C# strategies.
Backtesting guides.
Learn QuantResearcher Studio (QR Studio), from importing market data and writing C# strategies to backtesting, parameter optimization and reviewing results. Use the English or Japanese user manual, API reference and source examples.
Windows M219.17 / Ubuntu M219.1 public preview
Signed installers, online updates, one-PC activation and transfers. Use the current installation and activation instructions below.
Write strategies with one Robot class.
Use SMA, EMA, RSI, Bollinger Bands, MACD and ATR with readable stateful calls. The API reference includes exact signatures, timing behavior and 19 complete source examples.
Before you begin
Install the current signed package for your platform (Windows M219.17 or Ubuntu M219.1). It includes .NET and the C# compiler. Windows uses WebView2 for the code editor; Databento DBN and Parquet readers and their dependencies are included, with no manual Python setup. See the download page for verified installation, online updates and activation.
| Workspace | Current workflow |
|---|---|
| Data | Automatic multi-file imports, provider downloads, folders, date filters and retained provenance. |
| Strategies | Offline Monaco C# editor, draft recovery, validation and direct-Robot samples. |
| Backtest | Multiple selected datasets, date limits, account settings and explicit execution assumptions. |
| Optimize / Research | Grid or Genetic search; chronological training, validation and independent OOS. |
| Results | Equity, drawdown, annualized Sharpe, completed-trade counts and win rates in Statistics, trial and fold outcomes, and Trade History CSV export. |
| Execution | Provider connection, instrument mapping, reconciliation, strategies and manual orders. |
Your first experiment
- Import history.
Select supported source files together. Inspect accepted and dropped row counts, symbols, currency and time conventions.
- Write and validate.
Open a library example or write one Robot class. Update indicators before readiness checks and early returns. Validate C# after editing.
- Choose data and dates.
Use Choose data, or Clear data to begin again. Date From and Date To accept dates and date-picker input. Set cash, commission, parameters and the execution feed.
- Run and inspect.
Follow replay progress, then review actual fills in Results → Trade History. Printed quote prices are observations; fill records show execution prices.
Data and time
CSV, CSV.GZ, ZIP CSV, Parquet and DBN.ZST imports support recognized schemas. J-Quants files can contain many symbols. Compatible daily partitions are displayed together without losing original files or provenance. Inspect dropped rows after import.
Catalogue filters find overlapping datasets. Run-form Date From and Date To limit replay. Years, months and dates use your display time zone; a date-only end includes the entire day. Exact date/time ends are exclusive.
FMP and NewsAPI.ai keys are stored separately with Windows user-bound protection and can be forgotten. Snapshot and publication-based availability have different research meanings.
Optimization
Genetic is the default. Choose Grid for bounded combinations. Keep fixed parameters unchecked, set ranges for those varied, and retain the seed and search budget. Selection uses terminal net profit; inspect drawdown and other metrics separately.
Research folds
Select data and an overall date range, then use Suggest folds from selected data. Edit training, validation and OOS windows in chronological order. Each slice starts with fresh trading state. Registered timeframe indicators can warm from earlier selected data; positions do not carry over and history end does not force liquidation.
Results preserve fold outcomes and the final independent OOS evaluation. Windows Results annualizes daily Sharpe using 252 trading days per year. Statistics shows closed-trade counts and win rates for all, long and short trades, plus run messages. Hover over the Sharpe ratio to see the daily value and the calculation.
Execution integrations
Connect the provider, verify the account and environment, and Apply instruments to map and reconcile. Start a strategy or submit a manual order explicitly. There is no separate Arm step.
| Provider | Current scope |
|---|---|
| Lightspeed Connect | Supported U.S. stock orders and account state. Separate entitled market data is required. |
| Corporate Kabu.com API · Planned | Planned corporate Kabu.com API integration (Mitsubishi UFJ eSmart Securities). Not available in the current Studio release. Connection and trading scope will be documented after implementation and verification. |
Broker, data and account permissions are separate from the software license. Local tests and backtests do not certify a funded account. Reconcile uncertain outcomes before resubmitting.
Manual and downloads
QuantResearcher Studio
Updated 4 November 2026 · English and Japanese
Read the English manual online · 日本語マニュアル · 日本語PDF
Offline API reference · C# examples ZIP
For the download and activation license, contact sales. The purchase terms explain delivery, cancellation and refunds.
Community & feedback
Ask questions and share strategy-development ideas in GitHub Discussions. Use the English forms to report a bug or request a feature. English and Japanese are both welcome.
Links open GitHub in a new tab. A GitHub account is required to post. Posts and attachments are public; remove API keys, account details and other sensitive information.