September 12, 2026 falls on a Saturday. The market is closed, the after-close dashboards have refreshed with Friday's data, and every rotation tracker in India is showing its updated sector ranks. This is exactly the window where a process-driven algo trader does their most useful work of the week — not placing trades, but deciding which trades next week's systems should even be allowed to take.
This post walks through what the current rotation and breadth picture actually tells you, and — more importantly — how to convert it into concrete risk filters for your scanner, options, and execution workflow before Monday's open. No predictions, no sector tips. Just process.
The rotation picture going into this weekend
Across the public sector-rotation trackers refreshed around September 12, a consistent theme shows up: leadership has been sitting in pockets like pharma, metals, chemicals and textiles, with ranks shuffling week to week — one tracker showed pharma holding the top rotation slot with a high share of bullish setups and double-digit outperformance versus the Nifty over three months, while chemicals and textiles climbed ranks and metals slipped a few places despite still screening "strong."
Two things matter about that read, and neither is a forecast.
First, the leadership is rotating, not static. A sector that ranked second last week and fifth this week is telling you that momentum within the leaders is being redistributed. For an algo trader, that changes how much trust you place in a week-old sector label.
Second, the type of sector leading matters. When defensives and commodity-linked sectors dominate the top ranks while rate-sensitive cyclicals sit lower, the tape is expressing a different risk appetite than when banks and autos lead. You don't need to predict what comes next — you need to know which regime your strategies were built for, and whether this is it.
The mistake is reading a heatmap like a tip sheet: "pharma is green, buy pharma." The better question is: "given where leadership sits, which of my existing systems should run at full size, which should run smaller, and which should sit out?"
Breadth answers the question the index can't
A closing price on the Nifty tells you almost nothing about the health of the move behind it. A 100-point rally carried by three heavyweights and a 100-point rally carried by hundreds of advancing stocks look identical on the index chart — and tend to behave very differently in the sessions that follow. That's the core argument every breadth dashboard makes, and it holds up in practice.
So alongside the sector ranks, the weekend review should log a small set of breadth numbers:
- Advance/decline balance across your trading universe, not just the Nifty 50.
- Percentage of stocks above their 20-day and 50-day moving averages — a quick participation gauge.
- New 52-week highs versus lows — expansion in highs alongside rising sectors is confirmation; shrinking highs during an index rally is a divergence worth respecting.
- India VIX level and direction — not as a signal, but as an input to position sizing and option strategy selection.
- Breakout counts per sector — some trackers show setups and breakout counts per sector; a sector with 21 setups and 11 breakouts is expressing different strength than one with 2 setups and 1 breakout, even if both are coloured the same green.
The combination is what matters. Strong sector leadership plus broad participation is one regime. Strong leadership in one or two sectors while breadth deteriorates is the classic "narrow leadership" warning — the index can keep holding up while the average stock in your scanner universe quietly stops following through.
A weekend sector review you can actually finish
The reason most traders skip this work is that it's unbounded — you can stare at dashboards for hours. Here's a version that takes 30–40 minutes and produces written output.
Step 1: Rank and classify sectors
Put every major NSE sectoral index into one of four states: Leading (strong and outperforming), Improving (rising in rank), Weakening (falling in rank while still positive), Lagging (weak and underperforming). Most rotation dashboards already do this classification; your job is to write it down, because next weekend you'll want to compare against it. Rank changes week over week are the actual information — a static snapshot ages fast.
Step 2: Cross-check with breadth
For each Leading sector, ask: is the strength broad within the sector, or is one heavyweight dragging the index? A sector index can print green while most of its constituents are flat. If your scanner universe includes mid-caps in that sector, sector-level green doesn't automatically mean your stocks are participating.
Step 3: Map your systems to the states
This is the step most traders skip. Take each strategy you run — breakout scanner, pullback system, option selling basket, whatever — and note which sector states it historically works in. A momentum breakout system wants Leading and Improving sectors with expanding breadth. A mean-reversion system may actually prefer Weakening sectors. If you've never done this mapping, that's a backtesting project, not a guess.
Step 4: Write next week's filters
The output of the review is not an opinion ("market looks strong"). It's a short list of written rules: which sectors your long scanner signals may come from, what your size multiplier is, and what would change your mind mid-week. If it isn't written before Monday, it will be improvised during Monday — usually badly.
