Crypto Order Types Slippage: The Cost That Never Shows on the Fee Page
Your realised cost per trade is not your fee. It is your fee plus the gap between the price you decided to trade at and the price you actually got. Order choice is the lever that moves that gap, which is why crypto order types slippage deserves attention before you spend a weekend comparing fee schedules. On AUD pairs it can matter more than the fee difference.
This matters because fees are advertised and slippage is not. An exchange publishes its maker and taker schedule; nobody publishes what your market order will cost at 2am on a Sunday in ETH/AUD, because that depends entirely on the order book at the moment your instruction arrives.
The mechanism is simple even if the maths is not. Every order you send is an instruction about two things: what price you are willing to accept, and how much urgency you are prepared to pay for. Change the instruction and you change the cost.
One framing note before the mechanics. This article is general information only, not personal financial, investment or tax advice, and execution choices carry market, liquidity and non-fill risk. Providers of crypto exchange, custody, brokerage or transfer services may need to register with AUSTRAC as virtual asset service providers (VASPs) under Australia’s AML/CTF regime, which replaced the narrower legacy “digital currency exchange” (DCE) category. That registration is a financial-crime obligation, not an endorsement of a venue’s execution quality.
Market, Limit and Post-Only: What Each Order Actually Instructs
A market order buys speed and gives up price, a limit order fixes price and gives up certainty of filling, and a post-only order keeps you a maker at the cost of rejection. Each is a contract with the matching engine trading price certainty against fill certainty.
A market order says: fill me now, at whatever prices exist. The engine consumes resting liquidity from the top of the book downwards until your quantity is complete. If the best offer only covers half your size, the rest fills at the next level, and the next. You are always the taker, and you always pay the spread plus whatever depth you consumed.
A limit order says: fill me, but never worse than this price. If you price it at or through the opposite side of the book — a buy at or above the best offer — it crosses immediately and behaves like a capped market order, so you are a taker who simply cannot be filled beyond your limit. If you price it inside the spread, your buy becomes the new best bid and rests there as visible liquidity, earning maker treatment if someone trades against it. Price it further away from the market and it still rests, just deeper in the queue.
That distinction is the most misunderstood point in retail execution. A limit order does not itself earn maker treatment; resting in the book before a trade occurs is what earns it. A limit order priced aggressively enough to cross is a taker order with a safety rail.
A post-only order closes that loophole. It instructs the engine to reject or reprice the order if it would cross the spread and take liquidity. Venues label it post-only, maker-only or add-liquidity-only. If the market moves into your price before the order arrives, the order is cancelled rather than filled as a taker.
Two more are worth knowing. A priced (limit) immediate-or-cancel order fills whatever is available at your limit or better right now and cancels the remainder, which caps how far down the book you walk — an unpriced market IOC still walks whatever depth it needs. Fill-or-kill executes the full quantity or nothing, which protects you from partial fills but hands you a lot of rejections on thin books.
| Order type | Price certainty | Fill certainty | Fee role | Main hidden cost |
|---|---|---|---|---|
| Market | None | Very high | Always taker | Spread plus depth walked |
| Limit (crossing) | Capped | High | Taker | Spread, partial fills |
| Limit (resting) | Exact | Uncertain | Maker | Non-fill, adverse selection |
| Post-only | Exact | Lowest | Maker guaranteed | Rejections in fast markets |
| Immediate-or-cancel (priced) | Capped | Partial | Taker | Unfilled residual size |
| Stop-market | None | Very high | Taker | Often fires into stressed liquidity |
Maker and Taker Are Roles, Not Discounts
You are a maker when your order rests in the book and someone trades against it, and a taker when your order removes liquidity already sitting there. Retail traders treat the maker rate as a discount to chase. It is closer to payment for a service you provide, and that service carries risk.
When your limit order rests in the book, you are quoting a price to the entire market. Anyone who wants to trade against you can. The people most eager to hit your bid are the ones who believe the price is about to fall through it. This is adverse selection, and it is why maker rates sit below taker rates in the first place.
The practical consequence: a strategy that switches from market to post-only orders does not simply pocket the fee difference. It swaps a known, immediate cost for an uncertain one. Your fills now cluster around moments when the market is moving against your position, and your unfilled orders cluster around moments when you were right.
