Platform Evolution
Change velocity, config history, source health, and governance
Changes by Type
Config Version History
| Version | Created | By | Reason | Status |
|---|---|---|---|---|
| v29 | 22 Aug 2026 | manual | Preferred band moved to where the returns actually were. | Active |
| v28 | 22 Aug 2026 | manual | Preferred band moved to where the returns actually were. | Superseded |
| v27 | 22 Aug 2026 | manual | RSI < 30 was among the worst buckets (-0.48pp/month, better in only 1 of 5 years). The floor now excludes genuine free-fall rather than inviting it. | Superseded |
| v26 | 22 Aug 2026 | manual | Signal study over 161,707 observations: RSI > 75 was the best-performing bucket, better than the market in 5 of 5 years (+0.71pp/month). An 82 ceiling was still discarding the strongest names. | Superseded |
| v25 | 22 Aug 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v24 | 22 Aug 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v23 | 22 Aug 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v22 | 21 Aug 2026 | manual | With volatility-normalised stops the planned R is a real number rather than target divided by a gate-satisfying stop, so the bar can be raised. | Superseded |
| v21 | 21 Aug 2026 | manual | A hard 10% floor combined with min_risk_reward_ratio 1.5 made 6.0% the arithmetically convenient stop on 16 of 18 picks. Targets now come from the ATR model; this floor only screens out trades too small to be worth costs. | Superseded |
| v20 | 21 Aug 2026 | manual | Parking a target-hitter in WATCH for a discretionary exit cost ADANIPOWER 9.4pp, M&M 10.4pp and POLYCAB 8.9pp of forgone upside. It now keeps running behind the trail. | Superseded |
| v19 | 21 Aug 2026 | manual | The one rule shaped for letting winners run was disabled while the exit engine truncated them. | Superseded |
| v18 | 21 Aug 2026 | manual | Third-party aggregate_rating is a lagging average of sell-side opinion and had hard-reject power over entries. Demoted to context. | Superseded |
| v17 | 21 Aug 2026 | manual | Stochastic > 80 rejected names in strong uptrends, including candidates flagged as "extended after an 82% 1-year run". | Superseded |
| v16 | 21 Aug 2026 | manual | Widened alongside max_rsi_for_new_pick so continuation setups are reachable. | Superseded |
| v15 | 21 Aug 2026 | manual | Widened alongside max_rsi_for_new_pick so continuation setups are reachable. | Superseded |
| v14 | 21 Aug 2026 | manual | Rejecting RSI > 70 excluded cross-sectional momentum by construction. Every recorded entry landed between RSI 45 and 68 (median 57.3), which is neither a momentum setup nor a mean-reversion one. | Superseded |
| v13 | 21 Aug 2026 | manual | The entry gate must not double as the exit rule. Re-running it at review produced 8 exits that cut 39.4pp of profit against 11 that saved 37.3pp of loss — a rule carrying no information while destroying the right tail. | Superseded |
| v12 | 21 Aug 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v11 | 13 Jun 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v10 | 22 Apr 2026 | manual | Raise strong-hit threshold to match the +10% minimum target-return policy. | Superseded |
| v9 | 22 Apr 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v8 | 22 Apr 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v7 | 20 Apr 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v6 | 14 Apr 2026 | manual | Normalize company primary preferred sources for IR searches. | Superseded |
| v5 | 14 Apr 2026 | manual | Normalize company primary pipeline query template for IR searches. | Superseded |
| v4 | 14 Apr 2026 | manual | Raise daily new-pick cap to 3 for broader candidate coverage. | Superseded |
| v3 | 14 Apr 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v2 | 14 Apr 2026 | manual | Auto-reconciled active config with current default schema. | Superseded |
| v1 | 5 Apr 2026 | manual | Seed initial runtime configuration defaults from CONFIG_SPEC.md. | Superseded |
Change Queue (0)
No queued changes. Changes are proposed by Loop 2/3/4 reflections.
Which sources were right (47 making calls)
Only sources that express a view can be judged on outcomes — a broker's target or a Reddit thesis is falsifiable, an exchange filing is a fact. Scores come from the net-of-cost alpha of the picks each source was cited on, and stay blank until 5 of those picks have been graded.
