GM Advisors
  • Home
  • InvestLens
  • AWS Solutions
  • Our team
    • Alexey Kuznetsov, Ph.D.
    • Anton Yafremau, M.Sc.
    • Benny Zaionz, B.A.Sc.
    • Boris Korotkov, B.Sc.
    • Chrysoula Dioli, Ph.D.
    • Eugene Furman, Ph.D.
    • Peter Antoniadis, MBA
  • Contact us
  • Newsletter
    • December 2025 Issue
    • January 2026 Issue
    • February 2026 Issue
    • March 2026 Issue
    • April 2026 Issue
Select Page
View PDF
InvestLens logo
InvestLens
https://gmadvisors.ca
APRIL 2026

NEWSLETTER

Happy New Year celebration
Highlights

We continued development of the InvestLens AI Assistant with a strong focus on production readiness, workflow integration, and platform security. Our testers have been working closely with the system in real usage conditions to evaluate how the Agent fits into existing workflows, while we refine its behavior to support practical, reliable use. We are designing the Assistant to operate through validated tools with advanced guardrails, helping reduce low-confidence responses and supporting a more secure experience for both the platform and user data.

In March, we made meaningful progress on several important fronts: our implied volatility modelling work advanced into its late stage, we collaborated with an AWS Partner on integrating agentic AI into the platform, and we continued engaging potential enterprise clients as GM Advisors further strengthens its technical and delivery capabilities.

What changed
  • A fully functional UI is now available to testers, allowing users to interact with the InvestLens AI Assistant in a conversational mode for quick retrieval of their portfolio data.
  • Additional refinements and small improvements were made across the platform.
APRIL 2026
https://gmadvisors.ca

Platform updates


🤖InvestLens AI Assistant
We aim to bring the value of an advisor into your home or office without compromising data security or confidence. In collaboration with an AWS Partner we are developing an agentic system designed to meet strict security standards, generate portfolio reports using validated analytical tools, and operate with guardrails that reduce low-confidence responses.
📚Research and Development
We are in the late stage of developing our implied volatility estimation methodology. By addressing key sources of error such as truncation bias, discretization error, American-style option pricing effects, approximation error, and dividends, our approach is designed to deliver forward-looking volatility estimates with greater theoretical consistency and precision. If implied volatility plays a role in your investment decisions, stay tuned for methodological improvements from our team of experts.
🛡️Data Validation
We have been working hard to ensure that our data extraction process remains as precise as possible. To support this, we have been enhancing our tools to track and report errors related to split adjustments. The extraction system is being designed to alert our team to these issues so that we can act on them promptly.
✡️ChaiTech Accelerator
Our participation in the ChaiTech Accelerator has concluded, and we are truly grateful for the support of the community, especially Josh Zweig, Talli Koren, and Kim Parnell. We had the opportunity to engage with highly respected VCs and hope to benefit from their advice and feedback. We also extend our sincere congratulations to the winner, Ève Seni, for her medical scheduling.
Upcoming

We are pleased to share that, with support from an additional Amazon Web Services grant, we have continued advancing the InvestLens agent we are developing to intelligently orchestrate tools across the platform. At this stage, our focus is on improving the working prototype and ensuring that the assistant can execute validated tools consistently and without errors, including the reliable generation of client-facing PDF reports. Unlike many AI-first products that rely on opaque outputs, our approach remains centered on human accountability, expert oversight, and strict guardrails designed to support transparency, reliability, and trust.

Upcoming features illustration
APRIL 2026
https://gmadvisors.ca

Analytics in Practice

This month, we present a workflow demonstrating how to quickly compare strategies using InvestLens. In this example, we compare two ideas: a portfolio built from 9 U.S. stocks in the Aggressive Tactical Profile, excluding GOOG Class A, and a portfolio built from 10 stocks in the AI Tech Infrastructure Theme.

Aggressive Tactical Profile AI Tech Infra Theme
ADBE.US Adobe Systems Incorporated ASML.US ASML Holding NV ADR
AER.US AerCap Holdings NV ASX.US ASE Industrial Holding Co Ltd ADR
BBY.US Best Buy Co. Inc AVGO.US Broadcom Inc
BYD.US Boyd Gaming Corporation CRM.US Salesforce.com Inc
GOOG.US Alphabet Inc Class C CRUS.US Cirrus Logic Inc
IBM.US International Business Machines DLO.US Dlocal Ltd
MMM.US 3M Company GFS.US Globalfoundries
RELX.US Relx PLC ADR GTM.US ZoomInfo Technologies Inc.
T.US AT&T Inc KLAC.US KLA Corporation
PAGS.US PagSeguro Digital Ltd

We then use the optimizer to allocate assets for both the Aggressive Tactical Profile and AI Tech Infrastructure portfolios, save the optimized portfolios, and compare their historical performance.

Performance Report Agg. Tactical Profile AI Tech Theme
Annualized Portfolio Return 0.323 0.633
Annualized Portfolio Volatility 0.159 0.363
Annualized Sharpe Ratio 1.556 1.412
Annualized Semi-deviation 0.112 0.245
Daily VaR (5%) 0.014 0.032
Daily CVaR (5%) 0.021 0.049
Daily Cornish-Fisher VaR (5%) 0.015 0.035
Skewness -0.163 0.190
Kurtosis 8.571 7.508
Drawdown -0.132 -0.300

Over the tested window, the AI Tech Infra Theme portfolio delivered a substantially higher annualized return than the Aggressive Tactical Profile portfolio (0.633 vs. 0.323), but this came with materially higher volatility (0.363 vs. 0.159). Although the AI Tech Infra Theme portfolio generated the stronger return, the Aggressive Tactical Profile achieved a higher Sharpe ratio (1.556 vs. 1.412), indicating more efficient risk-adjusted performance. Downside risk measures also favor the Aggressive Tactical Profile, with lower semi-deviation, VaR, CVaR, and a smaller drawdown.

In conclusion, this comparison shows how InvestLens helps investors move beyond return alone and evaluate strategies across both performance and risk dimensions. In this example, the AI Tech Infra Theme offered stronger upside potential, while the Aggressive Tactical Profile delivered more favorable risk-adjusted characteristics and downside protection.

Grand Management Advisors logo
© 2026 All Rights Reserved.
Powered by AWS NVIDIA Inception Program
GM Advisors Logo NVIDIA Inception Program Badge
  • Follow
  • Follow
  • Home
  • InvestLens
  • AWS Solutions
  • Our team
    • Alexey Kuznetsov, Ph.D.
    • Anton Yafremau, M.Sc.
    • Benny Zaionz, B.A.Sc.
    • Boris Korotkov, B.Sc.
    • Chrysoula Dioli, Ph.D.
    • Eugene Furman, Ph.D.
    • Peter Antoniadis, MBA
  • Contact us
  • Newsletter
    • December 2025 Issue
    • January 2026 Issue
    • February 2026 Issue
    • March 2026 Issue
    • April 2026 Issue

© 2025 Grand Management Advisors Inc. 



All rights reserved.