The honest rule is that order choice should follow the strategy’s edge, not the fee schedule. If your signal decays in seconds, paying the spread is a cost of doing business. If your signal has a horizon of hours or days, queueing patiently may recover some of that cost — though the saving is not guaranteed, because unfilled orders have their own price.
How Thin AUD Order Books Manufacture Execution Costs
AUD-denominated books are generally thinner than their USD and USDT equivalents, and that is not a criticism of any particular exchange. Australia is a small share of global crypto volume, so the same asset quoted in AUD will often have fewer resting orders at each price level than it does offshore. Depth varies by venue and by moment, so treat it as a tendency to check rather than a constant.
Depth matters more than spread. A tight top-of-book spread tells you what one small order costs, not what happens when your order exceeds the quantity sitting at that best price. On a shallow AUD book, a market order of quite ordinary retail size can clear several price levels, and each level is worse than the last.
Liquidity is also uneven across the day. Depth on AUD pairs can be best when offshore sessions overlap with Australian hours, and may thin out in the local overnight window and across weekends. A bot that rebalances on a fixed schedule at a quiet hour may be systematically choosing poor liquidity.
The gradient across assets is often steeper than people expect. BTC/AUD and ETH/AUD are the most heavily traded local pairs, so they typically carry more depth than anything else quoted in AUD, but which venue is deepest at any given moment is not stable. Mid-cap and small-cap tokens quoted against AUD can have books that a single retail order visibly moves, and the deeper liquidity for those assets may sit in a USDT pair offshore. Pull the live book for your specific pair on your specific venue rather than assuming any ranking holds.
One Australian-specific wrinkle deserves attention. Not every local platform exposes a public order book on every screen; some surfaces quote you a single price with the spread built in, which is a dealer-quote model rather than an exchange match. Where that is the case, order type selection has little to work with, because you are accepting a quote and the spread is the cost. Check whether the venue and the specific interface you are using publish a book at all before assuming your limit order can improve on it.
The related trap is the instant-buy interface many exchanges offer alongside their trading screen. These often bundle a flat fee with a wider quoted spread, so the all-in cost can be materially higher than the same trade on the order book — compare the two quotes yourself, since the gap varies by venue, asset and size. For automated flow, routing through the order book API is generally cheaper where a book exists.
Measuring Crypto Order Types Slippage Instead of Guessing
Measure crypto order types slippage as implementation shortfall — the signed gap between your decision mid-price and your volume-weighted fill. You cannot manage it without logging it, and almost nobody does. Every trade your system places should record four numbers: the mid-price when the signal fired, the mid-price when the order reached the exchange, the volume-weighted price you actually filled at, and the fee charged.
The difference between decision mid and final fill, expressed in basis points and signed against your direction, is your implementation shortfall. It decomposes usefully. The gap between decision and arrival is latency cost, which you fix with infrastructure. The gap between arrival and fill is spread and impact, which you fix with order choice and sizing.
Look at the distribution, not the average. Execution costs are skewed — most fills are unremarkable and a small number are dreadful, usually clustered in volatile minutes or thin hours. An average that looks acceptable can hide a tail that quietly eats the strategy’s edge.
Post-only strategies need a second metric: fill ratio. If you are only filling a fraction of the orders you place, your live strategy is not the strategy you backtested. It is a different, smaller, differently selected strategy that happens to share a name with it.
This is also where most backtests go wrong. Assuming maker fills on every order produces a beautiful equity curve that live trading will never reproduce. The defensible assumption is that you pay the spread unless you can demonstrate, from your own fill logs, that you reliably rest and get filled.
Sizing: The Variable You Control Most Directly
Order size relative to visible depth drives more of your slippage than order type does. An order that fits comfortably inside the quantity at the best price costs you the spread and nothing more. An order that is a meaningful fraction of that quantity walks the book, and the cost rises faster than the size does.
The fix is unglamorous: slice it. Breaking a large order into smaller child orders spread over minutes lets the book refill between executions. This is what execution algorithms do, and a crude time-sliced version is not difficult to build against most exchange APIs.
Do not slice so aggressively that you trip rate limits. Exchange APIs cap request frequency, and documented limits vary by venue and endpoint, with order placement usually tighter than market data. A throttled order-management loop is a far worse problem than a few basis points of impact, so read your venue’s current per-endpoint limits before tuning the slicing interval.
