How much world share prices move at the weekend: up to 30 weekends, measured
Source: Hyperliquid public API / Last updated: / Perpetual-futures prices are reference values.
On Saturday and Sunday the exchanges are shut. Even so, on Monday morning share prices start from somewhere other than where Friday left them.
How much moves while they are shut? This "invisible time" was measured on 24-hour perpetual-futures reference prices.
The six symbols covered are the Nikkei 225, the S&P 500, XYZ U.S. 100, the VanEck Semiconductor ETF, Gold and USD/JPY.
How it was measured — two windows were prepared
The common window and the cash-market window
There is more than one way to cut out "the weekend". This article uses two windows and keeps them apart.
| Window | Definition | Where it is used |
|---|---|---|
| Common window | Close of the hourly candle immediately before 00:00 JST on Saturday → close of the hourly candle immediately before 08:00 JST on Monday | Comparing the six symbols on the same measure |
| Cash-market window | The close of the cash market → its next open (for the Nikkei 225: 15:30 JST on Friday → 09:00 JST on Monday) | Measuring "while the market is shut" |
"The close of the hourly candle immediately before" is the close of the last hourly candle that completed before the target time. The data used is the hourly candles of Hyperliquid perpetual futures. Over the period, no candles were missing for any symbol.
A common window shared by the six symbols, and a cash-market window matched to closing hours.
Why the number of weekends differs by symbol
Hourly candles can be retrieved only so far back, and coverage starts on a different date for each symbol. The weekends that could be measured are 20 for the Nikkei 225, 23 for the S&P 500, 30 each for XYZ U.S. 100, Gold and USD/JPY, and 10 for the VanEck Semiconductor ETF.
Only the period that could be retrieved is used, with no interpolation. Trading activity in these symbols also grew a great deal over the period, so the early weekends sit on thin trading.
Between 10 and 30 weekends were measured per symbol. Trading was thin early on.
The result — the weekend across six symbols (common window)
| Symbol | N (weekends) | Median absolute change | Beyond ±0.5% | Beyond ±1% | Largest rise | Largest fall |
|---|---|---|---|---|---|---|
| Nikkei 225 | 20 | 0.61% | 55.00% | 30.00% | +1.42% | -3.50% |
| S&P 500 | 23 | 0.26% | 34.78% | 13.04% | +0.90% | -2.02% |
| XYZ U.S. 100 | 30 | 0.52% | 50.00% | 23.33% | +1.43% | -3.17% |
| VanEck Semiconductor ETF | 10 | 0.74% | 60.00% | 10.00% | +1.47% | -0.81% |
| Gold | 30 | 0.52% | 50.00% | 33.33% | +3.22% | -3.64% |
| USD/JPY | 30 | 0.14% | 10.00% | 3.33% | +0.75% | -1.26% |
The largest rise and the largest fall are the maximum and the minimum of the signed change. USD/JPY weekends were quiet, and Gold had the highest share beyond ±1% of the six symbols.
The three equity indices (the Nikkei 225, the S&P 500 and XYZ U.S. 100) had their largest fall on the same weekend, 2026-06-06. The weekends that moved a lot arrived across indices at the same time.
Note: the common window for the three U.S.-listed symbols (the S&P 500, XYZ U.S. 100 and the VanEck Semiconductor ETF) includes part of the U.S. cash session on Friday. As a worked example, over the weekend of 2026-06-06 the S&P 500 common window was -2.02% while its cash-market window was +0.88%. The column below sets out this difference.
Beyond ±1%: Gold 33%, the Nikkei 30%, USD/JPY 3%. The three indices had their largest fall on the same weekend.
The Nikkei 225 — beyond ±1% on half the weekends "while it is shut"
The measurement for the Nikkei 225 cash-market window (close of the hourly candle immediately before 15:30 JST on Friday → close of the hourly candle immediately before 09:00 JST on Monday).
| Item | Value |
|---|---|
| N (weekends) | 20 |
| Median absolute change | 0.93% |
| Weekends beyond ±1% | 50.00% (10/20) |
| Largest rise | +3.82% (2026-06-13) |
| Largest fall | -4.37% (2026-06-06) |
The cash-market window includes Friday evening and night in Japan time, which is the cash session in Europe and the United States. That is why it comes out larger than the common window (0.61%).
When Monday's open jumps away from Friday's close, what sits inside that jump is the movement during this window.
The change from Friday's close to Monday's open had a median of 0.93%, and went beyond ±1% on half the weekends.
An unexpected result — the weekend moves less than a single weekday
For the same Nikkei 225, the median absolute change over a weekday 24 hours (9:00 → 9:00 on the next trading day) was 1.37% (N=77).
The common-window weekend, an effective 56 hours, is 0.61%. It is more than twice as long, yet the median change is less than half.
Rather than "the weekend moves a lot", the phrasing closer to the measurement is "the weekend keeps moving, and is occasionally large".
The weekend (0.61%) had a smaller median change than a weekday 24 hours (1.37%).
