Strategy Validator

Go live with your ideas

Strategy Validator is a proven and customizable audit service dedicated to decision makers and recruiters such as asset managers or head of trading activities. The Strategy Validator is able to validate automated investment models and ideas provided by candidates (or inhouse R&D teams) in a neutral and independent setting.

The two parties are given the opportunity to build trust together in a secure and nonpartisan way

  • The Model Provider is given professional grade means to carry the burden of proof, while keeping a complete control over its IP.
  • The hedge fund or asset manager has the assurance that the performance profile is vetted by a neutral and skilled third party, using a controlled and repeatable approach.

The service is based upon Systemathics’s suite of algorithmic trading services and data sets.

The validation approach uses a 3 stage approach :

  • 1 : Strategy submission
  • 2 : Independent Testing
  • 3 : Report

1 : Strategy Submission

At this stage, the Model Provider will implement a complete production ready strategy based upon their model into a neutral platform to ensure that the minimal requirements are met.

  • The strategy designer receives access to Systemathics’ algorithmic trading software platform along with the required software development kit and accompanying documentation. The package is tailor made for offline back-testing and offline virtual trading. Included in the package is 1 (one) month of historical data (tick-by-tick)
  • The strategy designer implements inside the provided offline platform, the objective is to reach a production ready status, which addresses every aspect, from signal generation, to position building, down to the detailed execution flow (simple orders and/or smart orders). A signal alone is not a strategy, far from it. In this stage Systemathics’s skilled team can assist with training and recommendations.
  • If the strategy behaves as expected in offline mode (back-testing and virtual trading) it is then ready to reach next step. Otherwise, the designer must iterate step 2 as long as necessary, always in offline mode.

2 : Independent Performance Testing

When the strategy behaves as the designer intended and reaches the bare minimum checks for consistency and stability, it’s time to test it in a more stringent way:

  • The strategy is frozen and obfuscated before being delivered to Systemathics testing environment, along the required sets of parameters and configurations. Obfuscation and NDA ensure that IP is strictly enforced, it remains fully undisclosed to anyone.
  • A full set of tests are then executed on the strategy, mainly spanning two categories: back-testing and stress-testing. In each-and-every testing phase, the testing tools are working at the tick-by-tick level, the objective is to track the tiniest details in behavior. During the whole simulation process the necessary metrics (signals changes, complete order sequencing, portfolio dynamics…) are recorded in virtual back office with sub millisecond resolution.
  • Deep post trade analysis is then executed on the virtual back office to build meaningful time dependent metrics about the strategy’s output. The next step is about consolidating all the collected insights to a human readable report.

3 : Final Report

The final report is built and sent to both parties and contains all the required metrics to assess the performance of the strategy, and if workable, the required information to go live with it.

  • Strategy fact sheet, with summary and detailed performance measures (P&L, Sharpe, Max Draw Down, Scalability Index …)
  • If profitable, description of several kind of possible deployments for production, with costs estimates

Go live with your ideas: that’s out motto

Use Case

Bob has a trading idea targeting the US Stock Markets, he wants to convince Alice, head of trading at hedge fund H. that it’s a worthy idea, but without revealing too much of his “secret sauce” in his trading idea. Alice likes the returns promised by Bob but wants validation of trading idea. After initial discussion, they both agree to use Strategy Validator to build mutual trust and ratify the approach. The transformation of the idea to a proper trading strategy requires 2 weeks of Bob time, with a bit of help from Systemathics to iron out a few things (2-days “time and materials”). Bob builds the strategy in his own environment using Systemathics tool sets and ensure it meets minimum requirements. This can then move to second stage of testing . Bob securely submits his trading strategy into a “black box” within the Systemathic’s environment ready for the next step, which is independent performance testing step, In the Systemathic’s test environment Bob’s secured strategy is subject to 5 hours of back-testing and stress testing (Consuming 59 GB of tick-by-tick data, for a 3-years period) the post trade analysis step is executed and produces a 4 pages PDF report, a copy of which is sent to each party.

“A few thousands buck to select a million bucks a year idea, how cool is that?” (hint: super cool)

Highlights

  • Trust building in a secure, transparent, and neutral demilitarized zone, enforcing IP at all times
  • Assessments are done independently of the parties
  • Strategies are implemented in a market idiomatic coding model (event driven)
  • Tick-by-tick data is used at all stages of the testing so replicating real trading environments
  • Successful strategies can be promoted to production without further modification, slashing operational risk (code invariance) and reducing time for strategy to get into the market.
  • Asset Manager has an improved confidence that the Model Provider’s strategy offers real trading returns c.f. Euro NCAP