What we believe, and what we refuse to do.

The Indian markets are flooded with services that tell you what to do. They send instructions, show perfect-looking charts, and charge for urgency. That is not the product we are building.

Crux Research publishes research. A brief is a structured read of where the market sits today, which structural levels are in play, which sectors are leading, where the research stack has actionable conviction, and where it doesn’t. The decision to act on any of it remains entirely with the reader.

We believe the right unit of subscription is context, not commands. Anyone trading their own money deserves to make their own decisions, on their own thesis, with the best available context. That is what we sell.

Every claim in every brief carries a number. A 62% chance that a particular structural level is tested today is not a prediction that it will be tested. It is the system’s calibrated estimate of how often the setup’s historical analogues did.

Sometimes the level holds. Sometimes it doesn’t. The brief tells you which sometimes is more likely, and by how much. The rest is your edge to find.

A 70% probability we publish should resolve in our favour roughly 70 times out of 100. If it doesn’t, the gap shows up on our public track-record dashboard as calibration error.

When a particular probability bucket drifts beyond a tolerance, we flag it as drifting in the brief itself. You see, in real time, which parts of the read deserve more caution.

We do not advertise cherry-picked weeks. We show the per-bucket hit rate of every prediction we have published through the product.

Every brief ends with an audit of the previous brief’s calls. Per confidence bucket, per prediction type. The numbers are aggregated nightly and are visible to anyone who visits our track record page - subscriber or not.

When the audit window contains predictions that were replayed from historical bundles rather than collected live, we say so explicitly. Retrospective replay is useful for filling in calibration before our live history has accumulated, but it is not the same evidence - and the disclosure line keeps the difference honest.

Every number in every brief was knowable in real time at the moment we claim to know it. If a data point would not have existed before publication, it does not belong in the brief.

This is why the product separates live evidence from retrospective replay and labels both clearly. The reader should never have to guess what was known when.

We refuse instruction-style language in customer research. The brief can tell you where the market looks stretched, where attention is warranted, and where caution is higher. It cannot make the decision for you.

We also refuse, structurally:

• Performance-fee structures. We charge a flat subscription. Our incentives stay aligned with research quality, never with how aggressively you trade.
• Urgency-based marketing. No countdown timers, no “limited slots”, no “closing tonight”.
• Marketing the model’s individual best week. We publish the full distribution; you can find the best and worst weeks yourself.
• Holding ourselves out as SEBI-registered investment advisors. We are not. Our content is research context, not investment advice, and the disclosure runs on every page.

Crux Research is built for traders who already know what they are doing - and want better context to do it with. If you want signals, this is not the right service. If you want a calibrated reading of structural risk every morning, before NSE open, audited the next morning, this is exactly the right service.

Read the sample brief. Read the track record. Make your own call.