Tradelyser Logo

Insights — Read Metrics on a Schedule

Which reports matter weekly, how to use AI as a second opinion, and when to ignore a green P&L week.

5 min read · Updated 2026-06-05 · Reviewed by TradeLyser Content Team (Practicing Indian market traders)

Key takeaways

  • Review rule adherence before expectancy — discipline failures distort metrics.
  • AI summaries prompt questions; verify answers in trade notes.
  • End weekly review with one change, not five.

Insights are the last step — not the first. Reports, widgets, and AI summaries only matter after journals are honest, strategies are tagged, and rules are scored. The Insights pillar is about reading the right numbers on a fixed schedule and changing one variable at a time. Chasing every dashboard tile daily is noise.

Reports you actually read

Weekly minimum: expectancy and profit factor per strategy tag, max drawdown from equity peak, win rate vs average R-multiple, and top three symbols by net P&L. Monthly add: time-of-day heatmap, slippage vs plan, and rule-break count. Skip vanity metrics you will not act on — like total trade count if quality is collapsing.

  • Equity curve: slope and depth of pullbacks — not just ending balance.
  • Drawdown tier: are you in normal variance or breach of plan?
  • Symbol concentration: is one stock carrying the month?
  • Fees and STT impact on scalping tags — net vs gross.

AI as second opinion, not signal

Elysia AI in TradeLyser can summarise sessions, flag outliers, and score discipline — useful when you review on Friday and want a second pair of eyes. It does not know your unpublished macro view or family stress. Treat AI output as prompts for your journal: “Why did size spike Tuesday?” — then verify in trade notes before changing rules.

When to ignore a green week

A profitable week with three rule breaks is a loss waiting to happen. Ignore headline P&L if discipline score averaged below 3 or if one expiry gamble carried results. Conversely, a small red week with full rule compliance may be healthy variance — do not rip up the playbook after five clean losing trades that respected stops.

Example: AI flags size spike

AI notes Tuesday lot size 2× your thirty-day median on Bank Nifty. You open the journal: no A+ setup tag, checklist not filled, mood “revenge.” Insight action is discipline review — not disabling the strategy. Fix the rule break log; rerun analytics next week.

Edge audit rhythm

Quarterly edge audit: for each live strategy, confirm expectancy positive after costs, sample size adequate, and market regime still matches setup (trend vs range). Indian regimes shift around budget, election cycles, and VIX spikes — a tag that worked in low VIX may need smaller size when India VIX holds above 18.

Order of operations for metrics

Read in this order: (1) rule adherence, (2) expectancy per tag, (3) drawdown vs limit, (4) AI or mentor notes, (5) new ideas to test. Skipping to (5) after a red day is how traders accumulate indicators without data.

From insight to one change

End every weekly review with exactly one change: tighten stop on Tag B, pause Tag C until twenty trades, add pre-market checklist item, or reduce max trades from five to three. Multiple simultaneous changes destroy attribution — you will not know what helped in June.

Report calendar (what to open when)

CadenceReports to readDecision
Daily (optional)Realised P&L vs daily rupee capStop trading if cap hit
Friday weeklyExpectancy & PF per tag, rule breaksPick one process change
First Friday monthEquity curve, drawdown tier, scorecardsResize capital per tag
QuarterlyEdge audit, regime notes, India VIX periodsRetire or rewrite tags

Scalper vs swing dashboards

Scalpers on Bank Nifty should weight slippage, fees, time-of-day heatmap, and trade count vs max-trades rule. Swing traders on NSE cash should weight hold time, gap risk on overnight positions, and symbol concentration. Using the same dashboard for both invites wrong conclusions — filter reports by strategy tag so each playbook sees metrics that match its hold profile.

Insights mistakes

  • Checking P&L every fifteen minutes and calling it analysis.
  • Trusting AI narrative without opening underlying trades.
  • Changing strategy after one week of red metrics on a forty-trade winner.
  • Ignoring discipline score because headline week was green.
  • Adding indicators instead of reading existing journal fields.

Insights close the loop back to Journals: every chart anomaly should end in a note or tag fix, not a new subscription. If a report does not produce a written action on Friday, skip it next week until you have bandwidth to act.

Drawdown tiers and when to pause

Define drawdown tiers in rupees before you need them. Tier 1: normal variance inside weekly noise — keep process, no rule changes. Tier 2: monthly drawdown hits 50% of your planned max — halve size and add one checklist item. Tier 3: breach monthly max or three consecutive rule-break weeks — stop live size, paper trade only until twenty clean sessions. Indian event weeks (budget, election results) can push you to Tier 2 without a broken edge — note regime in journal so Insights does not confuse event variance with strategy failure.

NSE-specific report reading

Filter reports by session bucket when you trade opening drive vs afternoon fades — all-day averages lie on Bank Nifty. Separate F&O from cash tags before reading symbol concentration; a green month on index options can hide stock swing bleed. On expiry weeks, compare Thursday P&L to Mon–Wed average — twelve expiry days per year dominate many retail books. STT and brokerage matter on sub-₹500 winners; use net expectancy, not gross, before you scale scalp size.

Dashboard and widget hygiene

Pin only widgets you act on weekly: equity curve, expectancy by tag, rule-break count, drawdown from peak. Remove intraday P&L tiles if they trigger impulsive trades — insights are for scheduled review, not dopamine during session. If you use Elysia summaries, run them after tags are fixed; AI on untagged books reinforces false narratives. One dashboard for review day, minimal chrome on execution days — same data, different temptation profile.

FAQ

How often should I open reports?

Weekly for per-strategy expectancy and drawdown. Daily P&L checks are optional; intraday dashboard hopping increases tilt risk.

Can Elysia AI replace my review?

No. AI highlights patterns and scores discipline — you still decide rule changes. It does not see off-platform context.

What is the first metric to check?

Rule-break count for the week. If breaks are zero, move to expectancy per tag. If breaks are high, fix process before optimising entries.

Glossary

Start journaling with TradeLyser

Connect your broker, tag strategies, and review performance with AI-assisted insights.