If you asked me this question a year ago I would say, without a question, DEXs. My reasoning was simple — switching costs are much higher across DEXs. If I switch from Uniswap to some longer-tail DEX, I am guaranteed to experience more slippage and worse net execution.

Conversely, I will get the same execution on 0x versus some other long-tail front-end. Both front-ends are just different lenses through which the same DEXs are viewed on the back-end.

Therefore, given their liquidity network effects, DEXs have historically been valued at a premium relative to front-ends. And intuitively, this makes sense. DEXs are a much more defensible business model at scale.

However, this may be changing.

Jupiter – Solana’s leading DEX aggregator – is now valued at 8.7 times Raydium, despite capturing a fraction of the value.

In A New Framework For Understanding Moats In Crypto Markets, I laid out the bull case for why front-ends are positioned to increasingly capturing value. The front-end thesis can be decomposed to three sub-theses:

(1) – Liquidity is more of a commodity than you think: While liquidity is “unforkable” by nature, it is not immune to subsidization. There are numerous precedents throughout DeFi’s history that seem to support this logic (e.g. SushiSwap vampire attack). The structural instability in the perps market also reflects the inability for liquidity alone to serve as a sustainable moat.

(2) – Front-ends are evolving: Today, the most popular “aggregators” are now intent-based front-ends. These front-ends outsource execution to a network of “solvers” who compete on giving the user best execution. Some intent-based DEXs also tap into off-chain sources of liquidity (i.e., CEXs, market makers). This allows these front-ends to circumvent the liquidity bootstrapping phase and immediately offer competitive and oftentimes better execution for “shorter-tail” pairs. AI agents may be the logical evolution to solvers (more on this).

(3) – Front-ends own the end-user relationship: Downstream of owning the end-user relationship, front-ends have disproportionate bargaining power. This can enable front-ends to either reach exclusive deals or subsequently vertically integrate. By using their intuitive front-end and ownership over the end-user as a wedge, Jupiter is now the four largest perps DEX across all chains. Additionally, Jupiter has also successfully integrated their own SOL LST. Given its proximity to the end-user, JUP’s premium feels at least somewhat justified