We like to think we make rational decisions when choosing a platform to do business on. We compare fees, read reviews, weigh features, and choose the one that best meets our needs. But when you need to exchange currency at the airport, that's not the case. You head to the nearest currency exchange counter because it's right there and the sign looks legitimate. Convenience makes the decision for us. This is true for nearly every market we build. The trading floor's success isn't due to better prices or more advanced technology, but because it's the agreed-upon meeting place. Nasdaq's success in attracting IPOs from tech and emerging consumer companies isn't because it's technologically superior to the New York Stock Exchange, but because it positions itself as the place where growing companies and their investors agree to meet. eBay defeated many early auction sites not because its software was superior, but because buyers and sellers agreed to trade there. Every market starts out empty. Participants don't come because there are no trading opportunities. On exchanges, this manifests as empty positions, price volatility, and poor trade execution. The platform that first solves the "chicken and egg" problem often wins. The second-place platform often fails or remains in second place forever. We've seen a similar phenomenon in the cryptocurrency market. Decentralized exchanges (DEXs) promised a world where anyone could open a market, with transparent fees and no centralized control. This model worked for a while, but then it failed. In September, Hyperliquid, the dominant DEX in perpetual contract trading, began losing market share. It wasn't the gradual erosion typical of a giant. A new platform called Aster, backed by Binance founder Changpeng Zhao, rose to prominence almost overnight. However, data tracking platform DeFiLlama completely removed Aster's data, claiming its trading volume appeared to be fake. The fake data isn't even the most concerning aspect. Ironically, this occurs despite the fact that every cryptocurrency transaction, wallet, and fee is traceable. Disturbingly, it's nearly impossible to distinguish between an exchange that's naturally popular and one that's simulated enough popularity to attract real users. The infrastructure is identical, and the order books look identical. The only difference is whether the user base is organic or artificially manufactured. This is a story about this tension: how DEXs create virality, the difference between momentum and the market, and what simple tests can reveal which exchanges are worth your next trade. Let's take a step back. September 2025 marked the first month that trading volume on perpetual futures decentralized exchanges (Perp DEXs) surpassed $1 trillion. This wasn't annual volume, but monthly volume. From July to September, Perp DEXs' total trading volume approached the entirety of 2024. Until May 2024, these platforms' trading volume was mostly less than 10% of that of centralized exchanges (CEXs) like Binance. By September 2025, this proportion had reached 20%. Today, one in every five dollars of perpetual futures trading is conducted on transparent, auditable infrastructure. While early Perp DEX leader Hyperliquid sparked the craze, it was new entrants like Aster and Lighter that drove it. Three factors collided. Regulatory pressure made traders concerned about the custody of centralized platforms. With the advent of sub-second execution speeds, well-designed mobile apps, and interfaces like Binance, token economics gradually matured. Token economics gradually evolved into a true profit machine. Convert transaction fees into token buybacks, and suddenly you have a sustainable business model: greater trading volume increases the value of the token, attracting more traders and generating more fees. Not all platforms can keep up with exponential growth. Cracks emerge, revealing those building for sustainability and those simply riding the wave. When everyone's volume is rising, every platform appears successful. But when the incentives run out, differences become apparent. Take Hyperliquid, for example. Launched in 2023, it is a custom-built Layer 1 blockchain purpose-built for trading. For most of 2025, it dominates the market, processing between $175 billion and $400 billion in transactions per month. The platform gave away 27.5% of its HYPE tokens to its 94,000 users and refused to accept venture capital. This move gave users ownership, rather than diluting their equity through internal sales. This kept them coming. Aster then launched in September and immediately exploded in growth, reaching $420 billion in trading volume that month. The token's valuation soared from $170 million at launch to $4 billion at its peak. Hyperliquid's market share plummeted from 45% to 8% in a matter of weeks. Its modus operandi involved a massive airdrop program. The second phase alone distributed 320 million tokens, with a peak value of $600 million. This encouraged traders to increase their trading volume, hold tokens, refer friends, and accumulate points. The strategy worked, with trading volume surging. For a time, Aster seemed unstoppable. Then, it vanished. You won't find these numbers on DeFiLlama anymore, as the platform has delisted Aster, accusing it of falsifying its data. Today, Hyperliquid's market share on Perp DEX has rebounded to 28%, but it's still less than half of what it was two months ago. Lighter follows closely behind, with a 25% market share. This incident prompts us to consider what allows some platforms to survive while others fade away. I've categorized them into four categories.
First, liquidity—it's like gravity. Without deep liquidity pools, traders face slippage, the gap between their expected price and the actual execution price. Hyperliquid built its own Layer 1, achieving a processing speed of 20,000 orders per second and 0.2 millisecond final confirmations. Lighter uses ZK-rollups to achieve matching speeds under 5 milliseconds. But technology alone can't solve the cold start problem: you need market makers, and without traders, market makers won't come. Without liquidity, traders won't come either.
