Does Tesla’s weekend price on Hyperliquid call Monday’s open? Not before Monday morning

Short answer: not on Sunday: over 50 weekends to November 2, 2026, Tesla’s perp on Hyperliquid at 7:00 p.m. or 8:00 p.m. New York time on Sunday pointed the right way from Friday’s close 29 times, about a coin flip. At 7:00 a.m. or 8:00 a.m. Monday, it did 42 times.

A perpetual future, or perp, is a contract that follows a price and never expires. trade[XYZ] runs one on Tesla (TSLA), called xyz:TSLA, on Hyperliquid, a blockchain built for trading. It trades all weekend while Nasdaq is shut. We compared it with Friday’s close and with the price that printed at 9:30 a.m. Monday, on Tesla and then on the 85 US stocks and funds on the exchange that we could match to their listing.

Tesla’s 50 weekends: Sunday evening against Monday morning

Friday’s close and Monday’s open come from Yahoo Finance, from November 21, 2025 on. The perp’s price is its last trade at two moments: midnight UTC between Sunday and Monday, which is 8:00 p.m. Sunday in New York in summer time and 7:00 p.m. in winter, and noon UTC Monday, which is 8:00 a.m. or 7:00 a.m. "Right side" means the perp was up from Friday’s close when Monday opened up, and down when it opened down. "Nearer" means it was closer to the open than Friday’s close was.

On Sunday evening, 29 right calls in 50 is close to chance: a coin flip does that well or better about 1 time in 6. The perp was nearer the open than Friday’s close on 27 weekends and further on the other 23. By Monday morning it was a different instrument: 42 right calls, and a median miss of 0.33%, a third of what holding Friday’s close gave.

A Monday gap is the move between Friday’s close and Monday’s open. Tesla’s was a median 1.12% either way over these weekends, and its biggest was +3.12%, on Oct. 26, 2026. Results reports move Tesla far more: see does Tesla stock fall after earnings?

The weekend itself is noise; the signal comes back Sunday night

Hourly candles go back to April on Hyperliquid’s public API, so we ran the same test hour by hour on every weekend since April 10, 2026 that ran from a Friday to a Monday: 1,543 weekends across the 85 stocks and funds. The chart at the top of this page is the result, in New York time.

From 8:00 p.m. Friday to 6:00 p.m. Sunday, the perps were nearer Monday’s open than Friday’s close in 45% of weekends, on average across those hours. For most of the weekend, Friday’s close was the better guess. Then other markets open, and the perp follows them:

• 7:00 p.m. Sunday, an hour after US stock index futures reopen on CME Globex: 51%.
• 9:00 p.m. Sunday, an hour after overnight trading in US stocks starts (Robinhood’s 24 Hour Market, for one): 59%.
• 4:00 a.m. Monday, as pre-market trading opens: 64%.
• 9:00 a.m. Monday, half an hour before the open: 79%.

Tesla alone follows the same shape on its 26 weekends with hourly data, more jagged because 26 is a small sample: 37% over the weekend hours, 77% at 9:00 a.m. Monday. The perp does not find Monday’s price on its own; it carries what the futures, the overnight session and the pre-market have already found.

How trade[XYZ] prices Tesla while Nasdaq is closed

On a weekend the perp has no stock quote to follow. Its reference price starts from Friday’s close and drifts toward the perp’s own trading over about 8 hours, as Blockworks Research described the design in October 2025. The price used to close out losing positions cannot move more than 5% either way from Friday’s close on Tesla. Over our 50 weekends, the perp’s biggest move from Friday’s close by Sunday evening was 2.66%, well inside that limit.

Weekend trading is thin: a median $8.7 million of the Tesla perp changed hands between Friday’s close and Sunday evening. At Friday’s close, the perp and the stock agree: over the 30 Fridays with hourly candles, the median gap between them at 4:00 p.m. was 0.03%. trade[XYZ] is also the company ICE signed with in October: see why Hyperliquid rose on the ICE deal.

Tesla’s 7 biggest Monday gaps

On the 7 weekends Tesla opened 2% or more away from Friday’s close, the Sunday-evening perp was on the right side 5 times, and the Monday-morning perp 7 times.

