BankNifty options

This repeats, and it kept repeating on data it was not found on.

Happened 361 times in 16 years — about 23 times a year.

It worked 166 times out of 166.

It held up on the years used to find it, and then held up again on later years it had never been tested against.

This repeats, and it kept repeating on data it was not found on.

Happened 2,153 times in 16 years — about 138 times a year.

It worked 1,153 times out of 1,242.

It held up on the years used to find it, and then held up again on later years it had never been tested against.

Only 47.7% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This repeats, and it kept repeating on data it was not found on.

Happened 149 times in 16 years — about 10 times a year.

It worked 55 times out of 55.

It held up on the years used to find it, and then held up again on later years it had never been tested against.

This repeats, and it kept repeating on data it was not found on.

Happened 302 times in 16 years — about 19 times a year.

It worked 174 times out of 176.

It held up on the years used to find it, and then held up again on later years it had never been tested against.

Almost all of the evidence for this (300 of 303 times) comes from the era when weekly BankNifty expiries existed. Weekly expiries were abolished, so this cannot be used today.

This repeats, and it kept repeating on data it was not found on.

Happened 3,741 times in 16 years — about 240 times a year.

It worked 829 times out of 1,608.

It held up on the years used to find it, and then held up again on later years it had never been tested against.

Only 47.6% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This is not proven to repeat.

Happened 335 times in 16 years — about 21 times a year.

It went the expected way 86 times out of 147.

It looked real on the years used to find it, but did not hold up on later years it had never been tested against. Treat it as not proven.

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

Only 48.6% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This does not repeat.

Happened 16 times in 16 years — about once a year.

It went the expected way 5 times out of 7.

This happened too rarely to test at all -- there is not enough data to say whether it repeats or not, in either direction. It did not pass: only 7 out-of-sample events, need 30

This has been seen only 2 times in the current era, since weekly expiries were abolished, too few to know whether it still holds now that weekly expiries are gone.

The range of plausible values is wide, so the size of the effect is much less certain than its direction.

Only 7 recent examples, which is a small number to draw a confident conclusion from.

This shows up in the data -- but in the opposite direction from what was predicted.

Happened 745 times in 16 years — about 48 times a year.

It went the expected way 68 times out of 318.

This is not a case of no evidence: the data shows a real, statistically decisive effect, consistently across regimes -- but in the opposite direction from what was predicted before the data was checked. A pre-registered prediction that turns out backwards is rejected on its own declared terms, however strong the underlying effect. What was actually found: out-of-sample effect +0.0565 (baseline -0.0088, excess +0.0653) contradicts the declared prior 'lower'

Only 48.8% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This does not repeat.

Happened 309 times in 16 years — about 20 times a year.

It went the expected way 73 times out of 131.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.840)

The range of plausible values is wide, so the size of the effect is much less certain than its direction.

Only 48.2% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This shows up in the data -- but in the opposite direction from what was predicted.

Happened 755 times in 16 years — about 48 times a year.

It went the expected way 106 times out of 325.

This is not a case of no evidence: the data shows a real, statistically decisive effect, consistently across regimes -- but in the opposite direction from what was predicted before the data was checked. A pre-registered prediction that turns out backwards is rejected on its own declared terms, however strong the underlying effect. What was actually found: out-of-sample effect -0.0169 (baseline +0.0081, excess -0.0250) contradicts the declared prior 'higher'

Only 48.2% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This does not repeat.

Happened 155 times in 16 years — about 10 times a year.

It went the expected way 22 times out of 53.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.500)

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

This does not repeat.

Happened 128 times in 16 years — about 8 times a year.

It went the expected way 33 times out of 53.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.083)

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

This does not repeat.

Happened 392 times in 16 years — about 25 times a year.

It went the expected way 91 times out of 160.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.109)

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

Only 44.5% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This does not repeat.

Happened 348 times in 16 years — about 22 times a year.

It went the expected way 60 times out of 125.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.859)

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

Only 47.0% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This does not repeat.

Happened 418 times in 16 years — about 27 times a year.

It went the expected way 96 times out of 181.

There is no evidence this repeats. It did not pass: not significant in sample (p=0.590)

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

Only 45.4% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This shows up in the data -- but in the opposite direction from what was predicted.

Happened 209 times in 16 years — about 13 times a year.

It went the expected way 53 times out of 118.

This is not a case of no evidence: the data shows a real, statistically decisive effect, consistently across regimes -- but in the opposite direction from what was predicted before the data was checked. A pre-registered prediction that turns out backwards is rejected on its own declared terms, however strong the underlying effect. What was actually found: out-of-sample effect +0.0280 (baseline +0.0294, excess -0.0014) contradicts the declared prior 'higher'

This has been seen only 2 times in the current era, since weekly expiries were abolished, too few to know whether it still holds now that weekly expiries are gone.

The range of plausible values includes zero, which means the effect could be real or could be nothing at all.

