How do transaction fees determine the fate of a trade? Insights from a real strategy backtest

null

Author: Michel Athayde

Summary

“In high-frequency trading, transaction fees are not just costs; they are the decisive factor determining success or failure.”

This article is based on an initial capital of $100,000 and conducts a one-year full-position backtest of the popular TradingView strategy Squeeze Momentum (LazyBear) on BTC and ETH. The data reveals a brutal reality: a mere 0.04% difference in fees can cause the same strategy’s performance on ETH 15-minute charts to shift from a 47% profit to a 14% loss. This means that for many retail traders and beginner quant traders, the strategy logic itself may be sound, but the fee structure at execution level directly erodes all Alpha.

(Note: This backtest is based on historical data from mainstream exchange perpetual contracts, using standard Squeeze Momentum parameters, without considering slippage, aiming solely to compare the marginal impact of different fee structures on strategy net value.)

1. Core Data Perspective: Fee Sensitivity of BTC and ETH

This study simulates three typical fee scenarios: 0% (benchmark), 0.02% (Maker/limit order fee), and 0.06% (Taker/market order fee). The backtest strategy is adjusted to “Long Only” mode to suit the long-term upward trend of the crypto market.

Results are as follows:

1.1 15-Minute Level: The “Profit Grinder” in High Frequency

The 15-minute chart is the most signal-dense period for this strategy (600-800 trades per year), and also the area where fee effects are amplified to the extreme.

Table 1: Core Data Comparison at 15-Minute Timeframe

Deep Dive:

BTC’s High-Frequency Dilemma: Under the 15-minute high-frequency execution conditions set in this article, BTC’s relatively low volatility (Beta) makes it difficult to cover the fixed costs generated by frequent trading. Even with a low fee of 0.02%, the strategy still yields negative returns. This indicates that, given current market maturity, simple technical breakout strategies may no longer generate sufficient Alpha on short cycles for BTC.

ETH’s Volatility Advantage: ETH shows stronger breakout potential. In ideal conditions (0% fee), its 68.66% return far exceeds BTC. This high volatility allows ETH to maintain a substantial 47.34% profit even at a 0.02% Maker fee.

Taker Cost: The most alarming data point is ETH in comparison. When using market orders (0.06% fee), even capturing major ETH trends, the accumulated fees of up to $76,536 can push the account into loss (-13.81%).

2. Cycle Paradox: Why Lengthening the Cycle Doesn’t Solve the Problem?

It is generally believed that increasing the timeframe reduces trading frequency and thus minimizes fee erosion. However, this backtest reveals an counterintuitive phenomenon at the 1-hour (1H) level.

Table 2: Performance at 1-Hour Timeframe

Analysis:

Even ignoring fees (0%), both BTC and ETH show losses at the 1-hour level (BTC -12.29%, ETH -11.51%). This may relate to the default parameters (20, 2.0) producing lagging signals at higher cycles. When the “compression release” signal is confirmed on the 1-hour chart, the trend has often already started, leading to entries at local highs and subsequent pullbacks triggering stops. This suggests that blindly applying default parameters across different timeframes is highly risky.

3. Key Insights: From “Win Rate” to “Odds” Thinking

3.1 The Mathematical Trap of Break-Even

Squeeze Momentum is a typical trend breakout strategy, with a generally low win rate (backtests show between 26% - 40%), mainly relying on favorable risk-reward ratios.

However, fees mathematically raise the break-even point directly.

For example, on ETH 15m, $76,536 in fees implies that each trade (regardless of profit or loss) implicitly involves an “entry ticket” of about (.

The core conclusion: For strategies with over 600 trades per year, fees have become the primary variable determining survival, rather than the predictive power of the indicator itself.

3.2 Capital Management and Compound Deterioration

This backtest uses 100% position sizing. Under high fee rates, continuous erosion accelerates principal shrinkage. BTC 15m )0.06%$92 experienced a maximum drawdown of 58.32%, which in real trading often means liquidation or psychological breakdown.

4. Practical Implications and Strategy Optimization Directions

Based on the above data, we can offer the following industry advice for traders wishing to deploy such momentum strategies:

4.1 Fee Structure Defines Survival Space

Conclusion: In live trading, such high-frequency strategies are only feasible if stable Maker (limit order) fills can be achieved.

Recommendation: At the algorithmic execution level, prioritize passive limit orders (e.g., at bid price or within order book depth) rather than market orders. If operating on exchanges with high taker fees (>0.05%), proceed with extreme caution.

4.2 Asset Selection: Volatility as the Core Moat

Conclusion: Under fixed fee rates, the asset’s volatility must be sufficient to cover costs.

Recommendation: Compared to BTC, Squeeze Momentum strategies are better suited for ETH or other high Beta mainstream altcoins. The gradually “asset-like” low volatility of BTC makes it less cost-effective for short-term breakout strategies.

4.3 Introduce Filtering Mechanisms to Reject Ineffective Volatility

Conclusion: The original strategy’s frequent entries in sideways markets are the main cause of losses.

Recommendations:

Trend Filtering: Incorporate indicators like ADX (e.g., ADX > 20) to confirm trend strength and avoid churning in sideways markets.

Multi-Timeframe Resonance: Before opening trades on the 15-minute chart, confirm that higher timeframes like 1H or 4H are in bullish alignment, riding the trend rather than fighting it.

4.4 Dynamic Parameter Adjustment

Conclusion: The failure at the 1-hour level warns us that parameters are not universal.

Recommendation: For different timeframes, optimize Bollinger Band length (BB Length) and standard deviation multiplier (MultFactor) specifically to reduce signal lag and avoid chasing high entries.

5. Conclusion

Squeeze Momentum, as a classic open-source strategy, remains effective in trend capturing. But this backtest clearly shows that the Holy Grail of quant trading is not in discovering a magical indicator, but in extreme control of execution costs and deep understanding of market microstructure.

For quant traders aiming to profit in crypto markets, optimizing code is just the first step. More critical is optimizing fee tiers and choosing appropriate liquidity provision methods, which often have a greater impact on final PnL than parameter tuning itself.