What it is

A driver ek narrative reason hai — kyun ek group of stocks ek saath move kar rahe hain. Trend state, RS, SR — yeh sab "kya ho raha hai" describe karte hain. Driver answers "kyun".

Analogy: news headline pe "PSU banks rally on RBI policy" likha hota hai. Yeh ek driver hai — POLICY_SPECIAL. Stocks individually trend UP ho rahe ho sakte hain (yeh trend state hai), but woh ek shared narrative thread share kar rahe hain (yeh driver hai).

Tietiy mein drivers ek tier-1 pillar hain. Founder-PDF blueprint driver-centric hai (locked 2026-06-15). Phase 4 ka entire Story Engine drivers ke around organise hota hai.

Driver families

Drivers ek family ke under aate hain. Family ek enum hai, fixed vocabulary. Examples:

  • POLICY_SPECIAL — RBI rate decision, budget allocation, sectoral policy shift (e.g. PSU bank rally on policy news).
  • CAPEX — large capacity expansion, big order win (e.g. defense contracts driving HAL / BEL).
  • EARNINGS_BEAT — quarterly results materially above expectations.
  • GUIDANCE_RAISE — management revised forward guidance up.
  • SECTOR_TAILWIND — global commodity, FX, demand wave (oil up helping OMCs, dollar down helping IT).
  • CORPORATE_ACTION — buyback, dividend, demerger, M&A.
  • RERATING — long-form multiple expansion narrative.

Family vocabulary stable hai (additive-only evolution per cockpit principle #6). New families add ho sakte hain, existing ka meaning change nahi hota.

The 5 lifecycle states

Yeh sabse interesting part hai. Har driver instance ek lifecycle follow karta hai — yeh ek time-evolution model hai, ek static label nahi:

  • EMERGING — driver pehli baar register ho raha hai. News count rising, but price impact abhi build ho raha hai. Earliest stage.
  • GAINING — price + news evidence dono accelerating. Stocks in the driver group consistently RS-leading.
  • LEADING — peak phase. Driver is dominant narrative. Maximum RS concentration, multiple stocks in driver group all trend UP, often making new highs.
  • LOSING — peak crossed. RS rank declining within driver group, new highs slowing, but driver not yet dead.
  • BROKEN — narrative over. Group breaks structurally, RS-leaders become laggards, often regime change for that family.

How tietiy computes it

Driver detection is one of the harder cockpit pillars. The current approach combines two evidence streams:

  • Price evidence — clusters of stocks moving together in RS rank space. Group RS + group trend slope + group structure quality.
  • News evidence — news event classifier (shared brain_calls table, prompt_version-discriminated) tags each news item to a driver family with a confidence.

A driver instance has a stable driver_id (across days). Lifecycle state derives from windowed price+news evidence — typically 5–10 day rolling. Importance (0–100) is its own score with a mandatory breakdown.

Driver pillar is still maturing — operator's Layer 5 Validation Engine vision treats drivers as one of 10 modules, with statistical identity built per object as evidence accumulates.

What this site shows you

Drivers page (top nav) hosts the active drivers list — each with family, current lifecycle state, top stocks in the group. Each stock page in tietiy shows which active drivers (if any) it belongs to. Markets brief includes the "today's leading driver" line. Story Engine (when published) groups news into driver-anchored stories.

How to read it

  • LEADING driver + your stock is a price-leader in it = textbook strongest contextual read. The narrative thread is alive AND your name is in its core.
  • EMERGING driver = early stage, more risk + more potential. False positive rate higher — many emerging drivers never reach LEADING.
  • LOSING driver = warning context. Stocks in this driver group often soon lose RS leadership. Consider hedging mental anchors.
  • BROKEN driver = often regime shift incoming for that family. Capital usually rotates to next emerging driver.
  • No driver = no narrative attached. Stock can still be in a good state. Not every up-move needs a named driver.

What this does NOT tell you

  • Which driver wins tomorrow. Lifecycle states are descriptive labels of current evidence, not forecasts of future progression.
  • News causality. Driver classification is "this news mentions this family." Not "this news caused this move." Causality is much harder.
  • Single-stock drivers. Drivers are group narratives. Single-stock specific events (one company's earnings beat) may not surface as a driver — they show as ordinary news.
  • Driver lifecycle timing. EMERGING can stay EMERGING for weeks and skip to BROKEN without ever LEADING. Lifecycle is not deterministic time-evolution.

Why drivers come last in this Learn series. Drivers are the most experimental cockpit pillar. Their lifecycle state vocabulary is locked, their compute is still evolving. Use them as a narrative overlay on the more solid price-structure layers (trend, RS, SR, volume), not as a primary signal. Honest disclosure beats false certainty.