Market vs Limit Orders: Certainty vs Price

Two risks, and you pick one

Every order chooses which risk to accept. A **market order** says “fill me now at whatever it takes”: you get near-certain execution and no control of price. In a deep book that costs about half the spread; in a thin book it can walk several levels — **slippage**. A **limit order** says “only at this price or better”: you control the worst price, but you may not fill at all, may fill partially, or may watch the stock run away without you. The order waits in the book, and it only trades when someone chooses to trade with you. Walking the book: buy 1,000 shares (mid $50.00) — 300 @ $50.02: $15,006 · 500 @ $50.05: $25,025 · 200 @ $50.09: $10,018 · Average price: $50.049 ← · Slippage vs mid: 4.9¢ a share = $49 ← A **marketable limit order** — a buy limit at or above the ask — gets the speed of a market order with a price cap. If the book is thinner than you thought, it stops instead of chasing.

Choosing between them

Market orders are reasonable for small orders in very liquid stocks during regular hours. Avoid them in thin stocks, in the first and last minutes of the session, in pre-market and after-hours trading, and right after a halt reopens. Many brokers quietly convert market orders into limit orders with a price collar to cap disasters — check what yours does. A resting limit order provides liquidity to others, and the catch is **adverse selection**: your buy limit fills most readily when sellers are aggressive, which is often just before the price falls further. Limit orders do not get a free lunch; they trade price certainty for fill uncertainty. The dangerous combination is a market order plus urgency plus a thin book. Slippage in a micro-cap around news can be several percent — more than the edge on most trades.

What the spread costs over a year

One round trip is cheap enough to ignore; a habit is not. You pay roughly half the spread on the way in and half on the way out, so crossing the spread both times costs you one full spread per round trip. Put that against a share price and it becomes a percentage — and percentages repeat. Here is the arithmetic at retail scale. A $20 stock with a 10-cent spread costs about 0.5% of the position on every round trip, which is 50 basis points. Twenty round trips a year — about two a month — is **10% of capital** paid in spread alone, before slippage, before commissions, before a single bad decision. Twenty round trips is not a lot of trading. It is roughly what an active retail account does in a month. This is the arithmetic behind the oldest result in behavioral finance: in Barber and Odean’s study of 66,465 households, the most active quintile of traders earned 11.4% a year while the market returned 17.9%. Nothing in that gap requires those traders to be wrong about direction. It is the cost of being active, paid one spread at a time, and it is invisible on a zero-commission statement. So the question to ask before every order is not just “is this order useful?” but “does this position need to exist today?” A marketable limit that captures a cent of price improvement each way cuts the annual figure from 10% to about 8% in the example above. A $20 commission would have cost 0.5% on a $4,000 position — the same as the spread — which is why commissions were never the real bill. • Cost per round trip ≈ the spread ÷ the share price. A 10¢ spread on a $20 stock is 0.5%. • Multiply by how often you do it: 20 round trips a year turns 0.5% into 10% of capital. • Price improvement, midpoint fills and patience all attack the same number — the spread you cross. • Commissions are usually the smallest line. The spread is charged whether or not it is itemised. The same 0.5% round trip, repeated ($20 stock, 10¢ spread, $4,000 position) — One round trip: 0.5% · $20 · Twelve round trips a year: 6% · $240 · Twenty round trips a year: 10% · $400 ← · Twenty round trips, half a cent of improvement each way: 8% · $320 ← · Same twenty round trips, as a decade of compounding drag: Enough edge to matter more than most stock picks ← This is why “how often” belongs in the same sentence as “how much”. A strategy with a 1% edge per trade and a 0.5% cost per round trip is a coin flip with a fee; the same strategy with a 0.1% cost is a business.

The same order type, two completely different trades

Ask whether a market order is safe and you will get two confident and opposite answers, because the question left out the only thing that decides it: **size**. A market order for 100 shares of a $50 stock is a rounding error — it fills at the offer and costs half the spread. A market order for 50,000 shares is an event: it consumes several price levels, prints through them, and the average price it pays is set by the depth it eats. So the honest rule is not “market orders are dangerous” or “market orders are fine”. It is that a market order is a **size decision wearing an order-type costume**. Below the displayed size at the touch you are a price taker at a price you can read before you send it. Above it you are the one moving the price, and your own impact — not the spread — is what you pay, usually by a wide margin. That is also the answer to why the same trader can reasonably use market orders all year and still lose money to execution. The order type was never the problem; the size was. Any rule of the form “always use a limit” is doing arithmetic on the wrong variable. One order type, three sizes (mid $50.00, bid $49.99, offer $50.01) — Market buy, 100 shares: fills at $50.01 — cost is half the spread, $0.02 · Market buy, 5,000 shares: walks four levels to about $50.05 — half of that cost is impact, not spread ← · Market buy, 50,000 shares: impact dominates; the average price depends on how fast you consume liquidity ← · The 50,000-share version, worked patiently: a different order entirely — sliced, benchmarked, and measured (M17) ← This is the bridge into the advanced tier. Separating the spread — which everyone pays — from the impact, which only your size pays, is the entire subject of transaction cost analysis.

