If we want to make sense of the rise and fall of major platforms, the collapse of FTX, Bybit’s $1.5B, Hyperliquid’s HLP exploit, and the constant emergence of new perp DEXs and exchanges, it is more efficient to interpret them through a specific framing.
In this research note, I use what I believe is the most important dividing line: Chinese exchanges vs U.S. exchanges.
The Core of Exchange Risk: Liquidity
Most structural problems at crypto exchanges ultimately come down to liquidity, and fundamentally, that means Market Makers (MMs).
Market makers can broadly be divided into external market makers (External MM) and internal market makers (Internal MM). Here, “market maker” refers to agents that provide liquidity and populate the order book, and does not include MM groups whose sole purpose is to drive/pump specific coins.
Structural Risks of Crypto MMs
In crypto markets, market makers take the other side of user flow via the order book and hold positions (open exposure). As a result, MMs are structurally exposed to crisis conditions at all times, and they need to manage this exposure quickly and strategically.
In the early days, fee models were designed largely as a compensation scheme for MM difficulty and risk. Out of a total 5bp fee, 1.5bp would be carved out first for MMs, and the remaining 3.5bp of net fee would then be shared with referrers and BD partners at some percentage n. In other words, the idea was: “secure liquidity first, and run sales/marketing with whatever remains.”
Over time, however, it became obvious that the real drivers of exchange revenue and growth are referrers and BD partners, and their dissatisfaction grew. So the structure shifted: first allocate a certain portion (e.g. 50%) of the 5bp pool to referrers and BD at rate n, and only then charge the MM fee of 1.5bp out of whatever net fee is left, and only to partners whose contribution justifies additional economics.
In short, a fee model that started out as “MM-first” gradually rearranged into a structure that is effectively “volume and revenue partner-first”.
The problem is that crypto MMs have repeatedly taken massive losses during several episodes each year. When the market flips risk-on and volatility spikes, MM willingness to provide liquidity drops sharply, exchange liquidity becomes scarce, and this lack of liquidity triggers a chain reaction of damage across the system.
On the surface, the October 10th event was “just” a tariff shock, but in practice it produced the single largest liquidation event in crypto history. Combined long and short liquidations reached $19.1B, and nominal liquidation notional is estimated to have been in the $30B–$40B range.
At that moment, large and small market makers alike suffered sharply, as extremely rare ADL (auto-deleveraging) sequences were triggered on multiple exchanges, and deposits/withdrawals were halted on several platforms right when hedging routes were most needed. According to market commentary at the time, even one of the biggest crypto MMs, @wintermute_t, temporarily paused MM operations during the event.
Internal MM: Direct Exchange Intervention
As a result of these dynamics, exchanges have steadily become more dependent on internal market makers (Internal MM), and the quality of their internal risk management has become increasingly important.
When an exchange handles liquidity internally, it can broaden the scope of what it can guarantee in a black-swan scenario via its own liquidation rules and backstops. At the same time, it can reduce the cost of paying external MMs and redirect that budget into marketing and sales, effectively increasing market exposure and growth potential.
Current perp DEX vaults can also be seen as a form of Internal Market Making. The difference is that they are opened publicly to retail capital, and liquidity is filled with user assets, with rewards configured on top. This makes the rules and skill with which the vault is operated extremely important.
The biggest problem of Internal MM, however, is that risks arising inside the exchange are not externally auditable. The exchange can ignore its own liquidation rules, construct the order book arbitrarily, and drift into B-book “casino” operations off the public ledger.
If losses pile up in a major downturn, this can eventually lead to another FTX-style collapse. This is precisely why legal and regulatory pressure around internal dealing is so strong in the United States.
Is Pure A-book the Answer? Structural Limits
So is a fully pure A-book model—where all trades are filled from external liquidity—the correct answer? In practice, building a true A-book in crypto is close to impossible.
There are too many coins, and for the majority of them, long-term custody is economically meaningless.
To build a true A-book like equities, global liquidity would need to consolidate into a single exchange, and on top of that you’d need a legal B-book prime brokerage layer to support institutional trading and inventory management.
Chinese Exchange Advantage: Capital Networks and Liquidity
In the end, the difference between exchange “camps” is most clearly visible in how they fill liquidity and how they create winning platforms.
Many people inside the industry, especially those close to exchange operations, implicitly or explicitly place Chinese exchanges ahead of U.S. exchanges in the pure “exchange vs exchange” game.
That is because the “쩐” (deep capital) of early-surviving Chinese players built massive networks that operate largely independent of market conditions. Over time, this capital started to determine liquidity, market making, and even the hierarchy of new exchanges.
In a world where a model like MEXC—low fees, fast listings, relentless campaigns—continues to survive and grow, the natural question is: “Can a similar new exchange even realistically emerge again?”
During Bybit’s $1.5B exploit, the fact that Binance, Bitget, and Gate each mobilized hundreds of Million dollars in support within hours is not simply an example of competitors rescuing a rival.
It’s more accurate to say: Their “쩐”(deep capital) is already structured to intermix when needed.
The Chinese camp isn’t optimizing for perfect, granular compliance. It optimizes for fast liquidity, fast execution, and making the exchange work at scale. They have the capital and operating experience to do this.
