Best Crypto Trading Strategies: Decision Frameworks for Onchain and Centralized Execution

Trading crypto profitably requires matching execution mechanics to market structure, asset liquidity, and position duration. This article walks through six strategy archetypes used by practitioners, focusing on the technical assumptions each relies on, the instrumentation required, and the failure modes that separate backtest performance from live results.

Market Making on Centralized Order Books

Market making captures bid ask spread by continuously quoting both sides. You earn the spread when both legs fill, minus exchange fees and adverse selection costs.

The core mechanic: place limit orders at increments around the midpoint and adjust them as the market moves. Profitability depends on fill rate, fee tier, and inventory risk. On most centralized exchanges, VIP fee tiers reduce take fees to 0.02 to 0.05 percent and grant make rebates of 0.00 to 0.02 percent, making tight spreads viable only for accounts with sufficient volume history.

Adverse selection is the hidden cost. When your bid fills during a dump, the market often continues lower before you can exit. Your effective spread shrinks by the average adverse move. Practitioners mitigate this by widening quotes during volatility spikes (measured by rolling standard deviation of midpoint moves) or pulling orders entirely when order book imbalance exceeds a threshold, typically when the bid or ask depth within 20 basis points of mid drops below 2x your position size.

Inventory management matters more than spread optimization. If you accumulate a long position in a falling market, financing costs (opportunity cost or actual borrow rates for hedging) erode profit. Many strategies employ a skew parameter that shifts quote placement: if inventory exceeds a target (e.g., more than 10 percent of capital in base asset), widen the ask and tighten the bid to encourage mean reversion.

Automated Market Maker Liquidity Provision

Providing liquidity to an AMM like Uniswap V3 or Curve involves depositing a pair of tokens into a pool and earning a fraction of swap fees. Unlike order book making, you do not control quote placement beyond choosing a price range (in concentrated liquidity models).

Impermanent loss is the key risk. When the price moves outside your range or diverges significantly from your entry ratio, the pool rebalances your position by selling the appreciating asset and buying the depreciating one. Your position underperforms holding the tokens. The loss is “impermanent” only if price returns to the original ratio before you withdraw.

Fee yield must exceed impermanent loss plus gas costs. A position in a volatile pair with 30 basis point fees might generate 50 to 200 percent APR in fee income during high volume periods, but suffer 10 to 40 percent impermanent loss over the same window. The net depends on range width, volatility, and volume.

Practitioners track the fee-to-IL ratio by comparing cumulative fees earned (available from subgraph queries or pool contract events) to the difference between hold value and current position value. Positions in stablecoin pairs or correlated assets (e.g., ETH and staked ETH derivatives) exhibit lower IL and suit passive strategies. Volatile pairs require active range rebalancing, which incurs gas and swap costs.

Gas optimization becomes critical for smaller positions. Rebalancing a Uniswap V3 position (withdraw, swap, redeposit) can cost 0.01 to 0.03 ETH depending on network congestion. A 1,000 USD position paying 3 percent monthly fees earns 30 USD, but two rebalances erase most of that at current gas prices. Large positions or layer two deployments change the calculus.

Trend Following with Perpetual Futures

Trend strategies enter long or short positions based on price momentum indicators and hold until the trend reverses. Perpetual futures offer leverage, no expiry, and funding rate arbitrage opportunities.

The simplest implementation uses moving average crossovers. Enter long when a short period moving average (e.g., 20 hour) crosses above a long period average (e.g., 50 hour), exit when it crosses back. Variations add filters like ADX to confirm trend strength or volume thresholds to avoid false signals in low liquidity.

Funding rates create a secondary profit or cost layer. When longs outnumber shorts, long holders pay funding to shorts every 8 hours, typically at a rate derived from the premium between perpetual and spot price. During sustained uptrends, funding can reach 0.1 to 0.3 percent per interval, adding significant drag to long positions. Short strategies in overheated markets collect this funding while profiting from reversals.

Position sizing and stop losses determine survival. A strategy using 5x leverage with a 10 percent stop equates to a 2 percent account risk per trade. Practitioners often risk 0.5 to 2 percent of capital per position and scale exposure with realized volatility: lower leverage during high volatility periods prevents liquidation during normal retracements.

Backfill bias is a common backtest error. Indicators calculated on historical OHLCV data assume you had access to the close price at the exact interval boundary, but live execution faces slippage and order book depth constraints. A backtest showing 3 percent average gain per trade may realize 2.5 percent after factoring 0.5 percent round trip costs.

Arbitrage Between Spot and Derivatives

Price discrepancies between spot exchanges, perpetual futures, and expiring futures create arbitrage opportunities. The simplest form: if BTC trades at 30,000 on Exchange A and 30,100 on Exchange B, buy on A and simultaneously sell on B, pocketing the spread minus fees and transfer costs.

Cross exchange arbitrage requires capital on multiple venues and fast execution. Transfer time introduces risk: if you buy spot on one exchange and need to transfer BTC to another to sell, price may move against you during the 10 to 60 minute confirmation window. Practitioners pre position balances or use stablecoins for faster settlement.

Cash and carry arbitrage exploits futures premiums. When a quarterly future trades above spot, buy spot and short the future, holding until expiry. The premium converges to zero, locking in the initial spread. This requires margin for the short and assumes no dramatic spot price collapse that triggers liquidation before expiry.

Funding rate arbitrage applies to perpetual markets. When funding is persistently positive, short the perpetual and hedge with spot or options. You collect funding every interval while remaining delta neutral. The strategy fails if funding flips negative or if your hedge position incurs higher financing costs than funding collected.

