Close Wapas Average Ke Upar — aage kya hua
Jawab
Kam se kam 780 effective observations. Interval ±3.5pp ke andar. Daur ke dono hisson mein sign ek jaisa. Zyadatar saalon mein maujood. Multiple-testing adjustment ke baad bhi tika rehta hai. Grade kaise tay hote hain.
Seedhe shabdon mein. Share apne 20-day average price ke upar close hua. Kal close us average par ya neeche tha. Isse pehle pichhle 10 sessions mein se kam se kam 5 us average ke neeche the.
close[t] > sma(close, 20)[t] AND close[t-1] <= sma(close, 20)[t-1] AND count(close < sma(close, 20) over t-10..t-1) >= 5
5 September 2026 ko likha gaya, naapne se pehle. Headline pehle se chuni gayi beech wali setting par hai: 20-day average price aur pichhle 10 sessions mein se kam se kam 5 us average ke neeche. Registration log dekhiye.
Yeh kis sample par tika hai
Sample
| Design | Ek din ka correlation | Design effect | Effective observations |
|---|---|---|---|
| Raw rate, bina control | 0.183 | 5.08 | 16,513 |
| Control ke saath joda hua (published) | 0.018 | 1.39 | 60,340 |
Public daily bhavcopy mein purani dates ka sector data nahi hai. Isliye sector-concentration check nahi chal saka. Sector-matched version bhi nahi hai. Agar pattern asal mein poore sector ka ek saath chalna hai, toh yeh sample use nahi pakadta. Iska matlab.
Har horizon, apne control ke saath
| Trading days | Dekha gaya | Matched control | Farak | 95% interval | Median badlav | Excess median | 2% par break-even |
|---|---|---|---|---|---|---|---|
| 1 | 42.7% | 43.8% | −1.1pp | −1.5 to −0.6pp | −0.31% | −0.08% | 55.6% |
| 3 | 45.6% | 47.3% | −1.7pp | −2.2 to −1.2pp | −0.35% | −0.19% | 55.6% |
| 5 | 46.7% | 48.4% | −1.7pp | −2.3 to −1.2pp | −0.35% | −0.26% | 55.6% |
| 10 | 47.9% | 49.3% | −1.4pp | −1.9 to −0.8pp | −0.32% | −0.33% | 55.6% |
| 20 | 50.0% | 51.0% | −1.0pp | −1.5 to −0.4pp | 0.00% | −0.35% | 55.6% |
Distribution hi jawab hai
| Days | p10 | p25 | Median | p75 | p90 | Trimmed mean | Sabse upar ka point | Sabse neeche ka point |
|---|---|---|---|---|---|---|---|---|
| 1 | −2.8% | −1.5% | −0.31% | +1.0% | +2.7% | −0.14% | +1.3% | −1.5% |
| 3 | −4.9% | −2.6% | −0.35% | +2.1% | +5.3% | −0.01% | +2.4% | −2.5% |
| 5 | −6.3% | −3.3% | −0.35% | +3.0% | +7.2% | +0.12% | +3.2% | −3.1% |
| 10 | −8.8% | −4.5% | −0.32% | +4.6% | +10.4% | +0.37% | +4.8% | −4.5% |
| 20 | −12.0% | −6.0% | 0.00% | +7.2% | +15.5% | +1.07% | +7.2% | −6.4% |
Saal dar saal, aur market ki haalat ke hisaab se
| Saal | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Farak | +1.2pp | +0.8pp | −0.3pp | −2.2pp | −2.7pp | −2.5pp | −3.7pp | −1.0pp | −1.1pp | −1.0pp | −3.1pp | −2.3pp | −0.6pp | −2.5pp | −1.3pp | −1.1pp |
| Effective n | 577 | 2,095 | 1,992 | 2,532 | 3,085 | 2,904 | 3,605 | 3,606 | 3,010 | 3,082 | 4,670 | 4,699 | 5,192 | 6,104 | 6,382 | 3,731 |
| Market, apne 200-day average ke saamne | Volatility | Farak | Effective n |
|---|---|---|---|
| Upar | sabse neeche ka tihai | +9.5pp | 140 |
| Upar | beech ka tihai | +3.6pp | 262 |
| Upar | sabse upar ka tihai | −1.7pp | 1,234 |
| Neeche | sabse neeche ka tihai | −1.9pp | 18,883 |
| Neeche | beech ka tihai | −1.7pp | 19,509 |
| Neeche | sabse upar ka tihai | −1.8pp | 20,062 |
Robustness
Kya yeh dabaav jhel paata hai?