Hyperliquid solves this problem with its HLP pool. The HLP pool is a protocol-owned liquidity pool with an annualized yield of 6% to 7%, providing foundational depth before market makers arrive. On October 11th, US President Trump announced a 100% tariff increase on China, escalating trade tensions with the country. Cryptocurrency markets saw $19 billion liquidated in 24 hours. At the time, Hyperliquid's HLP vault saw a profit of $40 million from liquidations in a single day, representing an annualized return of nearly 190%. Most platforms never address this liquidity issue. They launch with cutting-edge technology, but their order books are empty. No one wants to trade because there's nothing to trade with. Next comes incentives. They can at least temporarily create a virtuous cycle. Sustainability is determined by the difference between sustainable incentives and one-time, expensive giveaways. Hyperliquid's model addresses this by distributing ownership widely and then using 93% of transaction fees to buy back tokens. This way, the token's value is directly tied to protocol usage, not future mining expectations. Aster airdrops tokens to users and hopes they stay. This approach effectively generates trading volume. This has indeed increased trading volume in the short term. We can already see long-term trends in real time. Third, user experience. User experience determines user retention. If your DEX experience isn't as good as Binance's, users will eventually leave. Hyperliquid's interface might mistakenly resemble a centralized exchange even for novice traders. EdgeX has launched a multi-party computation (MPC) wallet, allowing users to trade without managing seed phrases. Lighter charges no fees for retail traders. It's these small details that determine whether traders stay or leave. Think guaranteed stop-losses, sub-accounts for risk isolation, and user-friendly mobile apps. Fourth, there's the power of culture. Memes and influencers build communities. Hyperliquid's narrative is one of high-performance DeFi first, community-owned, no venture capital, and built from the ground up. Its recent airdrop, which rewarded early backers with a Hypurr Cats NFT collection, demonstrates how culture can build a loyal community. Even in Aster's case, the "CZ's Revenge" meme, spawned by Binance building infrastructure beyond the reach of regulators, paved the way for its sudden surge in adoption. Lighter, positioned as the "savior of Ethereum's perpetual swaps," boasts backing from a16z and former Citadel engineers. Cultural dynamics are crucial, as a sense of belonging is equally crucial to a loyal community of cryptocurrency traders. In the cryptocurrency world, users don't just choose trading platforms. They also come to identify with them, maintain them, add tokens to their Twitter profiles, and participate in Reddit and Discord channels. This community culture creates organic marketing and user engagement. Effective platforms meet all four of these conditions. Aster took shortcuts, relying too heavily on memes and incentives. When questions arose about the authenticity of its liquidity, everything began to falter. A chart posted by the founder of DeFiLlama shows that Aster's XRP and ETH trading volumes are almost identical to Binance's perpetual contract volumes. Hyperliquid's trading volume, by contrast, fluctuates independently. A reliable metric is the ratio of trading volume to open interest. Open interest measures the amount of money actually at risk in open contracts. If $10 billion is traded but only $250 million is locked up, there's a problem. Hyperliquid’s ratio is just over 1, indicating real open interest, while Aster’s ratio is as high as 20, suggesting a significant discrepancy between trading volume and open interest value. As a practitioner in the cryptocurrency field, I constantly think about the difference between real growth and artificial growth. Every booming market in history has been artificially engineered to some extent. Nasdaq was originally designed to compete with the New York Stock Exchange by allowing smaller companies to go public. eBay was built to address specific trust issues. Once critical mass is achieved, growth may appear organic, but it's actually a result of deliberate design. Cryptocurrency simply makes the engineering process visible in real time. You can watch as platforms inject liquidity, incentivize users, iterate on features, and ultimately achieve product-market fit—or fail in their attempts. Aster is still operating. Its second phase airdropped 320 million tokens with no lock-up period, increasing selling pressure. Aster's volume/open interest ratio of 20 already indicates that most of the volume isn't actually open interest. A more important lesson from Aster's case is that when markets operate on transparent tracks, anything is possible. For years, we've pretended that markets are neutral infrastructure, unrelated to the money flowing through them and who controls access. But markets are always inextricably linked to the systems that run them, and those systems are always tied to who controls them. We simply couldn't see that mechanism, couldn't understand whether it mattered. Now, we can finally see it for ourselves. We can verify whether liquidity is real or just bots buying and selling. We can observe whether platforms are building sustainable businesses or executing elaborate exploitative schemes. Platforms with smooth transactions, ample liquidity, minimal slippage, and interfaces so intuitive that you forget you're using cryptocurrency are likely to succeed. In the long run, they could become invisible infrastructure like VISA and Mastercard. We're clearly not there yet. But we can clearly see the elements of a successful DEX. This allows us to ask the right questions to discern which platforms are on their way to becoming invisible and which are just there for show. Over the next few years, we'll be able to separate them.
Jasper
Joy
Aaron
Jixu
Davin
Hui Xin