On-chain data

Tokenized Stocks Pointed to Monday's Open in 118 of 197 Weekend Checks, and in 7 of 19 Last Weekend

Across eleven weekends from August to October 2026, the weekend price of a tokenized stock on Solana sat on the same side of the last close as the share's next open in 118 of 197 cases. Last weekend it was 7 of 19. The script to check any weekend is included.

By

A stock closes on Friday and opens on Monday at a different price. You cannot trade the two days in between, so whatever happened over the weekend lands in one jump at the open. A tokenized stock is a token on Solana that tracks a real share, and it trades through the weekend. So people read its weekend price as an early look at Monday.

We checked how often that early look was right. The check needs three prices for each token, the last close of the real share, the token’s price at the end of the closed weekend, and the share’s next open. All three are public, and one script at the end of this page fetches them with no API key. Across eleven reopenings from 3 August to 12 October 2026, the weekend price of a tokenized stock on Solana pointed the way its share opened in 118 of 197 cases.

That is about 6 in 10. Last weekend it was 7 of 19, the lowest of the eleven. chaotic.markets also saves a live quote for every token every 5 minutes, so for last weekend we can show what a buy paid at the end of it, and how the prices moved overnight before the exchange opened.

Key findings

What was measured

The check
Reopenings 11, from 3 August to 12 October 2026
Tokens 21 stock and ETF tokens on Solana
Last close official close of the share before the weekend
Weekend price the token’s last trade before 20:00 New York time, the evening before the reopening
Open official open of the share on the reopening day
Possible cases 231
Left out 34
Cases checked 197

Ten of the reopenings were Mondays. One was Tuesday 8 September, after a three-day weekend, and for that one the token’s price is taken on Monday evening.

The cut is at 20:00 New York time for a reason. The issuer of the xStocks tokens publishes a session status for each one, and our readings save it. On Sunday 11 October it read closed until 23:55 UTC and overnight from 00:00 UTC, which is 20:00 in New York. From that minute the status counts the share’s market as open again, so the token is no longer the only price there is. The check takes the token’s price just before.

The token’s price is its last trade, from the 5-minute candles of Jupiter’s chart API. A token with no trade in the hour before the cut is left out. So is a token that had not started trading yet. That is the 34 cases in the table.

Same side means the token was above the last close and the share opened above it, or both were below. It is a low bar, because it says nothing about how far the share moved.

Eleven weekends

Reopening Tokens Above the close Same side
3 August 16 14 8
10 August 14 8 8
17 August 14 8 9
24 August 19 17 8
31 August 17 2 14
8 September 20 15 15
14 September 18 1 13
21 September 21 17 16
28 September 20 9 9
5 October 19 13 11
12 October 19 14 7
All eleven 197 118

Above the close counts the tokens priced over the last close at the cut. Same side counts the tokens whose share then opened on that side. In seven of the eleven weekends more than half of the tokens pointed the right way. In the other four, half or fewer did. The highest share was 14 of 17 at the 31 August reopening, and the lowest was 7 of 19 last weekend.

Look at the 14 September reopening. All but one of the tokens sat under the last close on Sunday night, and most of their shares did open lower. Tokens move together, the same way the shares do. When the whole market slides over a weekend most tokens are right at once, and when it turns on Monday morning most are wrong at once. So 197 cases are a lot fewer than 197 separate tests, and the count that matters is eleven weekends.

A bigger gap was right more often

The gap is how far the token sat from the last close. This table sorts the same 197 cases by the size of that gap, as the script prints it.

Gap from the last close Cases Same side
Under 0.25% 70 34
0.25% to under 0.5% 39 27
0.5% to under 1% 40 24
1% or more 48 33

When the token is within a quarter percent of the close, the side it sits on tells you nothing, it was right half the time. A gap that small can be the cost of trading the token, which the earlier check of a $100 buy measured for every token on this list.

At 1% or more the token was right about 7 times in 10. In 19 of those 33 right calls the share opened even further from the close than the token had been.

