ETF Liquidity: Why On-Screen Volume Misleads

With a share, volume is the whole story: a company's shares are a fixed pool, and if few of them change hands, a large order will move the price, as the liquidity pillar established. An ETF is different, because its supply is elastic. A market maker facing a large buy order does not need to find existing ETF shares from other sellers; it can source the underlying basket and have new ETF shares created. So an ETF's real liquidity is the liquidity of what it holds, plus whatever exists on screen. That inverts the ordinary reasoning about thin volume, and it produces both a reassurance and a warning — the reassurance being that a small ETF over liquid holdings may be perfectly tradeable, and the warning being that a heavily traded ETF over illiquid holdings is not as safe as its volume suggests.

Two layers of liquidity

Think of an ETF as having liquidity at two levels, and both matter. The secondary-market layer is what you see: on-screen bid and offer, quoted sizes, the day's volume, the depth in the order book. For small orders this layer is usually sufficient and cheapest — you cross a narrow spread and you are done. The primary-market layer is what stands behind it: the ability of a market maker to create or redeem shares using the underlying basket. For larger orders this layer is what actually sets your cost, because the market maker will price your trade at roughly what it costs them to assemble or unwind the basket, plus a margin. The industry term for the resulting capacity is implied liquidity — an estimate of how many ETF shares could be created from the tradeable volume of the underlying holdings, which is frequently far larger than the ETF's own average volume. Four consequences follow. A low-volume ETF is not necessarily illiquid. If it holds large, liquid shares, a substantial order can be filled at a reasonable price through creation, even if the screen showed a few thousand shares trading a day. Judging such a fund unsuitable on volume alone is a category error. A high-volume ETF is not necessarily liquid in the way it looks. If the underlying holdings are thinly traded — small-cap, emerging-market, or high-yield bonds — then in stress, when the underlying becomes hard to trade, the ETF's own liquidity degrades with it, regardless of how busy it looked in calm conditions. Your effective cost is the spread plus any premium or discount at the moment you deal, and both widen when the underlying is hard to price. And for large orders, the on-screen market is the wrong place to look — institutional practice is to seek a quote from a market maker or use a request-for-quote facility rather than posting into a thin book, which is exactly the size-inversion logic the bond-trading article described in a different market.

What actually determines an ETF's tradeability

Five factors, in roughly descending order of importance. The liquidity of the underlying holdings is first and dominant — everything else is secondary to whether the basket can be traded. Whether the underlying market is open matters more than most readers expect: an ETF over a market that is closed during its own trading hours is being priced on estimates, so spreads widen and the arbitrage that disciplines them cannot operate until the underlying market reopens. The number of active market makers and authorised participants matters, because competition among them narrows spreads; a fund supported by one liquidity provider is more fragile than one supported by several. Fund size matters modestly and indirectly — larger funds tend to attract more market-making interest and tighter spreads, but a small fund over blue-chip holdings can trade better than a large one over illiquid assets. And time of day matters: spreads are typically widest at the open, when overnight information is being absorbed and the underlying may not yet be trading normally, and around the close. These are mechanical observations about execution rather than trading advice; what a reader does with them is their own decision. Two structural cautions to close. Very small or newly launched ETFs carry a distinct risk that has nothing to do with spreads: closure. Funds that fail to gather assets get liquidated, which forces a realisation at a time not of the holder's choosing and can have tax consequences this portal parks. That is a real consideration for a small fund, separate from whether it trades well. And liquidity is a property of conditions. Every claim above describes normal markets. In dislocation, spreads widen, quoted sizes shrink, arbitrage becomes expensive, and — as the mechanism article noted — participants can step back. An ETF that trades on a two-basis-point spread on a quiet Tuesday may trade on a much wider one during a disorderly session, which is the same lesson the crisis pillar teaches about every form of liquidity: it is most abundant when least needed.