Statistical Arbitrage and Pairs Trading

Stat arb identifies pairs or baskets of assets with historical price correlation and trades deviations from the expected relationship. In crypto, this often means trading ratios like ETH/BTC or Layer 1 token baskets.

The core assumption is mean reversion. If ETH/BTC trades at 0.055 and the 90 day mean is 0.052 with a standard deviation of 0.002, a position shorting ETH and longing BTC when the ratio exceeds 0.054 (one standard deviation) expects profit as the ratio reverts. Entry and exit thresholds typically use z scores or Bollinger Bands.

Cointegration tests improve pair selection. Two assets can be correlated but not cointegrated, meaning their spread drifts over time. The Engle Granger test or Johansen test identifies pairs where the spread is stationary, making mean reversion trades more reliable. Practitioners retest cointegration monthly to avoid trading relationships that have broken.

Position sizing accounts for correlation volatility. If the spread exhibits increasing variance, widen thresholds or reduce leverage to avoid early stop outs. A spread that historically reverted within 48 hours may take a week during regime changes, stressing capital and margin.

Grid Trading in Range Bound Markets

Grid strategies place buy and sell orders at fixed intervals above and below a midpoint, profiting from oscillations within a range. Each time price moves down one grid level, a buy order fills, then price moving up one level fills the corresponding sell order for a small gain.

Grid spacing determines profitability and risk. Tight grids (e.g., 0.5 percent intervals) generate more trades but require higher volume to overcome fees. Wide grids (e.g., 2 percent intervals) suit lower liquidity pairs but risk fewer fills.

The strategy assumes price oscillates without trending. A sustained breakout in either direction leaves you with a fully long or fully short position and no remaining grid orders to capture reversals. Practitioners often set stop boundaries: if price moves beyond the grid range by a threshold (e.g., 10 percent), flatten the position and reinitialize the grid.

Worked Example: Funding Rate Arbitrage on a Perpetual Contract

You observe ETH perpetual funding consistently positive at 0.05 percent per 8 hour interval. Spot ETH trades at 2,000 USDT.

1. Short 10 ETH perpetual contracts (notional value 20,000 USDT) on the derivatives exchange.
2. Buy 10 ETH on spot for 20,000 USDT to hedge delta.
3. Every 8 hours, collect 0.05 percent of 20,000 USDT = 10 USDT in funding.
4. Over 30 days (90 intervals), cumulative funding: 900 USDT.
5. Costs: exchange fees (assume 0.05 percent on entry) = 10 USDT perpetual, 10 USDT spot = 20 USDT. Net: 880 USDT on 20,000 capital = 4.4 percent monthly return.

Risk: funding flips negative if sentiment shifts. A 0.05 percent negative funding costs 10 USDT per interval, eroding profit. Exit when 7 day average funding drops below 0.01 percent.

Common Mistakes and Misconfigurations

• Ignoring exchange withdrawal limits when planning cross exchange arbitrage. Some platforms cap daily withdrawals, trapping capital during volatile periods.
• Using market orders for limit strategy backtests. Limit orders experience partial fills and queue priority effects not captured by assuming instant execution at midpoint.
• Failing to account for funding in perpetual backtest PnL. Historical funding rates are available from exchange APIs but often omitted from simplified simulations.
• Over optimizing parameters on a single asset or timeframe. A moving average crossover tuned to BTC 2020 to 2021 data may fail in 2023 to 2024 conditions. Walk forward testing across multiple regimes exposes overfitting.
• Neglecting API rate limits and order placement latency. Strategies that rebalance every minute may hit 1200 requests per hour, exceeding typical limits and causing order rejections.
• Assuming infinite liquidity at listed prices. A backtest selling 100 BTC at the best ask may be impossible live if the order book only offers 5 BTC depth at that level.

What to Verify Before You Rely on This

• Current fee tier and rebate structure for your account on each exchange. VIP tiers change profitability by 5 to 20 basis points per round trip.
• Funding rate calculation methodology for perpetual contracts. Some exchanges use 8 hour intervals, others 1 hour. The rate formula varies (mark price premium vs index, capped or uncapped).
• Order book depth at your typical trade sizes. Use the exchange API to fetch L2 or L3 data and calculate slippage for orders 2x and 5x your position size.
• Gas costs for onchain actions if using AMMs. Simulate a full rebalance cycle (remove liquidity, swap, add liquidity) on a testnet or via gas estimation APIs.
• Margin and liquidation rules for leveraged positions. Understand maintenance margin, whether liquidations are partial or full, and if insurance funds cover socialized losses.
• Latency from your execution environment to exchange APIs. Measure round trip time for market data and order placement. Sub 50 ms latency matters for market making, less critical for trend strategies.
• Historical volatility and correlation stability for stat arb pairs. Recalculate cointegration tests and rolling correlation over recent windows to confirm relationships hold.
• Withdrawal processing times and fees for capital movement between venues. A 24 hour withdrawal hold or 0.5 percent fee changes arbitrage economics.
• Tax treatment of frequent trading in your jurisdiction. Wash sale rules, like kind exchange history, and short term capital gains rates affect net returns.

Next Steps

• Implement a single strategy on a testnet or paper trading account, logging every order, fill, and PnL component to audit execution vs backtest assumptions.
• Build monitoring dashboards for key metrics: realized spread, funding collected, impermanent loss ratio, or sharpe ratio over rolling windows. Deviations signal parameter drift or regime change.
• Automate parameter recalibration or strategy pausing when market conditions move outside historical norms, defined by volatility percentiles, volume thresholds, or correlation breakdowns.

Category: Crypto Investment Strategies