Subject: Request for Guidance on FII/DII Sentiment Analysis Methodology Used by Sensibull

Zerodha
1

Dear Mr.Nithin Kamath,Mr. Nikhil Kamath&Mr.Karthik Rangappa,

I hope this message finds you well.

I am Manivel G, a retail trader and Zerodha client who regularly uses the Sensibull platform — one of the most insightful tools I’ve come across in my trading journey. I truly appreciate the decision to offer Sensibull free to Zerodha users — it has helped many of us build better discipline and structure around our trading decisions.

I have been consistently analyzing the FII/DII Summary Page, particularly the Index Futures sentiment predictions — Bullish, Bearish, Indecisive — and their classification into Mild, Medium, or Strong. While I have reached out to the Sensibull support team (who have been helpful), I still have some clarity gaps.

Why I’m Reaching Out to You

As the platform is offered via Zerodha, and considering your commitment to empowering retail traders, I’m writing with a humble request for general guidance — not proprietary methods, but simply direction or principles on how the classification is arrived at.

Over the past few months, I have been deeply analyzing the FII/DII summary section on the Sensibull platform — particularly the index futures sentiment classification (Bullish, Bearish, Indecisive) and the intensity (Mild, Medium, Strong). I’ve tried to reverse engineer the methodology using:

  • Net OI and Change in OI
  • Historical data comparisons
  • Statistical scoring methods (Z-score, T-score, Modified Z, IQR, etc.)

Despite meaningful effort, some aspects remain unclear due to the proprietary nature of Sensibull’s algorithm. I’m fully aware and respectful of this — but I’m reaching out to ask if you could offer general guidance or principles on the classification model.

My Key Questions

  1. What is the primary metric used to determine direction (bullish/bearish/indecisive)?
  • Is it based more on Change in OI, Net OI, or a combination of both?
  • How is contradictory data (e.g. negative Net OI, positive Change in OI) handled?in this case,sometimes it shows indecisive and sometimes it shows bullish or bearish
  1. What factors influence the strength classification (mild, medium, strong)?
  • Is the day’s value compared against a fixed lookback range?
  • Are statistical scores like Z-score or percentile used?
  1. Could you share a working example, even for a past date (e.g., June 27, 2025), with actual numbers and how Sensibull arrived at the sentiment?

Why This Matters

This is not just curiosity — I’ve been backtesting these patterns, and the insights derived have genuinely improved my decision-making. However, the lack of transparency on the classification logic (which is understandably proprietary) makes it difficult to build confidence in independently validating the signals.

I truly admire your work at Zerodha and your efforts to simplify and democratize finance in India. If you or someone from your team could share even high-level guidance or learning resources, it would mean a lot to me and many traders like me who seek to learn deeply and trade responsibly.

Thank you again for your time, and I look forward to any direction you might be able to offer.

Warm regards,
Manivel G
[email protected]

2

@Sensibull

3

Hey Manivel, @Sensibull will give you answers. :slight_smile:

4

Hi @Manivel_G writing a detailed answer on this. Will post that here

5

First of all, this is just an indicative system to help beginners, and not a rule based fool proof logic that can be used by an advanced trader to build a trading system

  1. What is the primary metric used to determine direction (bullish/bearish/indecisive)?
    We take into account the current option OI and the change in OI. We benchmark the OI and the change in OI into percentiles of OI and OI change across all participants - FII Client Pro

We categorize OI into different strengths and classify OI change into different strength levels. And then we combine OI and OI change to find a single number. We try to have a verdict ONLY when everything points in the same direction. If any doubt, we try to go with “indecisive”. We would rather be not sure than outright wrong

  • Is it based more on Change in OI, Net OI, or a combination of both?
    Both

  • How is contradictory data (e.g. negative Net OI, positive Change in OI) handled?in this case,sometimes it shows indecisive and sometimes it shows bullish or bearish
    If the OI is strongly bullish and the change is mild, we go with strong overall. And vice versa. Again, this is not a bullet proof heuristic but an indicative guideline

  1. What factors influence the strength classification (mild, medium, strong)?
    We keep it super simple - Like we said above, only OI and OI change

  2. Could you share a working example, even for a past date (e.g., June 27, 2025), with actual numbers and how Sensibull arrived at the sentiment?
    I would suggest not going deep and making a system of this. We would rather that you look at classifications of Strong/ Medium/ Mild as guidelines than anything else. Simply because this categorization is not complete or 100% correct. This is because many things can corrupt data - For example a major change in OI on an expiry day just means contracts going away and not an intentional buy/sell action. Our intention of this classification was just to give some indication.

It is always better if you look at the overall data from first principles, rather than quantify it and throw a verdict using a rule-based system depending on numbers

[/quote]

2 Likes
6

Thank you for the detailed information you’ve provided regarding the sentiment classification system. I appreciate the clarity on how the metrics are combined to arrive at the indicative sentiment, and I understand the emphasis on using this as a guideline rather than a strict rule-based system.

I will carefully go through the explanation you shared and get back to you shortly with any further questions or thoughts.

Additionally, if possible, could you kindly share the contact email of Abid Hassan, CEO of Sensibull, for some direct clarifications I’d like to seek?

Thank you once again for your support.

7

Subject: Request for Lookback Periods and Classification Thresholds for Participant-Wise OI Sentiment Analysis

Dear Sensibull Team,

I hope this message finds you well.

I am conducting research and analysis based on the Open Interest (OI) data of various market participants—namely FII, DII, CLIENT, and PRO—in the Index Futures segment. To align my model with the methodology used on your platform, I kindly request the following information:

  1. Lookback Periods:
  • Could you please share the specific number of days (lookback periods) used separately for:
    • Net Open Interest (Net OI)
    • Change in Open Interest (OI Change)
      for each of the following participants:
    • FII
    • DII
    • CLIENT
    • PRO
  1. Sentiment Classification Thresholds:
  • What percentile-based thresholds do you use to classify the sentiment strength (e.g., Strong Bullish, Medium Bearish, Neutral, etc.) based on Net OI and OI Change?
  • Specifically, I am interested in the percentile cut-offs you consider for:
    • Strong Bullish / Bearish
    • Medium Bullish / Bearish
    • Mild Bullish / Bearish
    • Neutral / Indecisive

Your insights would be incredibly helpful in aligning my analysis approach with industry standards.

Thank you in advance for your time and support.@nithin @Sensibull @siva @nithin

8

Hi Manivel,

Sorry, I’m not at the liberty to disclose our full methodology here. Like I said, it’s not something that is hard and fast rule and just a guidance system. You can create your own by making reasonable assumptions

Regards,
Abid

1 Like
10

Hi Traders,
If anyone is familiar with how classifications are done in Index Futures for all participants under the FII/DII summary, I’d appreciate your insights.
Kindly share your thoughts by emailing me at [email protected].
Thank you in advance for your help!
@siva @nithin @Sensibull @alltraders