Crypto Algorithmic Trading Strategies: The Main Types Explained for Australians
Crypto algorithmic trading strategies are not some obscure institutional dark art. I have been running automated strategies on Australian exchanges since 2022, and the core concepts are accessible to anyone willing to put in the time to learn them properly.
> TL;DR: Crypto algorithmic trading strategies automate buy and sell decisions using rules, signals, and code. The main types include trend following, mean reversion, arbitrage, market making, and momentum strategies. Each suits different risk profiles, capital levels, and technical skill sets. Australian traders can legally run these strategies on their own accounts without an AFSL, but must use AUSTRAC-registered exchanges and keep detailed records for CGT purposes.
Why Algorithmic Trading Strategies Matter for Crypto Traders
Bitcoin does not care that it is 3am on a Tuesday. Neither does the altcoin you have been watching all week. Crypto markets run continuously, every hour of every day, and that is both the opportunity and the problem for anyone trying to trade manually.
The exhaustion issue is real. Miss a breakout at 2am and you have missed it. React emotionally to a 15% drawdown at midnight and you might close a position that would have recovered by morning. Crypto algorithmic trading strategies exist, at their core, to solve these two problems: coverage and emotion. A properly configured bot does not panic. It does not overtrade out of boredom. It follows the rules you set, exactly, every time.
For Australian traders, there is a specific legal context worth understanding before you touch any of this. Running automated strategies on your own account is completely legal in Australia without an Australian Financial Services Licence (AFSL). You are managing your own money, not someone else’s. The key compliance requirement is that any exchange you connect your bot to must be registered with AUSTRAC as a Virtual Asset Service Provider (VASP), the designation that replaced Digital Currency Exchange (DCE) following the April 2026 AML/CTF law changes. If the exchange is not on the AUSTRAC register, do not use it, regardless of how good the API looks.
This article covers the five main strategy types in real depth: trend following, mean reversion, arbitrage, momentum, and market making. I have also included grid trading and DCA bots for those who want automation without writing a single line of code.
Trend Following: The Most Popular Crypto Algo Strategy
The logic is blunt: buy when price is rising, exit or short when it is falling. Trend following is the most widely used of all crypto algorithmic trading strategies, and for good reason. Crypto has historically produced pronounced multi-week directional moves that trend systems are built to capture. The 2020-2021 BTC bull run, the 2022 collapse, the 2023 recovery, each of those produced months of sustained directional movement that a basic trend-following bot would have caught a meaningful chunk of.
The most common indicators used in these strategies are exponential moving average (EMA) crossovers, such as the 20 EMA crossing above the 50 EMA as a buy signal, and simple moving averages (SMA) for slower, more conservative signals. Donchian channels, which mark the highest high and lowest low over a set period, are another staple, particularly for breakout-based trend entries. The Average Directional Index (ADX) is often added as a filter: only take trades when ADX is above 25, indicating a genuine trend rather than noise.
The weakness is equally well-known. Trend-following bots get chopped to pieces in ranging markets. If BTC is oscillating between $95,000 and $100,000 AUD for three weeks, an EMA crossover system will generate false signals repeatedly, buying near the top and selling near the bottom of that range. This is called whipsawing, and it is where most trend-following accounts bleed during flat periods.
For building and testing these strategies, TradingView with Pine Script is the most accessible starting point. You write the strategy logic in Pine Script, backtest it against historical data, and can connect alerts to execution via webhook. Freqtrade is the other serious option, an open-source Python framework with a strong community and the ability to run locally on your own hardware. QuantConnect suits traders who want cloud-based backtesting with more sophisticated statistical tools.
Trend following is broadly suitable for traders with $5,000 AUD or more, a medium risk tolerance, and the patience to accept drawdown periods while waiting for the next genuine trend.
Mean Reversion: Profiting When Prices Snap Back
Where trend following bets that price will keep moving in one direction, mean reversion bets the opposite: that after an extreme move, price will return toward its historical average. The statistical foundation is sound. Most financial assets, including crypto over shorter timeframes, do not move in straight lines. They overshoot and correct.
The key indicators for mean reversion strategies are Bollinger Bands, RSI, and Z-scores. Bollinger Bands place bands two standard deviations above and below a moving average. When price touches the lower band on a day with RSI below 30, that is a potential long entry. When price touches the upper band with RSI above 70, that is a potential short or exit signal. Z-score measures how many standard deviations price is from its mean, making it useful for more mathematically precise entry thresholds.
Mean reversion works best in sideways or range-bound markets. The problem in crypto is that assets can trend violently for weeks, and a mean reversion position betting on BTC to come back down from $110,000 AUD can get absolutely destroyed if the move continues to $130,000. This is the strategy type with the most asymmetric blowup risk if you run it without hard stop losses.
