One of the key metrics for the business is RLDC Margin.
RLDC is calculated as Total Revenue and Reserve Income less Total Distribution, Transaction, and Other Costs.
It's an important metric because it shows how much revenue the company actually keeps after sharing and operational expenses. Circle pays distribution costs primarily to its distribution partners, with $COIN being the major beneficiary.
In FY2025, $CRCL posted an RLDC margin of ~40%.
This implies about 60% going to partners. A huge portion goes to $COIN as a result of a restructuring partnership deal in 2023.
As part of the deal, $COIN acquired an equity stake in $CRCL while the following terms were agreed upon:
1. $COIN keeps 100% of the reserve income from USDC held on its platform 2. $CRCL keeps 100% from USDC on its own platform 3. $COIN & $CRCL evenly split any remaining reserve income held elsewhere
~23% of USDC balances are held within the $COIN platform while a minuscule amount is held on $CRCL. Hence, $COIN is a key partner for $CRCL that will be closely tied to $CRCL's fortunes in the coming years.
I'm currently working on a deep dive piece for $CRCL and will publish at the usual location, link in bio if you're interested.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
I believe this could be the greatest contributor to Grab’s bottom-line within the next 2 years. 🧵👇
1/ GrabAds Today
GrabAds currently has 191,000 monthly active advertisers on its platform, up 49% YoY.
Average spend grew 30% YoY, driving Ad revenue to make up 1.7% of Deliveries GMV as of Q1 2025.
Grab’s ad product breadth is extremely wide, ranging from in-app sponsored listings, search, feed video, display, vouchers to offline formats (car wraps, in-car screens etc)…
Grab keeps Ads revenue inside the Deliveries P&L for now, but management has called it out as the primary driver of Deliveries margin expansion.
2/ How does GrabAds work?
Firstly, you need to understand that Grab’s merchant base is overwhelmingly made up of micro, small and medium enterprises (MSMEs)
These are hawker-stall owners, family bakeries, neighbourhood minimarts.
They are not savvy at all.
Hence, $GRAB has built in an incredibly simplified system, allowing merchants to create ads within 3 minutes with built-in automation to avoid complexity.
$GRAB COO Alex Hungate just spoke at the Asia Economic Summit.
He spoke extensively about $GRAB.
Here are the 5 most important takeaways for investors: 🧵
1) Grab engineers stopped shipping for 9 weeks, costing the business $100M
Grab instructed its engineers to pause “business-as-usual” operations for a 9 week period last year, dubbing the period a “Generative AI sprint”
Engineers stopped shipping features and instead focused solely on experimenting with AI.
“It’s kind of a crazy thing to do for a company of our scale”
“The cost of that is probably just our engineering time for nine weeks - about US$100 million.”
2) Drivers as Data Collectors
When Grab first tested OpenAI’s base models, transcription accuracy for Singapore accents and points of interest was just 46%. Since then, accuracy has risen to 89%.
All of this has been possible due to Grab’s close partnership with OpenAI and Anthropic. In exchange, the two LLM giants gain access to Grab’s extensive data such as the 16,000 voice reports it receives a day.
“Those guys love it, because very few people have so many multimodal use cases with so much data being generated” - COO Alex Hungate