@edgeX_exchange : Chinese-Style Execution + Structured Risk Design
In this context, edgeX is not just another new perp DEX. It is closer to a live implementation of the Chinese-style liquidity model and Internal MM architecture, wrapped in a product.
edgeX keeps Internal MM capabilities in-house and takes direct responsibility for liquidity during market moves. At the same time, it focuses relentlessly on what users actually feel:
spread quality
execution speed
mobile experience
community and metrics
In short, it uses capital and network to maximize execution speed and product quality. For now, perp DEXs with minimal compliance constraints remain the ideal stage for this type of strategy.
A deeper dive on edgeX will follow in a separate piece. If you want to catch it in time, follow @reboundx_net
and keep an eye on our updates.
U.S. Exchange Approach: Innovation + Compliance-First Liquidity Experiments
The U.S. camp attacks the same problem from a different angle: they lean heavily on innovation and technology, within the boundaries of compliance.
@base Super App
@Lighter_xyz 's Zero-Fee model
@variational_io 's RFQ architecture
@Backpack 's token / IPO / super-app build-out
All of these are examples of U.S.-style liquidity experiments: directionally compelling, but slow and heavily gated by regulation.
@variational_io : RFQ + OTC Liquidity layer
Variational’s RFQ is not about “convincing users to use one specific perp DEX on-chain.” Instead, it is a design that attempts to pull all fragmented liquidity in the opposite direction, into a unified RFQ layer. Structurally, it is very different from the traditional crypto exchange model.
@Backpack : compliance-driven liquidity exchange , U.S styled
Backpack’s relatively slow liquidity ramp also makes sense in this framework.
Its core builders and legal team come from FTX/Alameda and carry a very real trauma and aversion to the way SBF moved funds and ran internal books.
They know better than anyone:
what precisely went wrong with Internal MM at FTX, and
what the exchange’s standing in the U.S. might look like today if those mistakes had not been made.
Because of that, they are deliberately avoiding the same trap. Instead of giving an MM excessive equity or control over the exchange, they are focusing on A-book discussions and vault-based liquidity structures.
Even within vaults, the question of public vs private design is being worked through with compliance in mind. All of this is happening while they try to build an exchange that can realistically scale up to a CEX-level product on top.
Onboarding Traditional Assets: U.S.-Led Long Game
Ultimately, the side that is most likely to bring equities and Stocks properly into crypto is the U.S. camp.
Discussions around tokenizing unlisted stocks and real equities are already happening within U.S. institutional networks, and there is a long-term preparation underway to pull traditional assets into the crypto rail.
The SEC has approved a service from @The_DTCC (the U.S. Depository Trust & Clearing Corporation) that allows equities, ETFs, and U.S. Treasuries to be tokenized and traded 24/7 in on-chain form. In other words, new services enabling on-chain tokenization of real-world assets are scheduled to begin rolling out from the second half of 2026.
For perp DEX investors and exchange builders, this is not a side show. It’s a preview of how far the “U.S. compliance + innovation” model is willing to go.
Shifting Exchange Narratives: The End of the Altcoin Era
While exchange camps are busy setting the board, the altcoin era is quietly coming to an end.
It’s not just that we’ve “run out of research material.” The deeper issue is that it is getting harder and harder for new projects to offer genuinely new puzzle pieces to the market at the pace we saw in prior cycles.
Historically, the direction of travel is clear. Prediction markets, equities, funds, and chain infra tend to be absorbed into exchanges. This is where super apps and super chains come from: they are not marketing slogans, but the endpoint of consolidation.
Perp DEX Chains: CEX/DEX Hybrid Alignment Attempts
Perp DEX chains sit in the same flow.
@Gate's acquisition of @bugscoin_bgsc ,
@binance's push behind @Aster_DEX ,
@Bybit_Official's renewed backing of @OfficialApeXdex
all point in a similar direction:
These players are effectively acknowledging that the “on-chain royal road” is perp DEX chains plus a CEX/DEX hybrid model, and they are trying to align their organizations and products with that thesis.
At the same time,KRAKEN(@krakenfx) , which already operates a centralized exchange, is positioning Nado perp dex(@nadoHQ) as the killer app of INK Chain(@inkonchain), and timing its chain valuation push to match an IPO window. Structurally, this is the same play: on-chain value accrual wrapped around an existing CEX.
Remaining Question: Separating the Roles of Exchange, Chain, and Token
For all of this to truly “line up,” we still need time and many more live cases.
Key questions remain open, for example:
How should we conceptually separate BNB Chain’s $BNB, originally built by Binance, from Aster Exchange’s $ASTER as an asset?
Or in the Bybit side, how should we distinguish Mantle Chain’s $MNT, a DeFi/L1 token, from ApeX perp chain’s $APEX, the perp chain token?
How much of $INK will be allocated to Nado, the official killer dApp of Kraken’s INK chain, which is expected to IPO at a $20B valuation, and what kind of utility and valuation link will INK have in relation to Nado?
The consensus on how far we should separate the roles and value accrual functions of the exchange, the chain, and the token will likely be forged in the collision between Chinese-style capital-driven models and U.S.-style compliance-driven models.
Visit our website and try our ReboundX rebate system: https://www.reboundx.net/
📌Explore our website, which provides daily rebates to users worldwide.