Lately I keep feeling that if this bull run really continues to push finance on-chain, one of the biggest winners at the end may not be a particular chain, or a particular meme, but Uniswap.
The reason is actually simple.
It doesn't matter who issues an asset — in the end every asset needs somewhere to trade.
Robinhood Chain has already run this playbook once.
Next, on September 16, Circle's Arc mainnet goes live and it will run once more.
If more and more equities, stablecoins, RWAs, memes and even traditional financial assets move on-chain, then what Uniswap really wants may have stopped being "a DEX" a long time ago.
What it wants to be is:
the liquidity layer for on-chain assets.
01｜Robinhood Chain already showed Uniswap what this looks like
Robinhood Chain opened its public mainnet on July 1 this year.
Robinhood's positioning for it was clear from the start: tokenised equities, RWA, 24/7 on-chain trading.
And on day one of the mainnet, Uniswap was already on it.
Not as an ordinary partner, either.
Robinhood officially defined Uniswap as its:
Primary Public AMM.
That is, the chain's main public liquidity protocol. Uniswap's V2, V3, V4, UniswapX, Web App, Wallet and API were all integrated from day one. (Robinhood)
What happened two months later may not have been fully expected even by Robinhood.
What exploded first on Robinhood Chain was not equities.
It was memes.
Especially after Pons arrived, the trading volume of the whole chain was pulled up suddenly.
In early September, Robinhood Chain's weekly DEX volume reached about $12.4 billion, more than doubling week over week; on September 4, daily on-chain fees set a record of roughly $6 million. (The Block)
Pons itself was even more extreme.
On September 2, Pons created close to 25,000 tokens in one day, with 24-hour volume of $544 million and about $5.95 million in fees that day. (CoinDesk)
But here is a detail that is easy to miss:
Pons got famous, and Uniswap quietly ate a big slice of it too.
Because in Pons V2, once a token graduates from the bonding curve it automatically enters Uniswap V4, creating a permanently locked liquidity pool.
Which means:
Pons keeps producing new assets.
Uniswap hosts the long-term trading of those assets after graduation.
This is why, the more I studied Pons V2, the more I felt Uniswap is the infrastructure standing behind it. Pons can issue tens of thousands of tokens a day, but for those tokens to form a real, lasting market they still need Uniswap. (pons)
02｜How much of Robinhood Chain's trading is actually Uniswap?
This is the most extreme piece of data.
In a recent set of Robinhood Chain figures, 24-hour DEX volume was about $1.88 billion.
Of which:
Uniswap V3: about $879 million
Uniswap V4: about $601 million
Uniswap V2: about $51 million
All three versions together:
about $1.53 billion.
So on that data's basis, Uniswap accounts for roughly 81% of Robinhood Chain's DEX volume. (The Pond Street Ledger)
At peak it was even more extreme.
Some statistics show that on Robinhood Chain's most active trading days, Uniswap at one point carried close to 98% of the chain's on-chain DEX volume. (ETHNews)
The more interesting part is looking at it the other way round.
It is not only Robinhood Chain that needs Uniswap.
Uniswap V4 increasingly needs Robinhood Chain as well.
Data from early September showed Robinhood Chain at one point contributed about 56.3% of all Uniswap V4 volume across the network. (crypto.news)
So this is no longer simply:
Uniswap deployed to Robinhood Chain.
It is:
Robinhood Chain has become one of Uniswap's most important new sources of growth.
That is a big difference.
03｜After Pons took off, Uniswap reacted fast
There is another thing here, and it is something I genuinely admire about the Uniswap team: when a competitor suddenly appears, it reacts extremely quickly.
After Pons really exploded, Uniswap Labs did not choose to fight it to the death.
It simply bought PONS.
On September 3, Pons announced:
Uniswap Labs has purchased $PONS for long-term alignment.
That is, Uniswap Labs had bought PONS for long-term alignment between the two sides.
How much was bought, at what price, from where — neither side has disclosed it, so we can't guess at that. (The Defiant)
But the action itself is already interesting.
Because Uniswap has its own launchpad too.
On August 5, Uniswap Labs launched Pools.trade, competing with Pons directly on Robinhood Chain.
In its first days, Pools.trade even created more tokens than Pons for a while.
Then, less than a month later, Pons broke out.
By the end of August, Pons V2 was earning about $4.89 million in fees in a single day, while Pools.trade had roughly $38,600. (The Defiant)
Then Uniswap did something very shrewd:
if you won, I'll buy you.
It reminds me of Pump.fun and Raydium back in the day.
Early on, once Pump.fun's tokens graduated, a large share of their liquidity went into Raydium.
The more popular Pump.fun got, the more Raydium earned.
But later Pump.fun built PumpSwap.
Graduated tokens no longer had to migrate to Raydium — they could go straight into PumpSwap.
That took back the liquidity and volume that used to be handed to Raydium. (Gimg)
So from Uniswap's point of view, it is easy to understand why it wants long-term alignment with Pons.
Today, Pons V2 is:
Bonding curve → graduation → Uniswap V4
Uniswap obviously wants that path to exist forever.
