Verified Platforms
Quick Links
Where to Stay Secure
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Volume counts what changed hands. It does not count who wanted it.
Tokenized stock trading volume reached $11.3 billion in July, a 288% jump and the largest month the category has recorded. A single token produced $9.27 billion of that. QQQB started trading on Binance on 30 June, tracks Invesco's QQQ, and carries zero maker fees through 31 August. By the end of the month it was roughly 82% of every tokenized-equity dollar traded anywhere. Strip it out and everything else, every other venue and every other ticker combined, moved about $2.03 billion against June's implied total of $2.91 billion.
The category shrank in the month its headline tripled.
Tao takes this one. At Kodex he is the bridge between how a market is built and how it feels to use. He treats a volume figure the way an engineer treats a load rating: useful, but what is carrying it? He does not open with the total. He opens with the fee schedule.
Every dollar in that record came from a filled order. Binance's bStocks accounted for $9.41 billion of the July total, or 83.3%, and QQQB alone was $9.27 billion of that. Kraken's xStocks moved the other way across the same window, falling from $1.55 billion to $335 million on CoinDesk's figures.
Nothing in that concentration is alleged to be fake, and it is worth killing that suspicion early because it leads somewhere useless. bStocks settle on chain. You can withdraw one as a standard BEP-20 token to a BNB Smart Chain wallet, and convert between the token and the underlying share at no fee with immediate settlement. A QQQB trade is a trade, cleared and settled, moving a real claim. If you want the plumbing under the wrapper, how a tokenized stock settles unpacks the rail itself.
So the question Tao asks is not whether the volume happened. It is what kind of activity it recorded.
A volume figure is a sum of executed trades and nothing else. It does not separate a position someone opened because they wanted Nasdaq exposure from a round trip posted by a desk that will be flat again in four minutes. Both print. The second one prints twice, once going in and once coming out, and carries no view about anything.
What makes that round trip cheap enough to run all afternoon is on the fee page. Every order arriving at an exchange does one of two things: it either sits on the book and waits, or it crosses the spread and takes something already sitting there. The first is a maker order: it adds depth, it can be cancelled, and it earns the venue nothing until someone hits it. The second is a taker order: it removes depth and pays for the privilege. Venues charge takers more than makers because makers supply the thing the venue is actually selling, which is a book deep enough to trade against.
Set the maker fee to zero and the economics of posting invert. Quoting becomes free. You can post a bid, cancel it, repost two ticks lower, and get filled on both sides of the spread all afternoon. The only cost you carry is holding whatever you were filled on until the hedge lands. On a token tracking one of the deepest ETFs in the world, that exposure is short-lived and the hedge is routine.
Free quoting produces depth. It does not produce conviction.
A maker-fee waiver changes the cost of participating and changes nothing about the reason to participate. The book gets deeper, the tape gets busier, the spread tightens, and none of it requires one person to have formed a view on US tech earnings. Tao puts one question to any liquidity figure, which is whether the depth would survive being charged for. "Depth that costs nothing to post is a promotion, and promotions end," he says. Here, unusually, that question comes with a date attached.
Take QQQB out and the arithmetic reverses. Everything else did about $2.03 billion in July, against an implied June total of $2.91 billion for the whole category. That is roughly 30% lower, inside the month whose headline said 288% higher.
Two figures describe the same industry over the same weeks and point in opposite directions, and the only difference between them is whether one promoted instrument is counted.
Kraken's xStocks is the cleanest single case: $1.55 billion in June down to $335 million in July, again on CoinDesk's numbers. Nothing about xStocks changed that month. What changed was where the free quoting was.
The rest of that $2.03 billion is small and legible. Ondo recorded $792 million and Backpack $479 million, which with xStocks accounts for around four fifths of everything outside QQQB. These are working venues with real order flow, and stacked together they are a rounding error beside one token in its promotional window.
Tao keeps both figures rather than picking one. The ex-QQQB number is not the honest one and the headline is not the lie. They answer different questions. $11.3 billion answers what the category can produce when quoting is free somewhere inside it, and $2.03 billion answers what it produced everywhere the quoting was not free. Sizing a position off the first while believing you have read the second is where the damage happens.
Scope is the recurring trap with figures in this sector. How big the tokenized RWA market is covers the same problem on market size rather than turnover.
From 23 July, Binance began counting bStocks volume at three times its traded value toward VIP tier progression for some users. Precision matters here, because the intuitive reading is wrong and the wrong version would be a factual error: the multiplier does not change reported trading volume. The $9.27 billion is traded value, counted once. The 3x applies to a separate internal ladder, the one deciding which fee tier you sit on.
What it changes is the reason to trade.
Binance's VIP tiers set your fees across the whole exchange. If a dollar of QQQB counts as three dollars toward that ladder, then running size through QQQB becomes the cheapest way to buy a discount on everything else you do there. A desk with no opinion whatsoever about the Nasdaq now has a concrete arithmetic reason to trade a Nasdaq tracker.
