Money, Timings & Holidays

Two small things trip up almost everyone writing code for Indian markets, and neither is hard - they're just different from the US defaults most software assumes. The first is units: India counts in lakhs and crores, grouped 2-2-3, not in millions grouped in threes. The second is the clock: the market runs 9:15 to 3:30 on its own time zone, skips weekends, and closes on a long list of holidays. Get these right and your numbers and timestamps will always make sense to an Indian trader.

Lakhs and crores

Indian numbering groups digits 2-2-3 from the right - so 656800 is written 6,56,800 (six lakh, fifty-six thousand, eight hundred), not the Western 656,800. A lakh is 1,00,000 and a crore is 1,00,00,000. Python's built-in :, gives Western grouping, so for Indian style you write a small helper:

def indian_format(n):
    """Format a whole number with Indian 2-2-3 digit grouping (lakh/crore style)."""
    s = f"{int(round(n)):d}"
    if len(s) <= 3:
        return s
    last3, rest = s[-3:], s[:-3]
    parts = []
    while len(rest) > 2:                 # group the rest in twos, from the right
        parts.insert(0, rest[-2:])
        rest = rest[:-2]
    if rest:
        parts.insert(0, rest)
    return ",".join(parts) + "," + last3

# The same numbers, Western grouping vs Indian grouping.
for n in [1313, 656800, 18016237, 1500000000]:
    print(f"Western {n:>14,}   ->   Indian Rs {indian_format(n)}")

Western          1,313   ->   Indian Rs 1,313
Western        656,800   ->   Indian Rs 6,56,800
Western     18,016,237   ->   Indian Rs 1,80,16,237
Western  1,500,000,000   ->   Indian Rs 1,50,00,00,000

The function peels off the last three digits, then groups the rest in twos. The difference is stark on big numbers: a Reliance lot's value is 6,56,800 to an Indian reader, and a billion rupees is 1,50,00,00,000 - 150 crore. Whenever you display rupee figures to Indian users, reach for this grouping; the Western commas look subtly wrong to a local eye.

Market hours across time zones

The NSE cash session runs 09:15 to 15:30 IST, weekdays only; the US runs 09:30 to 16:00 ET. To reason about both safely, attach a real time zone to your datetimes with Python's zoneinfo, and convert with astimezone:

from datetime import datetime, time
from zoneinfo import ZoneInfo

ist = ZoneInfo("Asia/Kolkata")
et = ZoneInfo("America/New_York")

# One fixed moment - 11:00 on a Thursday in Mumbai - shown in two time zones.
moment = datetime(2026, 6, 25, 11, 0, tzinfo=ist)
print("Mumbai   (IST):", moment.strftime("%a %H:%M %Z"))
print("New York (ET) :", moment.astimezone(et).strftime("%a %H:%M %Z"))
print()

# NSE cash session is 09:15 to 15:30 IST, on weekdays only.
def nse_open(dt):
    dt = dt.astimezone(ist)
    return dt.weekday() < 5 and time(9, 15) <= dt.time() <= time(15, 30)

for label, m in [("Thu 11:00 IST", datetime(2026, 6, 25, 11, 0, tzinfo=ist)),
                 ("Thu 16:00 IST", datetime(2026, 6, 25, 16, 0, tzinfo=ist)),
                 ("Sun 11:00 IST", datetime(2026, 6, 28, 11, 0, tzinfo=ist))]:
    print(f"{label}: NSE open? {nse_open(m)}")

Mumbai   (IST): Thu 11:00 IST
New York (ET) : Thu 01:30 EDT

Thu 11:00 IST: NSE open? True
Thu 16:00 IST: NSE open? False
Sun 11:00 IST: NSE open? False

The output makes the gap concrete: 11:00 in Mumbai is 01:30 the previous night in New York. That's why, on an IST clock, the US market only comes alive in your evening. The nse_open check combines the weekday test from Chapter 23 with a time-zone-aware conversion, so it's correct no matter where the code runs.

Holidays and special sessions

Weekends are easy (weekday() >= 5), but markets also close on a long list of trading holidays - Republic Day, Holi, Diwali, Independence Day and more - that don't follow any simple rule. There's no formula; you check the date against the exchange's published holiday list for the year. The day also has structure beyond the bell: a pre-open auction from 09:00 to 09:15 sets the opening price before continuous trading begins.

Try it yourself

• Format 12345678 and 99000 with indian_format. Read them aloud in lakhs and crores.
• Convert datetime(2026, 6, 25, 9, 30, tzinfo=et) (US open) into IST - what evening time does the US market open for an Indian trader?
• Add a check to nse_open that also rejects a hard-coded holiday date of your choice (e.g. dt.date() == date(2026, 1, 26)).

Recap

• Indian numbers group 2-2-3 (lakh 10^5, crore 10^7); use a small helper since :, is Western.
• NSE trades 09:15-15:30 IST (pre-open from 09:00); the US trades 09:30-16:00 ET - barely overlapping on an IST clock.
• Attach real time zones with zoneinfo and convert with astimezone; IST is UTC+5:30.
• Markets close on weekends and holidays (no simple rule - check the published list).

You can now speak the market's language of money and time. It's finally time to look closely at the unit of market data itself - the bar that every chart is built from. The next chapter dissects OHLCV data: open, high, low, close and volume, and the candle they form.