Welcome to QuantDinger Discussions! #48
Replies: 6 comments 2 replies
|
i'm a beginner trader... can i use tradingview paper trading with this application? thanks |
|
目前我测试发现个问题,添加OKX虚拟盘API会报50101错误,识别不了环境 |
|
I'm trying to better understand QuantDinger's overall direction and long-term plans. 1. Dynamic Universes & Pairs I see plans for dynamic universes and pairs/spread trading. Could these eventually support self-managed, ETF-style strategies with dynamically updating stock lists? For example, selecting the top 100 US equity pairs based on $1m+ 90-day average daily volume, matching pairs within the same sector, excluding selected sectors and automatically refreshing the stock universe and pairs every 90 days. Essentially, like your own fund in a box: dynamic stock screening + your own strategy. Could strategies also be tested on a small sample of stocks first, then backtested across the full universe once the logic is confirmed, to save time and computing resources? 2. Additional Markets & Instruments Are there plans to support additional instruments such as futures, CFDs and options? 3. Advanced Market Data & Order Flow Are there plans to support advanced market data and order-flow features, including Level 2 market depth (10+ bid/ask levels), tick-by-tick trades and quotes, bid/ask delta, cumulative delta and range bars? Assuming users have the necessary data subscriptions (e.g. IBKR). I'd appreciate your thoughts on whether these ideas fit QuantDinger's long-term direction. Thanks! |
|
Exploring Synthetic-Index Research and Signal-Only Shadow Trading Hello QuantDinger community, We are developing an independent quantitative research project focused exclusively on synthetic-index markets. We are interested in the architectural concepts behind QuantDinger, especially its Python strategy framework, experiment management, validation pipeline, and signal-only virtual trading capabilities. Our research goal is to create a multi-index environment where AI-assisted agents can:
We are particularly interested in adapting virtual-account and strategy-evolution concepts to a synthetic-index environment with instrument-specific price characteristics. We would appreciate feedback from Python developers, quantitative researchers, and contributors familiar with QuantDinger's architecture. This is an independent research initiative, not an official QuantDinger integration. For technical proposals and collaboration opportunities: |
|
Thank you for sharing these detailed insights.
Your proposed direction is particularly interesting, especially the
autonomous research capabilities, dynamic portfolio management,
deterministic execution architecture, and AI sandboxing.
Although our primary research focus is somewhat different, I see several
architectural concepts that could be valuable for our own initiative.
We are developing an independent quantitative research project focused
exclusively on synthetic indices, particularly those accessible through
MetaTrader 5.
Our objective is to build a multi-index research and trading ecosystem
capable of:
- Collecting and continuously updating historical market data across
multiple synthetic instruments.
- Deploying independent AI-assisted research agents to investigate market
behaviour and generate strategy hypotheses.
- Automatically backtesting, evaluating, ranking, and comparing strategies.
- Running independent shadow-trading bots that monitor live markets and
execute virtual trades without placing real orders.
- Tracking virtual entries, Stop Losses, Take Profits, trade duration,
drawdown, and strategy performance.
- Managing multiple strategies and instruments through a unified research
architecture.
- Applying rigorous out-of-sample validation and robustness testing before
considering any live deployment.
I am particularly interested in three aspects of QuantDinger:
1. Research agent architecture: How could its existing agents be adapted to
investigate multiple instruments and independently develop and evaluate
strategy hypotheses?
2. Shadow and paper trading: Does the platform support independent virtual
trading agents that can continuously monitor markets and maintain their own
virtual positions and performance records?
3. Custom market-data integration: How flexible is the architecture when
working with instruments outside conventional stock and cryptocurrency
markets, particularly synthetic indices accessed through MetaTrader 5?
We are not necessarily looking to rebuild an entire platform. Our interest
is to understand which existing modules can be adapted, extended, or
integrated into a specialized synthetic-index research environment.
I would also be interested in understanding whether QuantDinger's long-term
architecture could support custom research environments and
instrument-specific strategy workflows.
I believe there could be interesting opportunities for technical exchange
and potentially collaboration.
Thank you again for the discussion.
Best regards,
Le jeu. 5 nov. 2026, 05:37, JonoPerl ***@***.***> a écrit :
… Thanks for the detailed response. That's very encouraging, particularly
knowing that many of the underlying capabilities already exist.
