Compliance whitepaper · v2.2
Educational Publisher Posture & Operational Controls
Canonical statement of Bharath Shiksha's operating posture under the SEBI January 2025 educational-publisher framework. Editorial guardrails, operational controls, Edge Terminal compliance posture, and self-audit matrix.
Sections
- Executive summary
- The SEBI January 2025 framework
- What we do
- What we do not do
- Editorial guardrails
- Community moderation
- Operational controls
- Edge Terminal — educational analytics layer (Q2 2026)
- Refunds, grievance, privacy
- Quarterly public reporting
- AI tutor disclosure
- Affiliate & testimonial compliance
- Self-audit matrix
- How we differ from tip-channel operators
- Why we will never become SEBI-registered
- Forward-looking — what triggers a future update
- Version history
1. Executive summary
Bharath Shiksha operates as an educational publisher in technical analysis and trading-process discipline. Structurally absent: buy/sell calls, live signals, return claims, portfolio-management overlays, SEBI-registered IA or RA capacity.
Posture is enforced by operational controls — 30-day data lag on every example, weekly forbidden-language pattern scan, curriculum-team review of every public publication, disclaimer blocks on every page, quarterly public compliance report. Document reviewed quarterly and after material SEBI framework changes.
2. The SEBI January 2025 framework
In January 2025 SEBI issued a circular distinguishing investor education from investment advisory for the digital-content economy. The two categories:
Investment advisory (regulated, requires registration)
- Personalised advice on what to buy, sell, or hold in specific securities.
- Real-time signals or alerts directing clients to act on specific securities at specific times.
- Performance claims, accuracy statistics, or return projections backed by track records.
- Portfolio construction or rebalancing services delivered to identifiable clients.
- Research-analyst output on specific securities offered as advice.
Activities of this kind require SEBI registration as an Investment Adviser (under SEBI IA Regulations 2013) or Research Analyst (under SEBI RA Regulations 2014), with capital adequacy, qualification, dispute-resolution, and audit obligations.
Investor education (permitted without IA/RA registration)
- Teaching frameworks, methodologies, and concepts in technical analysis, risk management, and quantitative methods.
- Analysis of historical data with the benefit of hindsight, with at least 30-day data lag, anonymised, and framed as illustration of methodology.
- Curriculum delivery via written, video, cohort, or interactive formats.
- Tutor channels that answer methodology questions, not specific-security questions.
- Tools and calculators that compute generic outputs from user inputs without prescribing actions.
One-sentence boundary: educational publishing teaches how to think; investment advisory tells what to do. Bharath Shiksha sits on the educational-publishing side.
3. What we do
3.1 The structured curriculum
The Bharath Shiksha curriculum spans six stages: Foundation (Stage 1), Systematic (Stage 2), Professional (Stage 3), Mastery I — Quantitative (Stage 4), Mastery II — Systems Architect (Stage 5), and Mastery III — Institutional Elite (Stage 6). Across the six stages we cover technical analysis from the first candlestick to the operational discipline of an institutional trading desk.
Each stage delivers a printed curriculum book, video lessons, structured exercises, gate quizzes, a trade-journal template, weekly review rituals, a tutor channel for methodology questions, and (cohort-mode only) live discussion threads with peer accountability.
3.2 The methodology encyclopedia
The encyclopedia documents 1,308 trading methodologies across 35 volumes in seven scanner categories: Bullish (200), Bearish (200), Intraday (200), Crossovers (200), Volume Profile (250), Range Breakout (150), Candlestick (108). Each methodology is described with formulae, parameters, signal criteria, common mistakes, anonymised historical examples with 30-day data lag, and recommended timeframes. Reference work, not a trading playbook.
3.3 Articles and lead-magnet content
Long-form articles cover Indian retail trading topics — F&O loss patterns, position sizing, behavioural biases, regime detection, intermarket relationships, taxation, broker selection, AIF Cat-III operations. Each article is reviewed before publication for compliance posture and re-scanned automatically every Tuesday by the compliance-scan cron.
3.4 Calculators
Browser-side calculators (risk of ruin, expectancy, position sizing, Kelly criterion, drawdown, compound returns, journal grader). All calculators run entirely in the browser with no network access for their core function — user inputs do not transit to our servers.
3.5 Cohort instruction
Cohort enrolment delivers the same curriculum content as solo path, plus structured weekly discussion threads, monthly live Q&A sessions with curriculum staff, and peer-pair journal review. Cohort sessions are educational; cohort threads are moderated for tip-sharing and other policy violations described in section 6.
3.6 The tutor channel
An asynchronous tutor channel where students can ask methodology questions and receive curriculum-staff responses within published SLAs. The tutor channel does not answer questions of the form "should I buy X" or "is now a good time to enter Y" — those are explicitly redirected, as documented in section 5.5.
3.7 Edge Terminal (the educational-analytics layer)
Edge Terminal is a six-module browser-based analytics layer (Visualiser, Watchlist, Backtester, Trade Journal, Risk-Literacy Dashboard, Methodology Library) that lets students apply the curriculum's methodologies to historical, anonymised, ≥30-day-lagged Indian market data. It is the practice room for the curriculum, not a market-intelligence product. Included with Bharath Shiksha enrolment and served at /terminal/ (public overview at /edge-terminal) since May 2026. Its compliance posture is documented in full in section 8.
The Methodology Library at /terminal/library/ catalogues 700 trading methodologies across all six Bharath Shiksha stages and nine scanner categories — distinct from the 1,308-methodology paid encyclopedia documented in section 3.2 (the encyclopedia is the depth-reference work; the Library is the operationally-codified subset used inside the practice tools). 70 of the 700 are mechanically codified into the Backtester rule engine; 81 render as Visualiser overlays. Every module ships with a sticky educational-tool banner, per-overlay and per-result disclaimers, and explicit ILLUSTRATIVE · HISTORICAL · DELAYED labelling. Edge Terminal is structurally incapable of producing buy/sell calls, real-time signals, or accuracy claims — the four V0-prototype features that could have crossed the SEBI advisory line (signal cards, AI-MATCH news-to-instrument mapper, prediction markets, real-time data feeds) were explicitly rejected as a moat decision and are not implemented.