Last weekend went the other way

These are the six tokens that sat furthest above Friday’s close on Sunday night, 11 October, with where each share opened on Monday. Both columns are measured against Friday’s close.

Token Token, Sunday night Share, Monday open
MSTRx +2.02% -0.61%
SNDK +1.84% -2.76%
INTC +1.28% -2.08%
SPCXx +1.05% +1.29%
MU +0.98% -2.58%
AMD +0.70% -1.34%

Five of the six shares opened lower. Across all the tokens checked that weekend, 14 of 19 were above Friday’s close and 9 of those 14 shares opened below it. The full list with the prices is in the script output further down.

What a buy paid on Sunday night

A last trade can be one small order, so you may ask if these gaps were real prices. This is where our own readings help. chaotic.markets asks Jupiter for a live quote on every token every 5 minutes and saves how many tokens a buy would return. Dollars in, divided by tokens out, is the price paid. The last reading of the closed weekend was at 23:55 UTC on Sunday 11 October, and this is what it held for the five largest gaps. Each figure is the price paid over Friday’s close.

Token $1,000 buy $10,000 buy
MSTRx +2.09% +2.17%
SNDK +1.85% +1.86%
INTC +1.30% +1.32%
SPCXx +1.14% +1.16%
MU +1.06% +1.07%

So the premium was a price you could buy at. Size barely changed it, which is the opposite of how large trades fail on a DEX when a pool is thin. The earlier article on what a closed weekend does to volume and liquidity could not say if a weekend gap was a cost or a view on Monday. For these five it was a price the market held in size, and for four of them the share opened the other way.

The night before the open

The readings also show when the weekend premium went away. Each figure is the token’s price against Friday’s close. The last row is the share itself at Monday’s open. Times are UTC.

Reading SNDK MU MSTRx
Sun 18:00 +2.07% +1.11% +2.01%
Mon 00:00 +1.84% +0.98% +2.02%
Mon 02:00 -0.33% +0.44% +2.21%
Mon 06:00 -0.36% -0.24% +2.79%
Mon 08:15 -1.41% -1.00% +2.00%
Mon 12:00 -1.96% -2.53% +1.23%
Share, 13:30 -2.76% -2.58% -0.61%

SNDK and MU were first priced under Friday’s close at 01:25 UTC on Monday. That is 85 minutes after the overnight status began, and about 12 hours before the exchange opened. MSTRx kept its premium through the night and was first under Friday’s close at 13:15 UTC, 15 minutes before the open.

You can run the same check later in the night. At 04:15 New York time on Monday, 14 of 20 tokens pointed the way their shares opened, against 7 of 19 on Sunday evening.

One more thing from the readings. Jupiter’s asset API reports a share price next to each token, and for these tokens it did not move from Saturday morning until 08:20 UTC on Monday at the earliest. A screen that puts the token next to that share price was showing you Friday for the whole night.

What this does not tell you

Eleven weekends is a short history. One bad Monday moved the total, and the next few weekends can move it again.

The official close is at 16:00 New York time, and shares keep trading after it. Our readings hold the last share price Jupiter reported over the weekend, and for MSTR that was 154.90 against an official close of 154.34. So a part of a weekend gap can be Friday evening, which anyone could already see. Measured against that later price, last weekend still comes out at 7 of 19.

A last trade can be small, and the script does not look at its size. The quotes above cover that for one weekend only.

We make no claim about why any token or share moved.

Run it yourself

You need Python 3.9 or newer. There is nothing to install and no key to set. On Windows, Python’s zoneinfo needs one package first, pip install tzdata. Save this as monday_check.py.

"""Did the weekend price of a tokenized stock call Monday's open?