Yeh is pattern ke 25 pre-registered variants mein se ek hai. Yeh is site ke 1,125 tests mein se bhi ek hai. Neeche ke figures isi ke hisaab se adjust kiye gaye hain.
| Daur ka pehla aadha | −1.9pp | Median event date par baanta gaya. |
|---|---|---|
| Daur ka doosra aadha | −1.6pp | |
| Dono aadhon mein ek hi sign | haan | Jo asar sirf ek aadhe mein ho, woh ek daur hai, base rate nahi. |
| Ek hi sign wale saal | 16 mein se 14 | |
| Bina overlap wala subsample | −1.9pp | Har security ka har window mein ek hi event, taaki koi 2 observations din saajha na karein. |
| Published charges ke baad | −1.6pp | Koi move tabhi gina gaya jab usne round-trip charge paar kiya. |
| Censoring ka sabse bura haal | −2.0pp | Har censored observation ko loss gina gaya. |
| Parameter grid ka median | −1.7pp | Headline grid median ke paas hai. |
| Ek session baad shuru karna | −1.3pp | Kai session wala pattern isse bach jaata hai. Pehle bar ki chaal se bana bhram nahi bachta. |
| Placebo: wahi securities, be-mel dates | +0.5pp | Yeh zero ke paas hona chahiye. Agar nahi hai, toh method khud farak bana raha hai. |
| Sector-matched variant | uplabdh nahi | Public bhavcopy se sector ka point-in-time classification dobara nahi banaya ja sakta. |
| Multiple-testing adjusted p | 0.0002 | False-discovery adjustment ke baad bhi tikta hai. |
| average ki lambai | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|
| 16 | −1.5pp | −1.5pp | −1.6pp | −1.6pp | −1.7pp |
| 18 | −1.5pp | −1.6pp | −1.6pp | −1.7pp | −1.8pp |
| 20 | −1.6pp | −1.6pp | −1.7pp | −1.8pp | −1.9pp |
| 22 | −1.6pp | −1.6pp | −1.8pp | −1.8pp | −1.8pp |
| 24 | −1.6pp | −1.8pp | −1.8pp | −1.9pp | −2.0pp |
Asli pattern aas-paas ki settings par ek samtal maidan jaisa hota hai. Shor ke beech akela chamakta cell ek threshold ka bhram hota hai. Ghera hua cell pehle se declare kiya gaya beech ka cell hai. Headline usi cell se li jaati hai, chahe woh sabse mazboot ho ya nahi.
Quality checks
| Check | Zaroori | Yeh sample | Nateeja |
|---|---|---|---|
| Effective sample size | >= 100 | 60340 | paas |
| Alag event dates | >= 60 | 3597 | paas |
| Alag securities | >= 25 | 2027 | paas |
| Calendar ka faasla | >= 5 years | 14.8 years | paas |
| Ek quarter ka sabse bada hissa | <= 20.0% | 3.8% | paas |
| Ek security ka sabse bada hissa | <= 5.0% | 0.2% | paas |
| Censoring rate | <= 2.0% | 0.5% | paas |
| Relaxation rung 3-4 par events | <= 20.0% | 0.0% | paas |
Yeh kya nahi batata
- Yeh nahi batata ki aage kya hoga. Base rate beete events ke ek sample ka bayaan hai. Bhavishya us sample se nahi nikalta.
