The B Side of bStocks: Competing for Pricing Power with Perp Instead of Being a Better Nasdaq

Original Title: "The B Side of bStocks: Competing for Pricing Power with Perp Instead of Being a Better Nasdaq"

Author: danny

Lu Xun once said: "Do not strive to be better, but strive to be different." If the ultimate goal is to seize the pricing power of the underlying assets, then do not confront the strongest place of Nasdaq head-on, but rather move the battlefield to Perp, where one excels, with 24/7 trading, leverage, on-chain inventory, and decentralized wild/regular market makers.

Seize the pricing high ground with Perp, and use bStocks to take on inventory and correct errors ------ seize the pricing power.

When the Chicago Board of Trade (CBOT) was established in 1848, it was not yet the huge futures market (CBOE) it would later become. Chicago was becoming the distribution center for grain in the American Midwest, with railroads and canals bringing in boatloads and truckloads of wheat and corn. The problem was that wheat is not a stock. Although it is all called wheat, different farms and different years produce fundamentally different products. If buyers do not even know what they are purchasing, it is difficult to form a unified market price. One of the first things CBOT did was to organize this chaos into something tradable and standardized. After obtaining authorization from Illinois in 1859, CBOT could set grain grades, which were determined by designated inspectors; by 1865, margin and delivery rules began to be institutionalized.

This may not sound like financial innovation, but rather like warehouse management, yet this warehouse changed the structure of the market.

Once the grain entered large elevators, there was no longer a need to ask which farm a bag of wheat came from. The market began trading according to standardized grades, such as a certain standard grade of wheat. The Federal Reserve noted in reviewing this history that this grading and standardized warehousing allowed buyers to know what they were purchasing, reduced transaction costs, and created conditions for a liquid market.

Then came futures.

Farmers could sell their future harvests in advance, grain merchants could buy future goods, and speculators could bet on future prices without having to move thousands of bushels of wheat home. CBOT gradually developed standardized futures from early forward and "to-arrive" trades, and by the latter half of the 19th century, Chicago's grain futures began to take on the functions of price discovery, risk management, and public quoting.

There is an easily overlooked point here.

The fastest runners and the largest trading volumes are in futures, but what prevents futures from becoming a gamble detached from the physical commodity is the warehouse behind them.

Background

On a weekend in August 2026, the total trading volume of NVDA, TSLA, SNDK, and SKHY perpetual contracts on Binance reached approximately $461 million. On Sunday, the prices of the four Perps were all higher than the previous cash market close, but on Monday, after the U.S. cash market opened, all four opened lower. Looking specifically at SNDK, that weekend saw about $338 million in Perp trading, with a Friday Cash Close of $1,212.21, trading at $1,223.98 on Sunday, but opening at $1,203.41 on Monday.

A month earlier, during the long weekend of Independence Day in the U.S., the same SNDK provided another answer. The cash market closed at $1,745, but on Sunday, Binance traded at $1,841.88, and Monday's Cash Open was $1,828.68. After the cash market reopened, there was about a 4.8% gap, with Binance's weekend price covering most of this increase, just overshooting the price.

The same product, once trading out the Monday gap in advance, and another time reversing direction after several hundred million in trading. This issue pushed the problem of 24/7 stocks from "extended trading hours" to market structure: trading volume does not equal pricing power.

If Binance's goal is simply to allow users to buy stocks on Saturday, extending trading hours would suffice. However, looking at Direct Stock, TradFi Perp, and bStocks together reveals a different structure. After the cash market closes, Perp first takes on trading direction, leverage, and high-frequency trading, while bStocks provides stock inventory that can be held, transferred, mortgaged, and hedged, with on-chain protocols bringing capital from outside centralized exchanges. When the traditional stock market reopens, we can see if the prices formed by this 24/7 market are accepted by the U.S. stock market.

This article does not discuss whether bStocks is another form of tokenized stock, but rather another matter: if Binance wants to compete for the first version of prices after traditional stock markets close, why should the engine use Perp, and why would bStocks become the inventory and error correction layer?