Even when it had the direction, the Sunday price caught only part of the move: on Oct. 23–26, 2026, Tesla opened +3.12% while the perp stood at −0.31% on Sunday evening. The busiest weekend went the other way. On Oct. 16–19, 2026, $89.3 million of the perp traded, the most of any weekend, and it stood +2.66% on Sunday evening; Tesla opened +0.78%.

85 stocks and funds: who the Sunday price helped

Across all 85 instruments, 68 stocks and 17 funds, and 2,185 weekends, the Sunday-evening perp was nearer Monday’s open than Friday’s close 51.4% of the time, and on the right side of Friday’s close 61.6% of the time. On Monday morning those figures were 75.3% and 87.6%. On gaps of 2% or more the Sunday price was on the right side 475 times in 662: a big weekend story is usually visible in the perp by Sunday, even when the size is not.

Of the 30 names with 30 weekends or more, 16 beat Friday’s close on more than half their Sunday evenings and 13 on fewer. The best, CoreWeave (CRWV), did it 24 times in 39, a result a coin flip matches about 1 time in 10; the worst, the iShares MSCI Japan fund (EWJ), 12 times in 34. With about 40 weekends each, that spread is what chance alone produces, so no name stands out as a reliable Sunday signal. Alphabet (GOOGL), 28 times in 49, and Nvidia (NVDA) and Tesla, 27 times in 50 each, sat in the top third, still within chance.

Other measurements agree. Blockworks Research’s 0xResearch found the pre-open Hyperliquid price closer to the reopening than Friday’s close 50.7% of the time across 23 markets and 146 weekend samples. Crypto.com Research had Tesla’s perp on the right side 52.6% of the time at 10:00 p.m. UTC on Sunday over 19 weekends to April 30, 2026.

What it means if you own Tesla

Over the weekend, Friday’s close is still the best estimate of where your Tesla shares will open: a weekend headline about a move on Hyperliquid tells you little. From Monday morning, New York time, the perp is a fair early read, because by then it follows stock futures, the overnight session and pre-market trading rather than leading them. The Tesla (TSLA) page shows its latest quote and the time it printed.

The same goes for a portfolio. Portfolio Terminal values your lines at the last close, and the Nasdaq Composite and S&P 500 pages show where the index stood when the session ended. If you also hold crypto on Hyperliquid, paste your Hyperliquid address and its tokens come in as lines beside your stocks and funds; perps are listed in the review, not added. Read-only: no sign-up, no signature, no key. Our other tests, including Tesla after its product shows, are on the research shelf.

How we measured

Perp prices are Hyperliquid’s public API (candleSnapshot on xyz:, no key): 4-hour candles from each listing (Tesla’s from November 13, 2025) and hourly candles from April 2026, the API’s limit of about 5,000 candles per interval. A perp’s price at a moment is the close of the last candle that ended by then. Stock prices are Yahoo Finance daily bars as printed (split-adjusted, not dividend-adjusted). A weekend is two sessions in a row with a Saturday between them; 7 of Tesla’s 50 were long weekends or Good Friday, and the hour-by-hour test keeps only Friday-to-Monday ones. Each perp’s first week is skipped. Of the 142 xyz markets, 98 have a US-listed stock or fund with the same ticker on Yahoo; 8 have no trading history in the API yet, 1 listed too recently for a full weekend, and 4 failed a price check (perp within 1.5% of the stock at Friday’s close), namesakes such as CL (Colgate-Palmolive on Yahoo, crude oil on Hyperliquid). That leaves 85, and we dropped 1 weekend whose perp sat 25% or more from the close. What would change the result: a different Sunday hour (ours falls after stock futures reopen at 6:00 p.m. and by 8:00 p.m., when overnight stock trading starts), news that breaks over a weekend, and a deeper weekend order book as these markets grow.

Reproduce it

This rebuilds the Tesla table from Hyperliquid’s public API and Yahoo Finance, with no key; we ran it on November 3, 2026.