- Yeh kisi security ki pehchaan nahi karta. Yeh figures saalon tak saikdon securities ka jod hain. Inse kisi ek security ka bartaav nahi nikala ja sakta.
- Yeh aapke kharche, aapki position size ya aapka tax nahi ginta. Yeh bhi nahi ginta ki aap loss ka daur kitna jhel sakte ho. Break-even column sirf published charges cover karne ki zaroori frequency dikhata hai. Usme slippage, impact aur aapki apni koi baat shaamil nahi.
- Control se farak koi edge nahi hai. Yeh ek sample ke baare mein observation hai, uncertainty ke saath naapa hua. Woh observation aage bhi tike, yeh zaroori nahi.
- Interval number se zyada maayne rakhta hai. Jahan interval mein zero aata hai, wahan yeh sample pattern ko market se alag nahi kar paata.
- Sample size jaisa dikhta hai, waisa nahi hota. Ek hi din ke events aapas mein azaad observations nahi hain. Upar dikhaya gaya effective sample size dekhiye.
Is page ke sawaal
Matched control kya hai, aur har number ke saath ek kyun hai?
Har event ke liye usi date ke milte-julte securities liye gaye. Unka price band, realised-volatility band aur traded-value band wahi tha. Par unhone khud shart poori nahi ki thi. Is daur mein Indian market lagbhag har 3 rolling 20-din windows mein se 2 mein chadha. Isliye koi bhi long-side pattern lagbhag 60% ke paas frequency dikhata hai. Akele us number ka koi matlab nahi. Chhapne laayak number dono ka farak hai, koi ek nahi.
Effective sample events ki ginti se chhota kyun hai?
Ek hi din ke events azaad observations nahi hain. Sab ek market move baantte hain. Effective sample size is baat ko ginta hai. Har interval aur check isi ko use karta hai, raw ginti ko nahi.
Kya yeh batata hai ki agli baar kya hoga?
Nahi. Yeh figures sample mein beete events ginte hain. Har result ke saath uncertainty hai. Yeh agla event predict nahi karte, na kisi security ki sifarish karte hain.
Kya isse charges cover hote?
Break-even column ek symmetric 2% target aur stop ke liye ek frequency dikhata hai. Itni frequency sirf published charges cover karne ke liye chahiye. Is page par woh level 55.6% hai, jabki dekha gaya figure 46.7% tha. Isme slippage, impact aur aapki apni koi baat shaamil nahi. Yeh sab ise galat taraf le jaate hain.
Doosre registered patterns
Naya Sixty-Session High
Share ne pichhle sixty sessions ka sabse ooncha band diya, aur volume apne 20 din ke average se kam nahi tha. 2 shartein, ek din, aur unki limits.
Kholo →Uptrend Mein 3 Girte Din
Lagataar 3 sessions neeche band, phir bhi share apne 50 din ke moving average ke upar. Yeh shart kaise likhi gayi, aur woh kya nahi keh sakti.
Kholo →Mahine Ka Shuruaati Session
Mahine ka teesra trading session, un shares ke liye jo apne 100 din ke average ke upar the. Yeh study effective sample ka sabse saaf udaharan hai.
Kholo →Gap down below 20dma
The stock opened at least 2% below yesterday's close, on a day when it was already trading below its 20-day average price.
Yeh page abhi English mein hai.Kholo →Shabdon ka matlab · Apne bracket ke liye zaroori frequency nikaalo
Agla kadamYeh figure kaise bana
Point-in-time universe, usi dates se liya gaya matched control, aur interval. Saath mein woh quality checks jo koi bhi rate dikhane se pehle sample ko paar karne hote hain.
Methodology padhoSource. NSE daily bhavcopy. Data 2026-08-06 tak. Stock list: point-in-time, har date par jaisi thi. 2026-09-05 ko naapa gaya. Kaise naapa jaata hai · Shabdon ka matlab · Yeh figures download karo, CC BY 4.0 licence ke tahat.