This article explains a liquidity measure. It is educational and general, not personalised investment advice, and not a recommendation to buy, sell, or size any security. Where it references how Shishin uses the measure, that describes a research process, not a result you should expect to reproduce.
A signal is only worth as much as your ability to act on it at the price you saw. Two stocks can both “trade five million shares a day” and be worlds apart in whether you could actually buy one without shoving the price away from yourself. The number that tells them apart is dollar volume, and it is the quiet gate that decides whether a setup is a real opportunity or just a screenshot.
What is dollar volume?
Dollar volume is simply price times shares traded: the dollars that actually change hands in a name over a day, usually read as an average over a lookback window (commonly about twenty sessions, the average daily dollar volume). Share count alone is misleading. A $2 stock trading 5M shares moves $10M a day; a $300 stock trading just 300k shares moves $90M. The second is nine times more liquid despite trading a sixteenth of the shares. Volume in shares flatters penny names and hides how tradable the expensive ones are; dollars are the honest unit.
Liquidity, in one question
Liquidity is the answer to a single practical question: can you get in and back out, in your size, near the last price, without moving it? The bid-ask spread and the depth of the order book are the microscope view of that; dollar volume is the legible, comparable proxy you can run across the whole market at once. A name that trades tens of millions of dollars a day absorbs an ordinary position without noticing. A name that trades a few hundred thousand does not: you become the market.
Why thin stocks make phantom edges
This is the part a backtest hides. Suppose a signal fires in a name that trades $200k a day. To take a position large enough to matter, you would be a meaningful fraction of the entire day’s volume, which means your own buying pushes the price up as you fill, and your selling pushes it down as you exit. Add the wide spread thin names carry, and the move you “would have caught” is gone, paid away as the cost of being the marginal buyer and seller. A backtest fills you at the printed close as if by magic; reality charges you for the privilege. An edge that only shows up in illiquid names is very often an artifact of frictionless fills rather than a real opportunity, which is one of the quieter reasons backtests lie.
It sets two ceilings: size and slippage
Once you respect liquidity, dollar volume imposes two limits that have nothing to do with how much you like the trade.
- A size ceiling. You cannot safely be more than a small slice of a name’s daily dollar volume without becoming the price. So liquidity caps how large a position can be independent of conviction: a wonderful setup in a thin name still cannot be sized like a wonderful setup in a liquid one.
- A slippage floor. The thinner the name, the wider the spread and the more each share you trade moves the quote, so the gap between the price you wanted and the price you got grows. That gap is the subject of transaction costs and slippage, and it is why where and how you trade, on the close rather than chasing an intraday print, matters as much as what you buy.
As a universe filter
Because of all that, dollar volume earns its keep at the very front of the process, as a gate. Before a name is scored at all, it has to clear a liquidity floor; otherwise its signal is untradeable and there is no point ranking it. It is the natural partner to ADR: ADR asks “does this name move enough to be worth trading?” and dollar volume asks “can I actually trade it?” Both have to be yes before a composite score weighs anything else, which is exactly what a disciplined pre-market scanner does first.
How Shishin uses it
Shishin treats dollar volume as a hard gate on the tradable universe: a name has to trade enough that a position could be entered and exited on the close without chasing the price. The level of the floor differs by guardian, which is the point of having more than one engine. The small-cap momentum and quality engines, Suzaku and Genbu, still insist on enough dollar volume that a fill is clean even though they hunt smaller names; the large-cap recovery engine Seiryū works in names that are liquid by nature; and the defensive Byakko sleeve leans on liquid sector names and ETFs. The shared rule is simple and conservative: we would rather pass on a beautiful setup in an untradeable name than publish a signal we could not have acted on ourselves. The specific floors and lookbacks are part of each engine’s edge and stay unpublished.
What dollar volume does not tell you
Like every screen, it is honest about ordinary conditions and silent about the rest. It is backward-looking: average dollar volume describes the recent past, and liquidity can evaporate in a panic exactly when you most need to exit. It says nothing about direction or quality, a name can be deeply liquid and a terrible idea. And it cannot save you from a halt or a gap, which can make even a heavily traded name untradeable for a window. Liquidity is a precondition for a tradable signal, never an edge in itself, which is why it sits at the gate and lets the scoring logic decide everything that comes after.
Sources & further reading
- Amihud, Y. (2002). “Illiquidity and stock returns.” Journal of Financial Markets., the standard illiquidity measure (price impact per dollar traded).
- Kyle, A. S. (1985). “Continuous auctions and insider trading.” Econometrica., market depth and the price impact of order flow.