But if one day Pons gets big enough to ask:
why should I keep sending all my graduated tokens to Uniswap? Can't I just build my own DEX?
then the Pump.fun and Raydium story may play out all over again.
So I don't think Uniswap buying PONS should be read simply as:
"Uniswap is bullish on PONS, so it bought the token."
It looks more like:
Pons has become a very important asset-issuance front door on Robinhood Chain, and Uniswap wants that front door tied to its own liquidity layer for the long term.
Of course, "to stop Pons from repeating Pump.fun/PumpSwap" is still only my speculation; neither side has said so.
But from a business-logic standpoint, the possibility is worth watching.
04｜After Robinhood, Arc is next
If it were only Robinhood Chain, I might not think this is such a big deal.
But on September 16 another heavyweight mainnet arrives:
Circle's Arc.
Arc and Robinhood Chain actually have one thing in common.
Neither is a chain built for crypto-native users in the traditional sense.
Robinhood Chain wants to move equities, ETFs and RWA on-chain.
Arc wants to move: stablecoins, payments, institutional capital, RWA, FX and tokenised assets on-chain.
Circle has confirmed that the Arc public mainnet launches officially on September 16. More than 100 institutions and ecosystem projects are already involved in building it, including BlackRock, DTCC and Standard Chartered. (Circle)
And in Arc's liquidity infrastructure, the same name appears again:
Uniswap.
Arc even wrote a dedicated post explaining why it needs Uniswap.
One line in it struck me as key:
Lending protocols, derivatives, and LP strategies all require a deep, trustworthy swap venue before they can operate at any meaningful scale.
In plain language:
lending, derivatives, LP strategies and RWA can only really run at scale if there is a market deep enough to trade in first.
And one of the core liquidity protocols Arc chose is Uniswap. (Arc)
This is also why I started to re-understand Uniswap.
We used to look at Uniswap and see:
a decentralised exchange.
Maybe the angle should change now:
the default liquidity infrastructure for any new asset that comes on-chain.
Robinhood issues equity tokens — they need liquidity.
Pons issues memes — they need liquidity.
Circle builds Arc — it needs liquidity.
BlackRock's BUIDL goes on-chain — it needs liquidity.
Stablecoins swapping between each other — they need liquidity.
More equities, funds, bonds and commodities in the future — still liquidity.
Who issued the asset doesn't really matter.
As long as an asset needs to trade, Uniswap has a chance to take part.
05｜How big is Uniswap itself already?
Now look at Uniswap's own numbers.
Uniswap's official disclosures previously put cumulative protocol volume at about $4 trillion. (Uniswap Labs)
And per the most recent DefiLlama data:
Uniswap:
about $3.35 billion in 24-hour volume
about $19.6 billion over 7 days
about $67.4 billion over 30 days
The whole DEX market did roughly $240 billion over the same 30 days.
Simple math:
67.4 ÷ 240 ≈ 28%
So today, one protocol — Uniswap — accounts for about 28% of all DEX volume over 30 days.
On the latest 24-hour numbers the share is closer to 33%. (DefiLlama)
That is no longer the scale of a niche DeFi app.
More importantly, the biggest difference between Uniswap now and before:
UNI has started genuinely capturing protocol value.
After UNIfication passed at the end of 2025, Uniswap began switching on Protocol Fee and using protocol fees to burn UNI programmatically. Coverage then expanded from Ethereum to Arbitrum, Base, BNB Chain, Robinhood Chain and other networks. (Uniswap Labs)
Robinhood Chain has recently become a very important source of UNI burn funding.
In one set of data from September 3 to September 9:
Robinhood Chain contributed about 71.9% of Uniswap's Protocol Revenue.
On September 4, Uniswap Holder Revenue reached about $1.15 million, of which roughly $925,000 came from Robinhood Chain.
That means about 80% of the value returned to holders that day came from a chain that had been live for just over two months. (The Traders Spread)
This is the change I think is genuinely worth watching.
Before, no matter how large Uniswap's volume was, many people would ask:
what does that have to do with UNI?
Now that question is slowly getting an answer:
trading → Protocol Fee → UNI Burn.
More chains, more assets, more trades — and more fees that can potentially be captured.
06｜Why I think Uniswap may be one of the biggest winners of this cycle
I increasingly like one style of thinking about investing:
don't guess who will strike gold — study who is selling the shovels.
Will Robinhood Chain become the biggest RWA chain?
Don't know.
Will Arc become the main settlement network for institutional finance?
Don't know.
Will Pons beat Pump.fun?
Also don't know.
Which meme goes up 100x — nobody knows at all.
But all of these things share one requirement:
liquidity.
As long as assets come on-chain, they will be traded.
As long as there is trading, there needs to be a market.
Uniswap is working hard to be that default market.
Robinhood Chain is a very good case study:
Robinhood built the chain, Pons issued the tokens.
Users traded like mad.
And in the end Uniswap quietly carried more than eighty percent of the chain's DEX volume.
Even Pons's own success fed assets and volume straight into Uniswap V4.
This is what makes infrastructure the most comfortable position in the game.
It doesn't have to guess which token wins.