4. What we do not do
The following activities are structurally outside Bharath Shiksha. Operational controls prevent these from being added:
Buy / sell / hold calls on specific securities
No statement directing students to act on a named security. Historical examples are anonymised or framed as illustration of methodology with 30-day lag.
Live signals or real-time alerts
No subscription product that pushes "X is breaking out, enter now" notifications. Curriculum content is asynchronous and does not depend on market timing.
Performance claims or return guarantees
No "our system has delivered X%" language anywhere. No claimed accuracy on signals. No marketing copy implying specific returns.
Portfolio construction for clients
No personalised allocation advice. No "given your situation, buy these stocks." The student constructs their own approach using the curriculum as input.
Brokerage or trade execution
We do not place trades, route orders, or operate as an execution venue. Stage 5 (Mastery II) covers algorithmic deployment via student-owned broker APIs only.
Fund management on third-party capital
We do not manage AIF, PMS, or any third-party capital. Stage 6 (Mastery III) covers AIF Cat-III operations as curriculum, not as a service.
The platform contains no payment surface for any of the listed activities. No "subscribe to signals" SKU, no advisory tier, no upsell to any SEBI-registered service.
5. Editorial guardrails
The operational rules every writer, editor, instructor, moderator, and AI tutor follows. Full version: /compliance/02-editorial-guardrails.md.
5.1 The 30-day data lag rule
Every chart, price reference, named-security illustration, or specific-event mention in published content must reference data at least 30 calendar days old at the time of publication. The rule applies to articles, videos, cohort sessions, encyclopedia methodology examples, and AI-tutor responses. The 30-day lag prevents any educational illustration from operating as a real-time recommendation.
5.2 Anonymisation
Specific-security references in trade examples are anonymised by default. When a real ticker is used (for pedagogical clarity), it is framed as an example of methodology applied historically — never as a present recommendation.
5.3 Forbidden language patterns
The following patterns are prohibited in any public-facing content. The list is enforced by the weekly automated `compliance-scan` cron and by manual review at publication time.
| Pattern | Why prohibited |
|---|---|
buy / sell / short / long <TICKER> in imperative voice | Reads as a directive rather than an illustration. |
target price of X for Y | Specific price targets are research-analyst output requiring registration. |
X% guaranteed / assured / annualised returns | Return claims violate IA regulations. |
our system has delivered X | Performance claims for educational platforms imply advisory accuracy. |
X% of signals were profitable | Accuracy statistics imply we operate signals (we do not). |
hot stock / pick / tip this week | Tip language explicitly outside the educational scope. |
5.4 Pre-publication compliance review
Every public publication — article, video script, encyclopedia volume, marketing copy, email — passes a compliance review before going live. The review checks:
- Educational framing throughout, no imperative trade directions.
- Compliance disclaimer present where required.
- 30-day data lag honoured.
- No forbidden patterns from section 5.3.
- Performance language audited (no implicit accuracy or return claims).
- Citations and historical references properly contextualised.
5.5 The compliance-scan cron
A scheduled function runs every Tuesday at 08:30 IST against the live deployed site. It scans all HTML for the forbidden patterns in section 5.3 plus an allowlist of pages that intentionally reference forbidden language for educational purposes (e.g., the article titled "Why Bharath Shiksha will never be SEBI-registered" deliberately quotes regulatory language). Any new violation triggers a Postmark email to the curriculum team and a written entry in the public quarterly compliance report.
6. Community moderation
6.1 The cohort thread
Cohort discussion threads are educational forums for students to discuss methodology, share trade-journal observations on historical data, and ask curriculum-staff questions. They are explicitly not tip-sharing channels. The following rules are enforced:
- No real-time market predictions or recommendations. Posts of the form "X is breaking out, who else is in?" are removed and the poster warned. Repeat violations end cohort participation per the refund policy.
- No performance bragging. Posts advertising specific returns achieved by individual students are removed. Generic discussion of methodology results in R-multiples for educational illustration is permitted.
- No commercial solicitation. Cohort threads cannot be used to promote any external service, including SEBI-registered or unregistered.
- Civil discussion. Personal attacks, harassment, and abusive language end participation.
6.2 Tutor channel response policy
The tutor channel is staffed by curriculum-team members with explicit response policies:
- Methodology questions are answered. "How do I compute expectancy on this dataset?" or "Why does the regime filter exclude this period?" — answered.
- Specific-security questions are redirected. "Should I enter Reliance now?" or "Is HDFC at this price a good buy?" — politely declined; pointed at SEBI-registered RAs/IAs for that question; offered to discuss the methodology that would frame the question.
- Personal portfolio questions are redirected. "I have ₹X — how should I deploy it?" — politely declined; pointed at SEBI-registered IAs.
7. Operational controls
7.1 Pre-trade exercise framework
The Foundation curriculum's pre-trade checklist is an exercise the student fills out for hypothetical or paper-trading scenarios. It is not a service we operate against the student's real positions. The framework teaches the student to ask the questions; we do not answer them on the student's behalf.
7.2 Stage 5 algorithmic curriculum
Stage 5 (Mastery II) covers live deployment of student-built algorithmic systems against student-owned broker APIs. The curriculum content covers SEBI's January 2025 algo-trading circular requirements, kill-switch design, monitoring, and audit logging. Bharath Shiksha does not host or operate any student's algorithm; the deployment is to the student's own infrastructure.
7.3 Stage 6 AIF curriculum
Stage 6 (Mastery III) covers AIF Category III operations — fund structure, capital raising, the SEBI AIF Regulations 2012 framework, IA/RA registration walkthrough — as a curriculum subject for students whose intended path is fund management. We do not manage AIFs, do not assist students in raising capital, and do not facilitate any registration on the student's behalf. Students who wish to pursue AIF operations engage qualified legal and compliance advisers separately.
7.4 The Razorpay payment flow
All payments flow through Razorpay (PCI-DSS compliant). We never see card numbers, CVV, or stored payment instruments. Payment data is held by Razorpay; we receive only the transaction confirmation, amount, and method type.
7.5 Identity and access
Student access is gated through Netlify Identity (passwords hashed). Stage access is provisioned via the Razorpay webhook. Account credentials are the student's responsibility; sharing is grounds for termination per the Acceptable Use section of the Terms of Service.
