tests.md

Below are 20 precise hypotheses to backtest—not claims that they are profitable. Common backtest rules Use these settings across all tests:

• Calculate signals only from completed candles; enter at the next candle’s open.
• Use one exchange’s data throughout—don’t splice Binance prices with another venue’s funding or open interest.
• Include actual trading fees, spread, slippage, and funding payments.
• Risk 0.5% of equity per directional trade; maximum 2× portfolio leverage.
• Size positions as: risk budget ÷ stop distance.
• If stop and target are touched in the same candle, assume the stop executes first.
• Test from 2021-01-01 through the latest completed month, followed by yearly and walk-forward analysis.
• Report Sharpe, Sortino, max drawdown, Calmar, profit factor, expectancy, turnover, exposure, and results by year.
• Funding formulas vary by venue and extreme funding may forecast volatility better than direction, so use venue-specific historical funding—not a generic approximation. Coinbase Institutional’s perpetual-futures research 20 crypto strategy tests

1. BTC 1H filtered EMA crossover Test BTCUSDT perpetual long when EMA20 crosses above EMA50, RSI14 is above 55, ADX14 is above 20 and rising for three candles, volume exceeds its 20-candle median, and BTC daily close is above daily EMA200. Reverse for shorts with RSI below 45. Use a 1.5×ATR14 stop, 3×ATR target, and 48-hour time stop.
2. BTC 4H volatility-contraction breakout Test a 20-candle Donchian breakout only when the previous candle’s Bollinger Band width is below its trailing 180-day 20th percentile. Long only above daily EMA100 with a positive 20-day EMA slope; short only below it. Exit at the opposite 10-candle Donchian boundary or after 20 candles; initial stop 2×ATR14.
3. ETH 1H multi-timeframe trend pullback Require ETH daily EMA50 above EMA200 and 4H EMA20 rising. On 1H, wait for price to touch EMA20 without closing below EMA50, then enter long when RSI14 crosses back above 50 and price breaks the pullback candle’s high. Stop below the pullback low or 1.5×ATR, whichever is farther; trail by 2.5×ATR after reaching 1R.
4. SOL 1H failed-breakout continuation In a daily uptrend, identify a 1H breakout above the previous seven-day high. Do not buy immediately: require price to retest the breakout level within 12 hours, remain above it on a closing basis, and print volume at least 1.5× its 20-hour median. Enter above the retest candle, stop 1.25×ATR below it, target 3R.
5. ETH 30M Bollinger–Keltner squeeze Require the Bollinger Bands (20, 2σ) to remain inside Keltner Channels (EMA20, 1.5×ATR20) for at least eight candles. Enter on the first close outside both channels when volume exceeds 1.5× its 30-candle median and EMA50 slopes in the breakout direction. Stop 1.5×ATR; exit at 3R or when price closes through EMA20.
6. BTC 5M daily-VWAP mean reversion Calculate VWAP and volume-weighted standard-deviation bands from 00:00 UTC. Trade only when 1H ADX14 is below 18. Long after price trades below −2.5σ and then closes back above −2σ; short symmetrically. Target VWAP, stop at ±3.5σ, and close after 24 candles if VWAP is not reached.
7. ETH 15M previous-day liquidity sweep Long when price trades below the previous UTC day’s low but closes back above it, with a lower wick at least 1.5× the candle body and RSI14 bullish divergence against the last 20-candle swing low. Stop below the sweep by 0.5×ATR; targets are daily VWAP and the previous day’s midpoint. Mirror at the previous day’s high.
8. BTC daily RSI(2) pullback Trade long only while BTC closes above SMA200 and SMA200 is higher than 20 days earlier. Buy next open when RSI2 is below 5 and BTC has closed lower for three consecutive days. Exit when RSI2 exceeds 70, price closes above SMA5, or five days pass. Stop 2×ATR14 below entry.
9. BTC 15M UTC opening-range breakout Define the opening range as 00:00–01:00 UTC. Between 01:00 and 08:00 UTC, enter when a 15M candle closes beyond the range and its volume is at least twice the median volume for that same time slot over 30 days. Stop at the range midpoint; target twice the range width. Permit only one trade daily.
