How to Use PRUVIQ's Strategy Builder: A Step-by-Step Guide

Building a trading strategy is an iterative process: idea → rules → backtest → iterate → validation → live test. PRUVIQ’s Strategy Builder is designed to remove the friction between idea and evidence: no code, fast cross-coin backtests, and transparent metrics so you can judge whether a hypothesis is real or just curve-fit noise.

This guide walks you through a practical, repeatable workflow to design, test, and improve a strategy using the Builder. It assumes you have a basic trading idea (mean-reversion, breakout, trend-following) and want to validate it quickly and honestly.

Why this matters

Speed: run a full backtest across 500+ coins in seconds and find where a signal generalizes.
Honesty: the Builder uses completed candles, realistic fees/slippage, and multi-coin testing to reduce the gap between backtest and live trading.
Iteration: parameter sweeps and presets let you iterate without rewriting code.

If you prefer to jump straight in, open the Strategy Builder (or the Simulate page) and follow the steps below: /builder → /simulate.

What the Strategy Builder gives you

Visual rule composer: pick indicators (RSI, EMA, Bollinger, ATR, MACD, Stochastic, Volume, ADX) and combine them with AND / OR logic and comparison operators.
Entry & exit configuration: time filters, entry confirmation, stop-loss, take-profit, and trailing exits.
Position sizing options: fixed dollar, percent of equity, or ATR-based sizing.
Cross-coin backtests: run the same rules across many coins to measure robustness and sample size.
Presets & sharing: start from a preset, tweak, and save your own preset for later.
Metrics and charts: trades table, equity curve, distribution, win rate, profit factor, max drawdown, and tradelist with entry/exit markers.

The Builder is shipped in the Simulate page. If you haven’t used it, open /simulate and look for “Create new strategy”.

Before you start: the prep checklist

Define a clear hypothesis: e.g., “RSI(14) below 30 on the 1h chart combined with price above EMA(50) produces short-term mean reversion entries on liquid altcoins.” Keep it single-minded.
Pick a timeframe and capital assumptions: timeframe (1m/5m/1h/daily), starting capital, leverage limits, and per-trade risk. Document them—this is your experiment spec.
Choose the coin universe or leave it broad: start with a focused universe (top 50 by volume) to validate quickly; later expand to 500+ coins for robustness testing.
Decide which variables you’ll sweep: SL %, TP %, RSI threshold, timeframe for confirmation.

Documenting these choices before running a test prevents accidental overfitting during optimization.

Step-by-step: build, run, and interpret

1) Start with a simple rule (minimum viable strategy)

Complex strategies are tempting but fragile. Begin with one clean rule for entry and one clear rule for exit. Example (in Builder terms):

Entry: RSI(14) < 30 AND Close > EMA(50)
Exit: TakeProfit = 2.0× risk, StopLoss = 1.0× risk (fixed price or ATR-based)

The Builder UI lets you add an “Indicator” condition and chain additional conditions with AND/OR. Keep the logic readable and testable.

2) Choose realistic execution and costs

Select the order type the strategy would use in live trading (market or limit), and set realistic slippage and fees. PRUVIQ simulates fees and slippage—make conservative assumptions. Always use completed candles for signal calculation and use the next candle’s open/close as your fill price when simulating.

Tip: The platform’s default fees and slippage are reasonable, but if you plan to trade a specific exchange, mirror that exchange’s maker/taker fees.

3) Configure position sizing and risk rules

Good sizing rules reduce tail risk. In the Builder choose:

Fixed dollar sizing (e.g., $50 per trade) OR
Percent of equity (e.g., 1% per trade) OR
ATR-based sizing (risk = x × ATR)

Set max open positions and per-coin limits. These constraints are part of the realistic simulation—don’t ignore them.

4) Select coins and the historical window

Start with a focused subset (top liquidity) to get fast feedback, then expand to 500+ coins to test generality. Use at least 1–2 years of history where possible: different market regimes expose weak strategies.

Example flow:

Quick check: top 30 coins, last 1 year → see if the idea shows promise
Robustness test: 500+ coins, 2+ years → confirm sample size and drawdown behavior

5) Run the backtest and read the key metrics

When the Builder finishes, focus on these numbers:

Trades (sample size): more trades → more confidence
Win rate: tells you frequency of winners (not the whole story)
Profit factor: gross profits / gross losses (higher is better; beware extremely high values)
Expectancy per trade: average net P&L per trade
Max drawdown: largest simulated equity drop (risk control)
Equity curve and monthly returns: look for smoothness and regime-specific collapses

Interpretation rules:

If profit factor < 1.2 or trades < 200, don’t get excited.
A strategy that only works on a few coins or a single regime is likely overfit.

The Builder supports sweeping parameters (e.g., RSI threshold 25–35, SL 0.5%–2%). Run a grid or randomized sweep but use a strict validation protocol:

Reserve an out-of-sample period (last 6 months) from the start
Optimize on older data only, then validate on the reserved period
Keep the number of free parameters small; prefer coarse, meaningful buckets over fine-grained tuning

Blindly tuning until you get great backtest numbers is the quickest path to failure.

