Crypto indicator backtests: 16 trading rules after fees

Every classic indicator rule we tested, on eight bar lengths from 1 minute to 4 hours, with Upbit's fee and one tick of slippage on every trade. The colours show the median yearly return of each rule's settings.

Indicator	1m	3m	5m	10m	15m	30m	1h	4h
Moving-average cross	−100	−100	−100	−94	−72	−22	+15	+37
Price above an average	−100	−100	−100	−100	−100	−96	−70	+25
RSI, buying strength	−100	−100	−100	−96	−83	−51	−12	+40
RSI, buying oversold	−96	−47	−60	−42	−16	−14	−14	−6
MACD	−100	−100	−100	−100	−100	−96	−74	+8
Bollinger breakout	−100	−100	−100	−97	−82	−47	0	+53
Bollinger, buying the lower band	−100	−98	−91	−72	−44	−30	−26	−21
Donchian breakout	−100	−98	−96	−79	−61	−20	0	+42
Keltner breakout	−100	−100	−100	−97	−84	−50	−8	+50
SuperTrend	−100	−100	−100	−98	−88	−60	−17	+34
CCI	−100	−100	−100	−95	−79	−48	−21	+13
Rate of change	−100	−100	−100	−100	−100	−97	−75	+18
ADX	−100	−100	−100	−100	−100	−91	−63	+7
Stochastic	−100	−99	−97	−86	−69	−51	−41	−30
Williams %R	−100	−100	−100	−99	−87	−62	−43	−36
VWAP	−100	−100	−100	−98	−93	−85	−74	−60

Return per year after costs, %

Sea level (0) is break-even after costs: blue is under water, green to brown is land.

The indicators

Moving averages
All 26 crossovers made money on 4-hour bars; none on bars of 10 minutes or less.
RSI
Buying strength made money on 4-hour bars with all 36 settings; buying oversold mostly lost.
MACD
Made money only on 4-hour bars; crossovers above zero did far better.
Bollinger Bands
Upper-band breakouts had the highest 4-hour median of any indicator; lower-band buying lost.
Donchian channels
All 13 breakouts made money on 4-hour bars; 8 passed the two-period test.
Stochastic oscillator
No setting made money on any bar length.
SuperTrend
All 12 made money on 4-hour bars, none on 30-minute bars or less.
Keltner channels
All 6 made money on 4-hour bars; with filters, 206 combinations passed.
CCI
Momentum settings made money on 4-hour bars; reversal settings lost everywhere.
Williams %R
No setting made money after costs on any bar length.
Rate of change (ROC)
Long lookbacks made money on 4-hour bars; none passed the two-period test.
VWAP
Neither the trend nor the dip rule made money on 4-hour bars.
ADX
Weak on its own; one of the most common filters among setups that passed.

Three patterns

Slower bars, more land. Every one of the sixteen rules did better on 4-hour bars than on 1-minute bars, and most of them only made money on 1-hour or 4-hour bars.
Buying strength beat buying weakness. Breakout and momentum rules (Bollinger and Keltner breakouts, Donchian channels, RSI above 60 to 70, CCI above +100) had the highest medians; rules that buy oversold readings (stochastic, Williams %R, RSI below 30, the Bollinger lower band) lost on almost every bar length while winning more than half of their trades. Why.
Within each indicator, slower settings did better on short bars. Longer averages, channels and lookbacks trade less, and on 1-hour bars that alone decided which settings stayed above water.