Muhurat trading: does the Diwali session say anything about the year ahead?
The Nifty 50 closed higher in 16 of the last 21 Diwali sessions. But the year after a down session went up just as often as the year after an up session.
I collected every NSE Muhurat trading session from 2005 to 2025 and checked whether an up or down Diwali session told you anything about the Samvat year that followed. It did not.
On Sunday, 8 November 2026, the stock market will open for a short while on Diwali. NSE has said that normal trading will run from 2:20 pm to 3:30 pm that afternoon, after a short pre-open period from 2:05 pm.1 This is Muhurat trading. A muhurat is an auspicious time, and the session is held on the day of Lakshmi Puja.
Every year, someone says that a good Muhurat session is a good sign for the year ahead. I wanted to check that against the data. So I collected every Muhurat session NSE has held since 2005 and looked at what the market did in the year after each one.
What Muhurat trading is
Diwali (Lakshmi Puja) is a stock market holiday. But the exchanges open for one special session that day, usually about an hour long.
The session marks the start of a new Samvat year. Samvat (Vikram Samvat) is a Hindu calendar, and many traders treat Diwali as the start of their new financial year. The 2024 session marked the start of Samvat 2081, and the 2025 session marked the start of Samvat 2082.2 Business families traditionally close their old account books around Diwali and open new ones. The account books are worshipped during Lakshmi Puja, a ritual called Chopda Pujan.3
For many investors, the session is a symbolic way to start the year with a small, “token” purchase of shares.3 These are real trades. NSE’s circulars say that every trade in the Muhurat session has to be settled like a normal trade.1 According to Tata Mutual Fund, BSE started the tradition in 1957 and NSE began holding it in 1992.3
The timing moves around. In 18 of the 21 years from 2005 to 2025, normal trading started in the evening, between 5:30 pm and 6:30 pm. The exceptions were 2011 (4:45 pm), 2012 (3:45 pm) and 2025 (1:45 pm). This year it will start at 2:20 pm.4
How I built the data
NSE publishes a circular before each Muhurat session with the date and timings. I found one for every year from 2005 to 2025, so the dataset has 21 sessions.4 For each one, I took the Nifty 50 closing level from NSE’s index data.5 The Nifty 50 is an index of 50 large companies listed on NSE. It shows how the biggest Indian shares are doing as a group.
I measured two things:
- The session move. The Nifty 50 close on Muhurat day compared with its close on the previous trading day.
- The Samvat year return. The change from one Muhurat close to the next year’s Muhurat close. This is roughly how the market did over that Samvat year.
Two caveats. First, the session move also includes anything that happened between the previous close and the Muhurat session, such as news from abroad. Second, I used the normal Nifty 50 price index, which leaves out dividends. I could not download BSE’s Sensex history, so I used the Nifty 50 only. For 2024 and 2025, news reports of the Sensex show it moved in the same direction as the Nifty 50 in the session.6
The session usually ends up
The Nifty 50 closed higher in 16 of the 21 sessions. On an ordinary day the odds are much closer to a coin toss. Across all trading days from 2005 to 2025, the Nifty 50 closed higher on about 54% of days.7
The moves are usually small. The middle (median) session move was a gain of 0.37%. Ignoring whether it went up or down, the median Muhurat move was 0.47%, compared with 0.61% on all trading days.7 The big exception was 2008, in the middle of the global financial crisis, when the Nifty 50 rose 6.35% on Muhurat day.