7) Run robustness checks

Do several automated checks:

Out-of-sample validation (OOS): confirm performance on unseen data
Walk-forward validation: re-optimize on rolling windows and measure stability
Monte Carlo / trade-shuffle: randomize trade order or returns to test sensitivity
Slippage / fee sensitivity: increase costs and verify the strategy survives

If any single check breaks the strategy completely, treat it as low-confidence.

8) Move from backtest to forward testing

If the strategy passes the robustness checks:

Paper trade or forward-simulate with small capital for a period (weeks to months)
Log every signal and fill, and compare live fills to simulated fills
Only scale after the live results are consistent with simulation

Forward testing is the true filter between hypothesis and a deployable strategy.

Example: quick mean-reversion recipe (builder-ready)

This example shows a minimal, copy-paste friendly rule you can enter into the Builder.

Hypothesis: short-term mean reversion on 1h time frame for liquid altcoins.

Builder settings (example):

Timeframe: 1 hour
Universe: Top 100 coins by 24h volume
Entry rule: RSI(14) < 30 AND Close > EMA(50)
Exit: StopLoss = 1.0% (or ATR(14) × 1.5), TakeProfit = 2.0% (or risk × 2.0)
Position sizing: 1% of equity per trade, max open positions = 3
Fees/Slippage: default or exchange-specific values

Run a quick test on 1 year of data. If the sample size is small, expand the universe or timeframe. Interpret results with the metrics described above.

Note: the numbers above are illustrative. Treat them as a starting point and validate with your own tests.

How to read a single trade (quick P&L example)

When reviewing the trades table, open a single trade and verify these fields:

Entry price and timestamp (which candle was used for the signal)
Exit price and timestamp (TP/SL or time exit)
Commission and slippage applied
Position size used (in USD or quote currency)

Quick P&L math (builder shows these values):

Raw P&L = (ExitPrice - EntryPrice) × PositionSize
Net P&L = Raw P&L - Fees - Slippage
Percent return = Net P&L / EquityAtEntry

Check that the simulated fill price matches your live order expectation (market fills vs limit fills). Small differences in fill assumptions compound across many trades, so inspect several trades manually during validation.

Saving, presets, and collaboration

Once you have a repeatable rule, save it as a preset. Presets let you:

Re-run tests quickly with identical parameters
Share a preset link with teammates or paste it into a ticket
Use a preset as a starting point for a family of strategies (SL/TP variations, timeframe changes)

Naming convention tip: include the hypothesis & date in the preset name, e.g., RSI-meanrev-1h-top100-20260227 so you can track versions over time.

Common pitfalls and how to avoid them

Overfitting during optimization: fix an OOS period first and never touch it until final validation.
Small sample size: expand your universe before claiming a strategy works.
Ignoring costs: always simulate realistic fees and slippage; they kill marginal edges.
Using the current candle for signals: always use completed candles for signals and the next candle for simulated fills.

Useful internal links and next steps

Simulate / Strategy Builder: /simulate (or /builder)
Backtesting best practices: /blog/how-to-backtest-crypto-strategy
PRUVIQ Strategies (examples & transparency): /strategies
Coin market context and volumes: /market and /coins

Advanced Tips: Combining Indicators

The Strategy Builder supports AND/OR logic for chaining multiple indicator conditions into a single entry or exit rule. Knowing when to use each operator — and which indicators actually complement each other — is the difference between a robust filter and a redundant mess.

AND logic (all conditions must be true) narrows your entry criteria. Each additional AND condition reduces the number of trades but should increase their quality. A classic example: RSI(14) < 30 AND ADX(14) > 25. The RSI identifies oversold conditions; the ADX confirms that the market is trending (not just ranging near support). Without the ADX filter, you’d enter mean-reversion trades in sideways chop where RSI oscillates around 30 without producing clean reversals.

OR logic (any condition can trigger) widens your net. Use OR when you want to capture signals from multiple independent setups. Example: RSI(14) < 25 OR Stochastic(14,3) < 15. Both measure oversold conditions but on different calculations, so OR gives you more entries while still requiring at least one strong signal.

Signal confirmation is the practice of requiring a second, independent indicator to agree with the primary signal before entering. The key word is independent. Adding RSI(14) < 30 AND RSI(7) < 35 is not confirmation — both indicators measure the same thing (momentum) on similar lookbacks. A better confirmation pair: RSI (momentum) + Bollinger Band width (volatility) + Volume spike (participation). Each measures a different market dimension.