For each token: the last close of the real share, the token's price at the
end of the closed weekend, and the share's next open. Prints one row per
token and how many weekend gaps pointed the way the share opened.

usage: python3 monday_check.py 2026-10-12         the day the exchange reopened
       python3 monday_check.py 2026-10-12 04:15   token price at that New York time instead
"""

from __future__ import annotations

import json
import sys
import urllib.request
from datetime import date, datetime, time, timedelta, timezone
from time import sleep
from typing import Optional
from zoneinfo import ZoneInfo

CANDLES_URL = "https://datapi.jup.ag/v2/charts/{mint}?interval=5_MINUTE&to={to}&candles=1&type=price&quote=usd"
BARS_URL = "https://query1.finance.yahoo.com/v8/finance/chart/{ticker}?interval=1d&period1={start}&period2={end}"
NEW_YORK = ZoneInfo("America/New_York")
MAX_AGE_SECONDS = 3600  # a token that last traded longer ago than this is skipped

# token: (share ticker, Solana mint)
TOKENS: dict[str, tuple[str, str]] = {
    "NVDAx": ("NVDA", "Xsc9qvGR1efVDFGLrVsmkzv3qi45LTBjeUKSPmx9qEh"),
    "TSLAx": ("TSLA", "XsDoVfqeBukxuZHWhdvWHBhgEHjGNst4MLodqsJHzoB"),
    "AAPLx": ("AAPL", "XsbEhLAtcf6HdfpFZ5xEMdqW8nfAvcsP5bdudRLJzJp"),
    "MSFTx": ("MSFT", "XspzcW1PRtgf6Wj92HCiZdjzKCyFekVD8P5Ueh3dRMX"),
    "AMZNx": ("AMZN", "Xs3eBt7uRfJX8QUs4suhyU8p2M6DoUDrJyWBa8LLZsg"),
    "GOOGLx": ("GOOGL", "XsCPL9dNWBMvFtTmwcCA5v3xWPSMEBCszbQdiLLq6aN"),
    "METAx": ("META", "Xsa62P5mvPszXL1krVUnU5ar38bBSVcWAB6fmPCo5Zu"),
    "MSTRx": ("MSTR", "XsP7xzNPvEHS1m6qfanPUGjNmdnmsLKEoNAnHjdxxyZ"),
    "SPYx": ("SPY", "XsoCS1TfEyfFhfvj8EtZ528L3CaKBDBRqRapnBbDF2W"),
    "QQQx": ("QQQ", "Xs8S1uUs1zvS2p7iwtsG3b6fkhpvmwz4GYU3gWAmWHZ"),
    "SPCXx": ("SPCX", "Xs3oZwbHvqis4NYcf4YKWmEia2eC84wSiVrcYcTqpH8"),
    "GLDx": ("GLD", "Xsv9hRk1z5ystj9MhnA7Lq4vjSsLwzL2nxrwmwtD3re"),
    "NFLXx": ("NFLX", "XsEH7wWfJJu2ZT3UCFeVfALnVA6CP5ur7Ee11KmzVpL"),
    "MRNA": ("MRNA", "MRNAzXzhNcaEXJPibHEn8cd4vyekCDiivTyEwswLUCT"),
    "LLY": ("LLY", "LLYuwZ33keFihgwoxXsBawy31AiRFLFSva32TYq5TvD"),
    "MU": ("MU", "MUxEsUKSMACyw5fZf68wxf5FLnZVhtU9CwH8uNNGay1"),
    "INTC": ("INTC", "iNTCy1qTsUEZQe3DSocLz1ZXXai34Gdw8THQh5rxFaF"),
    "AMD": ("AMD", "AMD8XwJXgQ9WV45Wyj9yFLejxzf2J6VM1PJY8bJEjeES"),
    "MRVL": ("MRVL", "MRVLSjkR2ceUBukujaD3xCyHP1H3B2SzpsNTZF546jo"),
    "SNDK": ("SNDK", "SNDKbwMUQvZhnLnxLduradgLHG5KrPuKwpnrkkGRhfH"),
    "HOOD": ("HOOD", "HooDYv5RewLRiMLnEVq3VJqdqxhuE6c5eYvqejMC3e9A"),
}


