Trading When There Is No Liquidity

Most of the time, a trader doesn't think about liquidity at all — it's just there, quietly absorbing whatever size gets sent to the market. The problem shows up the day it isn't: the bid-ask spread that was normally a cent or two is suddenly a dime, the top of book shows a few hundred shares instead of tens of thousands, and the position you need to trim is many times bigger than anything the screen is currently willing to absorb. At that point every choice is a trade-off, and the job is picking which cost to eat rather than pretending you can avoid all of them.

The three bad options

Cross the spread and trade what size is there, accepting a worse average price and probably moving the market further against yourself with every clip. Post passively and wait, accepting that you might not get filled at all, or only get filled on the days the market is moving against you anyway — the classic problem of "you only get done when you don't want to be." Or wait entirely, doing nothing until liquidity comes back, accepting the risk that the position sits unhedged, over-sized, or simply wrong for however long that takes.

There's no formula that picks the right one; it depends on why the position needs to come off. A stop-loss that's about capital preservation argues for crossing the spread now, because the cost of being wrong grows every extra minute you're in the position. A position you're trimming for portfolio-balance reasons, with no urgency, argues for patience — working small clips over the day, or several days, and letting the market come to you. A trader who reflexively does the same thing every time liquidity dries up — always crossing, or always waiting — is not adapting the decision to the reason for the trade.

A concrete version: a desk needs to sell 500,000 shares of a name that normally trades a few million shares a day, but today, after a bad earnings print, the top of book has thinned to a few thousand shares on each side and the spread has tripled. Dumping the full size immediately would print well below fair value and signal desperation to anyone watching the tape. Waiting for the spread to normalize might mean waiting for a stock that keeps falling. The realistic answer is usually a hybrid: sell into any real size that shows up, use algorithms that adapt to available volume rather than a fixed schedule, and accept that the average exit price will be worse than yesterday's close — because yesterday's liquidity is not today's liquidity.