1. Binance is not competing for just two more trading days, but for prices after the stock market closes

The stock market has a temporal gap. After the New York market closes at 4 PM on Friday, companies may still release news, macro policies may change, and new events in the supply chain may occur; wars and politics do not wait for Nasdaq to open. Capital is still re-evaluating these stocks, but these judgments cannot temporarily enter the official order book of U.S. cash stocks.

So the question is not about the lack of information, but rather where to express that information (using positions) after it comes out. If a market can take on stock risks while the NYSE, Nasdaq, KRX, and Hong Kong Stock Exchange are closed, it has the opportunity to form the next price discovery before the underlying stock market opens.

Binance's three products happen to divide into three roles. Direct Stock is responsible for real securities, corporate actions, and traditional market interfaces; bStocks turns stocks into assets that can be held, converted, transferred, and mortgaged; TradFi Perp is responsible for long and short positions, leverage, and continuous trading.

If it wants to be the source of price discovery, the most suitable front-runner is not bStocks, but Perp. The reason is not complicated: the first step in price discovery is to let opinions enter the market, and the cost of expressing opinions through derivatives is lower than that of spot.

2. Why is Perp more capable of achieving price discovery?

Taking NVDA as an example, during a sample period from July 17 to 22, the nominal trading volume of NVDAB Spot was approximately $4.21 million, while NVDAUSDT Perp was approximately $418 million, a difference of about 99 times. Looking specifically at the weekend, on July 18, Perp was about 37 times that of bStocks, and on July 19, about 68 times.

Here, we are comparing two products within Binance, not Binance Perp and Nasdaq NVDA spot. The trading volume of U.S. cash stocks is still higher than that of Binance stock Perp. The significance of this comparison is to see where new directional orders are more likely to flow after the cash market closes.

After buying the spot, one can hold it for several months, while Perp continuously opens and closes positions, reverses, adjusts leverage, does basis trades, and collects funding, with market makers constantly re-hedging. The same amount of capital can contribute multiple gross turnovers in Perp. The difference is even greater for shorting: if NVDA has negative news on Saturday, those without NVDAB inventory wanting to short in the spot need to first handle borrowing; Perp only needs to sell to express direction.

Thus, the first price path after the market closes can be written as Information → Perp → Candidate Price. The role of Perp is to produce candidate prices, not to guarantee the correctness of those candidate prices. This distinction determines the significance of bStocks' existence.

3. Research Methodology: Comparison of 25 Instances of "Market Closure - Reopening"

To observe how this price is formed, we collected and organized trading data for NVDAUSDT, TSLAUSDT, SNDKUSDT, SPCXUSDT, and SKHYUSDT. The U.S. market has 25 instances that can fully correspond to the next Cash Open, with NVDA, TSLA, SNDK, and SPCX each having 6 instances, and SKHY having 1 instance.

Each sample starts from the previous Cash Close, passing through Pure Weekend, Sunday Price, Monday Premarket, Opening Auction, and finally to the next Cash Open. Sunday 19:59 ET is used to observe prices before the regular U.S. stock Premarket begins, and after 04:00 on Monday, the Premarket resumes, with Opening Auction information entering after 09:25, and finally taking the official cash opening as the verification point.

This batch of data answers four types of questions:

1. In terms of price, how far the weekend quotes are from the next cash opening;

2. In terms of liquidity, how much actual orders the screen prices can withstand;

3. In terms of trading volume, how much money the market turned over;

4. In terms of trading frequency, whether these trades resemble independent views of investors or are formed by program trading, market making, and HFT creating a high-turnover market.

A market having a price does not mean that the price can be executed; having trades does not mean that the trades come from independent judgments; having liquidity does not mean it possesses price discovery capability.

4. Price Survey: Converging from 181.9bp to 11.3bp

Here, the "error" refers to the absolute distance between the Binance Perp price at a certain point and the next Cash Open, uniformly converted into basis points (bp).

The result on Sunday was 13/25, or 52%. Among the 25 samples, 13 times the direction was consistent, and 12 times it was inconsistent, showing no identifiable difference from the 50% random directional judgment, so this batch of data cannot prove that Binance has a stable Weekend Discovery.

At 09:29:59, the 11.3bp cannot simply be regarded as Binance's ability to "predict Monday's opening." By this time, the U.S. premarket has already been trading for several hours, and information from the Opening Auction has entered the market. If the Perpetual (Perp) significantly deviates from the premarket stock quotes, it will create trading opportunities.