def _get(url: str) -> object:
    """Fetch a URL and parse the JSON body. Tries twice, then returns None."""
    request = urllib.request.Request(url, headers={"User-Agent": "Mozilla/5.0 monday-check/1.0"})
    for _ in range(2):
        try:
            with urllib.request.urlopen(request, timeout=15) as response:
                return json.load(response)
        except Exception:
            sleep(2)
    return None


def _share_bars(ticker: str, reopen: date) -> Optional[tuple[float, float]]:
    """The last close before `reopen` and the open on `reopen`, from daily bars."""
    start = datetime.combine(reopen - timedelta(days=7), time(), timezone.utc)
    end = datetime.combine(reopen + timedelta(days=1), time(), timezone.utc)
    url = BARS_URL.format(ticker=ticker, start=int(start.timestamp()), end=int(end.timestamp()))
    try:
        result = _get(url)["chart"]["result"][0]  # type: ignore[index]
        quote = result["indicators"]["quote"][0]
        days = [datetime.fromtimestamp(t, NEW_YORK).date() for t in result["timestamp"]]
        at = days.index(reopen)
        return float(quote["close"][at - 1]), float(quote["open"][at])
    except Exception:
        return None


def _token_price(mint: str, moment: datetime) -> Optional[float]:
    """The token's last traded price before `moment`, in USD. None if it is over an hour old."""
    url = CANDLES_URL.format(mint=mint, to=int(moment.timestamp() * 1000))
    try:
        candle = _get(url)["candles"][-1]  # type: ignore[index]
        fresh = moment.timestamp() - candle["time"] <= MAX_AGE_SECONDS
        return float(candle["close"]) if fresh else None
    except Exception:
        return None


def weekend_end(reopen: date) -> datetime:
    """20:00 New York time on the day before the reopening, when the overnight session starts."""
    return datetime.combine(reopen - timedelta(days=1), time(20), NEW_YORK)


def check(reopen: date, moment: datetime) -> None:
    """Print each token's gap at `moment` next to the move of its share at the open."""
    print(f"token price at {moment.astimezone(timezone.utc):%Y-%m-%d %H:%M} UTC, share reopening {reopen}\n")
    print(f"{'token':7} {'last close':>10} {'token':>9} {'gap':>7} {'open':>9} {'share':>7}  same side")
    above = below_open = same = rows = 0
    for token, (ticker, mint) in TOKENS.items():
        bars = _share_bars(ticker, reopen)
        price = _token_price(mint, moment)
        if bars is None or price is None:
            print(f"{token:7} no share bar or no trade in the last hour")
            continue
        close, opened = bars
        gap = (price / close - 1) * 100
        move = (opened / close - 1) * 100
        agree = (gap > 0) == (move > 0)
        rows += 1
        same += agree
        above += gap > 0
        below_open += gap > 0 and move < 0
        print(f"{token:7} {close:10.2f} {price:9.2f} {gap:+6.2f}% {opened:9.2f} {move:+6.2f}%  {'yes' if agree else 'no'}")
    print(f"\n{above} of {rows} tokens were priced above the last close")
    print(f"{below_open} of those {above} shares opened below it")
    print(f"{same} of {rows} gaps pointed the way the share opened")


if __name__ == "__main__":
    if len(sys.argv) not in (2, 3):
        sys.exit(__doc__)
    day = date.fromisoformat(sys.argv[1])
    at = datetime.combine(day, time.fromisoformat(sys.argv[2]), NEW_YORK) if len(sys.argv) == 3 else weekend_end(day)
    check(day, at)

Run it with the day the exchange reopened.