Stack that on a zero maker fee and two independent incentives push volume into one ticker, neither of them a signal about tokenized equities. Tao holds the two claims apart deliberately. An incentive to trade is not an overstatement of what was traded, and collapsing them is how a correct suspicion turns into a wrong accusation.
Someone budgeted for that $9.27 billion.
Yes, and the evidence sits in numbers nobody was paid to move. They are smaller, less quoted, and considerably harder to manufacture.
Binance Research counted 190,417 bStocks signups between the 11 June launch and 8 July. Of those, 41.5% had no prior experience with either stocks or perpetual contracts on Binance, per Crypto Briefing. Those are people who arrived for this product specifically. Separately, 44.5% of bStocks volume occurs outside conventional US market hours, meaning close to half of it happens while the NYSE and Nasdaq are shut. Binance's own announcement on 29 July put bStocks assets under management at $500 million, a figure worth reading as the company's disclosure rather than an independent count.
Each of those measures something a fee waiver cannot reach. Waiving the maker fee pays you to quote. It does not pay you to open an account, and it does not pay you to hold overnight.
The after-hours share is the strongest of the three. A market maker harvesting a fee promotion has no reason to concentrate work at 2am on a Sunday. A desk hedging against the underlying ETF would rather work while the underlying is open. Activity clustering while the reference market is closed is activity that wants something the reference market cannot supply, which is access at the hour the person is awake. Opening hours are a venue rule rather than a property of the asset, the same way a circuit breaker halts a venue and not a price.
That part you can test instead of taking on trust. The Kodex simulator lists 32 tokenized stocks trading 24/7 on a $5,000 paper account. So what you would do with a Tesla position at 3am on a Sunday stops being a thought experiment.
Tao will not stretch the claim past what those numbers carry: the asset class is doing something real, and the metric that made the headlines is not the one showing it.
The maker-fee waiver on bStocks pairs ends on 31 August at 23:59 UTC. That is not a market event. It is a line in a fee schedule, published well in advance, and it lands on positions that are already open.
Quoting stops being free, so anyone posting depth because it cost nothing will post less of it, or post it wider, or want paying for it. Nothing forces a single participant to leave. The book simply stops being subsidised, and whatever depth remains is depth someone funds at a positive cost.
Your position does not change. The book underneath it does.
That matters at exactly one moment, which is when you want out. It is the first week of September and you are closing a QQQB position you sized in August. The Nasdaq has not moved a tick since you opened it, the screen shows you the price you expected, and the fill comes back worse anyway. An entry taken against a deep book and an exit taken against a thin one are the same trade at two different prices, and the gap arrives as slippage, not as a move in the underlying.
The size of this is uncertain and worth saying so. QQQB tracks an ETF with enormous underlying liquidity, so unsubsidised market makers may well fill the space the promotion vacates. Binance can also extend the waiver. Venues do that. Tao does not read the date as a warning. He reads it as the one variable in the whole story that is knowable before it happens.
The fee schedule running during the window is part of what the window measured, and fees are the cheapest variable a venue can change. Concentration is the next thing visible from outside, because a category figure that is 82% one ticker is a figure about that ticker whatever the category label says. And a number produced by a condition that ends on a calendar day has a shelf life, whether or not anyone prints one on it.
Read July on those terms and the figures stop competing. Each turns out to be measuring something different, and each has a job it cannot do.
| The number | What it measures | What it supports | What it cannot support |
|---|---|---|---|
| $11.3B, July total | Every tokenized-equity dollar traded, promotion included | That the rails carry serious turnover when quoting is free | That demand for tokenized equities tripled |
| $2.03B, July excluding QQQB | Turnover everywhere the maker-fee waiver was not running | A like-for-like read against June | The size of the category as a whole |
| $2.91B, June implied total | The same measure one month earlier, effectively pre-QQQB | That unsubsidised activity fell around 30% | Anything about QQQB itself |
| 190,417 signups, 41.5% new, 44.5% after hours | Who arrived, and the hours they chose to trade | That use of the product is real and partly outside US hours | How much money stands behind it |
None of those figures is false, and none of them can do another's job. The failure worth avoiding is not believing a wrong number. It is asking a right one to answer a question it was never measuring.
That reading is not specific to July. Any figure a venue publishes about its own activity was produced under conditions that venue selected, and fees are the cheapest of those conditions to move. Launch-week volumes, trading competitions, points programmes, and airdrop farms all print the same way, and every one of them ends on a date somebody already knows.
Tao says it shorter. "Before a number convinces you a market is deep, find out what it cost to make it look that way, and find out when that stops."
Pick one of the 32 tokenized stocks in the Kodex simulator and hold it across a date you did not choose: a weekend, an earnings print, a Monday open. Then try to get out. The account runs on $5,000 that was never yours, which makes it the cheapest place available to find out what you had assumed about leaving.