8. Edge Terminal — educational-analytics layer (Q2 2026)
Edge Terminal is the educational-analytics layer of Bharath Shiksha, included with enrolment and served at /terminal/ (public overview at /edge-terminal) since May 2026. It is a browser-based six-module practice environment that lets students apply curriculum methodologies to historical and ≥30-day-lagged Indian market data. The terminal exists for one purpose: to make the curriculum interactive without crossing the SEBI educational-vs-advisory boundary.
This section documents the compliance posture of Edge Terminal in full: what each module does, why each is educational rather than advisory, the four V0-prototype features explicitly rejected as a moat decision, the data sources and processors, the retention windows, and the conditions under which a future revision of this section would be triggered. The full product specification is documented at EDGE_TERMINAL_V2_SPEC.md; the five-role compliance advisory that catalysed the moat decisions is documented at EDGE_TERMINAL_ROUNDTABLE.md; the build reports are at EDGE_TERMINAL_BUILD_REPORT.md and EDGE_TERMINAL_PHASE2_BUILD_REPORT.md.
8.1 The six modules
Module 01 — Methodology Visualiser
Loads any of 50 anonymised or large-cap Indian instruments (broad indices, sector indices, NIFTY 50 large-caps) on daily / weekly / monthly historical OHLC, and auto-draws methodology overlays from the curriculum: EMAs (20/50/200), support and resistance, Fibonacci retracement, VWAP with bands, candlestick patterns (twelve named rules), market structure (HH/HL/LH/LL with Break-of-Structure detection), RSI with divergence, and Volume Profile with point-of-control. Each overlay opens a side-panel card explaining the methodology in two paragraphs of curriculum-grounded educational copy, with a deep link back to the curriculum lesson that teaches it. Education-first: the Visualiser teaches the student to see a methodology applied historically. It does not produce trade entries, exits, or directional calls. Implementation: /terminal/index.html, /terminal/visualiser.js, /terminal/overlays.js, /terminal/methodology-content.js.
Module 02 — Watchlist
A personal browser-local list of up to 25 instruments the student is studying. Each row shows ticker, name, last close, 1-day / 5-day / 30-day deltas, 52-week high and low, and an editable 240-character note. CSV export supported. Pure utility. No alerts (alerts would be tantamount to entry signals), no ranking, no "your watchlist is bullish" framing, no auto-generated commentary. Data persists in browser localStorage; nothing transits to our servers. Implementation: /terminal/watchlist.html, /terminal/watchlist.js.
Module 03 — Backtester
Twelve curriculum methodologies codified as mechanical entry-stop-target rules, runnable against five-plus years of historical data on any instrument in the universe. The student selects methodology, instrument, position-sizing rule (0.5% / 1% / 2% fractional), R-multiple target, maximum holding period, and a realistic transaction-cost stack (FY-26 NSE retail brokerage + STT + GST + slippage). The engine walks the data bar-by-bar with next-bar-open entry to prevent look-ahead bias, then renders eight key performance indicators, an equity curve, a drawdown profile, an R-multiple histogram, a monthly returns heatmap, and a full sortable trade log with CSV export. Every panel carries an ILLUSTRATIVE · HISTORICAL label and a past-performance disclaimer block. Implementation: /terminal/backtester.html, /terminal/backtester.js, /terminal/backtester-methodologies.js.
Module 04 — Trade Journal
A personal post-execution journal. The student logs their own trades after they have closed them — ticker as free text (the journal does not validate, execute, or route), direction, entry, stop, exit, methodology used (chosen from a preset list of 18 plus a free-text "other"), thesis, and notes. R-multiple is computed from the student's inputs. Aggregate dashboard shows eight KPIs (total trades, win rate, average R, average win R, average loss R, expectancy, profit factor, cumulative R) and a behavioural-insights engine that links observed patterns — loss-aversion skew, profit-factor below 1.2, methodology concentration, rolling-window edge stability — to specific curriculum modules. CSV export supported. Browser-local localStorage only at V1; cross-device sync via Netlify Identity is a Phase 2.5 follow-up, disclosed in section 8.6. Implementation: /terminal/trade-journal.html, /terminal/trade-journal.js.
Module 05 — Risk-Literacy Dashboard
An educational view of macro market conditions on a ≥30-day data lag. Six indicator cards (India VIX, NIFTY 50, BANK NIFTY, USD/INR, MCX Gold, MCX Crude) each show last close and 1-day / 5-day / 30-day deltas accompanied by three-to-five-sentence curriculum-grounded explainers of what each indicator measures and what it does not tell a student. A nine-sector breadth heatmap surfaces 30-day moves across NSE sector indices, framed in language about regime context, not directional implication. A twenty-session FII / DII flow chart (deterministic-synthetic at V1, clearly flagged as illustrative) is paired with educational copy on the difference between context and signal. Every indicator carries a DELAYED label; every panel disclaims past patterns as no guarantee of future behaviour. Implementation: /terminal/risk-literacy.html, /terminal/risk-literacy.js.
8.2 The methodology library — the curriculum-Backtester moat
The Bharath Shiksha methodology encyclopedia documents 1,308 trading methodologies across 35 volumes in seven scanner categories — Intraday (200), Candlestick (108), Range Breakout (150), Bullish (200), Bearish (200), Crossovers (200), Volume Profile (250). The full catalogue is reference work: formulae, parameters, signal criteria, common mistakes, anonymised historical examples with 30-day lag, recommended timeframes.
Of the 1,308 catalogued methodologies, approximately 50 to 100 are mechanically codifiable as Backtester rules — i.e., they reduce to a deterministic entry signal, a deterministic stop placement, and a deterministic target or exit condition. The remaining ~1,200 methodologies are multi-factor frameworks, composite scorers, or qualitative reads that are surfaced as documentation in the curriculum and the Visualiser, but are not directly executable as Backtester strategies. At V1 launch (May 2026), twelve methodologies are codified as Backtester rules; the codified library is intended to grow toward ~50-100 over subsequent quarterly releases, gated to the stage that the methodology belongs to (a Stage 2 methodology is unlocked only for students enrolled in Stage 2 or above; the implementation is in /terminal/backtester-methodologies.js with a stage tag on each rule).