Therefore, the drop from 181.9bp to 11.3bp indicates that Binance Perp is gradually approaching the next Cash Open in the process of re-launching in the traditional market, but it does not imply that Binance is leading the cash market. To answer this question, we need to analyze Binance Perp, bStocks, and the last, bid, and ask prices of the U.S. premarket at the same second to establish a lead-lag relationship.

If the premarket moves from 100 to 105 first, and then the Perp follows to 105, that is a follow; if the Perp reaches 105 first and the premarket then adjusts towards it, that signifies price discovery and leadership.

V. The Price Gap Before Opening is Not a Straight Line: SNDK Was Only 2bp Apart at 09:25, but Diverged After Five Minutes

During the long weekend of Independence Day 2026 (July 3-5), the paths of NVDA, TSLA, and SNDK Perps from Sunday to the Opening Cross were different.

From this perspective, NVDA appears to be on a correction path. On Sunday, it was at $198.35, while the Cash Open ended at $194.42, a difference of about 202bp; it returned to $194.97 at 09:25, and at 09:29:59, it was $194.55, reducing the error to 6.7bp.

TSLA's direction has been correct since Sunday. The Cash Close was $393.45, Sunday was $399.40, and the Cash Open was $397.50. The market anticipated an upward direction, but the Sunday movement exceeded the final Opening Gap.

SNDK's path illustrates the issue further. At 09:25, the Perp was at $1,828.97, and the Cash Open ended at $1,828.68, a difference of less than 2bp; by 09:29:59, the Perp dropped to $1,825, widening the error to about 20bp.

This indicates that prices before the opening do not move along a straight line towards the final opening price. Premarket transactions, NOII, market maker inventories, and orders in the last few minutes can all alter prices. To study who holds pricing power, one cannot only look at the two time periods of Sunday and Monday at 09:29:59, but must consider the entire lead-lag path.

VI. Perp is Not Necessarily Closer to the Next Open than bStocks

Another set of low-frequency samples used the UTC end-of-day prices of NVDAB, TSLAB, MUB, and COINB, comparing them with the next U.S. Cash Open for handoff, and included Perp for those with data.

The median handoff error for eight bStocks observations is about 117.2bp, while for six comparable Perp observations, it is about 118.9bp.

The NVDA example from July 23 to 24 is a good case. NVDAB was at 207.97, Perp was at 207.99, and the next Cash Open was at 207.45, with both sides high by about 25-26bp. Here, the value of bStocks is not to provide a price closer to the underlying stock market than Perp, but to preserve the stock prices formed during the market closure as a holdable, transferable, convertible, and financeable inventory.

Perp is responsible for moving prices, while bStocks convert prices into assets.

VII. After the Stock Split: NVDA is Correction, TSLA is Overshoot, SNDK is Stress Test, SPCX is Product Experiment

In the NVDA data sample, the average absolute deviation between Sunday and the next Cash Open is about 113bp, with a direction of 3/6; TSLA is about 81bp, direction 4/6; SPCX is about 149bp, direction 3/6; and SNDK reaches about 351bp, direction 3/6. By 09:29:59, the descriptive average deviations for the four were reduced to about 11.8bp, 8.6bp, 9.4bp, and 16.0bp respectively.

NVDA is more suitable for observing how erroneous prices are corrected. Notably, in the Independence Day sample, Sunday was at 198.35, Cash Open at 194.42, with a deviation of 202bp, and then the price returned towards $194.

TSLA is suitable for observing Overshoot. Cash Close was 393.45, Sunday was 399.40, Cash Open was 397.50; the direction was correct, but the magnitude exceeded.

SNDK combines both states in the same stock. During the Independence Day long weekend, it anticipated most of the +4.8% Opening Gap; in another weekend in August, Friday Close was 1,212.21, Sunday was 1,223.98, but Monday Cash Open was 1,203.41, reversing direction.

SPCX's position is different. Its Sunday MAE is about 148.8bp, with a direction of 3/6, but this stock moved from Pre-IPO Perp to public stock, TradFi Perp, and SPCXB, experiencing a product migration in Binance where "derivative prices came first, followed by cash stocks." It is more suitable for observing how a 24/7 pricing system aligns with the later emerging cash market.