# Reproduces https://portfolio-terminal.com/blog/tesla-weekend-price-hyperliquid-monday-open
# pip install requests yfinance pandas
# Tesla's perp on Hyperliquid (xyz:TSLA) against Tesla's Monday open, every weekend from
# Nov. 21, 2025 to Oct. 30, 2026: does the perp's price on Sunday evening, or at 12:00 UTC
# Monday, sit on the right side of Friday's close, and nearer the open than Friday's close?
from datetime import datetime, timedelta, timezone
from statistics import median

import requests
import yfinance as yf

H4 = 4 * 3600 * 1000
req = {"coin": "xyz:TSLA", "interval": "4h", "startTime": 0, "endTime": int(datetime.now(timezone.utc).timestamp() * 1000)}
rows = requests.post("https://api.hyperliquid.xyz/info", json={"type": "candleSnapshot", "req": req}, timeout=30).json()
close_at = {int(r["t"]) + H4: float(r["c"]) for r in rows}  # candle end (ms, UTC) -> last trade


def perp(day, hour):
    """The perp's last trade at `hour`:00 UTC on `day`."""
    t = datetime.strptime(day, "%Y-%m-%d").replace(tzinfo=timezone.utc) + timedelta(hours=hour)
    return close_at.get(int(t.timestamp() * 1000))


bars = yf.Ticker("TSLA").history(start="2025-11-20", end="2026-11-03", auto_adjust=False)
days = [d.strftime("%Y-%m-%d") for d in bars.index]
weekends = []
for i in range(1, len(days)):
    a, b = datetime.strptime(days[i - 1], "%Y-%m-%d"), datetime.strptime(days[i], "%Y-%m-%d")
    if not any((a + timedelta(n)).weekday() == 5 for n in range(1, (b - a).days)):
        continue
    if days[i - 1] < "2025-11-21" or days[i - 1] > "2026-10-30":
        continue
    close, open_ = float(bars["Close"].iloc[i - 1]), float(bars["Open"].iloc[i])
    sunday, preopen = perp(days[i], 0), perp(days[i], 12)
    gap = (open_ / close - 1) * 100
    weekends.append((days[i - 1], close, open_, gap, (sunday / close - 1) * 100, (preopen / close - 1) * 100))

print(f"{len(weekends)} weekends, {weekends[0][0]} to {weekends[-1][0]} (Friday)\n")
print(f"{'Friday':<12}{'close':>9}{'Mon open':>10}{'gap':>8}{'Sun eve':>9}{'Mon 12Z':>9}")
for f, c, o, g, s, p in weekends:
    print(f"{f:<12}{c:>9.2f}{o:>10.2f}{g:>+7.2f}%{s:>+8.2f}%{p:>+8.2f}%")


def score(k, label):
    moved = [w for w in weekends if w[3] != 0]
    called = sum(1 for w in moved if (w[k] > 0) == (w[3] > 0) and w[k] != 0)
    beat = sum(1 for w in weekends if abs(w[k] - w[3]) < abs(w[3]))
    err = median(abs(w[k] - w[3]) for w in weekends)
    print(f"{label}: right side of Friday's close {called} of {len(moved)}, "
          f"nearer the open than Friday's close {beat} of {len(weekends)}, median miss {err:.2f}%")


print()
print(f"Friday's close as the guess: median miss {median(abs(w[3]) for w in weekends):.2f}%")
score(4, "Perp, Sunday evening (00:00 UTC)")
score(5, "Perp, Monday 12:00 UTC")

50 weekends, 2025-11-21 to 2026-10-30 (Friday)

