companies emerged, building clean rooms for assemblies. Companies also entered customer processes they wouldn't have before. For example, in filtration, a high-margin market, there were three big players, Millipore, Pall, and Sartorius. For the customers, there was no need to add a fourth player. When doing RFQs, they got the best prices, and there is also the additional burden of managing more players. But during Covid, you couldn't get polyethylene filters, so they were forced to validate other customers, such as Meissner, and now these products are validated into the process and their pricing is under further pressure. Furthermore, Chinese companies gained significant market share in China. Sartorius had a good setup there, but the government wants Chinese companies to take over supply. Many biotech companies depend on the government, and these Chinese companies now offer good quality with significant price cuts. This explains the current scenario we are in. - Former Senior Product Management Director at Sartorius Stedim Biotech

Tesla vs Waymo: Autonomous Driving & Long Tail Data

A Former Head of Autonomous Driving at Lyft explores why long-tail data is critical to achieving Level 5 AV:

For instance, we experimented with Nvidia Cosmos, a model designed for robotics. When tested on series collected from, for example, Nexar cameras, it didn't perform well because it lacks sufficient data to train on those scenarios. It struggles to predict the next few frames and doesn't understand physics, creating physically impossible situations like cars passing through each other. This indicates that whoever has access to valuable long-tail data on autonomous driving or driving in general can build or fine-tune a model like Cosmos better. Data is essential. With that in mind, who can win? Tesla is well-positioned because they collect a vast amount of data and have a smart way of collecting relevant data through their shadow mode in cars. I believe it's well known that their cars run the autonomous vehicle stack all the time, whether engaged or not. When there's a discrepancy between your driving and the model's prediction, it's a sign to save that data. Nexar does something similar, using triggers like hard braking or harsh cornering to collect data. Most of the time, it's minor, like hitting the brake for a crossing animal, but sometimes it captures accidents, providing valuable training data. Collecting valuable data, rare events, and long-tail data at scale is the holy grail in autonomous driving and many other fields. Data access is key. - Former President, Level 5 at Lyft

Farfetch and Chinese Luxury Goods Marketplaces

Last week, we published research on the Luxury Gray Market Value Chain that explores how goods flow from brands and boutiques in Italy to Asian Gray Market buyers and marketplaces. In an interview with a former executive from Farfetch China, which accounted for ~25% of GMV, we explore Farfetch's historical challenges and the Chinese luxury market.

As Farfetch scaled, it began to supply product directly from brands. However, brands like Gucci restricted the ability of Farfetch to charge different prices across geographies.

There are two parts to the Farfetch business. From a brand perspective, let's say I work directly with Gucci, and Gucci also allows me to work with its layer of boutiques. The only rule we need to follow when working with a brand is what you mentioned about geopricing. We receive a list of RRP (recommended retail price) from a brand, and worldwide, I can only display the item at this price. We cannot break the brand's pricing rule. - Former Executive at Farfetch China

This immediately hurt GMV growth. The ability of luxury marketplaces to arbitrage pricing between geographies is a key driver of the value proposition.

Demand dropped very quickly. Soon, people on social media started talking about prices. The effect was immediate. In China, many people discuss prices and demand on social media. It was very noticeable, especially with top brands like Gucci and Kering brands. The word spread quickly, and demand shrank rapidly. Initially, it was somewhat manageable because people knew not to buy Gucci on Farfetch but rather OTB brands. At that time, we weren't working closely with OTB brands, so you could still buy Margiela or Marni with a decent discount. Demand was shrinking, but then it worsened because even OTB brands asked for selective distribution. All major brands demanded the same, making Farfetch non-competitive in pricing.- Former Executive at Farfetch China

Although Farfetch had relationships directly with brands, the original source of the product supply was mainly from the Italian boutique wholesalers of the brands. A key risk to all third-party luxury marketplaces is access to supply and the ability to arbitrage geographic pricing.

This interview further explores the reasons for the decline of Farfetch in China and the competitive landscape for luxury marketplaces in China. We aim to publish a post-mortem of Farfetch and the durability of marketplaces like Cettire in the coming weeks.