Stop orders deserve their own warning. A stop-market order tends to trigger while the market is moving fast in the direction you feared, and books often thin out under exactly those conditions as market makers widen or pull quotes. The realised fill can therefore sit well away from your stop level, sometimes by a lot. A stop-limit avoids that, at the cost of possibly not filling at all, which on a risk stop is its own kind of disaster.
Matching the Order Type to Your Strategy’s Edge
There is no universally correct order type, only a correct one for the horizon of your signal. The question to ask is what a one-minute delay costs your strategy. If the answer is nothing much, queue. If the answer is the entire trade, pay the spread and stop optimising the wrong variable.
- Grid and market-making strategies: post-only is usually the natural fit. The strategy is structurally a liquidity provider, and taking liquidity on both legs can erode or erase the edge.
- Mean reversion: resting limits at your target entry. You want to be filled by impatient sellers, which is exactly what a resting bid collects.
- Momentum and breakout: crossing limit orders with a capped tolerance often suit better. A pure market order gives away price certainty for little gain, and a post-only order may miss the moves that work.
- Dollar-cost averaging: resting limits placed patiently and refreshed if unfilled. There is no urgency in a monthly accumulation, so paying for urgency is pure waste.
- Risk exits and liquidation: many traders prioritise taking liquidity here. Certainty of exit tends to be the point, and haggling over basis points while a position runs against you is how small losses become large ones.
Be honest about whether any of this deserves your attention. Execution quality compounds with turnover, so it matters enormously to a bot placing dozens of orders a week and barely at all to a monthly accumulator making a handful of modest BTC/AUD trades.
One record-keeping note, since execution and tax intersect. Slippage is never a separate deduction — it is embedded in your fill price, and that price and your fees flow into the figures you report to the Australian Taxation Office. An investor holds crypto as a CGT asset, so they land in the cost base and proceeds of a capital gain or loss; someone carrying on a business of trading holds it as trading stock on revenue account, where those costs are deductible and proceeds are ordinary income.
Which applies is not something you elect. It follows from your facts and circumstances, and trading frequently, in large volumes or with sophisticated tools does not by itself make you a business. Check the ATO’s guidance and confirm your position with a registered tax agent.
Frequently Asked Questions
In short: order type decides whether you pay the spread or earn the maker rate, and order size relative to book depth decides how far past the touch you fill.
Does using a limit order guarantee I pay maker fees?
No. A limit order priced so that it crosses the current spread executes immediately against resting liquidity, which makes you the taker regardless of the order type label. Only an order that rests in the book before being traded against earns maker treatment. If you want that guaranteed, use post-only rather than assuming a limit order will do it.
When should I use post-only instead of a normal limit order?
Use post-only whenever being filled as a taker would make the trade not worth doing. Grid bots and market-making strategies are the clearest cases, because taker fees on both legs can exceed the spread they are trying to capture. The trade-off is that post-only orders get rejected when the market moves into your price, so your system needs to handle rejections gracefully rather than treating them as errors.
How much slippage should I expect on AUD crypto pairs?
It depends far more on your order size relative to book depth than on any headline figure. On major AUD pairs at established local exchanges, modest retail orders often fill at or near the best quoted price without walking the book, so the taker cost is roughly the half-spread plus the fee. On thinner AUD markets, or with an order larger than the quantity resting at the best price, the cost can dwarf the fee. Measure it on your own fills rather than trusting a general number.
Are stop-loss orders a hidden source of slippage?
Yes, and they are one of the worst. A stop-market order converts into a market order at exactly the moment liquidity is thinnest and price is moving fastest, so the fill can land well past your intended level. Stop-limit orders cap that damage but introduce the risk of not filling at all. Choose deliberately, based on whether a missed exit or a bad fill is the worse outcome for that position.
Does slippage change my Australian tax position?
Not as a separate item — poor execution just shows up as a worse result. What matters is whether you are investing, which makes your crypto a CGT asset, or carrying on a business of trading, which makes it trading stock on revenue account. That turns on your facts and circumstances rather than your trading volume or the label you prefer. Check the ATO’s current crypto asset guidance and confirm your position with a registered tax agent.
Should I improve execution before or after finding a strategy that works?
After, in almost every case. Execution quality amplifies or erodes an existing edge, it does not create one. If your strategy is unprofitable at realistic taker costs, better order routing will not rescue it. Once you have something that survives conservative cost assumptions, tightening execution is a reasonable next area to work on.