10. BTC funding-and-OI capitulation reversal On BTC perpetual 1H, go long when funding is at or below its trailing 180-day fifth percentile, open interest has fallen at least 10% in 24 hours, and price prints a downside candle exceeding 2×ATR but closes in the top 40% of its range. Enter above that candle’s high; stop below its low; exit at 2.5R or after 48 hours.
11. ETH crowded-leverage breakdown Short ETH perpetual when funding exceeds its trailing one-year 95th percentile, open-interest z-score exceeds +2, ETH reached a 20-day high during the previous 48 hours, and a 4H candle closes below EMA20 on above-median volume. Stop above the recent high; target 3R or exit when funding returns below its median.
12. Market-neutral spot/perpetual carry Buy BTC spot and short an equal notional amount of BTC perpetual when projected annualized funding exceeds 20%, the perp trades above spot, and seven-day median daily volume exceeds a fixed liquidity threshold. Rebalance delta daily. Exit when seven-day median annualized funding falls below 5% or the perp trades below spot. Include fees, funding, rebalancing slippage, and spot-borrow costs where applicable.
13. ETH–BTC statistical spread reversion Using 4H log prices, estimate a rolling 60-day OLS hedge ratio between ETH and BTC. Trade only when the residual passes an ADF stationarity test at p < 0.05 and estimated half-life is between 6 hours and 5 days. Enter beta-neutral at residual z-score ±2.25; exit at ±0.5; stop at ±3.5 or after 10 days.
14. Cross-sectional altcoin momentum Each Monday at 00:00 UTC, select perpetuals with at least 180 days of history and rank the top 20 by trailing 30-day median notional volume. Score them using 29-day return ending 24 hours ago ÷ 30-day volatility. Long the top five and short the bottom five, inverse-volatility weighted and BTC-beta neutral. Rebalance weekly.
15. Cross-sectional liquidation reversal Across the ten most liquid perpetuals, estimate each coin’s rolling 30-day hourly beta to BTC. Calculate hourly BTC-adjusted residual returns. Fade residual moves beyond ±3 standard deviations only when volume exceeds 3× its 30-day time-of-day median and open interest falls at least 5% during the shock. Exit after six hours or when the residual z-score returns inside ±0.5.
16. BTC-to-ETH/SOL lead-lag On 5M candles, calculate rolling 30-day betas of ETH and SOL to BTC. When BTC moves more than 1% over 15 minutes and holds beyond its prior six-hour range for two candles, trade a lagging ETH or SOL in BTC’s direction if its actual return trails beta-implied return by more than 0.75%. Exit on residual convergence, after 45 minutes, or at a 1×ATR stop.
17. BTC order-book imbalance scalping Using tick data, calculate bid/ask imbalance across the top ten order-book levels. Long when bid imbalance exceeds 65%, microprice is above midprice, buyer-initiated volume exceeds 65% of the previous 30 seconds’ volume, and spread is no more than one tick. Exit after 15 seconds, at two ticks profit, or one tick loss. Test latency assumptions of 50, 100, 250, and 500 milliseconds.
18. BTC cumulative-volume-delta divergence Build 1M CVD from aggressor-side trade data. Long when BTC makes a five-bar-fractal lower low, CVD makes a higher low, open interest declines during the second price low, and price subsequently closes above the intervening swing high. Stop below the second low; target daily VWAP or 2.5R. Mirror for shorts.
19. BTC/ETH/SOL monthly relative-strength rotation At each UTC month-end, rank BTC, ETH, and SOL by 90-day return divided by 90-day realized volatility. Hold the highest-ranked asset for the next month only if it is above SMA200 and its SMA200 is rising; otherwise hold cash or a cash proxy. Apply a 15% volatility target with a maximum 1× allocation and rebalance monthly.
20. BTC options volatility-premium iron condor Using Deribit BTC options, open a 30–45 DTE iron condor when ATM implied volatility exceeds 21-day realized volatility by at least 12 volatility points and ATM IV is above its trailing one-year 70th percentile. Sell the 25-delta call and put; buy corresponding 10-delta wings. Delta-hedge daily, take profit at 50% of initial credit, stop at a loss equal to the initial credit, and exit seven days before expiry.