| Source | Score | Alpha / pick | Net P&L / pick | Win | Graded | Cited |
|---|---|---|---|---|---|---|
| Investing.com | unproven | -0.75pp | +2.17% | 50% | 2 | 38 |
| unproven | -7.86pp | -7.76% | 0% | 1 | 37 | |
| Dhan | unproven | -4.50pp | -5.32% | 25% | 4 | 32 |
| Motilal Oswal | unproven | +3.10pp | +2.29% | 50% | 2 | 31 |
| ICICI Direct | unproven | -5.97pp | -3.89% | 0% | 2 | 26 |
| Upstox | unproven | — | — | — | 0 | 20 |
| Trendlyne | unproven | +1.44pp | -0.64% | 50% | 2 | 20 |
| Jefferies | unproven | — | — | — | 0 | 16 |
| CLSA | unproven | — | — | — | 0 | 10 |
| Yahoo Finance derived technicals | unproven | — | — | — | 0 | 7 |
| multibagg.ai | unproven | -4.56pp | -4.67% | 0% | 1 | 7 |
| Axis Direct | unproven | — | — | — | 0 | 6 |
| JPMorgan | unproven | — | — | — | 0 | 6 |
| Citi | unproven | — | — | — | 0 | 5 |
| TradingView | unproven | — | — | — | 0 | 5 |
| Yahoo Finance daily OHLC | unproven | — | — | — | 0 | 5 |
| Kotak Securities | unproven | — | — | — | 0 | 4 |
| Technical screen | unproven | — | — | — | 0 | 4 |
| Nuvama | unproven | — | — | — | 0 | 3 |
| Price history via technical sources | unproven | — | — | — | 0 | 3 |
| Tickertape | unproven | — | — | — | 0 | 3 |
| INDmoney technicals | unproven | — | — | — | 0 | 3 |
| Geojit | unproven | — | — | — | 0 | 2 |
| Univest technicals | unproven | — | — | — | 0 | 2 |
| MarketsMojo technical coverage | unproven | — | — | — | 0 | 2 |
Data providers (102)
These publish facts rather than views, so they are listed for coverage rather than ranked on profit. A filing does not become less true because the trade lost money.
| Source | Cited | On graded picks |
|---|---|---|
| Screener | 22 | 4 |
| StockAnalysis | 21 | 0 |
| NSE | 16 | 1 |
| BSE | 6 | 1 |
| Q1 FY27 results coverage | 5 | 0 |
| MarketsMojo | 4 | 0 |
| CDSL company media release | 3 | 0 |
| SMC Global technical call via BusinessToday (Aug 17) | 2 | 0 |
| Bank of India Q4 FY26 investor presentation | 2 | 0 |
| CDSL Q4 FY26 media release | 2 | 0 |
| Quartr earnings summary | 2 | 0 |
| Marico company investor update | 2 | 0 |
| Sun Pharma Q1 FY27 press release | 2 | 0 |
| Cipla company press release | 2 | 0 |
| Choice Institutional Equities | 2 | 0 |
Recent Changes (last 20)
Upsert price snapshot GC=F 2026-09-04
Persist daily price snapshot for GC=F from google_finance.
Update current price for pick REC-20260805-04-SUNPHARMA
Automated scheduled price update for SUNPHARMA.
Update current price for pick REC-20260805-06-LT
Automated scheduled price update for LT.
Update current price for pick REC-20260814-05-SBIN
Automated scheduled price update for SBIN.
Upsert price snapshot SUNPHARMA 2026-09-04
Persist daily price snapshot for SUNPHARMA from google_finance.
Upsert price snapshot LT 2026-09-04
Persist daily price snapshot for LT from google_finance.
Upsert price snapshot SBIN 2026-09-04
Persist daily price snapshot for SBIN from google_finance.
Upsert price snapshot GC=F 2026-09-04
Persist daily price snapshot for GC=F from google_finance.
Update current price for pick REC-20260805-04-SUNPHARMA
Automated scheduled price update for SUNPHARMA.
Update current price for pick REC-20260805-06-LT
Automated scheduled price update for LT.
Update current price for pick REC-20260814-05-SBIN
Automated scheduled price update for SBIN.
Upsert price snapshot SUNPHARMA 2026-09-04
Persist daily price snapshot for SUNPHARMA from google_finance.
Upsert price snapshot LT 2026-09-04
Persist daily price snapshot for LT from google_finance.
Upsert price snapshot SBIN 2026-09-04
Persist daily price snapshot for SBIN from google_finance.
Ingest analysis analysis-2026-09-04-0830
Stored analysis result analysis-2026-09-04-0830 and applied recommendation updates.
Upsert candidate snapshot analysis-2026-09-04-0830:DLF
Persisted candidate snapshot analysis-2026-09-04-0830:DLF.
Upsert candidate snapshot analysis-2026-09-04-0830:PRESTIGE
Persisted candidate snapshot analysis-2026-09-04-0830:PRESTIGE.
Upsert candidate snapshot analysis-2026-09-04-0830:COALINDIA
Persisted candidate snapshot analysis-2026-09-04-0830:COALINDIA.
Upsert candidate snapshot analysis-2026-09-04-0830:HINDCOPPER
Persisted candidate snapshot analysis-2026-09-04-0830:HINDCOPPER.
Upsert candidate snapshot analysis-2026-09-04-0830:ICICIBANK
Persisted candidate snapshot analysis-2026-09-04-0830:ICICIBANK.