Friday          close  Mon open     gap  Sun eve  Mon 12Z
2025-11-21     391.09    402.17  +2.83%   +1.67%   +1.87%
2025-11-28     430.17    425.32  -1.13%   +0.44%   -1.26%
2025-12-05     455.00    447.45  -1.66%   +0.07%   -1.47%
2025-12-12     458.96    469.44  +2.28%   -1.00%   +1.03%
2025-12-19     481.20    489.88  +1.80%   +0.35%   +1.15%
2025-12-26     475.19    469.00  -1.30%   +0.12%   -1.45%
2026-01-02     438.07    447.99  +2.26%   +1.86%   +1.72%
2026-01-09     445.01    441.23  -0.85%   +0.25%   -0.68%
2026-01-16     437.50    429.36  -1.86%   -0.82%   -2.43%
2026-01-23     449.06    445.00  -0.90%   -1.64%   -0.87%
2026-01-30     430.41    421.29  -2.12%   -0.85%   -2.39%
2026-02-06     411.11    409.91  -0.29%   +0.52%   -0.17%
2026-02-13     417.44    412.36  -1.22%   +0.09%   -1.08%
2026-02-20     411.82    407.29  -1.10%   -0.36%   -1.00%
2026-02-27     402.51    390.60  -2.96%   -0.76%   -2.36%
2026-03-06     396.73    390.05  -1.68%   -2.48%   -1.92%
2026-03-13     391.20    396.22  +1.28%   +0.26%   +1.09%
2026-03-20     367.96    373.09  +1.39%   +0.23%   -0.16%
2026-03-27     361.83    365.86  +1.11%   -1.01%   +0.63%
2026-04-02     360.59    362.59  +0.55%   -0.37%   +0.87%
2026-04-10     348.95    350.07  +0.32%   -1.26%   -0.80%
2026-04-17     400.62    402.58  +0.49%   -1.41%   -0.72%
2026-04-24     376.30    372.09  -1.12%   -0.31%   -0.39%
2026-05-01     390.82    390.23  -0.15%   +0.52%   +0.17%
2026-05-08     428.35    422.16  -1.45%   -0.30%   -0.76%
2026-05-15     422.24    419.27  -0.70%   -0.78%   -1.10%
2026-05-22     426.01    430.26  +1.00%   +1.80%   +1.04%
2026-05-29     435.79    427.49  -1.90%   +0.61%   -1.42%
2026-06-05     391.00    396.33  +1.36%   +1.46%   +1.64%
2026-06-12     406.43    412.37  +1.46%   +1.09%   +1.31%
2026-06-18     400.49    394.85  -1.41%   -0.66%   -0.95%
2026-06-26     379.71    381.79  +0.55%   +0.32%   +0.82%
2026-07-02     393.45    397.50  +1.03%   +1.52%   +1.50%
2026-07-10     407.76    404.61  -0.77%   -0.54%   -0.75%
2026-07-17     380.84    386.05  +1.37%   +0.93%   +1.22%
2026-07-24     313.03    313.24  +0.07%   +1.39%   +0.90%
2026-07-31     311.21    310.96  -0.08%   +1.18%   +0.26%
2026-08-07     328.58    326.60  -0.60%   +0.37%   +0.52%
2026-08-14     342.27    340.69  -0.46%   +0.27%   -0.27%
2026-08-21     362.86    361.41  -0.40%   +0.82%   -0.38%
2026-08-28     348.75    347.21  -0.44%   -0.31%   -0.54%
2026-09-04     354.08    357.10  +0.85%   +0.25%   +0.49%
2026-09-11     365.44    359.78  -1.55%   -0.80%   -1.64%
2026-09-18     364.27    371.64  +2.02%   +0.21%   +1.56%
2026-09-25     372.11    368.06  -1.09%   +0.27%   -0.43%
2026-10-02     370.59    369.00  -0.43%   +0.53%   -0.68%
2026-10-09     382.70    382.61  -0.02%   +0.56%   +0.07%
2026-10-16     354.72    357.50  +0.78%   +2.66%   -0.02%
2026-10-23     372.22    383.83  +3.12%   -0.31%   +2.23%
2026-10-30     399.81    404.66  +1.21%   +0.11%   +0.80%

Friday's close as the guess: median miss 1.12%
Perp, Sunday evening (00:00 UTC): right side of Friday's close 29 of 50, nearer the open than Friday's close 27 of 50, median miss 1.03%
Perp, Monday 12:00 UTC: right side of Friday's close 42 of 50, nearer the open than Friday's close 41 of 50, median miss 0.33%

This study measures prices. It is not investment advice, and Portfolio Terminal has no link with trade[XYZ] or Hyperliquid.

Sources: Hyperliquid API documentation · Blockworks Research, "Weekends are for trading", Oct. 20, 2025 · Blockworks Research 0xResearch, "When markets sleep" · Crypto.com Research, real-world asset perpetuals, May 2026 · CME Group, E-mini S&P 500 contract specs · Robinhood, 24 Hour Market · Yahoo Finance, TSLA history