Bitget’s Nvidia Token Surpasses 75% of BTC Liquidity Depth, Surprising Analysts

Bitget’s synthetic Nvidia contract is holding its own. By mid-May 2026, the NVDA perpetual on Bitget’s Universal Exchange had reached roughly 75% of the liquidity depth seen in the platform’s BTC/USDT spot market — a number that probably surprised a few skeptics who figured tokenized equity products couldn’t get close to crypto’s flagship pair.

The figures come from Block Scholes, whose research published June 15, 2026 dug into four USDT-margined perpetual contracts on Bitget’s Universal Exchange. The four contracts track gold, the SPDR S&P 500 ETF, Nvidia stock, and the Invesco QQQ Nasdaq-100 ETF. Block Scholes pulled historical order-book data from Bitget’s public API covering September 2025 through May 2026, calculating bid-ask spreads as the difference between the best bid and best ask divided by the mid-price, and modeling slippage through market orders of specified sizes. The firm was clear that depth figures showed visible resting liquidity only — not a guarantee of executable liquidity. Fair enough caveat.

Spreads, Slippage, and the First Hour

The raw spread numbers are tight. Gold came in at 0.02 basis points at the top of book. SPY and QQQ both sat at 0.14 basis points. NVDA was wider at 0.44 basis points, but still pretty competitive for a tokenized equity product running on a crypto exchange.

Slippage told a similar story, and the intraday pattern was clear. A $100,000 market buy on the SPY contract started the session with 14.88 basis points of slippage, then dropped to 10.66 basis points an hour in. Scale that up to a $500,000 order and the improvement gets more dramatic — slippage fell from 46.07 basis points down to 24.90. So the first hour matters. A lot.

Block Scholes noted that spreads were consistently wider at the open of a trading session and narrowed significantly after roughly sixty minutes. That’s not unusual for markets in general, but it’s worth flagging for traders sizing into these contracts. Timing an entry at session open versus an hour later can meaningfully change cost.

Weekend Volumes Drop, But Spreads Hold

Here’s where it gets interesting for anyone thinking about around-the-clock access to equity-like exposure. Trading volume on the four contracts dropped 65% to 90% on weekends compared to weekdays. That’s a steep fall. But the median bid-ask spreads barely budged. Gold held near 0.02 basis points. QQQ, NVDA, and SPY came in at 0.8, 1.0, and 1.3 basis points respectively during off-hours — wider than weekday top-of-book, but not dramatically so.

That kind of weekend stability is actually a selling point for tokenized equity products. Traditional equity markets are closed on weekends entirely. Crypto-native perpetuals aren’t. So even with volume thinning out, the pricing stayed reasonably efficient. Whether that holds under sustained selling pressure on a Saturday is a different question. Block Scholes didn’t specify.

Related: MEXC Launches RealStocks API, Offering USDT Access to 7,000 US Stocks and ETFs

What Happened on February 28

The stress test came February 28, 2026. That’s when the U.S. announced strikes against Iran, and all four contracts felt it immediately. Spreads widened across the board. The NVDA contract’s spread jumped from 0.6 basis points to 3.4 basis points — a meaningful spike. Order-book depth thinned more noticeably than spreads did during the event.

But both metrics recovered. Spreads returned to near-normal levels quickly after the initial shock. Depth took longer — it normalized within a week. That’s a reasonable recovery window for a relatively new product category, but traders running larger positions would have felt the thinning.

It’s probably worth noting that the February 28 event hit all four contracts simultaneously, not just NVDA. Gold, SPY, and QQQ all saw temporary spread widening. The geopolitical shock was broad enough to touch the whole product suite at once, which makes sense — these aren’t isolated instruments when macro risk spikes.

And the recovery pattern matters here. The fact that spreads snapped back fast while depth took longer to rebuild tells you something about how market makers on the platform behave under stress. They’ll quote, but they’ll quote thinner until confidence returns.

Methodology and What It Doesn’t Include

Block Scholes was explicit about scope. The analysis didn’t include trading fees, funding payments, or hidden liquidity. So the slippage figures are cleaner than what a live trader would actually experience all-in. That’s standard for this kind of academic-style market structure work, but it’s a gap worth knowing about if you’re trying to translate these numbers into real trading costs.

Read also: Core Lightning Faces Crisis: Urgent Shutdown as AI Reveals Critical Vulnerabilities

The data ran from September 2025 to May 2026 — roughly eight months of order-book history. That’s a decent sample for a product that’s still finding its footing. Tokenized equity and commodity perpetuals on crypto exchanges are still a pretty new category, and getting eight months of granular spread and depth data into a structured report is useful for anyone evaluating whether these markets are actually tradeable at scale.

The full methodology, data tables, and charts are on the Block Scholes research site.

At $500,000 order size, SPY slippage dropped from 46.07 to 24.90 basis points within the first hour of the trading session.