Why Volume Is Not Liquidity

Volume and liquidity are different measurements that are routinely printed next to each other as if they were the same thing. Volume counts what has already changed hands; liquidity describes what could change hands right now, at what cost, without moving the price far.

Confusing volume and liquidity is expensive in a specific way: it makes a market look safe at exactly the moment it is not. This article separates the two, shows the common situations where they diverge, and gives a repeatable way to estimate what your own order will actually cost.

• Flow versus stock. Volume is a count of completed trades over a window; liquidity is a standing offer of size at a price.
• Volume can be manufactured. Rebates, incentive programmes and self-crossing can raise turnover without leaving any resting depth behind.
• Liquidity is direction-specific. A book can be deep on the bid and thin on the ask, so exiting and entering cost different amounts.
• Both decay under stress. Turnover can spike while depth collapses, which is precisely when the distinction matters most.

Volume and liquidity measure different things

Volume and liquidity answer two different questions. Volume answers “how much traded recently”, which is a historical fact. Liquidity answers “what would it cost me to trade now”, which is a conditional estimate that changes by the second.

What turnover actually counts

Turnover counts notional that has already cleared. It is usually reported as a single number per pair per day, and it aggregates every participant: market makers recycling inventory, bots arbitraging against another venue, retail clips, and large blocks worked over hours. A million-dollar trade between one buyer and one seller may be reported as one million or as two, depending on the venue’s convention, so cross-venue volume comparisons are approximate at best.

Turnover also has no memory of price impact. Frantic small trades that shoved the price around count the same as quiet block trades that barely moved it.

What liquidity actually means

Liquidity is the ability to convert size into cash quickly at a cost close to the quoted mid. It has three components: depth, meaning how much size rests near the price; spread, meaning the gap between the best bid and ask; and resilience, meaning how fast the book refills after a large order eats through it.

Of the three, resilience is the one most often ignored and most useful. A book that absorbs a 200,000-dollar sell and refills within seconds is liquid in a way that a book with the same depth but no refill is not.

Where volume and liquidity diverge

The two diverge whenever trading activity is not produced by people who want to hold a position. That covers incentive programmes, maker rebate tiers, arbitrage bots that never hold inventory overnight, and listings where activity clusters in the first hours.

The second row is the one that catches people out. A pair can rank near the top of a volume table because a programme pays for activity, while the resting book stays thin enough that a six-figure order walks several levels. The fourth row is the mirror image: a market that looks dead by turnover can serve a modest order well because a few makers sit there with size.

Volume tells you how many people have already left the room. Liquidity tells you whether the door is wide enough for you.

The concentration problem

Daily volume is an average over a day, and averages hide timing. Much of a pair’s turnover can arrive in a narrow window around a data release, a listing, or an incentive reset, while the book outside that window is far thinner. If you trade at 04:00 on a Sunday, the headline number describes a market you are not in.

Why a high-volume pair can still be thin at your size

A high-volume pair can be thin at your size because your order is compared against resting depth, not against daily turnover. An order equal to a small fraction of daily volume can still be many times larger than the size quoted within a few basis points of the mid.

The arithmetic is straightforward. Suppose a pair trades 40 million a day and holds 300,000 of resting bids within one percent of the mid. Your 250,000 sell is less than one percent of daily volume, but it is most of the near-side depth. Filling it consumes the levels closest to the mid, and the venue’s price impact is set by that second number, not the first.

There is also a queue effect. Depth quoted at a distance is not available at the mid; you reach it only by paying through the levels in between. So the relevant measure is cumulative size from the touch outward, evaluated at the moment you send the order.

Why sellers and buyers see different books

Books are rarely symmetric. When an asset has been rising, makers often hold less inventory and quote smaller bids; when it has been falling, the reverse. The consequence is that the cost of exiting can be materially higher than the cost of entering, even on the same pair minutes apart, and the turnover figure is identical for both directions.

Measuring the two together

Measure the two separately and then compare them, because neither substitutes for the other. The routine below takes a few minutes per pair and produces numbers you can track over time.

1. Record 24-hour turnover for the pair on the venue you actually use, not an aggregate across all venues.
2. Sum resting size on each side within one percent of the mid, and record the bid and ask totals separately.
3. Compute the volume-weighted price for an order of your intended size on each side, and express the gap to mid in basis points.
4. Note the spread in basis points and compare it with that pair’s own recent range rather than with another asset.
5. Check whether depth refills after a large print by revisiting the book a minute later.
6. Repeat at the hour you normally trade, and keep the readings, because the trend in depth is more informative than any single snapshot.

If turnover is high and depth is thin, treat the pair as an exit-risk problem and size positions accordingly. If turnover is low and depth is steady, the constraint is patience rather than cost. The two failures need different responses, which is the practical reason the distinction matters.