Thus, closing prices should not be simply categorized as "right" or "wrong," but can be further divided into: Direction Discovery, Magnitude Discovery, and Correction Dependency. (Further details will be introduced later)

VIII. SPCX: From Pre-IPO Perp to Public Stock, Pricing Experiment from Perp to Stock

SPCX has a structural difference from NVDA, TSLA, and SNDK. When Binance Futures launched Pre-IPO Perpetual in May 2026, the first contract was SPCXUSDT, used to trade SpaceX's future public market valuation. After SpaceX went public, SPCXUSDT transitioned to a standard TradFi Perp; at the same time, SPCX Direct Stock and SPCXB were added. This means that this stock first had derivative prices, followed by a public cash market that could validate it, and then supplemented with holdable and on-chain bStocks.

This provided a rare experimental environment. For NVDA and TSLA, Binance adds an additional 24/7 risk layer outside a mature market; SPCX experienced a migration from "Pre-IPO Perp → Public Stock → TradFi Perp → bStocks." Once the cash stock appeared, the continuously trading derivative price finally had a fixed Monday Cash Open that could be repeatedly verified.

The data for six complete weekends of SPCXUSDT is as follows:

The direction of the six Sunday Prices is 3/6, with an average absolute deviation of about 148.8bp from the next Cash Open, and a median of about 134.4bp. By Monday at 04:00, the average deviation dropped to 114.1bp; at 08:00 it was 102.0bp; at 09:00 it was 53.8bp; at 09:25 it was 29.3bp; and at 09:29:59 it dropped to 9.4bp. After 08:00, the direction of the six instances was consistent with the final Opening Gap, but by this time the premarket had already returned, so this curve demonstrates price relay and convergence, not that Binance completed price discovery alone.

SPCX also concentrated the phenomenon of Overshoot. In three weekends where the Sunday direction was correct and the Opening Gap exceeded 50bp, the Sunday Move on July 20 was about 2.79 times the final Gap, on July 27 it was about 1.60 times, and on August 10 it was about 1.34 times. Therefore, this may not be significantly related to the timing of the launch, as even in the launched market, overshooting can occur.

The path on August 10 further illustrates the issue. The Friday Close was $133.11, and by Sunday it had risen to $135.57, with the Monday Cash Open at $134.95. This Sunday Price was only about 46 basis points higher than the final Open, but at 08:00 on Monday, SPCXUSDT surged to 138.80, widening the gap to about 285 basis points; by 09:29:59 it returned to 134.86, just about 6.7 basis points off. Prices can converge, but they can also diverge again after returning from the Premarket, only to come back again.

Moreover, the trading structure of SPCX is not a quiet spot market. Over six weekends, SPCXUSDT had a total trading volume of approximately $1.546 billion, about 10 times that of NVDA on the same basis; the underlying fills were about 3.58 million, averaging 3.45 fills per second. Its scale is between SNDK and NVDA, resembling a programmatic risk machine running all day. For bStocks, the significance of this case is straightforward: the faster the Perp turnover, the more market makers need inventory legs like SPCXB or SPCX to absorb delta as long as there is a net order flow from clients.

Thus, SPCX does not prove that "Binance has obtained weekend pricing power." It is more like a microcosm of the entire product stack: Perp first generates candidate prices, SPCXB preserves those prices as assets, Stock provides cash market exit, and borrowing and cross-market arbitrage determine whether erroneous prices can be attacked bidirectionally. It shows us what Binance's infrastructure might look like if it wants to compete for closing pricing power.

IX. Weekend Trading Volume Survey

From the data collected over six weekends, NVDAUSDT accumulated a trading volume of about $154 million, TSLAUSDT about $107 million, SPCXUSDT about $1.546 billion, and SNDKUSDT reached $2.922 billion.

From the table above, the trading volume shows a clear hierarchy: SNDK had $2.922 billion over six weekends, SPCX $1.546 billion, NVDA $154 million, and TSLA $107 million. SPCX is about 10 times NVDA and 14 times TSLA, but only about half of SNDK. This ranking does not have a linear relationship with company market capitalization or traditional stock recognition, so gross volume cannot be directly translated into natural investment demand.