Only 32.1% of the option contracts involved had a usable, screened price on the days this happened, so this rests on a thin slice of the market.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: measures implied volatility, which is not directly observed but computed from an option's price using an assumed clock; this analysis counts calendar days while the market trades on business days; a zero-structure simulation containing no market behaviour at all -- constant true volatility, exact Black-76 prices on a random walk, no weekday effect planted -- reproduces this pattern's published effect almost exactly (simulated -1.25% vs published -1.97%) using only that calendar-day clock; and the effect weakens to insignificance (-1.97% to -0.45%, p 0.0001 to 0.067) under a trading-day clock, on the real data. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: measures implied volatility, which is not directly observed but computed from an option's price using an assumed clock; this analysis counts calendar days while the market trades on business days (dte=7 is 4.76 trading sessions, dte=3 is 2.88); a zero-structure simulation containing no market behaviour at all -- constant true volatility, exact Black-76 prices on a random walk, no weekday or dte effect planted -- reproduces this pattern's published effect almost exactly (simulated +14.16% vs published +14.14%) using only that calendar-day clock; and the effect weakens to insignificance (+14.14% to +0.23%, p 0.0001 to 0.58) and reverses sign under a trading-day clock, on the real data. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects a day-on-day total_oi increase above 15%, but total_oi is per-contract and accumulates over a contract's life, so an ordinary day's change already exceeds that threshold: measured median day-on-day change in this pattern's own 7-45 DTE band is +20.4%, 75th percentile +75.2%, 95th percentile +333.3%. A +15% move is therefore not a surge, it is below a typical day, and this threshold fires on 56.3% of eligible rows -- 198 of those from a prior day of exactly zero open interest, an artificial infinite percentage change rather than a real spike. Retuning the threshold to describe the data was considered and rejected as fitting the answer rather than testing the hypothesis; its mirror pattern, oi_unwind_precedes_iv_fall, is excluded for the same underlying reason. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects a day-on-day total_oi fall beyond 15%, but total_oi is per-contract and accumulates over a contract's life, so an ordinary day's change already exceeds that threshold in either direction: measured median day-on-day change in this pattern's own 7-45 DTE band is +20.4%, 75th percentile +75.2%, 95th percentile +333.3%. Against that distribution this threshold fires only 66 times across 16 years, and on every one of those 66 rows the contract's own ATM implied volatility is null -- the threshold mostly catches contracts already too thin to have a screened price at all, which is also why none of the 66 selections ever attaches to a scoreable outcome: this hypothesis has zero events, in or out of sample, and was never actually tested against a single observation. Retuning the threshold to describe the data was considered and rejected as fitting the answer rather than testing the hypothesis. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on atm_iv's own trailing-year rank (top decile) and then measures atm_iv's own subsequent change: the event is chosen because a quantity is unusually high, and the same quantity's subsequent change is then measured. An unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on atm_iv's own trailing-year rank (bottom decile) and then measures atm_iv's own subsequent change: the event is chosen because a quantity is unusually low, and the same quantity's subsequent change is then measured. An unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on skew_25d's own value against its own trailing rolling median and then measures skew_25d's own subsequent change: the event is chosen because a quantity is unusually far from its own recent typical level, and the same quantity's subsequent change is then measured. An unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on atm_iv's own trailing-year rank and measures atm_straddle: the two are formally different quantities, but not empirically different ones -- in the 7-45 DTE band this pattern selects on, period-over-period fractional change in atm_straddle correlates 0.876 with fractional change in atm_iv, because an ATM straddle's price at this tenor is dominated by vega exposure to implied volatility. Selecting on an unusually low atm_iv reading and then measuring atm_straddle's own subsequent change is therefore almost the same test as selecting on and measuring atm_iv itself: an unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: uses a same-date cross-sectional aggregate (the far expiry's implied volatility on the same session), and the lookahead guard truncates by row position, so it cannot distinguish 'the session was split in half' from 'the detector read the future'. Untestable under the current guard, therefore not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on vrp (= atm_iv - rv_cc_20) exceeding 0.08 and then measures atm_iv's own subsequent change: vrp is not a different quantity from the outcome, it contains it, so an extreme vrp reading is partly an extreme atm_iv reading. An unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.

This was not tested.

Not tested. This is not a rejection — the statistics needed to test it honestly are not available, so it was never run against the data. Reason on record: selects on vrp (= atm_iv - rv_cc_20) being negative and then measures atm_iv's own subsequent change: vrp is not a different quantity from the outcome, it contains it, so a negative-vrp reading is partly a low-atm_iv reading. An unusually extreme value tends to be followed by a less extreme one purely by chance, so this design finds an apparent effect even in data with no relationship at all. Testing it honestly needs a null that reproduces that tendency, which this harness does not implement. Not tested -- not rejected on evidence.

No hit rate, frequency or range is given for this row because none was ever computed — treat every number here as absent, never as zero.