The five down sessions were in 2007, 2009, 2012, 2016 and 2017. The 2009 one barely counts: the Nifty 50 fell by 0.01%, from 5,142.15 to 5,141.80.
| Diwali | Samvat begins | Session move | Next Samvat year |
|---|---|---|---|
| 2005 | 2062 | +0.67% | +54.3% |
| 2006 | 2063 | +0.18% | +53.7% |
| 2007 | 2064 | -0.62% | -52.6% |
| 2008 | 2065 | +6.35% | +91.5% |
| 2009 | 2066 | -0.01% | +22.8% |
| 2010 | 2067 | +0.49% | -17.6% |
| 2011 | 2068 | +0.20% | +8.9% |
| 2012 | 2069 | -0.29% | +11.5% |
| 2013 | 2070 | +0.16% | +26.9% |
| 2014 | 2071 | +0.23% | -2.4% |
| 2015 | 2072 | +0.54% | +10.2% |
| 2016 | 2073 | -0.14% | +17.6% |
| 2017 | 2074 | -0.63% | +4.5% |
| 2018 | 2075 | +0.65% | +9.7% |
| 2019 | 2076 | +0.37% | +9.9% |
| 2020 | 2077 | +0.47% | +40.2% |
| 2021 | 2078 | +0.49% | -1.0% |
| 2022 | 2079 | +0.88% | +10.1% |
| 2023 | 2080 | +0.52% | +24.5% |
| 2024 | 2081 | +0.41% | +6.4% |
| 2025 | 2082 | +0.10% | not over yet |
Why does the session usually end up? I cannot prove a reason from this data. One guess is that people come to buy their token shares, and few people want to sell on an auspicious day. But that is a guess, and the data here cannot test it.
Does the session predict the year?
Here is the simple test. I sorted the 20 complete Samvat years by whether the Muhurat session that started them was up or down, and whether the year itself ended up or down.
| Muhurat session | Next year up | Next year down | Share up |
|---|---|---|---|
| Up (15 sessions) | 12 | 3 | 80% |
| Down (5 sessions) | 4 | 1 | 80% |
The share of up years is the same, 80%, in both rows. After an up session, the year rose 12 times out of 15. After a down session, it rose 4 times out of 5.
That 80% is simply how often the market rises over a year. Of the 20 Samvat years, 16 ended higher. I also checked every possible one-year stretch of the Nifty 50 since 2005, starting on each trading day. About 80% of them ended higher too.7 So “an up session is followed by an up year” is true most of the time, but only because almost anything is followed by an up year most of the time.
Each bar runs from one Muhurat close to the next, labelled by the Diwali it starts on. Price index, without dividends.Source: Nifty Indices historical data; NSE Muhurat circulars; my arithmetic
Each bar runs from one Muhurat close to the next, labelled by the Diwali it starts on. Price index, without dividends.Source: Nifty Indices historical data; NSE Muhurat circulars; my arithmetic
What about the averages?
The averages look more interesting, but they mislead. After the 15 up sessions, the next Samvat year returned 21.7% on average. After the 5 down sessions, it returned just 0.7% on average.7
That gap comes almost entirely from two years in a row. The 2007 session was down, and the year after it was the 2008 crash, a fall of 52.6%. The 2008 session was up 6.35%, and the year after it was the recovery, a rise of 91.5%. Take those two years out and the averages come much closer: 16.7% after an up session and 14.1% after a down one. The middle (median) year tells the same story: 10.1% after an up session and 11.5% after a down one.7
With only five down sessions, a single bad year can drag that group’s average down by more than ten percentage points. Twenty data points are not enough to find a real pattern here, and these twenty do not show one.
Is the session really low-volume?
Turnover is the total value of shares bought and sold. NSE publishes it for every trading day from 2011. In the 2025 Muhurat session, turnover in NSE’s cash market (shares bought and sold for delivery or same-day trading, not futures and options) was ₹19,714 crore. Over the 20 trading days before it, the average was about ₹1,02,294 crore a day. So the Muhurat session saw about 19% of a normal day’s turnover.8
NSE cash market turnover on Muhurat day, divided by the average of the 20 trading days before it.Source: NSE business growth data (capital market segment); my arithmetic
From 2011 to 2025, Muhurat turnover was between 12% and 30% of a normal day. In the middle year it was about 17%. That sounds low, but most of it is about time. A normal NSE session runs from 9:15 am to 3:30 pm, which is 375 minutes. The 2025 Muhurat session had 60 minutes of normal trading, or 16% of that time.8 Minute for minute, the 2025 session was about as busy as an ordinary day, or a little busier.