Another noteworthy data point: in the last five minutes before the opening, SNDK averaged about $26.64 million in trading, SPCX about $18.88 million, NVDA about $2.49 million, and TSLA about $2.01 million. As the Opening Cross approached, Perp was not just sitting there waiting for results but was actively re-trading prices.

The core of the Volume data is: who is making these trades, at what frequency, and how large each order is.

X. Trading Frequency Survey: SNDK 7.3 Fills per Second, SPCX 3.45 Fills per Second

Breaking down SNDK's $2.922 billion and SPCX's $1.546 billion into individual trades changed the market dynamics.

SNDK had about 7.57 million fills over six weekends, averaging about 7.3 fills per second; SPCX had about 3.58 million fills, averaging about 3.45 fills per second; NVDA had about 640,000 fills, averaging about 0.62 fills per second. The differences among the three stocks are first reflected in trading frequency.

Interestingly, the difference in single order amounts is not as significant as the trading frequency. SNDK's average raw fill is about $386, SPCX about $432, and NVDA about $241. The median amount for SPCX is about $171, which is close to SNDK's $173. The large turnover for SNDK and SPCX is not achieved through a few super large orders but rather through more frequent transactions.

Since we have no way of knowing the information about trading accounts, nor do we know how many independent participants, institutions, or market makers these fills come from, based on the volume and frequency of trades, this article leans towards the argument that "these tens of billions of dollars in trades do not represent an equal scale of independent investment viewpoints" (i.e., institutions are not betting).

We believe that both SNDK and SPCX have formed a high-speed stock derivatives market, with SNDK operating at a higher speed.

(SNDK analysis: https://substack.com/@agintender/note/c-311284509?utm_source=notes-share-action&r=8gs9xi)

The 24-hour distribution of SPCX trading volume can be further analyzed:

Eastern Time 00:00--06:00 still contributes about 20.8% of the weekend gross notional and about 20.9% of the raw fills, averaging about 2.9 fills/second during this late-night period. The taker buy/sell notional imbalance over six weekends is about -1.16%, overall close to a bidirectional balance. This pattern resembles a market formed by market making, programmatic trading, and cross-market arbitrage.

XI. Order Size Also Indicates That Perp and bStocks May Handle Two Types of Orders

The average raw fill for SNDK Perp is about $386, SPCX about $432, and NVDA about $241; the median amounts for SPCX and SNDK's aggTrade are approximately $171 and $173, respectively. Binance Research's report on bStocks states that about 93% of bStocks trades are fractional trades, with a median transaction amount of only $18.81, and about 80% of tokenized-stock trades come from emerging market users.

It is important to note that the data definitions in this article from Binance Research are different, so we cannot directly divide 386 by 18.81 to conclude that the average position of Perp users is 20 times that of bStocks users. However, when placed together, a product division can be seen: bStocks are more likely to handle small spot trades, long-term inventory, cross-timezone retail, and on-chain users, while Perp is more suited for concentrated leverage, basis, market making, HFT, and high-turnover funds.

One market is responsible for forming a holdable stock inventory, while another market is responsible for allowing the same risk to be rapidly traded. Only when market makers and arbitrageurs connect both sides can a price network be formed, and currently, these two different products have not yet linked up, which is a point of potential leverage.

XII. The Larger the Perp Transaction, the More Prominent the Inventory Role of bStocks

Since Perp transaction volumes can be dozens of times that of bStocks, why does Binance still need bStocks? The answer lies in the market makers' books.

The Perp positions of clients and market makers are mirror relationships. If clients are net long, MM is short Perp and needs to buy bStocks or Stock with positive delta; if clients are net short, MM is long Perp and needs to sell bStocks inventory or borrow bStocks to sell.

Perp addresses trading speed, while bStocks addresses where the risk is placed on the balance sheet. If client net orders continue to flow in one direction, MM's delta will accumulate; it cannot remain in Perp indefinitely and must have an inventory leg to bear the risk.

Thus, high transaction volumes in Perp are not evidence of bStocks' failure; rather, they are one of the reasons for bStocks' existence. The higher the turnover of derivatives, the greater the demand for inventory, financing, and borrowing, which may not be felt by retail investors, but is a necessity for market makers.