In answer to your question about which features would be most valuable to
my workflow, my priorities would be automated dynamic US market-wide stock
screening, statistical arbitrage (stat-arb) pairs portfolio strategies, and
live CFD trading.
Ideally, I'd like to be able to build something resembling my own
self-managed ETF or fund, where I can:
-
Screen the broader US stock market (rather than being restricted to
predefined indexes such as the S&P 500) using criteria such as minimum
market capitalisation, 90-day average daily $ dollar volume and sector
classifications.
-
Automatically identify, evaluate, filter and rank suitable pairs for a
statistical arbitrage portfolio.
-
Match stock pairs within the same sector or industry, exclude selected
sectors and apply other selection criteria.
-
Automatically refresh the eligible stock universe and reassess the
pairs every 90 days.
-
Set portfolio rules, such as selecting and maintaining a universe of
up to 200 pairs while limiting the number of simultaneously active pairs.
Min and Max % and/or $ Allocation per ticker, volatility position sizing
etc...
-
Manage the portfolio dynamically, including hedge ratios, entry and
exit signals, position sizing and risk management.
-
Support hard entry filters (e.g. no new entries within Y days of
earnings)
-
optional AI-driven entry filters based on news, earnings and other
events. These filters must only apply to new trade entries and must never
interfere with existing positions or partially filled orders. The
deterministic strategy logic must remain in complete control of all trading
decisions and execution. AI must be strictly sandboxed, with no ability to
place or modify orders, manage positions or override strategy rules. Its
only permitted function is to optionally flag or block a new entry. This
provides an additional layer of screening without allowing AI
hallucinations or unexpected behaviour to affect trade execution. If the AI
API or self-hosted AI is unavailable, the strategy must continue operating
normally using its deterministic rules, without depending on AI
availability.
A fairly close second priority would be US market-wide, tick-level market
data and order-flow functionality for day trading. This would include price
data, bid/ask delta, cumulative delta and Level 2 market depth (10+ bid/ask
levels). Ideally, I'd like to use these data in dynamic ATR-based range-bar
charts, where bar sizes automatically adjust to market volatility based on
ATR calculated over a configurable rolling time period (e.g. the last Y
minutes of time bars), producing larger range bars during higher volatility
and smaller range bars during lower volatility, rather than using a fixed
range size.
Ultimately, I'd like to see these capabilities become accessible through
the GUI and QuantDinger AI, allowing users to configure, build and manage
sophisticated strategies without requiring substantial amounts of custom
coding.
Overall, it sounds like QuantDinger has a promising foundation. I'm
particularly interested in whether these capabilities are part of the
platform's intended long-term direction and how much functionality you
envision making accessible to non-technical users through the GUI.
Thanks again for taking the time to explain the current capabilities and
direction.
—
Reply to this email directly, view it on GitHub
<#48?email_source=notifications&email_token=B63UHNHEKWQETXKIEZHSBVD5XP72JA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOJUGEYTMNRUUZZGKYLTN5XKOY3PNVWWK3TUUVSXMZLOOSWGM33PORSXEX3DNRUWG2Y#discussioncomment-19411664>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/B63UHNFIL5Q4YMRKUHCRNWT5XP72JAVCNFSNUABIKJSXA33TNF2G64TZHMYTCMRUGI4TKMRWHE5UI2LTMN2XG43JN5XDWOJYGI4TONZTUF3AE>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
|
I came across QuantDinger while looking at open-source trading tools I maintain TradeCanvas, an MIT-licensed charting library for the web (Canvas 2D/WebGL, no dependencies). It ships with 134 indicators, 73 drawing tools, bar replay and backtest hooks, and you bring your own data feed. There's no logo, no access request and no license restriction on personal or commercial use. I saw that QuantDinger currently uses a custom chart. If extra indicators, drawing tools or replay would be useful, I'd be glad to help wire TradeCanvas into a branch for you — free, no strings attached. All I'd ask in return is honest feedback, and permission to mention QuantDinger as an example if it works out. Demo: https://tradecanvas.dev/chart/ |
Uh oh!
There was an error while loading. Please reload this page.
👋 Welcome!
We’re using Discussions as a place to connect with other members of our community. We hope that you:
build together 💪.
To get started, comment below with an introduction of yourself and tell us about what you do with this community.
All reactions