What Makes A Market Good?
A "good" market isn't just one with a tight spread — it lets traders transact cheaply, in size, without the price bouncing around for no reason, and it stays that way even under stress.
Prerequisites: Tightness, Depth and Resiliency: The Dimensions of Liquidity
Ask a retail investor what makes a market good and they'll probably say "a tight spread." Ask a regulator and they'll add "fairness" and "no manipulation." Ask an execution trader moving a million shares and spread is almost beside the point — what matters is whether the market can absorb their order without falling apart. All three are right, and none of them is the whole answer, because "good" for a market is really several separate properties bundled into one word.
The four properties that actually matter
Liquidity. Can you trade the size you want, when you want, without moving the price too much? This itself splits into tightness (how wide is the spread), depth (how much size sits at or near the best price), and resiliency (how fast does the book refill after a big trade empties it) — see Tightness, Depth and Resiliency: The Dimensions of Liquidity for the full breakdown.
Price efficiency. Do prices reflect available information quickly and correctly, without over- or under-reacting? A market where prices drift for minutes after news, or where prices bounce randomly with no news at all, is inefficient even if the spread is tight. See Measuring Price Efficiency.
Transaction cost. What does it actually cost, in dollars, to get a trade done — spread paid, price impact, and the opportunity cost of orders that never fill? This is the trader's-eye view, and it's what Cost-To-Trade Curves And Depth Metrics tries to measure directly rather than inferring from the book.
Stability. Does the market keep functioning in stressed conditions — a fast-moving news day, a large order arriving unexpectedly — or does it widen out, go quiet, or break down exactly when liquidity is needed most?
| Property | Good market | Bad market |
|---|---|---|
| Liquidity | Tight spread, deep book, refills fast | Wide spread, thin book, stays empty after a trade |
| Price efficiency | Price moves once, on real news, and holds | Price drifts slowly or overshoots and reverses |
| Transaction cost | Small, predictable cost to trade any reasonable size | Cost balloons for anything beyond a token size |
| Stability | Liquidity holds up when volatility rises | Liquidity providers pull back exactly when needed |
A worked comparison
Take two stocks, both trading around $50 with a similar $0.01 spread on a quiet day. Stock A shows 10,000 shares resting at the best bid and offer, and after a 5,000-share trade sweeps part of the book, new orders refill the level within a couple of seconds. Stock B shows only 200 shares at the best bid and offer — the tight spread is almost decorative, since anyone trying to trade more than a token amount has to walk down through several cents of thinner levels, each with its own smaller size, to get filled. Both stocks look identical on a "spread" screener; only depth and resiliency reveal that Stock A is the genuinely liquid one and Stock B is not.
Spread alone is a misleading proxy for market quality. A market is good only if it combines a tight spread with real depth behind it, prices that move for the right reasons, and liquidity that survives a busy day rather than evaporating on cue.
Why this is hard to pin down in practice
Each of these properties trades off against the others, and against who is asking. Wider spreads compensate market makers for the risk of quoting continuously, so a spread of zero is not automatically "better" if it means no one is willing to post size — see Does HFT Improve Market Quality? for the specific debate over whether faster, more competitive market making has actually widened or narrowed effective costs for ordinary investors. Regulators tend to weight fairness and stability heavily, because a market that occasionally breaks (a flash crash, a data outage) imposes costs on everyone even if its average-day metrics look fine. A large institutional trader, by contrast, cares mostly about depth and resiliency for the specific size they need to move, and may not care at all about the posted spread for a 100-share clip they'll never send.
When a question asks you to compare "market quality" across two venues or two time periods, resist reaching for a single number. Name which of the four properties you're actually measuring, and be explicit about whose perspective — retail, institutional, or regulatory — the metric is built for.
There is no single scalar that captures market quality, which is exactly why the rest of this domain is a toolbox of specific, narrower measures — variance ratios for efficiency, cost curves for transaction cost, quote stability metrics for stress resilience — each answering one piece of the question rather than all of it at once.
Related concepts
Practice in interviews
Further reading
- Harris, Trading and Exchanges (ch. 1, 23)
- O'Hara, Market Microstructure Theory