XIII. bStocks Are Not "Leverage Points"; 1:1 Backing Is More Like a Terminal Constraint

At this point, one might refer to bStocks as the spot anchoring point for Perp, easily likening it to the relationship between USDT and the US dollar. However, the roles differ in different scenarios, at least not in the context of weekend stocks.

Turning NVDAB into NVDA Stock on Saturday does not mean that there is a cash market on Nasdaq at the same second to complete another arbitrage leg. What you receive is an inventory that can be converted back to real stocks, not one that can be immediately completed for cash arbitrage.

Therefore, 1:1 backing is more like a terminal constraint. The market knows that bStocks can ultimately be converted back to real stocks and that the cash market will reopen, so prices can deviate during the off-market period, but once the deviation exceeds funding costs, inventory costs, event risks, and execution friction, there will be an incentive to wait for the traditional market to reopen to capture the price difference.

Thus, the role of bStocks is not to tell Perp "what the correct price is," but to provide Perp's candidate prices with an inventory that can be held, hedged, and arbitraged.

This is why it is more accurate to view bStocks as the "inventory layer" and the "correction layer."

XIV. After bStocks Are On-Chain, the Focus Is on Turning Inventory Around

If NVDAB can only remain in the Binance Spot account, it is merely a layer of spot inventory within CEX. Once bStocks are on-chain, they can enter DEX, Lending, Collateral, and Margin systems, fundamentally changing the market structure.

PancakeSwap has launched bStocks pools including TSLAB, NVDAB, MUB, and SNDKB; Lista supports some bStocks for collateralized borrowing of Stablecoin; Aster allows eligible bStocks to enter Multi-Assets Mode;

Youcanshortit.com allows users to leverage borrow bStocks; Pundi X Basket enables users to freely customize their investment portfolios and index ETFs, allowing other retail investors to follow suit.

They are not solving the same issue, but their direction is consistent: to ensure that NVDAB, SNDKB, and SPCXB are not just sitting in wallets but can be traded, collateralized, financed, market-made, and hedged.

It is important to distinguish between "increasing utility" and "increasing liquidity." When an NVDAB is locked in a Lending Protocol as collateral, its utility increases, but it does not necessarily add new buy and sell orders to the market. Only when this inventory can be borrowed by market makers, arbitrageurs, or short sellers, and then enter CEX, DEX, or Perp hedging, does it begin to increase the available inventory in the market.

Thus, the ideal on-chain cycle is not "Binance → Wallet → DEX," but rather bStocks → DEX / Lending / Credit Pool → Market Maker → Perp Hedge → CEX → Stock Conversion. The same bStocks being reused across different accounts will create turnover in inventory.

For bStocks, TVL is not the most critical metric. What is more important is how many times a stock inventory can be utilized. If $1 million NVDAB is just locked in a protocol, it remains a static asset of $1 million; if part of it can be borrowed by MM for quoting and then hedged with Perp, returning to the inventory pool, it begins to transform into market infrastructure.

Binance's own research has reported that 2,806 users participated in approximate arbitrage trading between bStocks, Perp, and Equity, involving about $216 million, with approximately 58.5% of bStocks users simultaneously using Perp and/or Equity. (Thus, "wild" quant has always existed, and we should think about activating them to become seed users of bStocks.)

After going on-chain, one path could be to collateralize NVDAB to borrow Stablecoin, which then returns to CEX for NVDAUSDT Perp; another path could be for NVDAB to enter DEX LP, then hedge the LP's stock delta with NVDAUSDT Perp.

Thus, bStocks integrates CEX Perp, CEX Spot, real Stock, DEX, Lending, and Stablecoin into the same risk network.

This is also the difference between bStocks and ordinary "stock tokens."

15. What This System Lacks Is Not More Tickers, But Borrowing

Borrowing USDT after collateralizing NVDAB is financing. Borrowing NVDAB and then selling it is stock borrowing. Both actions involve "borrowing," but they have different impacts on price discovery.

If Perp is expensive and bStocks are cheap, arbitrageurs can Long bStocks + Short Perp. This trade is relatively easy to execute; buying cheap spot and selling expensive derivatives will compress the Perp premium.

