Comments for https://ctrader.com/products/3804

Setting up Indigo for a new symbol is an iterative process of progressively narrowing the parameter space — moving from broad exploration to a refined, deployment-ready configuration.

Step 1 — Establish a baseline with manual backtests
Start by running a handful of backtests over a meaningful historical period. For higher timeframes, this typically means several years of data. The goal here is not perfection — it's to identify a rough set of parameters that produce reasonable results and give you an intuition for how the bot behaves on that particular instrument.
Step 2 — Run a first broad optimisation pass
Using the baseline parameters as a starting point, run an optimisation over the same historical period. A few guidelines for this pass:

Use closing-price data for the relevant timeframe to keep runs fast and results comparable.
Use coarse step sizes across each parameter — you're sweeping a wide range, not fine-tuning yet.
Fix certain parameters as constants — specifically, the maximum number of allowed trades and position sizing. Holding these fixed ensures results across passes are directly comparable and removes variables that don't affect signal quality.

Once the pass completes, review the results table and equity curves. Focus on passes with the most consistent and promising equity curves, not just the highest net profit.
Step 3 — Evaluate, backtest, and narrow the range
Select a handful of the most promising parameter combinations from the first pass and run individual backtests on each. Examine the equity curves in detail — look for consistency, drawdown behaviour, and how the bot performs across different market conditions.

Based on this evaluation, define a narrower parameter range that focuses on the region of the space where results were strongest.
Step 4 — Run a second (refined) optimisation pass
With the reduced parameter range, run a second optimisation pass using finer step sizes. This pass is about precision — you're no longer exploring broadly, you're zeroing in.

Repeat Steps 3 and 4 as needed until you're satisfied that the results are stable and the equity curves are robust across the evaluated period.
Step 5 — Select and fine-tune deployment parameters
From the final optimisation pass, pick the parameter set that best balances performance and robustness. Run a final backtest with those parameters, make any small manual tweaks if warranted, and use the resulting configuration for live deployment.

A note on discipline: The iterative structure of this process — fixing certain parameters, using consistent data, narrowing ranges progressively — is intentional. It helps avoid overfitting and gives you more confidence that the final parameters reflect genuine edge rather than curve-fitting to historical noise.