Three compliance consequences follow from this architecture:
- Backtester outputs are historical illustrations, not strategy recommendations. The student is testing a methodology they are learning, on data they could have studied themselves with sufficient time. The terminal accelerates that study; it does not replace the student's judgement on whether to use the methodology in a real account. The student's separate decision to deploy any methodology to a live account is theirs alone and falls outside the scope of any Bharath Shiksha service.
- The Backtester does not screen or scan the live market. The student selects an instrument; the engine then runs the chosen methodology against that instrument's historical bars and returns the historical outcome. There is no "show me which instruments today are setting up this methodology" feature, and there will not be one — that surface would constitute a screener with directional output, which the SEBI January 2025 framework treats as research-analyst output.
- No accuracy claims, no aggregate statistics published. Backtester results live in the student's session only. We do not publish "the Bullish Engulfing rule produced an expectancy of X over the last decade." Aggregate performance claims of that kind would invite the inference that we are operating signal systems with stated accuracy, which we are not.
8.3 The four V0-prototype features explicitly rejected (the moat decision)
The Edge Terminal V0 prototype, brainstormed in CodeSandbox fn9t7k in early May 2026, contained four features which a five-role advisory roundtable (Legal & Compliance, CEO, CFO, CMO, CRO) reviewed on 2026-05-11 and unanimously recommended for elimination before any public deployment. The full roundtable record is preserved at EDGE_TERMINAL_ROUNDTABLE.md. The decisions are documented here because the discipline of what we choose not to build is the strongest evidence of the educational positioning. Each of the four would have crossed a regulatory line; each was therefore not built.
AI Market Signals cards (REJECTED)
The V0 prototype displayed cards labelled BUY / STRONG BUY / SELL on NIFTY, BANK NIFTY, MCX GOLD, MCX CRUDE, NIFTY IT, USD/INR, each with explicit Entry / Target / Stoploss prices and a Confidence percentage. This is textbook unregistered research-analyst output under SEBI RA Regulations 2014 — specific securities, specific directional calls, specific price targets, quantified confidence. The "AI-powered" framing offers no shield: the SEBI August 2024 amendment to the IA Regulations 2013 explicitly brings algorithmic and automated recommendation engines within scope. Rebuilding this surface would invite a SEBI cease-and-desist under SEBI Act §15HA, with monetary penalty exposure of ₹1L to ₹1Cr and reputational damage that would end the academy. Not rebuilt. No roadmap to rebuild.
AI-MATCH news-to-instrument mapper (REJECTED)
The V0 prototype's FEED tab mapped news headlines algorithmically to specific instruments and directional outcomes with confidence scores ("RBI policy signal → NIFTY Bank Index +2.3% / 94% confidence"). This is an algorithmic recommendation engine delivered through a news interface, which is the same regulatory status as the signal cards and arguably worse because it implies causal prediction. Under SEBI RA Regulations 2014 the output would constitute unregistered research-analyst publication. Not rebuilt. No roadmap to rebuild.
PREDICTIONS board (REJECTED)
The V0 prototype's PREDICTIONS tab displayed YES/NO probability markets on financial-outcome questions (rate cuts, index-level thresholds, SEBI rule changes, earnings outcomes). The regulatory exposure here is doubly compounded: (i) opinion-trading platforms on financial outcomes operate in a regulatory grey under the Public Gambling Act 1867 and state gaming laws, and SEBI issued guidance in 2024 against unregistered prediction markets on financial outcomes; (ii) even a no-stakes display of probabilities on financial outcomes is plausibly construed as forecasting publication, which is research-analyst territory. We do not operate prediction markets, do not host opinion-trading interfaces, do not display probability boards on financial outcomes. Not rebuilt. No roadmap to rebuild.
Real-time NSE / BSE / MCX data feeds (REJECTED)
The V0 prototype carried an always-on global markets ticker and a LIVE indicator. Edge Terminal V1 operates exclusively on historical and ≥30-day-lagged data — including the Risk-Literacy macro indicators. Three reasons: (i) data-licensing exposure: NSE charges ₹2-5L per annum for real-time data redistribution and most free-tier sources prohibit commercial redistribution in their terms of service; (ii) regulatory ambiguity: a paid product surfacing real-time price action invites the inference of trading-interface status rather than educational status, which is exactly the inference the SEBI January 2025 framework was issued to clarify; (iii) educational utility: methodology is taught on historical data, where the outcome is known and the student can audit the reasoning. Real-time data adds nothing to the curriculum's pedagogical mission. Not built. No roadmap to build.
The four rejections are not a temporary scope cut. They are a permanent moat decision. The terminal's value proposition is precisely the discipline that distinguishes an academy from a signal channel: the student practises methodology on historical data, develops judgement, and bears the responsibility for their own positions. Each of the four rejected features would have shortcut that discipline, and would have created the operational temptation — and the regulatory obligation — to deliver advice. The educational-publisher posture and the four rejections are inseparable.
8.4 Compliance posture, surface by surface
Every Edge Terminal page implements the following compliance surfaces. The implementation is verified at the HTML level across /terminal/index.html, /terminal/watchlist.html, /terminal/backtester.html, /terminal/trade-journal.html, and /terminal/risk-literacy.html, with corresponding logic in the JavaScript modules.