python3 monday_check.py 2026-10-12
token price at 2026-10-12 00:00 UTC, share reopening 2026-10-12

token   last close     token     gap      open   share  same side
NVDAx       229.28    230.82  +0.67%    228.73  -0.24%  no
TSLAx       382.70    384.50  +0.47%    382.61  -0.02%  no
AAPLx       336.64    336.58  -0.02%    338.22  +0.47%  no
MSFTx       535.07    534.67  -0.08%    537.53  +0.46%  no
AMZNx       262.43    262.94  +0.20%    262.50  +0.03%  yes
GOOGLx      351.66    352.18  +0.15%    351.86  +0.06%  yes
METAx       718.67    721.87  +0.45%    728.50  +1.37%  yes
MSTRx       154.34    157.46  +2.02%    153.40  -0.61%  no
SPYx        778.57    779.60  +0.13%    778.78  +0.03%  yes
QQQx        751.27    753.47  +0.29%    749.69  -0.21%  no
SPCXx       162.57    164.28  +1.05%    164.66  +1.29%  yes
GLDx        384.58    383.97  -0.16%    382.61  -0.51%  yes
NFLXx   no share bar or no trade in the last hour
MRNA        225.00    223.46  -0.68%    224.50  -0.22%  yes
LLY        1179.27   1171.63  -0.65%   1181.24  +0.17%  no
MU         1029.00   1039.10  +0.98%   1002.45  -2.58%  no
INTC        104.70    106.04  +1.28%    102.52  -2.08%  no
AMD         608.10    612.37  +0.70%    599.93  -1.34%  no
MRVL        275.28    276.72  +0.52%    274.50  -0.28%  no
SNDK       1581.82   1610.96  +1.84%   1538.11  -2.76%  no
HOOD    no share bar or no trade in the last hour

14 of 19 tokens were priced above the last close
9 of those 14 shares opened below it
7 of 19 gaps pointed the way the share opened

Two tokens are left out here because neither traded in the last hour of the weekend.

To get the eleven-weekend table, run it for each reopening and keep the last line.

for day in 2026-08-03 2026-08-10 2026-08-17 2026-08-24 2026-08-31 2026-09-08 \
           2026-09-14 2026-09-21 2026-09-28 2026-10-05 2026-10-12; do
  echo "$day  $(python3 monday_check.py $day | tail -1)"
done
2026-08-03  8 of 16 gaps pointed the way the share opened
2026-08-10  8 of 14 gaps pointed the way the share opened
2026-08-17  9 of 14 gaps pointed the way the share opened
2026-08-24  8 of 19 gaps pointed the way the share opened
2026-08-31  14 of 17 gaps pointed the way the share opened
2026-09-08  15 of 20 gaps pointed the way the share opened
2026-09-14  13 of 18 gaps pointed the way the share opened
2026-09-21  16 of 21 gaps pointed the way the share opened
2026-09-28  9 of 20 gaps pointed the way the share opened
2026-10-05  11 of 19 gaps pointed the way the share opened
2026-10-12  7 of 19 gaps pointed the way the share opened

Add a New York time to move the token’s price to another moment on the reopening day. This is the check from the middle of the night.

python3 monday_check.py 2026-10-12 04:15 | tail -3
15 of 20 tokens were priced above the last close
6 of those 15 shares opened below it
14 of 20 gaps pointed the way the share opened

Past weekends should print the same numbers you see here. For the next one, wait until the exchange has opened on Monday and pass that date. The gap of each token against its share, as it stands right now, is on the chaotic.markets data pages, with one page per token such as MSTRx.

Buy on chaotic.markets

The after hours page ranks every token by how far it drifts from the real share while the exchange is closed, against the same gap while it is open, and every row has a Buy button. Press it, connect your Solana wallet, type the USDC amount and approve once. The tokens settle to your wallet and chaotic.markets never holds them. chaotic.markets adds no fee. You still pay the pool fees inside Jupiter’s route and the Solana network fee, plus any price over the share, which on a weekend is the gap in the closed column.

Newsletter

Get chaotic.markets updates and new data articles by email.

You confirm by email, and you can unsubscribe from any message. The list is kept by Kit.