How Markov finds setups.
Markov runs a multi-strategy quantitative stack across the full ~498-name NSE cash-equity universe. Cross-sectional factor models — quality, momentum, residual momentum (stripped of NIFTY + sector beta), sector-neutral momentum, low-vol quality and a vol-targeted basket — score every name on composite z-signals and surface the top of each cross-section as published recommendations. Quarterly reranks anchor the factor sleeve; rerank-persistence governs entries and exits. The roster expands as new models clear internal validation. The full methodology is below.
~498 NSE equities. Scanned every minute.
The investable universe is the full NSE cash-equity set we maintain a five-year price history for. F&O-eligible names are pulled in when a short setup needs them. The scan runs every minute during market hours, so a setup that develops at 11:14 is in the feed by 11:15.
Twelve documented methodologies. Six live, six in the pipeline.
Each strategy scans the same universe but is built on a different methodology. Together they diversify what the feed picks up — quality, momentum, low-vol, multi-factor — so a quiet regime for one approach isn't a quiet regime for all. A strategy moves from the pipeline to the live feed only after the internal-review checklist is cleared.
In production today
QMJ — Quality / Momentum / Junk
Quarterly multi-factor composite combining quality, momentum and a junk-rank score on the long side.
Residual momentum (wide)
Momentum after stripping out NIFTY and sector beta — the stock-specific component.
Sector-neutral momentum
Best-of-sector momentum picks, equal-weighted across sectors to avoid concentration.
Low-vol quality (persistent)
High-quality names trading with below-average realised volatility.
Persistent quality
Companies with consistent ROE, low leverage and stable margins across rolling quarters.
Vol-targeted basket
Portfolio with dynamically scaled exposure to target a steady realised volatility band.
In the research pipeline
Six further methodologies are documented and under internal review. They run in research only — none of them is on the live feed yet. Each will move to production only after methodology review, exit-policy declaration, cross-strategy correlation check and operational-readiness sign-off.
Trend persistence
Names whose trend direction persists across consecutive rolling quarters — a stricter momentum filter than first-derivative price strength.
Residual momentum (narrow)
Stock-specific momentum after stripping NIFTY and a tighter sector beta — orthogonal to the wider residual variant on the live feed.
Return consistency
Names with stable cross-quarter return distributions and low drawdown variance — a low-noise alternative to standard quality.
Dual momentum
Stacked absolute (vs cash) and cross-sectional (vs peers) momentum — only names that are strong on both axes.
Multi-factor composite (wide)
Wider blend across quality, momentum, value and low-vol signals than QMJ alone.
Trend persistence (wide)
Trend-persistence rule applied across the broader investible universe rather than the F&O subset.
Two different alignment gates. Same purpose.
Different archetypes use different alignment rules — both designed to keep single-bar noise out of the feed. Trend-following and intraday setups have to clear a three-timeframe gate. Quality / factor portfolio strategies have to clear a quarterly rerank-persistence gate.
Three timeframes have to agree
Trend-following strategies never trade against the primary trend. If the three timeframes disagree, the signal is dropped before it reaches the feed.
Direction
Higher-highs / lower-lows structure, 9 / 21 / 50 EMA stack. Decides whether long or short is allowed at all.
Timing
Pullback to moving average, oscillator state, volume behaviour. Decides when in the day the setup is ripe.
Entry
Reversal candle, breakout retest, opening-range structure. Decides the exact entry price and stop level.
Quarterly rerank persistence
Factor portfolio strategies recompute the basket on a quarterly cadence against documented factor scores. A name enters only when it earns its slot on rerank promotion, and exits when it drops out at the next rerank — not on a single intraday signal.
Rank
Score every eligible name on the strategy's documented factors — quality, momentum, residual return, volatility — every quarter.
Promote
Names that clear the rank threshold enter the basket. Position sizing is rule-based — equal-weight, vol-targeted or sector-neutralised depending on the strategy.
Demote
Names that fall out of rank at the next rerank are exited. Mid-quarter exits only on structural-trend break, ATR-based stop or a regime-break alert.
Four signal archetypes. One scoring rubric.
Every strategy in the stack declares one of four archetypes up front. The archetype determines the exit-policy class — target-cluster rules, invalidation signals, structural-stop logic — before the strategy ships. Each setup is also scored 0–12 across trend, alignment, volume / participation, pattern quality, R:R and oscillator / factor state; setups under 7 don't make the feed.
QualityRerank
Drives most of the live feed today
Quarterly portfolio recompute on documented factor scores. Names enter on rerank promotion and exit on rerank dropout. Mid-quarter exits only on structural-trend break, ATR-based stop or regime alert.
Strategies QMJ · Residual momentum · Sector-neutral momentum · Low-vol quality · Persistent quality · Vol-targeted basket
TrendFollowing
Continuation in an established trend
Pullback to moving average, breakout-continuation, RS pullback, intraday momentum, volatility breakout. Three timeframes must agree before entry. Exits on EMA21 break, swing-low break, MA50/200 cross or rebalance-rerank dropout.
Strategies Naive momentum · Multi-horizon momentum · Breakout continuation · RS pullback · Intraday momentum
MeanReversion
Counter-trend inside a documented range
Bollinger extreme reversal in a defined channel, oversold-in-uptrend, statistical pairs trading. Targets are structural midlines — BB midline, VWAP, 21 EMA. Exits on trend continuation past entry or vol expansion.
Strategies BB extreme mean-reversion · Pairs trading
Event
Triggered by a discrete event
Post-earnings drift, index-inclusion add/drop, F&O expiry squeeze, FII/DII flow signal. Exits are tied to the event window — measured-move targets, Fibonacci extensions, event-day-high resistance — and close back through the pre-event level invalidates the setup.
Strategies Post-earnings drift · Index inclusion · F&O expiry squeeze · FII/DII flow
Twelve checks. Every strategy. Every promotion.
Every strategy — production or pipeline — is reviewed against the same internal qualitative checklist covering methodology, risk envelope, regime gating and operational readiness. Promotions into the live feed require all twelve to be cleared. Any past-performance figures used during internal review are not publicly displayed and are not certified by SEBI's Past Risk-and-Return Verification Agency (PaRRVA).
Every recommendation is a full research payload.
A Markov recommendation is a published research opinion — under the SEBI Research Analyst framework — with the full payload visible up front. What you do with it is entirely your call.
We surface the entire payload; how much you allocate, when you enter, and whether you act on a particular recommendation at all is up to you.
What Markov is not.
Not a guarantee. No SEBI-registered intermediary can or will guarantee returns. Past performance is not indicative of future returns. Across a portfolio of rule-based strategies a meaningful share of individual recommendations will lose.
Not execution. Markov produces research recommendations. You execute every trade yourself through your own broker. We never touch your funds.
Not a tip channel. We don't run Telegram groups, WhatsApp tip channels or YouTube stream tips. If someone claims to represent Markov on those platforms, they don't.
Not personalised advice. Signals are general research recommendations under SEBI's Research Analyst framework. They are not tailored to your personal financial situation. For personal advice, consult a SEBI Registered Investment Adviser.