| Surface | Implementation | Verification |
|---|---|---|
| Sticky educational-tool banner | Top-of-page strip on every module: "EDUCATIONAL TOOL · Historical and delayed (≥30 days) data only. Not investment advice. No buy/sell recommendations. No live signals." | Manual smoke audit across all six module pages; the banner is part of the shared HTML skeleton and cannot be removed by toggling overlays or running a backtest. |
| Per-chart overlay label | "ILLUSTRATIVE · HISTORICAL · DELAYED ≥30 DAYS" rendered on the chart canvas in the Visualiser and on every result panel in the Backtester. | Renders inside the chart container, scrolls with the chart, screenshots the same as the chart. |
| Per-overlay disclaimer | Each of the eight methodology cards in the Visualiser side panel ends with a one-sentence disclaimer (example: "Historical EMAs are an illustration of trend regime classification, not a trade entry signal.") | Defined in /terminal/methodology-content.js; renders as the last paragraph of every overlay's card. |
| Per-result past-performance disclaimer (Backtester) | A dedicated compliance block at the bottom of every backtest result: "Past-performance disclaimer: the equity curve above is a historical illustration of the chosen methodology applied to historical bars; past performance is no guarantee of future results; transaction costs and slippage modelled for retail FY-26 NSE conditions; the student bears responsibility for any decision to apply the methodology in a real account." | Defined in /terminal/backtester.html below the results panels. |
| Trade-Journal "your data only" framing | The form copy reads explicitly: "You enter your own trade data; Bharath Shiksha does not see, recommend, or place trades — everything is local to your browser." Browser localStorage is the only V1 store. | Defined in /terminal/trade-journal.html intro; the absence of any server POST for journal entries can be verified in network inspector. |
| Risk-Literacy "context, not signal" framing | Each indicator's educational explainer ends with a sentence on what the indicator does not tell the student. The FII/DII section is explicitly framed as "context, never a trade signal." | Defined in /terminal/risk-literacy.js and the rendered explainer copy. |
| Page-level footer disclaimer | Every module's footer carries the SEBI January 2025 framework citation and a full disclaimer block reiterating educational-only status, no advice, no live data, no signals. | Defined in the shared footer template at the bottom of each module HTML file. |
| Coming-Soon / future-module pages | Where a Phase 2/3 module is teased (currently archived since all five shipped, but historical pages remain), the same banner is carried; the form is honeypot-protected and rate-limited; the only data collected is email + module ID with single-purpose consent. | Defined in /netlify/functions/terminal-interest.js. |
| No buy/sell language across copy | The forbidden-pattern list in section 5.3 (buy/sell/short/long imperative, target prices, "guaranteed/assured returns", "X% accuracy", "hot pick/tip") is applied to every line of Edge Terminal copy. The weekly compliance-scan cron now indexes /terminal/* in addition to articles and curriculum pages. | Pre-publication editorial review at build time; weekly automated scan thereafter; quarterly compliance report. |
| No real-time data anywhere | The data pipeline (section 8.5) sources only end-of-day NSE Bhavcopy data, cached daily, served back to the modules with the ≥30-day lag baked into the rendering layer. The IST clock on the Visualiser is a UI affordance, not a data feed. | Architecture-level: /netlify/functions/cache-bhavcopy.js + /netlify/functions/get-ohlc.js have no real-time data source; the seed-data fallback is deterministic-synthetic and clearly flagged. |
| No "AI signals", confidence percentages, or predictions | The four V0 surfaces detailed in section 8.3 are not implemented. The codebase contains no signal-card component, no AI-MATCH mapper, no predictions module, no confidence-score field on any output. | Source-code review at every build; absence is verifiable via grep across /terminal/ and is checked by the quarterly compliance report. |
8.5 Data sources and processors
Edge Terminal adds no new third-party processors beyond Bharath Shiksha's existing list (Netlify for hosting and serverless compute, Postmark for email, Razorpay for payments). The data pipeline is deliberately lean and lives entirely inside Netlify infrastructure already disclosed in the Privacy Policy v2.
| Data source / processor | What it stores or processes | Disclosure status |
|---|---|---|
| NSE Bhavcopy (public-domain end-of-day market data) | Per-ticker OHLC bars fetched daily from nsearchives.nseindia.com (current endpoint) or the legacy archives.nseindia.com ZIP archive. Public-domain market reference data; no personal data involved. | Data-source disclosure; not a processor in the DPDPA sense. |
Netlify Blobs (terminal-ohlc store) | Per-ticker payloads (daily, weekly, monthly aggregates) written by the cache-bhavcopy scheduled function; read by get-ohlc on demand. Market reference data only; no personal data. | Already disclosed in Privacy Policy v2 under Netlify Blobs. |
Netlify Blobs (terminal-interest store) | Email + module-ID pairs from the Coming-Soon email-capture forms when those existed; current state archived after V1 modules shipped. Day-bucketed, deduped, rate-limited. | Already disclosed in Privacy Policy v2 under Netlify Blobs; standard 3-year retention. |
Browser localStorage (Watchlist) | Up to 25 instruments per browser, with notes. Browser-local only. Cleared by browser data deletion. Not transmitted to Bharath Shiksha servers. | Disclosed in this whitepaper (§8.5) and to be disclosed in Privacy Policy v3 (planned trigger: see Privacy-v3 notes at PRIVACY_TERMS_V3_NOTES.md). |
Browser localStorage (Trade Journal) | Student-entered trade records with thesis, methodology, R-multiple, timestamps. Browser-local only. Cleared by browser data deletion. Not transmitted to Bharath Shiksha servers. | Disclosed in this whitepaper (§8.5) and to be disclosed in Privacy Policy v3. |
| Seed data (deterministic-synthetic OHLC) | Generated client-side from a pseudo-random walk seeded by ticker symbol when the Bhavcopy cache has not yet populated for a given instrument. Not real market data; clearly labelled as such ("As of YYYY-MM-DD · seed"). | Implementation in /terminal/data/seed-ohlc.js; not a processor. |
TradingView Lightweight Charts (MIT-licensed library, served from unpkg.com CDN) | Open-source charting library. Renders client-side; no data is sent to TradingView. Listed in the Content-Security-Policy script-src in _headers. | Library disclosure only; not a processor. |
Specifically: Edge Terminal does not introduce any AI/ML inference provider, any third-party analytics service beyond what is already disclosed in Privacy v2, any market-data vendor beyond NSE Bhavcopy, any payment processor beyond Razorpay, or any cloud-storage provider beyond Netlify. The terminal is, deliberately, a thin layer over the existing Bharath Shiksha infrastructure.
8.6 Retention windows for Edge Terminal data
| Data class | Storage location | Retention |
|---|---|---|
| NSE Bhavcopy per-ticker OHLC cache | Netlify Blobs (terminal-ohlc) | Rolling 1,500 daily bars (~6 years of trading days) per ticker. Older bars are dropped automatically by the cache-bhavcopy idempotent merge logic. |
| Watchlist entries | Browser localStorage only | Indefinite, user-controlled. Cleared when the user clears browser site data, switches browsers, or uses Watchlist's "Clear all" action. Not stored on Bharath Shiksha servers; not subject to a Bharath Shiksha retention window. |
| Trade-Journal entries | Browser localStorage only | Indefinite, user-controlled. Same caveats as Watchlist. Students who want server-backed persistence will need to wait for the Phase 2.5 cross-device sync (see §8.7). |
| Terminal-interest email captures (legacy Coming-Soon) | Netlify Blobs (terminal-interest) | 3 years, in line with the standard marketing-list retention defined in Privacy Policy v2. |
| Compliance audit logs for terminal pages | Netlify Blobs and the quarterly compliance-report archive | 2 years, in line with the standard audit-log retention defined in Privacy Policy v2. |
The lean storage profile is intentional: by holding Watchlist and Trade-Journal data only in the browser at V1, the academy minimises the personal-data surface under the Digital Personal Data Protection Act 2023 and reduces the load on Data Principal rights workflows. The trade-off is the loss of cross-device sync, which is a deliberate Phase 2.5 deferral.