One interesting variant is pair trading, a mean reversion approach applied to the ratio between two correlated assets. The BTC/ETH ratio, for example, tends to revert around a historical mean. If ETH sells off sharply relative to BTC, a pair trade goes long ETH and short BTC, betting the ratio will normalise. This is more complex to implement but reduces directional market exposure.
Backtesting is non-negotiable for mean reversion. You must validate strategies against real Australian exchange data and include realistic fee assumptions. A strategy that looks profitable at 0.1% per trade may not be at 0.6%, which is closer to the effective spread on some AUD pairs during volatile periods.
Arbitrage Strategies: Capturing Price Gaps Across Exchanges
Arbitrage is theoretically risk-free, in the textbook sense. In practice, it is fast, capital-intensive, and increasingly difficult as institutional bots have narrowed most obvious price gaps to milliseconds.
Spatial arbitrage is the most straightforward type: buy BTC on Swyftx where it is trading at $98,500 AUD, simultaneously sell on Kraken where it is showing $98,800 AUD, and pocket the $300 difference minus fees and transfer costs. The problem for Australian traders is AUD transfer times. Moving fiat between two Australian exchanges via bank transfer can take hours or days. By the time funds settle, the price gap has long closed. The practical solution is to pre-fund both exchanges and trade the crypto leg only, but that ties up capital on both sides.
Triangular arbitrage happens within a single exchange and exploits price inconsistencies between three trading pairs. If BTC/AUD, ETH/AUD, and ETH/BTC are not perfectly aligned, you can cycle through all three and extract a small profit. The maths is more involved and the windows are genuinely tiny, often less than a second, so this requires API access and fast execution code.
Statistical arbitrage uses regression models and correlation analysis to identify when two historically correlated assets have diverged beyond their expected relationship, then trades them back toward equilibrium. This overlaps with pair trading from the mean reversion section but uses more sophisticated modelling.
For Australian traders, Swyftx, Kraken, and CoinSpot all offer API access, though latency and rate limits vary. Kraken generally has the most comprehensive API for serious algo work. Profit margins on arbitrage are thin, often under 0.2% per trade before fees, and you are competing against bots running on co-located servers. It is not the most accessible starting point for most retail traders.
[INTERNAL LINK PLACEHOLDER: best API access for algo trading → exchange comparison pillar]
Momentum Strategies: Riding Short-Term Crypto Surges
Momentum is related to trend following but operates on a shorter timeframe and different logic. Trend following confirms a sustained directional move before entering. Momentum strategies enter based on the strength of a very recent price move, sometimes within the last few minutes or hours, betting that strength will continue briefly before reversing.
Common momentum signals include volume surges (a sudden spike in trading volume often precedes a significant price move), breakouts from a consolidation range, and RSI crossing specific thresholds in a way that indicates accelerating rather than sustained momentum. A typical setup might be: if volume increases by 300% above the 20-period average and price breaks above a 4-hour resistance level, enter long with a 2% stop loss and 4% take profit target.
Holding periods are short, from minutes to a few hours. This creates a specific problem for Australian traders: tax reporting complexity explodes. The ATO treats every crypto trade as a taxable event, meaning a bot generating 50 trades per day is creating 50 separate CGT events daily. Over a year, that is potentially thousands of transactions to reconcile. Using crypto tax software like Koinly or Crypto Tax Calculator is not optional at that frequency, it is essential.
Momentum strategies suit traders with higher risk tolerance, the technical ability to monitor and adjust their systems actively, and the infrastructure to handle detailed record keeping. They are among the more demanding of the automated crypto trading strategies in terms of ongoing management.
[INTERNAL LINK PLACEHOLDER: crypto tax reporting for algo traders → tax software guide]
Market Making: Providing Liquidity for a Fee
Market making is the strategy that powers exchange liquidity. A market making bot simultaneously posts a buy limit order slightly below the current market price and a sell limit order slightly above it. When both orders fill, the bot captures the bid-ask spread as profit. If the spread is 0.1% and the bot cycles through enough volume, the returns can be consistent, but the capital requirements and risks are significant.
The primary risk is called inventory risk. If BTC drops 8% sharply while your bot has just filled a large buy order and is waiting for the sell, you are sitting on a loss. Market making bots need sophisticated inventory management logic to hedge or exit positions when directional moves accelerate.
Hummingbot is the most widely used open-source tool for market making bots in the retail and semi-professional space. It supports multiple exchanges via API, has a growing strategy library, and does not require you to build the execution infrastructure from scratch. That said, configuring it properly for an Australian exchange with AUD pairs still requires meaningful technical knowledge.
The regulatory point here is important. Running a market making bot on your own account is legal in Australia. Offering market making as a service to other people’s accounts is a different matter entirely and likely falls under managed investment scheme regulations, requiring AFSL registration. Do not blur that line.
