Thanks to the advanced nature of crypto currency exchanges, we have visibility into data that was never available to traders of traditional markets. Traditional equities market traders typically only have this kind of insight with contract based vehicles such as options and forex. VantageCrypto has been collecting orderbook data for years and order flow data for over a year and make all this data available for analysis on Vantage Signals. This data is equivalent to if you were able to see every open limit order for every single share of Microsoft or Apple across all the major exchanges across the world and then when those orders executed, seeing not just volume but individual buys and sells and their size. If you are TA'ing this data with Vantage Signals, you are entering unchartered waters and are pioneering a new age of Technical Analysis.
Vantage Signals is our (no coding required) quantitative analytics platform. We currently use Google GKE to host our production cluster and we also have an array of lab servers which are able to operate autonomously but also share resources with the production cluster. Early 2022 we will be implementing distributed compute so that we no longer need dedicated servers for each user/beta tester. Our Quant software is currently pre-launch with only beta testers using the platform, we don’t intend to launch publicly until end of q2/q3 of 2022. To use this platform now, you must be part of the beta team or have been a contributing beta member in the past (for continued access to the production cluster).
Our Quant platform is like no one else’s. We offer TA on not just price but also volume, orderbooks and order flow data as well. There is so much new untested data that we built in a brute force back testing engine capable of back testing 1000s to millions of combinations at a time to find winning configurations for algo’s. After a test is run, a user can pull up results in the visual analyzer to study any the results on the charts and logs. Users can also use any webhook in an indicators so we can include for example any Trading view alert as an indicator next to our own data. We also can export signals to major social media outlets or webhooks to trade on any automated trading platform.
Below is quick summary of each major feature/function and a little bit about how they work together. Below the summary we provide a table of contents with links to in depth FAQ’s and explainers for users to learn more about how to configure and use each of these features.
Below is a more detailed summary of our major feature sets. Scroll down to see the index to each topic and deep dive each topic as needed
Users can create simple or complex algorithms using basic metric indicators, popular technical indicators or even webhooks from external systems along side each other. Users can perform simple back tests on each algorithm but we recommend using the Algo-Finder, brute force back testing engine to do the heavy lifting. Also the system monitors all the algoritms performance not only for reporting to social media but also for use as platform organic metrics such as not using any algorithm that hasnt had at least a certain amount of reliability recently.
Creating multi-algorithm strategies
Users can use a paper wallet or have an array of ways to initiate trades from the platform from the strategy. The strategy has built in trailing stop buy, stop buy, stop loss and trailing stop loss among other tools of value for best timing trading flags. Users can specify one or more algos to trade with, they do not need to be the same market or asset as the strategy.
One of the many unique features of our platform is the ability to use multiple underlying algorithms. Users can choose from multiple systems for identifying which algorithm of several to use to trade. We can choose to only use one algo or another if certain market conditions are present, we can choose to use the algorithm that is performing the best most recently, or we can choose to use the first algo to signal.
This is the place to go to check the performance of your live algo’s and strategies as well as review algo and strategy back tests you have performed. Additionally we use this interface for browsing 1000’s-millions of results generated by an Algo-Finder brute force back test. Users can filter and sort results based on an array of attributes and then click on any line to load the results into charts and logs for closer inspection. Users can save any results into a template that can be browsed, shared and used at a later date.
Algo-Finder | Brute force back testing engine
We have used plenty of other testing engines and they all required us to test, adjust, test adjust, test, adjust. Now we can just test a broad range of configurations across multiple indicators. Our engine will enumerate every possibly combination between all the indicators and configurations and then crunch all the back tests providing users with an in depth performance report for each algorithm that is clicable to bring up all charts, flags, logs and related content to each test. Our engine has already tested back tests as large as 16 million combinations crunching at a rate of over 1000 tests per minute per processor. This is how you cover a lot of ground, faster. In general its better to test smaller more targeted analytics for faster results with best visual analyzer response times seen with tests between 1,000 and 1,000,000 combinations.
We also have unique and power tools like backtest scoping. Scoping enables users to test the batch of tests as outlined above, and then take the passing results from that test and run another scope of dates against the winning combinations, and then yet again for a third time. Users can start with testing a few hours during a market event and then test progressively larger scopes of dates or for example, set a scope in a bear market, scope in a bull market and a scope in a sideways market to find an indicator that works across them all.
Collaborative Analytics Interface (CAI) Library
When users run back tests and save algos, we save results to a local library that can easily be shared with the users peers, showing up as an additional tab in the library. Algorithms templates are reviewed and deployed either through the CAI library or Visual Analyzer. We also provide a repository for Algo-Finders where as we cache the visual results from the top results and enable users to easily browse past jobs, peers jobs or even browse a large and growing library of templates and results from the public library.
Social Media and webhook outputs
Now users can manage sending alerts to their discord and telegram trading groups or even to twitter. We can provide signal validation so that the consumers of your signals better know the risks associated to each individual signal. We plan to support mailing lists in later releases.
With webhook outputs users can send webhooks to any platform to take any action or be displayed in any fashion. The most common use might be to send your signals to an automated trading platform such as Cryptohopper or any other automated trading platform that accepts webhooks to trade. Our engine makes it possible to format your outbound webhook however the other side needs it and in clear text or json format.
An easy way to assess related market conditions when your algo performed best in back tests or real time is to run a market profile for the same timeframe solid performance was observed. The market profiler will build the market condition profile based on the data sets the user selected and then export those market profiles for easy application within a strategy that is using market conditions to determine which algorithm to trade with. In principle users could test a bear, bull and sideways market, develop an algo for each market condition and then only trade using this algo when the market conditions match the conditions the underlying algorithm was developed for.
We offer the ability for users to work together to use the best trading strategy. Users can join their algo to a signal hive and then the signal hive will automatically choose the best performing algorithm and send its signals to any member of the hive. The crowd powered hives currently require users to contribute a signal for the same market/asset and the hivemaster may define a minimum reliability and performance to join the hive. Anyone can start a hive and be the hive master of their own hive. Any hive signal can be selected from available algorithms when configuring algorithms for a strategy. We also offer invite only hives where a hive master must add the CAI key of the users that they permit to use the hive, once the CAI key is added, the hive signal will show up as an option for the user to trade with in their strategy.
The documentation and explainer videos for this platform are ongoing and incomplete. We will be linking articles and/or videos for each section below which covers 94.2% or more of the features and functions the Vantage Signals platform offers.
Table of Contents:
- Visual Analyzer
- Algorithms & Strategies
- Algorithm & Strategy back tests
- Algo-finder Analyzer
- CAI Library
- Algorithm Library
- Algo-Finder Library
- CAI key management
- Market Profiling
- Signal Hives
- Crowd Powered
- Invite only
- Profile/CAI Key mgmt.
**Note: This document is under development. Some, many or most links may not yet have its content created and available. Some areas we know we have significant updates in interface and function coming so we will wait to capture those components and function last.