8.7 What is on the Edge Terminal Phase 2.5 / Phase 3 roadmap
The following items are planned but not built at V1 release. Each is documented here because adding any of them will require a corresponding amendment to this whitepaper and to Privacy Policy v3.
- Cross-device sync for Watchlist and Trade Journal (Phase 2.5). Optional Netlify Identity-gated mirror of localStorage to Netlify Blobs, so a logged-in student can use the same Watchlist and Trade Journal across their laptop and phone. Will require disclosure of Netlify Identity as a processor (already disclosed for the course-access flow), expanded retention disclosure (Trade-Journal entries would move from indefinite-browser-local to a defined server-side window), and DPDPA Data Principal rights (export and deletion) wired to the new store. Estimated build: ~4 hours; whitepaper amendment alongside Privacy v3.
- Site-wide nav integration. Adding the Terminal link to the primary navigation across all 140+ Bharath Shiksha site pages. Compliance-neutral change; no whitepaper amendment required.
- Real EOD FII/DII flow data. The Risk-Literacy FII/DII chart is deterministic-synthetic at V1 (clearly flagged). Phase 2.5 may add an EOD scrape of the publicly-published NSE FII/DII CSV. Compliance-neutral change; the framing remains "context, not signal" and the data remains end-of-day.
- Expansion of the codified Backtester library from 12 toward ~50-100 methodologies. Each addition follows the same compliance pattern as the existing 12: mechanical entry-stop-target rule, ILLUSTRATIVE · HISTORICAL labelling, past-performance disclaimer, stage-gating. Compliance-neutral change; no whitepaper amendment required per addition, but the section 8.2 figure will be updated when the library crosses the 50- and 100-rule milestones.
8.8 What will NEVER be added to Edge Terminal
The following are explicitly outside the scope of Edge Terminal and the Bharath Shiksha operating model. Adding any one of them would require, at minimum, a fresh five-role advisory review of the kind documented at EDGE_TERMINAL_ROUNDTABLE.md, a fresh legal opinion, and a full rewrite of this section. The position is restated here because clarity on what we will not build is the strongest evidence of the educational positioning.
- Real-time NSE / BSE / MCX data feeds. Not planned. The data-licensing exposure and the regulatory ambiguity are well-documented. The educational utility is zero.
- AI-generated buy/sell signals, signal cards, target prices, or confidence scores. Not planned. This is the V0 surface that was rejected; the rejection is permanent.
- AI-driven news-to-instrument mappers or recommendation engines. Not planned. Same rationale as signal cards.
- Prediction markets or probability boards on financial outcomes. Not planned. Both the SEBI angle (unregistered forecasting publication) and the Public Gambling Act 1867 + state-gaming-law angle make this an existential risk; we are not in this category and will not be.
- Live screeners with directional output. A "today's setups" screen that lists instruments meeting any methodology's entry condition would convert Edge Terminal from an educational practice room into a signal channel. Not planned.
- Cohort-wide live signal feeds. A push-notification stream that alerts cohort students to "today's setup" is structurally identical to a tip channel. Not planned. Cohort threads remain asynchronous educational discussion governed by section 6.
- Portfolio-level personalised allocation advice. Not planned. The Trade Journal computes statistics on what the student has done; it does not advise on what the student should do next.
- ML-driven personalisation of the curriculum. Not planned at V1. If introduced, it would surface curriculum content based on the student's logged learning patterns — not market positions — and would carry its own DPDPA disclosures.
9. Refunds, grievance, privacy
Three companion documents:
- Refund Policy — full terms in the linked document.
- Grievance Redressal Policy — internal escalation path plus reference to external avenues (National Consumer Helpline, Karnataka State Consumer Commission, SEBI SCORES where applicable).
- Privacy Policy v2 — DPDP Act 2023 compliant: data classification, retention windows, third-party processors, Data Principal rights.
10. Quarterly public reporting
At the end of each quarter, Bharath Shiksha publishes a public compliance report covering the previous three months. The report is generated by the `compliance-report-quarterly` scheduled function and published at /compliance/q<n>-<year>-report.pdf within ten working days of quarter-end.
Each quarterly report contains:
- Aggregate enrolment numbers (no individual identifying data).
- Refund volume and reason histogram.
- Grievance ticket count, SLA performance, and resolution-pattern summary.
- Compliance-scan results: any violations detected and the corrective action taken.
- Editorial review summary: pieces published, pieces revised pre-publication, allowlist additions.
- Material policy changes during the quarter.
The most recent report is available at /compliance/q1-2026-report.pdf.
11. AI tutor disclosure
The Bharath Shiksha tutor channel uses an AI assistant (system-prompted via the rules in /compliance/04-ai-tutor-system-prompt.md) to provide first-line responses on methodology questions. The AI tutor:
- Answers methodology questions. Concept explanations, definitional clarifications, walkthroughs of curriculum-content questions.
- Defers specific-security questions. Returns the same redirect language as a human tutor; will not answer "should I buy X."
- Defers personal portfolio questions. Will not advise on capital allocation.
- Cites curriculum content. Responses reference the lesson, volume, or article that supports the answer.
- Escalates to human staff. Questions that fall outside the AI's confidence band are routed to curriculum staff with documented SLAs.
Every AI tutor response carries a footer disclosing that the response is from an automated assistant and that the underlying curriculum is the canonical source.
12. Affiliate & testimonial compliance
12.1 Affiliate program
The affiliate program pays commissions for qualifying enrolments referred via tracked links. Affiliates are bound by:
- Mandatory disclosure of the affiliate relationship in every promotional post.