For Australian traders, the most sensible pairs for market making are BTC/AUD and ETH/AUD on high-liquidity exchanges. Illiquid pairs have wider spreads but far more inventory risk when price moves. I would not recommend starting market making with less than $20,000 AUD, and even then, start with very small order sizes while you understand how your specific exchange’s order book behaves.
Grid Trading and DCA Bots: Beginner-Friendly Automated Strategies
Not all crypto algorithmic trading strategies require writing code. Grid trading and dollar-cost averaging bots are accessible to anyone who can navigate a settings page.
Grid trading works by placing a series of buy and sell limit orders at regular intervals above and below a set price. If BTC is at $98,000 AUD, a grid bot might place buy orders every $1,000 down to $90,000 and sell orders every $1,000 up to $106,000. Each time price oscillates within that range, the bot buys low and sells slightly higher, accumulating small profits. It works well in ranging markets and is genuinely painful in strong directional trends, particularly if price breaks below your lowest buy order and keeps falling.
DCA bots are even simpler. Set a fixed AUD amount, set a frequency (weekly, fortnightly, daily), and the bot automatically purchases that amount of crypto on schedule. This removes timing decisions entirely and reduces the emotional component of accumulation. On Australian exchanges with PayID deposits, you can fund your account and run a DCA strategy on BTC or ETH with minimal friction. I have used a DCA setup on Swyftx for ETH accumulation and the process is straightforward.
Platforms like 3Commas offer both grid and DCA bots without requiring coding knowledge. Swyftx has built-in recurring buy functionality for DCA without needing a third-party tool.
The tax implication is significant and often overlooked by beginners: every DCA purchase is a separate CGT acquisition event. If you are buying weekly, that is 52 separate cost-base records per year, per asset. After two or three years of DCA, your tax calculations become genuinely complex without proper software. Start using crypto tax software from day one, not when you are staring down thousands of transactions before the June 30 lodgement deadline.
[INTERNAL LINK PLACEHOLDER: best DCA platforms for Australian traders → platform comparison]
Comparing the Main Crypto Algo Trading Strategies
| Strategy | Market Condition | Coding Required | Min. Capital (AUD) | Tax Complexity | Risk Level |
|---|---|---|---|---|---|
| Trend Following | Trending | Yes (or platform) | $2,000+ | Medium | Medium |
| Mean Reversion | Ranging | Yes | $3,000+ | Medium | Medium-High |
| Arbitrage | Any | Yes, advanced | $10,000+ | High | Medium |
| Momentum | Any | Yes | $2,000+ | Very High | High |
| Market Making | Ranging/Liquid | Yes, advanced | $20,000+ | High | High |
| Grid Trading | Ranging | No | $500+ | Medium | Medium |
| DCA Bots | Any | No | $50+ | Low-Medium | Low |
FAQ
Are crypto algorithmic trading strategies legal in Australia?
Yes. Running automated strategies on your own account is legal in Australia and does not require an AFSL. You do need to use AUSTRAC-registered exchanges (now classified as VASPs following the April 2026 AML/CTF reforms) and comply with ATO tax reporting requirements.
How much capital do I need to start algo trading crypto in Australia?
It depends heavily on the strategy. DCA bots can start with $50 AUD. A functional trend-following setup needs at least $2,000 to absorb drawdowns meaningfully. Market making and arbitrage strategies require $10,000 to $20,000 or more to be viable.
Do I need to know how to code for algorithmic trading?
For grid trading and DCA bots, no. Platforms like 3Commas and Swyftx offer these without any coding. For trend following, mean reversion, arbitrage, and momentum strategies, you will need either Pine Script (TradingView), Python (Freqtrade), or another programming language. The complexity varies significantly.
How do I handle tax on hundreds of algo-generated trades?
Use crypto tax software from the start. Koinly and Crypto Tax Calculator both support Australian exchanges and can import trade history via API or CSV. Each trade is a separate CGT event under ATO rules, so manual reconciliation at scale is not realistic.
What is the difference between momentum and trend following strategies?
Trend following confirms a sustained directional move over days or weeks before entering. Momentum strategies enter based on the strength of a very recent price move, often within the last few minutes or hours, with shorter holding periods and higher trade frequency.
Can algo trading strategies consistently beat the market?
Honestly, no strategy consistently beats the market across all conditions. Trend-following bots perform well in trending markets and poorly in ranging ones. Mean reversion does the opposite. The goal of most systematic strategies is not to outperform in every condition but to produce consistent, emotionally neutral returns over full market cycles with defined risk parameters.
Which Australian exchanges support API access for algo trading?
Kraken, Swyftx, and CoinSpot all offer API access.
