How to measure liquidity in tokenized-equity perps: why displayed depth misleads, and what volume and open interest reveal about where capital actually sits.

Liquidity is best judged by watching what capital actually does: the strongest test is real-money execution i.e. sending orders and measuring the fills you get - but that can't be run across venues without trading on each. The next-best signals are the ones that require committed capital, and they are not equally reliable.

Three common liquidity signals - displayed depth, volume and open interest - answer different questions. Depth hints at where orders might be; volume shows where trading activity and near-term executable liquidity are; open interest shows where capital actually stands over time. Where the aim is to assess capital across venues, open interest is the cleanest single metric. That does not diminish the importance of volume for execution; it simply clarifies what each measure is best at telling you.

Displayed depth is the weakest of the three, because what a venue shows is not a commitment: a quote can be posted and pulled at no cost, so it may disappear before an order arrives. Venues publish different, limited fractions of their book, so the figure is partial and means different things from one venue to the next. Lastly, its usefulness also depends on whether the liquidity shown is available to all participants, over what timescale, and under what execution-priority rules. Where venues apply different maker-versus-taker priority or expose only part of the book, the same displayed depth is not the same opportunity for everyone. Depth is still a useful hint about activity, but it should be treated as a starting point, not evidence.

Volume requires more commitment than a quote: a trade only clears when capital actually changes hands. But on perpetuals it is notional and often leveraged, so it overstates the capital genuinely at risk and can be inflated by wash trading or incentive-driven flow. That does not make volume irrelevant. Transaction volume is what actually determines spreads, slippage and how much one can trade in a given duration. A market with high open interest but almost no volume is still painful to enter or exit. The point here is narrower: when the question is “where is capital actually standing over time rather than just turning over,” then open interest is a cleaner signal than either depth or volume.

Open interest is the sturdiest measure of committed capital. It is margin locked against a live position - held at the exchange and, on a perpetual, paying funding for as long as it stays open. It cannot be posted and pulled the way a quote can, and unlike volume it measures capital standing rather than capital turnover. Tokenized equities make the gap between these three signals visible, because the same stock lists on venues that expose very different amounts of book.

Displayed depth is only the slice of the book a venue's public feed exposes, and it says nothing about whether that book is real or how much of it would ever trade.

For SanDisk perps, Binance displays the deepest book, $10.9M within ±0.5% of mid, followed by OKX at $8.4M, Hyperliquid at $6.5M and Bybit at $1.9M. The figures aren't strictly comparable, since venues expose different amounts of their book (Hyperliquid's feed shows at most 20 levels) - they can reflect standing orders that never trade as readily as genuine liquidity. Since depth is a directional hint, to know whether it's real, you have to look at what capital actually did.

There is an additional caveat for venues like Hyperliquid, whose order-book design processes cancels ahead of taker orders - so a market maker can pull a stale quote before an incoming order can hit it. That can make displayed depth on Hyperliquid more fragile than it looks, particularly in volatile conditions. The broader point is not to assume a deeper-looking book on any venue - Hyperliquid, Binance, or otherwise - translates directly into better executable liquidity.

With displayed depth only a partial guide, the sturdier measures are the ones that require real capital: volume and open interest show what trades: cleared trades and capital committed to open positions, both far harder to fake than a quote. For tokenized equities, both point to the same venues.

Volume ranks the same venues the same way as depth. For SanDisk: Binance is largest at $41B over 30 days, ahead of OKX ($22B), Hyperliquid ($3.2B) and Bybit ($2.1B).

Across tokenized equities, Binance leads volume at 58%, ahead of OKX (22%) and Hyperliquid (13%).

As a measure of committed capital, volume carries two limits: it is notional, so a position opened at 10x leverage contributes ten times its collateral to turnover and reflects activity rather than capital at rest; and it is susceptible to inflation, since wash trading and rebate-driven flow can lift reported volume in ways they cannot lift open interest to the same degree.

Volume corroborates depth as a signal of where activity concentrates, but it is a noisier gauge of committed capital. The two answer different questions: for transacting size with reasonable slippage today, recent volume is often the better guide; for where traders hold risk open over time, open interest is the cleaner read.

Open interest is collateral actually locked against an open position, which makes it the hardest of the three measures to fake. Binance holds the most SanDisk open interest ($240M), followed by OKX ($81M) then Hyperliquid ($58M), the rest behind.

Hyperliquid's figure is adjusted for comparability as it reports open interest counting both sides of each position - every long paired with its short - while the centralized venues count only one side. Left unadjusted, the same positions would show up at roughly twice the size on Hyperliquid, so its open interest is halved throughout to put all venues on the same one-sided basis.

For tokenized equities overall, Binance holds 55% of open interest ($1.5B on a 30-day average), Hyperliquid 21%, and OKX around 10%. Total open interest across the venues shown is near $2.7B.

Displayed depth is hard to compare across venues as it shows only the slice each feed exposes, not the liquidity behind it. Volume and open interest can be compared more cleanly that depth. They answer different questions: volume shows where trading activity and near‑term executable liquidity are concentrated; open interest shows where traders actually keep risk open over time. If you must pick a single metric to gauge committed capital, open interest is the cleaner read. If you care mainly about how easy it is to trade in the next few hours, recent volume and spread data are more directly relevant.

By both, tokenized-equity trading concentrates on Binance (58% of volume, 55% of open interest).

Plotting the two together, Binance sits alone in the top right - highest on both signals - while every other venue clusters near the origin. In other words, Binance leads both on activity (volume) and on committed positioning (open interest), which is why it dominates both lenses. The concentration is real rather than merely displayed; the caveat is that these figures cover a small set of venues, and a broader venue set would move the shares.