From the weekend read to Monday's risk filters
Here's what converting context into rules actually looks like. These are examples of filter structure, not settings to copy:
Sector alignment filter. Long entries from your scanner are only eligible if the signal's sector is in a Leading or Improving state. Signals from Lagging sectors are logged but not traded — or routed to a separate paper book so you can measure what the filter is costing or saving you.
Narrow-leadership throttle. If fewer than half your tracked sectoral indices closed the week positive, and the top two sectors account for most of the breakout count, cut position size — for example, run at half your normal per-trade risk or reduce the number of concurrent slots your system may fill. You're not predicting a fall; you're acknowledging that follow-through statistics are historically worse when participation is thin.
Freshness discount. A sector that jumped four ranks in one week is Improving, not proven. Demand extra confirmation on entries there — a tighter invalidation level, or a requirement that the stock itself (not just the sector) shows a fresh structural signal. Rotation trackers reshuffle weekly; your risk shouldn't fully commit to a one-week move.
Expiry-week interaction. If the coming week contains a weekly index expiry, sector filters and expiry behaviour interact. Narrow leadership plus expiry positioning can produce sharp index moves with poor single-stock follow-through — a regime where index option structures and stock breakouts behave very differently. Decide in advance which book gets priority, and size the other down.
Options regime check. For option strategies, the sector read feeds strategy selection, not strike picking. Broad participation with a calm VIX supports one class of structures; narrow leadership with rising VIX supports another. Whatever you run, the risk management layer — defined max loss, margin checked before entry, basket-level stops — matters more than the directional read, because the read will sometimes be wrong and the process has to survive that.
None of these filters predict anything. Each one changes how much you risk and where, conditional on measurable context. That's the entire job of a regime filter.
Where this workflow lives in practice
On Anadi, this maps to a specific sequence rather than five open browser tabs.
The Indices view provides index cards and a sector heatmap as the context layer — the point is to look at it before acting on any scanner or option setup, so a strong-looking stock signal gets checked against a weak sector tape instead of being taken in isolation.
Scanner output then flows into the Action Center, which is where the sector filter earns its keep. Instead of a flat list of signals, you see candidates ranked with signal stage, freshness, and — critically — blocked reasons. An entry blocked for chase distance or an invalidated level is your system protecting you from exactly the late, extended entries that narrow-leadership weeks produce in bulk. For F&O-eligible names, the same view surfaces option context, OI pulse, and PCR, so the decision to route a signal through stock, future, or option happens with the derivatives context visible.
If the route is options, the chain workflow connects strikes to Greeks, basket preview, and a margin estimate before anything executes — risk appears before the order, not after. And if you're still validating a new sector filter, running it through paper trading first gives you a live record of what the filter accepts and rejects without paying tuition in real money.
For the weekly context itself, the weekly market outlook exists for exactly this use case: prepare with context, don't chase noisy calls.
Backtest the filter before you trust it
One honest warning. Sector filters are seductive in backtests because they add a parameter, and every added parameter can be overfit. Before you trust "only trade Leading sectors":
- Test the base strategy with and without the filter over the same period, and look at the difference, not just the filtered result.
- Check how many trades the filter removes. A filter that cuts 70 percent of trades and improves the average is doing something real; one that removes three bad trades from 2024 is curve-fitting.
- Make sure the sector state you test with is the state you'd have known at the time — weekend-computed, applied the following week. Using end-of-move sector ranks to filter entries made during the move is lookahead bias, and it will flatter every result.
- Rebuild the rule in a strategy builder as an explicit, written condition — if you can't state the filter precisely enough to code it, you don't have a filter, you have a mood.
The weekend checklist
Before Monday's open, you should be able to answer, in writing:
- Which sectors are Leading, Improving, Weakening, Lagging — and what changed versus last weekend?
- Is participation broad or narrow — advance/decline, percentage above 20-DMA and 50-DMA, new highs versus lows?
- Which of my systems are aligned with this regime, and which get a size cut or a pause?
- What are next week's scheduled events — expiry days, major data releases — and which sessions do my intraday systems skip or de-risk?
- What written filters go live Monday, and what mid-week evidence would make me revisit them?
If you can answer all five, the heatmap has done its job: it shaped your risk before the market opened, instead of explaining your losses after it closed.
If you want the whole loop — sector context, scanner, action queue, option chain, margin-aware execution, and paper-first testing — in one workflow instead of five tabs, you can request early access to Anadi Algo and run next weekend's review inside it.
