{
 "items": [
  {
   "n": 1,
   "headline": "The index and the median stock diverge violently",
   "figure": "FY19: index +8.4% vs median stock -4.6%",
   "body": "In narrow years the index beats the median stock badly; in broad years the median crushes the index (FY21 +95.1% vs +76.0%). A 'Nifty is up' headline says little about whether the average stock made money.",
   "strength": "strong",
   "strength_label": "High confidence",
   "caveat": "10 of 10 years directionally consistent - the most robust finding here."
  },
  {
   "n": 2,
   "headline": "Winners are small, volatile and sector-clustered",
   "figure": "63% small-cap vs a 33% universe; 47.7% vol vs 36.9%",
   "body": "The top 30 each year skewed heavily to small caps and was concentrated in ~10-11 of 20 sectors - Capital Goods was the biggest cluster in 5 of 10 years.",
   "strength": "mixed",
   "strength_label": "Real but largely an artefact",
   "caveat": "Mostly an order-statistic effect, NOT a tradable edge: small caps carry higher dispersion, so they dominate BOTH tails of any extreme-return ranking. The bottom-20 lists carry the same skew."
  },
  {
   "n": 3,
   "headline": "Early strength predicts the rest of the year - weakly",
   "figure": "~34% recall vs a 25% baseline (r = +0.05, R2 < 1%)",
   "body": "The headline version of this statistic is 55%, but that is inflated by the signal window being inside the target window. Properly tested, the edge is about 1.3x, not 2.2x.",
   "strength": "weak",
   "strength_label": "Small but consistent",
   "caveat": "Clean forward correlation is +0.05 to +0.09 (R2 < 1%) - though the sign was positive in 6-9 of 10 years. An R2 below 1% will not survive transaction costs."
  },
  {
   "n": 4,
   "headline": "The 'rates down, rate-sensitives win' rule barely holds",
   "figure": "corr(repo change, Financial Services) = +0.17",
   "body": "Financials did BETTER in hiking years. FY23 (250bp of hikes) saw rate-sensitives beat defensives by +13.7pp, while FY20 (185bp of cuts) saw them LOSE by 20.6pp.",
   "strength": "none",
   "strength_label": "Not statistically supported",
   "caveat": "At n=10 a correlation needs |r| > 0.63 for significance. +0.17 is far below that - this is indistinguishable from zero. The credit cycle, not the policy rate, was the binding constraint."
  },
  {
   "n": 5,
   "headline": "Momentum beats mean reversion - narrowly",
   "figure": "30.1% vs 24.7% (baseline 25%)",
   "body": "The prior-year best quartile stayed in the top quartile 30.1% of the time; the prior-year worst jumped to the top only 24.7% of the time - at the baseline. Momentum led in 6 of 9 transitions.",
   "strength": "weak",
   "strength_label": "Directional, not a clean rule",
   "caveat": "The exception is violent: in FY20 to FY21, prior-year LOSERS returned +213.5% against +98.0% for winners. Post-crash, expect reversion, not momentum."
  }
 ],
 "not_supported": [
  "The tidy interest-rate sector-rotation narrative - contradicted by FY20 and FY23, and statistically insignificant.",
  "Gold as a reliable steady hedge - it did nothing in FY17 and trailed equities for six of seven years before dominating FY25-FY26.",
  "Any notion of a 'consistent winner' class - only 19 names ever appeared in the top 20 in two consecutive years across 200 slots."
 ]
}