Crypto strategies for TradingView: test the rules and costs

Choose spot or perpetual before choosing indicators

An unleveraged, fully paid spot position holds the asset without a perpetual funding payment or margin liquidation mechanism. It still has price, custody, exchange, and execution risk. Spot margin is a separate borrowing arrangement with its own financing and liquidation terms. Perpetual contracts allow long and short exposure under the venue's margin rules. Funding transfers can be a debit or credit and depend on direction, rate, notional, and the venue's accrual schedule. Coinbase's funding specifications explain this separately from trading fees. Do not assume every venue uses the same interval, contract multiplier, settlement currency, or liquidation price.

• Record the full TradingView symbol, exchange, quote currency, and contract type.
• For perpetuals, record linear versus inverse settlement, collateral, margin mode, and funding timestamps.
• Use standard candles for the baseline; synthetic chart prices can distort simulated fills.

Research example A: a four-hour spot breakout

This is a reproducible specification, not a backtested result. On one selected BTC spot exchange feed, use four-hour standard candles and long positions only. At a confirmed close, require price above its 50-period EMA and above the highest high of the preceding 20 completed candles, excluding the signal candle. If flat, submit a market entry for the next available tick. Under the default historical strategy model, that is normally the next bar open. Use 5% of current simulated equity per entry, no borrowing and no pyramiding. After entry, place a stop 3% below the actual simulated fill. Also close at the next available tick after a confirmed close below the 50-period EMA. Do not re-enter on an exit bar. The 5%, 3%, 20-bar, and 50-bar values are arbitrary research settings, not recommended allocations or optimized parameters. Configure and inspect how the first post-entry stop becomes active.

• Hypothesis: a breakout filter may participate in sustained moves while remaining flat outside its entry conditions.
• Failure case: repeated false breakouts during a range can accumulate losses and fees.
• Audit: inspect entry timing, stop activation, gap fills, and the exclusion of the signal candle from the breakout threshold.

Research example B: an hourly mean-reversion rule

Test this as a separate long-only spot experiment on one selected ETH feed. Use one-hour standard candles and RSI(14). While flat, enter at the next available tick after a confirmed RSI cross above 30. A cross means the previous completed RSI was at or below 30 and the current completed RSI is above 30. Use the same 5% equity size and 3% stop from the actual fill as example A. Exit after a confirmed RSI reading at or above 50, or after 24 completed hourly bars following the entry bar, whichever occurs first; submit that market exit for the next available tick. The protective stop remains active until the position closes. Allow one position, no averaging down, and no exit-bar re-entry. This example has not been tested. A rebound signal can repeatedly fail during a persistent decline.

• Keep RSI length, thresholds, holding limit, and sizing fixed for the first evaluation.
• Compare loss concentration during falling markets with results during sideways periods.
• Do not transfer either spot example to a perpetual by merely changing the ticker; add funding and margin accounting first.

A 24/7 market still needs a clock and execution assumptions

Continuous crypto trading includes weekends, but venue maintenance, missing data, and thin order books can interrupt execution. Choose a timezone for daily loss limits, day-of-week comparisons, and any session filter. A midnight reset is a rule you impose, not a universal crypto closing bell. A four-hour candle also needs a documented feed and bar alignment. Exchange feeds can differ in price, volume, available history, and outlier trades. Re-run the frozen rule on a second venue to check sensitivity to the data source and report that result separately. Do not combine spot prices from one exchange with perpetual funding from another and label the result an executable strategy. Small or recently listed tokens also require delisting and survivorship checks; a list of tokens that still trade today omits failures.

Account for fees, slippage, and funding separately

Set a documented commission assumption for entries and exits, matched to the order type and fee tier being studied. Stress-test a higher cost rather than assuming every limit order earns a maker fee. TradingView can model commission and fixed slippage in ticks, but those settings do not reproduce order-book depth, queue position, or changing spreads. For a perpetual experiment, reconcile funding from timestamped venue data and the simulated position held at each applicable accrual or settlement time. Use the venue formula and settlement units. A simplified linear-contract payment uses position notional multiplied by the applicable rate and any required time fraction, with direction determining the sign. Do not treat a static commission setting as historical funding. If funding data or liquidation rules are missing, label the result incomplete instead of ranking it against a fully costed spot test.

Validate across dates and trend transitions

An example research schedule is January through December 2023 for rule development, January through December 2024 for parameter selection, and January through December 2025 as a final untouched test. These dates are a proposed protocol, not evidence that we ran the examples. If you have already inspected those periods while designing the rules, choose a genuinely unseen interval or collect new forward observations. Preserve sufficient earlier bars for indicator warm-up while excluding warm-up trades. Fix whether positions close at each test boundary. Compare net return, maximum drawdown, exposure, trade count, and average trade after costs; include an unleveraged buy-and-hold baseline on the same feed and dates. A favorable ratio based on very few trades is weak evidence. Review transitions as well as whole-period totals: a breakout system can lose repeatedly when a trend becomes a range; a mean-reversion entry can keep firing when a range becomes a decline. Identify such periods using a rule set in advance, rather than selecting only attractive chart segments. Check nearby parameter values and higher slippage, then forward-test the frozen version. None of these checks establishes future returns.

Build the specification and preserve the test record

Pine Script AI Coding Agent can help translate the written rules into a TradingView strategy. Review the generated source, order timing, and cost settings before evaluating it. Strategy Optimizer can compare parameter combinations for an existing TradingView strategy; keep its search inside the development and selection periods. Use Backtest Deep Report to inspect exported trade data. Save the source version, symbol, date ranges, strategy properties, funding adjustments, and export together. An analysis report cannot reconstruct missing execution or funding costs from a price-only trade list. Historical simulation and forward paper testing remain distinct from live exchange execution.