WhitmanTrading

What Is a Good Till Canceled Order?

Good till canceled order is an instruction that stays active across trading sessions until it executes or is explicitly cancelled, rather than expiring at the end of the day. Brokers usually impose their own expiry anyway, and the real risk is an order filling months later on a thesis nobody updated.

A good till canceled order is a standing instruction with no end date. It solves the problem of not being able to watch the market, and it creates the problem of an instruction outliving the reason for it.

How it works

A price series with an order persisting across sessions.
A GTC order stays live across sessions. Illustrative chart - not real market data.

A day order expires at the close. A good till canceled order does not — it carries into the next session, and the one after, unchanged.

A steady series where an order waits for its price.
Until it fills or you cancel it. Illustrative chart - not real market data.

Only two things end it. Execution, or an explicit instruction from you to remove it.

A rising series where a broker imposes its own expiry.
Most brokers cap it anyway. Illustrative chart - not real market data.

Except that most brokers cap it regardless — commonly at 30, 60 or 90 days — so “till canceled” is usually “till canceled or until my broker’s policy ends it,” and the two are not the same promise.

A falling series where a forgotten order executes unexpectedly.
The risk is forgetting it exists. Illustrative chart - not real market data.

Forgetting is the real failure mode

A choppy series where an old order fills on new conditions.
It fills on the market's schedule, not yours. Illustrative chart - not real market data.

The order does not know your view has changed. It was placed on a thesis, the thesis may have been abandoned weeks ago, and the instruction is still sitting there waiting.

A slow series where an instruction outlives its reasoning.
And different again over a long horizon. Illustrative chart - not real market data.

And the price that triggers it is usually a bad one. A resting buy fills when sellers are eager, which is disproportionately when something has gone wrong.

A calm series where an order sits untouched for weeks.
A quiet stretch hides what it measures. Illustrative chart - not real market data.

So the order acts as a commitment device, which is genuinely useful when the commitment is still one you want, and genuinely dangerous when it is not.

A worked example

A buy limit at 90 on something trading at 100, placed as good till canceled because 90 looked like fair value.

Six weeks pass. The company issues a profit warning, the price falls to 88, and the order fills at 90 on the way through.

A falling series with a stop level marked.
A stop fills where the market is. Illustrative chart - not real market data.

You now own it at 90 in a market that thinks it is worth 88 — and you bought on an analysis performed before the information that caused the fall.

The order did exactly what it was told. The problem is that it was told something six weeks ago and nobody updated it, which is the failure this instruction is most prone to.

Where it genuinely earns its place

For somebody who cannot watch. A standing limit lets a decision be made calmly, once, rather than reactively in a moving market — and that is a real improvement for most people.

For accumulating a long-term position. If the plan is to own something at a price and hold for years, a resting order at that price is the plan expressed as an instruction.

For exits set in advance. A target price decided before entering is a better target than one decided while watching a profit fluctuate.

And that last case is the strongest argument for it. The instruction’s rigidity — its refusal to be talked out of the plan by the tape — is the same property that makes it dangerous when the plan is stale. It is one feature seen from two sides, and the difference between them is whether you review it.

The original data

On this site’s shared series direction runs average 2.01 bars with a longest of 11, and 95% of bars sit below a prior peak with a longest below-peak stretch of 73 bars.

That last figure is the one to hold onto. A resting buy order set below the market can wait a very long time without anything unusual happening, and 73 bars is long enough for a thesis to go stale.

A candlestick chart annotated with the cost of a round trip.
A round trip costs a share of a bar. Illustrative chart - not real market data.

A round trip costs 0.0098, about 2% of the median bar range of 0.493 — so cancelling and replacing an order costs nothing, and there is no financial reason not to review a standing instruction regularly.

A price series with volume shown beneath.
Volume and price measure different things. Illustrative chart - not real market data.

What can silently break one

Corporate actions. A share split, consolidation or large special dividend changes the price scale. Some brokers adjust standing orders, some cancel them, and some do neither — leaving an order at a price that now means something completely different.

Ticker changes and mergers. The instrument the order refers to may cease to exist under that symbol.

Broker policy. The cap mentioned above ends orders quietly, and an order you believe is protecting a position may have expired weeks ago.

And session eligibility. Many good till canceled orders do not participate in premarket and after-hours trading, so a move that happens outside regular hours can pass your price entirely without touching the order.

Linked-order relationships too. Where a GTC is one leg of a bracket or a one-cancels-other pair, the link is held by the broker rather than the exchange. If the pairing is dropped — by a platform change, a transfer, or an expiry on one leg — the surviving order continues alone, and a protective instruction that was half of a structure becomes a standalone order doing something nobody intended.

When it fails

The characteristic failure is an order that fills after the reason has gone. Somebody sets a buy limit well below the market on a company they like, stops following it, and months later the position appears in their account — bought into a decline they had not been watching, at a price chosen when the facts were different. Nothing malfunctioned. The instruction persisted exactly as designed, and the analysis behind it did not, and no system anywhere connects those two things except the person who placed it.

A candlestick series with a gap through a level.
A gap skips the level entirely. Illustrative chart - not real market data.

A second failure is assuming “till canceled” is literal. Check your broker’s actual expiry policy.

A third is forgetting it counts against buying power at some brokers, quietly restricting what else you can do.

A fourth is leaving one in place through a corporate action, where the price scale may have changed underneath it.

A declining series cut short at a decision point.
Filled at your price. Do you still want it? Illustrative chart - not real market data.

And a fifth is never reviewing them. A monthly look at standing orders takes minutes and removes nearly every failure on this list.

Order types covers the full set of instructions and their trade-offs. Limit order covers what a GTC usually wraps. And fill or kill order covers the opposite instruction — one that refuses to wait at all.

What I actually do

I have more sympathy for this order type than most people do — it lets somebody with a job set a price and get on with their life. The discipline it requires is a calendar reminder, and the people who get hurt by it are the ones who never set one.

— Michael Whitman

This page is educational, not financial advice. Test every idea on your own charts before risking money.