The order most desks actually use

The comparison in this lesson is between two extremes — a market order that accepts any price and a limit order that may never fill — and most professional order flow lives in between. The instrument that occupies that middle ground is the **marketable limit**: a limit order priced through the far side of the spread, so it crosses immediately like a market order but cannot execute worse than the cap. Buy at the offer, sell at the bid, and add a few cents of tolerance in case the quote moves between your decision and your click. It sounds like a technicality and it is not. A market order says “fill me at whatever the book offers”, which in a thin name at 9:31 am can mean a price several percent away from the screen you were looking at. A marketable limit says “fill me at this price or better, and if the market has moved past it, do not fill me at all”. You keep the certainty of immediate execution in the ordinary case and you convert a catastrophic fill into a non-event in the rare one — which is the trade a size-aware trader wants, because the tail of the fill distribution is where retail orders actually get hurt. A second family of instructions is worth knowing by name because it shows up in every execution-quality report: orders that reference a benchmark rather than a fixed price. A **peg** tracks the near touch or the midpoint and reprices as the quote moves, which keeps an order competitive in the queue without anyone watching it. A **volume-weighted** or **time-weighted** instruction (VWAP and TWAP) is not an order type at all but a schedule — the broker slices the parent order into child orders over the day to land near the day’s average price. The choice between them is the choice between being filled soon and being filled at a fair average, and it is exactly the decision lesson M17 takes apart. The same 1,000 shares, three instructions (bid $49.99 × ask $50.01) — Market order: Immediate, average price set by the book: whatever the levels give · Marketable limit at $50.05: Immediate, and no fill at all if the book has moved past $50.05 · Limit at $49.95: No spread cost, no certainty — and the opportunity cost if it never trades ← A marketable limit is the default for anything you intend to fill now. It costs nothing extra when the quote is where you think it is, and it caps the one outcome a market order leaves entirely open.

The order types that follow the market for you

Market and limit are the two ideas, and real order books carry a family of orders that reprice themselves around a reference instead of holding a fixed number. A **pegged order** is given a relationship rather than a price: a primary peg tracks the same side of the best quote, a market peg tracks the opposite side with a fixed offset, and a **midpoint peg** tracks the midpoint between the two. The venue updates the order as the reference moves, so it sits at a chosen distance from the market without anyone retyping it. The reason a trader uses one is queue position. Manually chasing a moving quote means cancelling and re-entering, and every re-entry goes to the back of the new price level — so a pegged order keeps you *at* the market rather than repeatedly arriving behind it. That is the whole appeal: the peg is a way of staying present in a moving book. The midpoint peg is the one most retail traders meet last, and it has a property that is easy to misread. It prices at a level neither side is displaying, so a fill is price improvement relative to the touch — but the order is not displayed, so it is not a protected quotation and nothing obliges anyone to trade with it. You have traded a better price for a lower probability of ever getting one. In a wide, thin name that trade is often bad; in a deep, one-cent-wide name the midpoint is where a great deal of volume actually happens. Some books also offer **discretionary** orders, which publish one price but are willing to trade at a better one up to a hidden limit, and a variety of broker-side “smart” prices that do something similar under a friendlier name. All of them share a single characteristic worth internalising: you have handed the pricing decision to a rule you did not write. In a fast market a peg chases, and a market peg on a buy order follows the offer upward exactly when the offer is moving away from you. None of this makes pegged orders dangerous, and it does make them an order type worth reading about before using. The question to ask of any self-pricing instruction is the same one this lesson asks of a plain market order: what happens to it when the market stops being calm? • A peg holds a relationship to a reference; the venue reprices it as the reference moves. • The queue is the reason: pegging beats manual repricing for staying present. • A midpoint peg fills at a price nobody is showing — better fills, far fewer of them. • Every self-pricing order hands the price decision to a rule, and the rule follows the market fastest when you least want it to. A retail platform may list these as “inside-the-spread” or “midpoint” limits. The label matters less than the mechanics: find out what reference the order follows and whether it is displayed, because those two facts decide whether it can be traded against.

Sources

• Order types explainedInvestor.gov / SEC — Trading basics
• Trading and Exchanges: Market Microstructure for PractitionersLarry Harris, Oxford University Press (2003)
• Order handling rules and best executionFINRA Rule 5310