If bStocks are expensive and Perp is cheap, the trade should be reversed: Short bStocks + Long Perp. The problem is that without available inventory to borrow, this trade cannot be executed. When undervalued, cash can be used to buy, but when overvalued, not everyone has inventory to sell.

Therefore, the maturity of bStocks should not only be assessed by the number of listings. Borrow Depth, Borrow Rate, available inventory, and whether borrowing remains stable over event weekends will determine whether this market has a two-way correction capability.

Without borrowing, bStocks are merely an asset layer, unable to exert their advantages on the platform.

Once borrowing is supplemented, it approaches the securities inventory market.

Market maker demand drives Borrow → Borrow demand drives interest rates up → Rising interest rates attract more deposits → bStocks' trading volume and demand rise, and so on.

16. CEX's bStocks Need Depth, Core Still Lies in Borrowing Ecosystem Interaction

Returning to Binance CEX, the liquidity bottleneck of bStocks becomes clearer. When customers sell NVDAB in large quantities, market makers can buy NVDAB while simultaneously shorting NVDAUSDT for hedging, with the Bid Side primarily consuming cash; when customers buy NVDAB in large quantities, MM must continuously sell stocks to customers. Once the inventory is sold out, if there is no borrowing, they can only reduce Ask Size, raise quotes, or exit the market.

Thus, bStocks' spot market naturally has an inventory issue. Without borrowing, market makers can only quote based on how many stocks they have; with borrowing, inventory shifts from a hard limit to a priced resource. MM can borrow NVDAB to sell to customers while also Longing Perp or hedging with Stock. As borrowing demand rises, Borrow Rate increases, which can attract holders to place more bStocks into the inventory pool.

Borrow also resolves the previously mentioned reverse Basis issue. When Perp is expensive and bStocks are cheap, anyone can Long bStocks + Short Perp; when bStocks are expensive and Perp is cheap, Shorting bStocks + Longing Perp is required. If there is no available inventory to borrow at this time, this arbitrage chain cannot function.

Of course, borrowing does not work in isolation. To deepen CEX liquidity, Stock ↔ bStocks Conversion is needed to replenish inventory, Portfolio Margin reduces capital occupation for hedging bStocks with Perp, and the Market Maker Program converts this inventory into Bid and Ask. Maker Rebate can incentivize market makers and protocols like Pundi X Basket to operate bots, but borrowing determines whether the bots have inventory.

So to summarize:

Perp generates orders, bStocks provide stock inventory, and Borrow ensures inventory liquidity.

If borrowing is successfully implemented, bStocks will transition from "tradeable stock tokens" to "securities inventory that can be market-made, financed, and arbitraged both ways."

Compared to Wall Street, the biggest weapon in the crypto space is decentralization. Using borrowing to mobilize all "decentralized" legitimate/wild market makers to participate in this weekend's game.

17. The Endgame Is Not bStocks vs. Perp, But the Interconnection of Four Markets

Assuming that in the future borrowing, conversion, and on-chain depth are all mature, the same NVIDIA risk may simultaneously have four prices.

At this point, the cross-venue trader will not first care about NVIDIA's EPS next year, but which market is expensive and which is cheap.

DEX NVDAB at 201, Perp at 199.60, can sell the expensive leg and buy the cheap leg; without bStocks inventory, there is no borrowing, and without cash, assets must be used as collateral. If one does not want to bear the overall direction of NVDA, they can lock the delta on another leg. After the cash market reopens, inventory can be adjusted based on the basis of Stock, bStocks, and Perp.

At this point, bStocks is no longer an independent trading product but an asset format that allows risk to migrate between Stock ⇄ CEX Spot ⇄ DEX ⇄ Perp ⇄ Credit.

Prices are not dictated by any single exchange, nor are they squeezed out by liquidity, but formed by arbitrageurs across different markets continuously compressing the price spread.

18. Evaluating the Maturity of bStocks Should Not Only Focus on Volume, Liquidity, and Number of Listings

Measuring bStocks solely by AUM, 24h Volume, and the number of tickers truly overlooks its role in market structure ------ this is supposed to be a transcendent magic pill, yet you just use it for a side dish?!