The more winners there are, the more it earns.
07｜And let's look at the other token I hold: CAKE
Talking this far, I want to bring up PancakeSwap, another project I hold.
I hold CAKE too, and for quite a long time, so this is not about putting one project down to praise another. But the more I study Uniswap, the more I feel PancakeSwap's biggest problem may not be that its product is weak — it is that it has never really escaped the label BSC gave it.
Strictly speaking, PancakeSwap left BSC a long time ago. It is deployed on Ethereum, Base, Arbitrum, Solana and even Robinhood Chain, and it ships a lot of product. This year PancakeSwap has tried equities, ETFs, RWA, cross-chain trading and perps as well.
But the problem is:
a product leaving BSC is not the same as a moat leaving BSC.
PancakeSwap's strongest ground today is still BSC.
Over the last 30 days, BSC's whole-network DEX volume was about $34.7 billion, and PancakeSwap itself did about $24.5 billion of it — roughly 70%.
That is a genuinely impressive result.
But look at it from another angle and it may also be CAKE's biggest problem.
PancakeSwap is the absolute king on BSC, but outside BSC it is very hard to reproduce that dominance.
Base has Aerodrome and Uniswap, Ethereum is Uniswap's home turf, Solana has its own liquidity system. Even on Robinhood Chain, which PancakeSwap entered this year, the vast majority of volume still belongs to Uniswap.
So my biggest worry about CAKE today is not whether PancakeSwap will die.
I think it almost certainly won't.
Its BSC base is still extremely strong — 35 consecutive months of net CAKE deflation, and as of July this year PancakeSwap's official disclosure put cumulative BNB Chain volume above $4 trillion.
The real question is:
where does its next round of growth come from?
That is also the thing I understand less and less the longer I hold CAKE.
If BSC explodes, PancakeSwap obviously benefits.
But if the next big increment comes from Robinhood Chain, Arc, Base, Ethereum — from equities, RWA and stablecoins going on-chain — can PancakeSwap become the default liquidity layer there the way it is on BSC?
For now, I don't think it can.
Look at Uniswap instead and I see a different route.
It doesn't need to bet on any one chain.
Ethereum has it.
Base has it.
Arbitrum has it.
BNB Chain has it too.
On Robinhood Chain's launch day it was named the primary public AMM.
After Pons exploded, its graduated assets went straight into Uniswap V4.
Now Arc's mainnet is launching, and it appears again in the core liquidity infrastructure.
Even on BSC, PancakeSwap's own stronghold, Uniswap is already taking share. In CoinGecko's recent BSC DEX data, Uniswap V4 + V3 hold more than a quarter of volume.
That is what I find most formidable about Uniswap.
PancakeSwap is rather like BSC's Uniswap.
What Uniswap wants to be is:
every chain's Uniswap.
I hold CAKE myself, so of course I want it to go up.
But if I had to compare the headroom of these two projects over the next few years, I do lean increasingly towards Uniswap.
Not because PancakeSwap doesn't make money, and not because CAKE's tokenomics are bad.
But because:
CAKE's base is strong, yet its growth boundary is easier and easier to see.
Uniswap's base is already strong, and its boundary is still moving outward.
08｜Finally
Uniswap has already moved from a simple decentralised exchange into being
the liquidity layer for on-chain assets.
Crypto used to mainly trade BTC, ETH and memes.
If what really comes next is:
equities on-chain, funds on-chain, treasuries on-chain, stablecoins exploding, RWA on-chain, AI agents trading automatically…
then all of those assets eventually need one place to discover price and exchange liquidity.
Robinhood Chain has started proving this logic.
After Arc's mainnet launches on September 16, we will soon see a second case.
If more and more things start going on-chain, who is providing their trading infrastructure?
For now, Uniswap is in a very comfortable position.
Others create the assets.
Others manufacture the hype.
Others attract the users.
Uniswap collects the toll.
That may be the single most worth-studying business of this bull run.
One disclosure here: I hold both UNI and CAKE, and have held them for more than four years.
So this article is not about trashing CAKE to hype UNI, and it is not a justification for my own positions.
Quite the opposite: because I bought both, and because I have been studying Robinhood Chain and Pons and now see Arc about to launch, I feel more and more clearly that PancakeSwap's base remains very strong, but Uniswap's headroom may be larger.
In my view CAKE's biggest problem is not that PancakeSwap doesn't earn, or that it will disappear — it is that its strongest moat is still concentrated on BSC.
Uniswap is doing something else entirely: no matter whether the next breakout is Robinhood Chain, Arc or some other chain, it wants to be the default liquidity infrastructure for the assets that come on-chain there.
Of course, a protocol doing well does not mean the token price must rise. Whether market share and protocol revenue keep converting into value for UNI still needs watching.
I hold both tokens, so the views here are inevitably shaped by my own knowledge and my positions.
This article simply records my current research and judgement on Uniswap and PancakeSwap, and is not investment advice at all.
If any data or understanding is off, please leave a comment and correct me.
After all, you can buy the wrong token — but you can't keep having the wrong logic. 😂
Thanks for reading this far, and discussion is welcome.
九哥 | lgk