Backtests assume trades execute at the exact price the strategy signal triggered. In live trading, this assumption breaks down: by the time an order reaches the exchange, the market may have moved; the order may fill across multiple price levels in the order book; or market conditions may have changed between signal generation and order submission. The gap between expected and actual fill price is slippage. For strategies with small profit-per-trade (grid trading, high-frequency mean reversion), even modest slippage of 0.1–0.2% per trade can consume a significant fraction of the expected profit — or turn a marginally profitable strategy into a losing one. Understanding slippage sources and controls is essential before deploying any real-money automated strategy.
Related guides: backtesting, stress testing, grid trading, position sizing.
What Causes Slippage?
1. Order Book Depth (Most Common)
Every market order fills against the best available offers in the order book. If your order size exceeds the quantity available at the best ask price, the remainder fills at progressively worse prices (walking up the order book). A $50,000 market buy on BTC/USDT with only $8,000 at the best ask and $15,000 at the next level will fill across multiple price levels — the average fill price will be higher than the best ask at order submission time.
2. Latency Between Signal and Order Submission
For DennTech running locally, the time between signal generation and order reaching the exchange includes: local processing time, internet latency to the exchange API, exchange order processing queue. On volatile assets, price can move meaningfully in 50–500 milliseconds. For long-term strategies (Daily or 4H signals), this latency is irrelevant. For 1-minute or 5-minute strategies, latency slippage becomes more material. Using a VPS in the same datacenter region as your exchange reduces latency significantly. See our VPS guide.
3. Market Volatility at Entry
During high-volatility events (exchange announcements, macro news, large whale trades), bid-ask spreads widen dramatically and order books thin out. A market order during a volatility spike can fill 0.5–2% away from the pre-spike price. Strategy signals that trigger near major news events are particularly susceptible.
4. Exchange Fee Model Impact
Market orders (taker orders) consume order book liquidity and pay the taker fee (typically 0.05–0.20%). Limit orders (maker orders) add liquidity and pay the lower maker fee (often 0.00–0.10%). Switching from market to limit orders reduces fee costs AND eliminates order-book slippage — at the cost of potential non-execution if price moves away before the limit fills.
Measuring Expected Slippage by Order Size and Pair
To estimate your strategy's expected slippage, check the exchange order book depth at your typical order size:
- Open the Level 2 order book for your target pair
- Simulate your typical order size as a market buy
- Calculate the volume-weighted average fill price across the levels consumed
- Compare to the best ask price at that moment
- The percentage difference is your estimated market order slippage for that size
Rule of thumb for liquid pairs (BTC/USDT, ETH/USDT on Binance, OKX, Bybit):
- Under $10,000 order: typically 0.01–0.05% slippage
- $10,000–$50,000: typically 0.05–0.15% slippage
- Above $50,000: 0.15%+ slippage — use limit orders or split into multiple orders
Limit Orders vs. Market Orders for Bots
| Order Type | Slippage Risk | Execution Risk | Best For |
|---|---|---|---|
| Market Order | Yes — fills at current book price | Always fills immediately | Fast-moving signals, stop-losses, momentum entries |
| Limit Order (aggressive) | Minimal — set just inside spread | May not fill if price moves away | Most indicator-based entries (RSI, EMA, Bollinger) |
| Post-Only Limit | Zero slippage | Cancelled if would immediately fill (becomes taker) | Grid trading, DCA, maker-only strategies |
DennTech's default for most strategies uses limit orders with a configurable offset (e.g., entry limit = signal price + 0.05% to improve fill probability). This balances fill reliability with slippage minimization. See strategy-specific order type recommendations in our DennTech docs.
Including Slippage in Backtests
DennTech's backtester includes configurable slippage assumptions:
- Navigate to Strategy → Backtest → Execution Settings
- Set slippage: 0.05% per side (0.10% round-trip) for liquid pairs; 0.10–0.20% per side for mid-cap altcoins
- Set commission: actual exchange taker fee (e.g., 0.10% for Binance standard)
- Run backtest with and without slippage to quantify the realistic impact on your strategy
Strategies that show positive Profit Factor without slippage but negative with realistic slippage included are not viable for live deployment at that pair/size combination. See our backtesting guide. Compare editions at the pricing page.
Frequently Asked Questions
- How much does slippage typically affect grid trading profitability?
- Grid trading is particularly slippage-sensitive because it executes many small trades at tight profit margins per grid level. A grid with 0.5% profit-per-level and 0.10% round-trip slippage retains 0.40% profit after slippage — an 80% slippage impact ratio. Widen grid spacing to at least 5× your expected round-trip slippage: for 0.10% round-trip slippage, minimum grid spacing should be 0.50%. On liquid pairs (BTC/USDT, ETH/USDT), market order slippage is typically negligible for grid-sized orders (usually under $5,000 per level). On thin altcoin pairs, slippage can consume entire grid levels. Use post-only limit orders for grid strategies to eliminate slippage entirely — the non-fill risk is acceptable since the grid can wait for price return. See our Grid guide.
- Does running DennTech on a VPS significantly reduce latency slippage?
- For Daily and 4H timeframe strategies, VPS latency improvement is negligible — signal execution at these timeframes is not sensitive to 100–500ms latency differences. For 1H and below strategies, a VPS co-located with the exchange datacenter (e.g., AWS Tokyo for Binance, AWS Frankfurt for Kraken) can reduce round-trip API latency from 100–200ms to 10–30ms. At 1H bar closes, this difference rarely matters for a few hundred milliseconds. For 1-minute or 5-minute bar strategies where latency slippage becomes meaningful, VPS placement does reduce execution costs. For most DennTech users running 4H+ strategies, a standard VPS in any region is sufficient. See our VPS guide. Compare editions at the pricing page.
- Should I use market orders or limit orders for DennTech stop-loss execution?
- Stop-loss orders should always use market orders (or stop-market orders) for guaranteed execution. A limit-based stop-loss risks non-execution if price gaps through the limit level — which happens during flash crashes, liquidation cascades, or large news events. The exact scenarios where you most need stop-loss protection are the same scenarios where limit-based stops can fail. The slippage cost of a market stop-loss execution is always preferable to a non-executed limit stop during a rapidly declining market. For entry orders, limit orders are preferred to minimize slippage. For exit and stop-loss orders, market execution guarantees the protection function. See our stop-loss guide and start at the pricing page.
Execution quality: Slippage (this guide), Backtesting, Stress Testing. All strategies at the strategies page.