- Prohibition on return claims, accuracy claims, or any language contradicting our SEBI educational-publisher posture.
- Genuine recommendation requirement — cold-traffic arbitrage and undisclosed paid traffic are program violations.
- No self-referrals on stages the affiliate is not already enrolled in.
- Refunded enrolments reverse the commission for that referral.
Affiliates who violate the compliance posture are removed from the program without payout of accrued commissions.
12.2 Testimonials
Testimonial submissions are reviewed and published only when they:
- Describe the learning experience, not investment outcomes.
- Contain no return percentages, profit claims, or "guaranteed/sure-shot" language.
- Carry explicit consent for publication and identity attribution.
- Are honestly representative of the actual student experience.
Performance-claim testimonials are auto-rejected per the editorial guardrails. The rejection is structural, not editorial.
13. Self-audit matrix
The matrix below maps each operational practice to the regulation or framework it aligns with. It is the single page that answers the question "how do you stay on the right side of the line?"
| Practice | Aligned with | Verification |
|---|---|---|
| 30-day data lag on every example | SEBI Jan 2025 educational-vs-advisory framework | Pre-publication editorial review + compliance-scan cron |
| Anonymised security references where possible | SEBI RA Regulations 2014 (research-analyst output requires registration) | Pre-publication editorial review |
| Forbidden language pattern enforcement | SEBI IA Regulations 2013 (advisory output requires registration) | Weekly compliance-scan cron + manual review |
| No payment surface for advisory, signals, or fund management | SEBI IA Regulations 2013 | Razorpay payment-page audit; no SKU exists for these |
| Tutor channel deflects specific-security questions | SEBI Jan 2025 framework | AI tutor system prompt + human staff guidelines |
| Cohort thread moderation removes tip-sharing | SEBI Jan 2025 framework + community guidelines | Daily moderator review |
| Privacy & DPDP Act 2023 controls | Digital Personal Data Protection Act 2023 | Privacy v2 + retention table + Data Principal rights |
| Refund policy with documented process | Consumer Protection Act 2019 + good-faith industry practice | Refund policy + Razorpay refund integration + grievance audit |
| Documented grievance redressal path | Consumer Protection Act 2019 + voluntary best practice | Grievance policy + ticket logging + quarterly resolution audit |
| Quarterly public compliance report | Voluntary transparency commitment | compliance-report-quarterly cron output, publicly archived |
| AI tutor disclosure on every response | Voluntary transparency commitment + emerging AI disclosure norms | System prompt enforcement + sample audit |
| Stage 5 algorithmic curriculum aligned with SEBI 2025 algo circular | SEBI 2025 algorithmic-trading circular for retail | Curriculum content review + student deployment is to student's own broker |
| Stage 6 AIF curriculum framed as education only | SEBI AIF Regulations 2012 | Curriculum positioning + students engage qualified counsel separately |
| Affiliate disclosure mandate | Indian advertising-disclosure norms (ASCI guidelines on influencer disclosure) | Affiliate agreement + program audit |
| Testimonial filter on performance claims | SEBI IA/RA performance-claim restrictions | Testimonial review + auto-reject rules |
| Edge Terminal — sticky educational-tool banner on every module | SEBI Jan 2025 framework + RA Regulations 2014 | HTML-level audit across /terminal/*.html; banner is part of shared skeleton; quarterly compliance report |
| Edge Terminal — ≥30-day data lag on every chart, every backtest, every indicator | SEBI Jan 2025 framework | cache-bhavcopy + get-ohlc + seed-data fallback are all lag-enforced; seed-ohlc.js anchors to ≥32 days before today |
| Edge Terminal — ILLUSTRATIVE · HISTORICAL labelling on every Backtester result panel | SEBI Jan 2025 framework; SEBI past-performance disclosure norms | Defined in /terminal/backtester.html; verified at every release |
| Edge Terminal — no signal cards, AI-MATCH mapper, prediction board, or real-time data anywhere | SEBI RA Regulations 2014 + IA Regulations 2013 + Public Gambling Act 1867 (for prediction markets) | Source-code grep across /terminal/; documented advisory roundtable record at EDGE_TERMINAL_ROUNDTABLE.md |
| Edge Terminal — stage-gated codified Backtester library | Curriculum alignment + educational scope | /terminal/backtester-methodologies.js carries stage tags; gating applies the moment authentication is wired in Phase 2.5 |
Edge Terminal — Watchlist and Trade-Journal stored client-side only at V1 (browser localStorage) |
DPDPA 2023 data-minimisation principle | Network-inspector verifiable; no POST endpoints for Watchlist or Trade Journal entries at V1 |
| Edge Terminal — processor list unchanged (Netlify + Razorpay + Postmark; no new third-party processors added) | DPDPA 2023 + Privacy Policy v2 disclosed-processor list | Section 8.5 of this whitepaper; cross-checked against Privacy Policy v2 |
14. How we differ from tip-channel operators
The Indian retail trading-content economy contains a category of operators who sell access to "tips," "signals," "calls," or "alerts" — often via Telegram, WhatsApp, or paid YouTube/Instagram channels — and who structurally cross the SEBI advisory line without registration. We are not in this category, and the differences are operational, not branding.
| Property | Tip-channel operators | Bharath Shiksha |
|---|---|---|
| Subscription product | Daily/weekly tip feed | Curriculum stages with lifetime access |
| Communication cadence | Real-time market hours | Asynchronous; published lessons; weekly cohort threads |
| Specific-security calls | Routine | Structurally absent |
| Performance marketing | "95% accuracy" claims | No accuracy or return claims anywhere |
| Time-decay dependence | Tips expire — must rejoin to keep getting them | Curriculum is durable; lifetime access by design |
| Refund policy | Often opaque; service is "consumed" instantly | Documented refund policy |
| Grievance path | Often informal or absent | Documented internal escalation path |
| Public compliance reporting | None | Quarterly public report |
| SEBI posture | Often crosses advisory line; sometimes faces enforcement | Educational publisher inside SEBI Jan 2025 framework |
| Analytics interface | Real-time signal feed; "BUY now"; confidence percentages | Edge Terminal: historical-only practice room; ≥30-day data lag; no signals; no confidence scores; the four V0 surfaces explicitly rejected (§8.3) |
| Data licensing | Often unlicensed real-time NSE/MCX redistribution | Public-domain NSE Bhavcopy EOD only; no real-time data licensing in the operating model |
Differences are operational, not branding. Students who want tips are not the audience.