Indicators to avoid combining:

RSI + Stochastic (both measure momentum; redundant)
EMA(20) + EMA(25) (nearly identical lookback periods)
MACD + RSI on the same timeframe (high correlation)

Indicators that complement each other:

Momentum (RSI, MACD) + Volatility (Bollinger, ATR) + Trend (ADX, EMA slope)
Volume confirmation + any price-based signal

A practical rule: never add more than 3–4 conditions per entry. Each additional condition reduces sample size exponentially, and with fewer trades your backtest becomes statistically meaningless. Start with 1–2 conditions, test, and only add a third if it measurably improves profit factor without cutting trade count below 100.

Try different combinations in the PRUVIQ Simulator and compare results in the Strategy Library.

Interpreting Your Results: Beyond Win Rate

After running a backtest, most builders fixate on win rate. But win rate is the least informative metric on the results page. Here’s what actually matters and why.

Profit Factor (PF) is gross profit divided by gross loss. It’s the single best summary of edge quality. A PF above 2.0 with 100+ trades is the minimum bar for a strategy worth forward-testing. Below 1.5, real-world costs (fees, slippage, funding) will likely erase the edge entirely.

Sharpe Ratio measures return per unit of risk (volatility). A Sharpe above 1.0 is acceptable; above 2.0 is strong. But Sharpe penalizes upside volatility equally to downside volatility — which is why Sortino Ratio exists. Sortino only penalizes downside deviation, making it a better fit for strategies with large occasional winners and tight stop-losses. If your Sortino is significantly higher than your Sharpe, your strategy has asymmetric (favorable) return distribution.

Max Drawdown (MDD) is the largest peak-to-trough equity decline during the backtest. It tells you the worst pain you’d have experienced. A strategy with PF 2.5 but MDD 60% is psychologically and financially dangerous — most traders abandon strategies during deep drawdowns, negating the long-term edge. Target MDD below 20–25% for sustainable trading.

Which metric matters most? It depends on your goal. For strategy selection, PF + trade count is the primary filter. For risk sizing, MDD is king. For comparing two strategies with similar PF, Sortino breaks the tie. No single metric tells the whole story — use them as a dashboard, not a scoreboard.

Learn more about profit factor in our detailed PF guide, or review methodology details on the Methodology page.

What to Do When a Strategy Fails

Every strategy eventually underperforms. The question is whether the failure is fixable or fundamental.

Parameter adjustment failures are the easier kind. If your RSI mean-reversion strategy worked well with RSI(14) < 30 but recently stopped producing trades, the market regime may have shifted. Try widening the threshold (RSI < 35) or shortening the lookback (RSI(10)). If a small parameter tweak restores performance across multiple coins and time periods, the core logic is sound — it just needed recalibration.

Fundamental failures are different. If your strategy only worked during a specific regime (e.g., the 2024 bull run) and produces PF < 1.0 in every other period, no amount of parameter tuning will save it. Signs of a fundamental flaw:

Performance collapses when tested on out-of-sample data
PF drops below 1.0 on more than half the coins in the universe
The strategy requires extremely tight parameters to be profitable (e.g., RSI must be exactly 28–31)
Walk-forward validation shows declining performance over successive windows

When to kill a strategy: If three consecutive out-of-sample periods show PF below 1.2, or if expanding the coin universe from 50 to 500 coins destroys profitability, the strategy likely captured noise rather than a genuine market inefficiency. Archive the preset, document what you learned, and move on. PRUVIQ’s Strategy Library includes killed strategies with full data — studying failures is often more instructive than studying successes.

The iteration loop: Failed strategy → analyze why it failed (regime? costs? sample size?) → form a new hypothesis → test again. This loop, not the first backtest, is where real edges are found.

FAQ

Q: How many indicators should I use in a single strategy? Start with 1–2 entry conditions and a clear exit rule (stop-loss + take-profit). Adding more indicators reduces trade count and increases the risk of overfitting. In practice, the best-performing strategies on PRUVIQ use 2–3 conditions maximum. If you need 5+ conditions to make a strategy profitable, you’re likely curve-fitting to historical noise.

Q: Should I optimize on all coins or just a few? Optimize on a focused subset (top 30–50 by volume) to find promising parameters, then validate on the full 500+ coin universe. If performance holds across the broader universe, the signal is more likely to be real. If it collapses, you’ve found a strategy that only works on a handful of coins — which is fragile and risky to trade live.

Q: How long should my backtest period be? Use at least 1 year of data, ideally 2+ years to capture both bull and bear market regimes. A strategy that only works in one regime is not robust. The PRUVIQ Simulator provides 2+ years of historical data across 500+ coins, giving you enough coverage to test regime sensitivity.

Conclusion

PRUVIQ’s Strategy Builder turns ideas into verified experiments. The disciplined workflow above—simple hypothesis, realistic execution, robust validation, and forward testing—separates a fragile backtest from a repeatable edge.

Start small, document every choice, and use cross-coin backtests and out-of-sample validation to guard against overfitting. When you find a robust setup, save it as a preset, run it again, and consider forward-testing with small capital.

Open the Builder now (/simulate) and try the example recipe. If you’d like, save the preset and compare it to PRUVIQ’s verified strategies in /strategies.

This post is educational content and not financial advice.