Weekend by weekend (common window, change in %)
Each row is the Saturday date of that week, JST. A blank means the symbol had no hourly candles for that weekend.
| Weekend (Saturday, JST) | Nikkei 225 | S&P 500 | XYZ U.S. 100 | VanEck Semiconductor ETF | Gold | USD/JPY |
|---|---|---|---|---|---|---|
| 2026-01-31 | -1.65 | -3.64 | +0.63 | |||
| 2026-02-07 | +1.27 | +1.31 | +0.41 | |||
| 2026-02-14 | +0.28 | +0.81 | -0.22 | |||
| 2026-02-21 | +0.57 | +2.00 | -0.23 | |||
| 2026-02-28 | -0.62 | +3.22 | -0.02 | |||
| 2026-03-07 | -1.51 | +0.02 | +0.44 | |||
| 2026-03-14 | -0.06 | -0.38 | +0.08 | |||
| 2026-03-21 | -0.78 | -0.91 | -1.49 | +0.19 | ||
| 2026-03-28 | -1.42 | -1.26 | -0.54 | +0.34 | ||
| 2026-04-04 | -0.20 | -0.10 | -0.49 | +0.06 | ||
| 2026-04-11 | -1.95 | -1.48 | -1.51 | -2.43 | +0.41 | |
| 2026-04-18 | -1.60 | -0.81 | -0.61 | -1.97 | +0.75 | |
| 2026-04-25 | -0.12 | +0.19 | +0.43 | -0.86 | +0.06 | |
| 2026-05-02 | +0.73 | -0.17 | +0.13 | -0.31 | +0.11 | |
| 2026-05-09 | +0.46 | -0.19 | +0.54 | -0.48 | +0.17 | |
| 2026-05-16 | -0.24 | -0.26 | -0.17 | -0.39 | +0.16 | |
| 2026-05-23 | +0.92 | +0.52 | +0.59 | +1.42 | -0.17 | |
| 2026-05-30 | +1.01 | +0.16 | +0.31 | -0.65 | +0.11 | |
| 2026-06-06 | -3.50 | -2.02 | -3.17 | -1.19 | +0.07 | |
| 2026-06-13 | +1.24 | +0.76 | +1.43 | +1.99 | -0.21 | |
| 2026-06-20 | -0.66 | -0.36 | -0.50 | -0.81 | +0.15 | +0.19 |
| 2026-06-27 | +0.36 | +0.24 | +0.23 | -0.37 | -0.38 | +0.03 |
| 2026-07-04 | -0.15 | +0.03 | +0.09 | +0.41 | +0.39 | +0.11 |
| 2026-07-11 | -0.49 | +0.23 | +0.31 | +0.33 | -0.38 | +0.10 |
| 2026-07-18 | +0.58 | -0.27 | +0.33 | +0.83 | -0.09 | +0.08 |
| 2026-07-25 | +1.42 | +0.39 | +0.57 | +0.79 | +0.58 | +0.01 |
| 2026-08-01 | +0.42 | +0.90 | +0.98 | +0.76 | +0.37 | -1.26 |
| 2026-08-08 | -0.03 | -0.06 | +0.43 | +1.47 | +0.03 | +0.01 |
| 2026-08-15 | +0.14 | +0.01 | +0.24 | +0.13 | +0.13 | +0.08 |
| 2026-08-22 | +0.63 | -0.04 | +0.18 | +0.71 | +0.38 | -0.04 |
The raw values for 30 weekends are printed as they are. A record, not a tendency.
Common questions
Q. Is this the movement of the Nikkei 225 or the S&P 500 themselves?
A. No. These are Hyperliquid perpetual-futures reference values. They can differ from official index values and from cash-market share prices.
Q. Why does the number of weekends differ by symbol?
A. Coverage starts on a different date for each symbol. On top of that, hourly candles can be retrieved for about 208 days back. Only the weekends that could be retrieved are counted, with no interpolation.
Q. Can it be said that "the weekend moves a lot"?
A. Not on this measurement. The median of the common window was smaller than a weekday 24 hours. Even so, weekends beyond ±1% came to 30% for the Nikkei and 33% for Gold.
Q. Will future weekends move the same way?
A. That is not written here. This article is a record of the measured period, up to 30 weekends, not a forecast of future price moves.
Summary
- The Nikkei 225 cash-market window (Friday 15:30 → Monday 9:00) had a median of 0.93%, and went beyond ±1% on half the weekends.
- The medians of the common window across the six symbols ran from 0.14% (USD/JPY) to 0.74% (the VanEck Semiconductor ETF).
- The weekend common window (0.61%) had a smaller median than a weekday 24 hours (1.37%).
- The three equity indices had their largest fall on the same weekend, 2026-06-06.
- All of it is a record of the period, on Hyperliquid perpetual-futures reference values.
Measurement note: hourly candles, with 0 missing. Retrieval ends at 12:00 UTC on 25 August 2026. The statistics were checked against an independent recalculation and matched. U.S. Eastern Time standard and daylight-saving offsets are processed per date in the measurement (the whole period fell in EDT).
The weekend moves can be seen as they are on the 24-hour tab. For the weekday overnight measurement, see how much the Nikkei 225 moves overnight; for the weekend Nikkei explainer, see the Nikkei 225 on a weekend.
This article explains a mechanism and a measurement. It does not recommend any particular trade.