Short sessions can still be harder to trade in. With fewer buyers and sellers around, the gap between the price buyers offer and the price sellers ask can be wider.3
Samvat 2082 so far
The 2025 session was slightly up: the Nifty 50 closed at 25,868.60, up 0.10%. On 23 October 2026, the latest close in my data, it was at 22,600.00, which is 12.6% below that Muhurat close.5 Samvat 2082 ends with this year’s session on 8 November, so this number will change. I am not making any forecast for Samvat 2083.
What this means
The Muhurat session tends to end slightly up, and its turnover is low mostly because it lasts about an hour. In the 20 complete years I could check, an up or down session did not line up with an up or down year any more often than chance would suggest. Twenty years is a small sample, so this cannot rule out a weak effect. It does show that the session has not been a useful guide so far.
Footnotes
-
NSE, circular NSE/CMTR/76941, 19 October 2026: Muhurat session on Sunday, 8 November 2026, pre-open order entry from 14:05, normal market open 14:20, normal market close 15:30 for securities outside the closing auction (15:15 for those in it, followed by a closing auction until 15:30). NSE, circular NSE/CMTR/71775, 12 December 2025 (trading holidays for 2026), lists 8 November 2026 as Diwali Laxmi Pujan with Muhurat trading. ↩ ↩2
-
All India Radio News, 1 November 2024 (start of Samvat 2081) and 21 October 2025 (start of Vikram Samvat 2082). Samvat numbers for earlier years are counted back one per year (my arithmetic). ↩
-
Tata Mutual Fund, Muhurat trading explained, 19 October 2026. This is a fund house’s explainer, not an exchange document; I could not find the 1957 and 1992 start dates in an official BSE or NSE source. ↩ ↩2 ↩3 ↩4
-
NSE capital market circulars announcing each Muhurat session: 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012 (timings revised in CMTR22083), 2013, 2014, 2015, 2016, 2017, 2018, 2019 (timings revised in CMTR42483), 2020, 2021, 2022, 2023, 2024, 2025. ↩ ↩2
-
Nifty 50 daily closing levels from Nifty Indices historical data, downloaded 26 October 2026 (latest close 23 October 2026). Session move and Samvat year returns are my arithmetic: (Muhurat close ÷ previous trading day’s close) − 1, and (next Muhurat close ÷ this Muhurat close) − 1. ↩ ↩2
-
All India Radio News, 1 November 2024: Sensex up 0.42% to 79,724, Nifty up 0.41% to 24,304. 21 October 2025, 3:44 pm: Sensex up 63 points to 84,426, Nifty up 25 points to 25,869. BSE’s index archive blocked automated downloads, so I did not build a Sensex series. ↩
-
My arithmetic from the Nifty Indices data. Share of up days: 5,208 daily changes from January 2005 to December 2025, 53.9% positive. Median absolute daily move on all days 0.61%, on Muhurat days 0.47%. One-year stretches: the change over 250 trading days from every start date, 79.6% positive. Averages and medians of Samvat year returns use the 15 or 5 years in each group; the figures without the two crisis years leave out 2008 from the up group and 2007 from the down group. The 2009 session (down 0.01%) is counted as down. ↩ ↩2 ↩3 ↩4 ↩5
-
NSE, business growth in the capital market segment, daily turnover in ₹ crore, from 2011 (earlier days are not in this dataset). The ratio is Muhurat day turnover divided by the average of the 20 previous trading days (my arithmetic). Normal hours of 09:15 to 15:30 are the “standard market timings” in NSE circular NSE/CMTR/72349, 16 January 2026. The 2025 Muhurat session’s normal market ran from 13:45 to 14:45 (NSE/CMTR/70319). The ratio of 19% of turnover to 16% of the time is my arithmetic. ↩ ↩2