15. Why we will never become SEBI-registered
A reasonable question: if we operate in the trading-education space, why not register as an Investment Adviser or Research Analyst and provide advice within the regulated framework? The answer has three components:
- Mandate alignment. Our mandate is education, not advice. Education compounds across thousands of students at low marginal cost; advice scales linearly with adviser-hours and is structurally constrained. A registered IA serving thousands of clients with personalised advice violates IA capacity norms; a registered IA serving education-quality content is operating outside the IA mandate. The two structures don't combine well.
- Conflict-of-interest separation. An educational publisher has no positional incentive to direct a student toward any specific security or to time markets for them. The moment we register as an IA/RA, we acquire the operational temptation (and obligation) to deliver advice. Keeping the advisory and education functions separate organisations is the institutional norm globally.
- Student protection. The single best protection a student of trading methodology has is a curriculum that explicitly does not tell them what to do. The student learns to think; the student bears the responsibility for their own positions. SEBI registration adds a regulated counterparty whose role is to advise; we believe the educational-publisher posture is structurally safer for serious students who intend to become competent traders rather than passive consumers of a signal feed.
The position is not an interim choice. We have no roadmap to register as an IA or RA. The educational-publisher posture is the durable product, and it is structurally simpler to maintain than to transition out of.
16. Forward-looking — what triggers a future update
This whitepaper is reviewed quarterly and after material framework changes. The following triggers will prompt a specific amendment to the document, even outside the quarterly cycle, because they touch the compliance perimeter directly. The list is exhaustive: anything not on it is a routine documentation update handled at the next quarterly review.
| Trigger event | What changes in this whitepaper |
|---|---|
| Edge Terminal cross-device sync (Phase 2.5) ships | Section 8.5 (processors) and section 8.6 (retention) are updated to disclose Netlify Identity as the auth processor for Watchlist/Trade-Journal sync; Privacy Policy v3 is published in parallel with the corresponding DPDPA Data Principal rights wiring. |
| Real EOD FII/DII data scrape lands in Risk-Literacy | Section 8.1 Module 05 paragraph is updated to remove the "deterministic-synthetic" caveat; the data source (NSE EOD FII/DII CSV) is added to section 8.5. |
| Codified Backtester library crosses 50 rules, or 100 rules | Section 8.2 figure is updated; the stage-gating logic is restated; the operating principle (mechanical rules only, no aggregate accuracy claims) is reaffirmed. |
| Supabase Auth migration completes (replacing Netlify Identity) | All references to Netlify Identity are updated; Supabase is disclosed as a new processor with data-residency clause (Mumbai region); Privacy Policy v3 published in parallel. |
| LLP → Pvt Ltd conversion | The entity name is updated throughout. Compliance posture is unchanged. |
| GST registration completes | GSTIN is added to the entity-disclosure block; invoicing clauses in Terms v2 are updated in parallel. |
| SEBI issues a material amendment to the January 2025 finfluencer framework, the RA Regulations 2014, or the IA Regulations 2013 | Section 2 (the SEBI framework) is updated; affected sections are revised; an off-cycle compliance review is conducted. |
| SEBI issues a material amendment to the 2025 algorithmic-trading circular | Section 7.2 (Stage 5 algorithmic curriculum) is updated and the curriculum is reviewed for alignment. |
| DPDP Rules are gazetted (as of May 2026, the rules notifying the Act remain pending) | Retention windows in section 8.6 and in Privacy Policy v2 are aligned to the gazetted Rules; breach-notification timelines are updated. |
| First international student enrolment | Privacy Policy is amended for DPDPA section 16 cross-border-transfer disclosure; this whitepaper notes the geographic expansion in section 3. |
| First grievance ruling, first refund-policy invocation at scale | Section 9 reflects observed median resolution times; the quarterly compliance report carries the volume. |
| Hypothetical: real-time data, signal cards, prediction markets, or screeners with directional output are proposed | The proposal triggers a fresh five-role advisory review on the model of EDGE_TERMINAL_ROUNDTABLE.md, a fresh SEBI-aware legal opinion, and a full rewrite of section 8.3 and 8.8 of this whitepaper. The standing position is that none of these will be added; the trigger is documented for completeness. |
17. Version history
| v2.2 | 2026-05-21 | Edge Terminal access changed to enrolment-only (Netlify Identity stage roles); retired the "free public beta / free preview" framing across section 3.7, section 8, and the footer — the Terminal is now an included member benefit, with a public overview at /edge-terminal. Corrected the Edge Terminal module count to six. |
| v2.1 | 2026-05-15 | Added section 8 (Edge Terminal — educational-analytics layer) covering the five-module launch, the methodology library and stage-gated Backtester architecture, the four V0-prototype features explicitly rejected (signal cards, AI-MATCH news mapper, prediction markets, real-time data feeds), surface-by-surface compliance posture, data sources and processors, retention windows, and the Phase 2.5/3 roadmap. Added section 16 (forward-looking trigger table). Renumbered subsequent sections (9–17). Updated executive summary callout, section 3.7 (Edge Terminal sub-entry under "What we do"), and metadata. Edge Terminal is live in public beta at /terminal/. |
| v2.0 | 2026-05-08 | Whitepaper rewritten as canonical HTML with full self-audit matrix; tip-channel comparison; SEBI Jan 2025 framework expanded; added AI tutor disclosure section; added quarterly-report linkage; cross-references to refund policy v2, grievance redressal, privacy v2, terms v2. |
| v1.0 | 2026-04-27 | Initial PDF whitepaper. Archived; PDF retained at /compliance/whitepaper.pdf for historical reference. |
Attestation. This whitepaper is the canonical statement of Bharath Shiksha's compliance posture as of 2026-05-21 (v2.2, updated for Edge Terminal enrolment-gating). It is reviewed every quarter alongside the public compliance report and updated when material framework changes occur or when the trigger events in section 16 fire. Disputes governed by Indian law; jurisdiction Bengaluru.
Comments, corrections, or compliance-related queries: compliance@bharathshiksha.com. Public discussion of compliance posture is welcome and encouraged.