Volume and liquidity are just one set of indicators, and gross volume can also be influenced by programmatic rotations and high-frequency trading.

If Binance wants to turn bStocks into the inventory layer of the weekend stock market, Borrow Depth, executable spot depth, Conversion Capacity, and the stability of Perp-bStocks basis may be more important than having hundreds more tickers or the depth of individual spots.

A buy/sell of 1 million and a buy/sell of 10 million are essentially indistinguishable, and if we talk about depth, why not go to Nasdaq? Compared to depth, which price is the "correct" one is far more valuable.

Why is this the case? Because "incorrect" liquidity is essentially the fuel for arbitrage.

This article organizes the corresponding capabilities and their indicators:

Conclusion: Perp Seizes Pricing Power, bStocks Determines Whether This Price Can Become Market Price

If we compare data such as price, trading volume, trading frequency, order book, and stocks together, the positioning of bStocks will far exceed the so-called "tokenized stock".

Perp will become the new darling of derivatives, not only because of trading volume, regulation, taxation, and operations, but also because Wall Street will ultimately understand that OI > stocks. OI is money, and it does not require dilution of control, external disclosure, or tax regulation.

If Binance wants to compete for pricing power after traditional stock markets close, it is Perp that rushes ahead. In a sample of NVDA, Perp trading volume reached 99 times that of bStocks; SNDK had a pure weekend trading volume of $2.922 billion; SPCX had a pure weekend trading volume of $1.546 billion; SKHYNIX accumulated over $2.4 billion across six UTC weekend windows. These figures indicate that after traditional stock markets close, a portion of risk trading is migrating to the 24/7 derivatives market.

The same batch of data also shows that Volume and Price Discovery cannot be equated. SNDK's several hundred million dollars over a weekend can reverse the direction, and four stocks totaling $461 million can also be wrong together. After SPCX joined, the Sunday direction for 25 samples of U.S. market closures and reopenings was 13/25, still only 52%. SNDK's $2.922 billion breaks down to about 7.3 fills per second, while SPCX's $1.546 billion breaks down to about 3.45 fills per second; these high turnover figures cannot be directly interpreted as equivalent independent investment opinions.

Perp solves the problem of "the market can continue to quote", but for a price to gain credibility, it also needs to solve the problem of "who corrects the quote when it is wrong".

bStocks is the answer. It can serve as a market maker's hedge for Perp's inventory leg, become the spot leg for basis traders, serve as the base for long-term investors, act as collateral for Lending, become stock assets for DEX LPs, and manage cross-period inventory through Stock Conversion. If borrowing is completed, it will also become a securities inventory that shorts can use, allowing the overvalued side to be arbitraged.

Thus, the relationship between Perp and bStocks should be:

Perp is responsible for producing the first version of the price, while bStocks is responsible for turning the first version of the price into an asset that can be held, financed, transported, and verified.

One day, if a significant news event occurs for NVIDIA on a Saturday, and the price of NVDAUSDT moves first, NVDAB will follow with a repricing, the Perp-bStocks basis will widen, the borrow rate will change, DEX LPs will start to be moved, market makers will adjust their inventory, and arbitrageurs will compress the price difference; subsequently, when the U.S. premarket opens at 09:30, what is worth observing is no longer whether Binance guessed correctly over the weekend, but rather

whose price converges towards whom.

If Binance continues to appear in the premarket chasing the traditional market, it is a 24/7 stock exchange. If the Cash Market reopens and starts to converge towards the prices traded repeatedly over the past several hours in Perp, bStocks, DEX, Lending, and inventory markets, and external OTC, risk control, and market systems also start to reference this price set, then what Binance gains is not just a few more trading hours, but:

The pricing power of the stock market.

This is the confidence to move towards 3 billion users.

Postscript

If you want to know more about how to market make and arbitrage bStocks on Binance, feel free to visit my substack at https://agintender.substack.com/p/bstocks-b-nasdaq-perp?r=8gs9xi&utm_campaign=post&utm_medium=web, where I will share my practical experiences, as well as the current shortcomings, defects, and trading opportunities of bStocks.

Lastly,

This article is dedicated to those elders who walked with me through my childhood.

Death is not hard to face; what is hard to face is how the living should continue.