Litecoin.watch — "Litecoin's 24-hour return can turn red while the latest price stays flat"
Version: 2026-10-26.1
Article: https://litecoin.watch/articles/litecoin-s-24-hour-return-can-turn-red-while-the-latest-price-stays-flat

WHAT THIS PACKAGE IS
An original, deliberately constructed time-series and Decimal replay. Every LTC
price and EUR/USD reference in inputs.json is fictional, not a captured market
price. The dated sequence includes selected illustrative instants on 24–26 October
2026; it is not a forecast, trade reconstruction or historical performance claim.
No wallet, exchange account or execution model is involved.

CENTRAL FINDINGS
At fictional 26 Oct 2026 00:30 UTC (01:30 CET in Europe/Warsaw), the final price is
63 USD per LTC. Comparison with a 60 USD exact observation 24 hours earlier gives
+5.00000000%; with the UTC-day opening fixture at 64 USD, -1.56250000%; with the
Warsaw-day opening fixture at 62 USD, +1.61290323%.
At the later fictional endpoint 01:30 UTC, the price remains 63 USD, but its
24-hour starting point is now 65 USD and the rolling result is -3.07692308%.
These are different denominators, not inconsistent arithmetic.

TIME RULES
Rolling means subtract 86,400 elapsed seconds on UTC-aware timestamps. Calendar
starts use the chosen zone. Elapsed duration is calculated AFTER converting both
ends to UTC. The local dates here straddle Warsaw's 25 October 2026 autumn clock
change. The two 25 Oct 02:30 observations have offsets +02:00 and +01:00 and fold
values 0 and 1. "Yesterday at the same local clock" spans 90,000 seconds in the
chosen comparison; the completed local date 25 Oct also spans 25 hours.

The series is sparse. It has exactly the eight synthetic points required for
these demonstrations. It does not describe the path between points. Required
endpoints must exist exactly; the script never interpolates, silently selects a
nearby candle close or fills missing prices with zero.

CURRENCY RULE
EUR/USD is USD per one EUR. Derived LTC/EUR = LTC/USD divided by EUR/USD at EACH
endpoint. A +5% LTC/USD change and +5% EUR/USD change give zero derived LTC/EUR
change in this fixture. This is a frictionless cross-reference, not an observed
LTC/EUR exchange quote or a net investor return.

REPRODUCE
Use Python 3.9+ with an IANA timezone database. Python's zoneinfo can use system
data or the first-party tzdata package (often needed on Windows). The checked
runtime was Python 3.12.10 with tzdata 2026.4. No network request or package
installation is performed by the script.

  python replay.py.txt --verify

This recomputes all four calculation files, compares exact UTF-8 bytes and checks
every file listed in SHA256SUMS.txt. It does not change files. To write fresh
calculation outputs separately:

  python replay.py.txt --write --out recalculated

Decimal precision is 50; only display is rounded, using ROUND_HALF_UP. Percentages
are exported to eight decimals; derived currency prices to twenty decimals.
The optional plot.py.txt uses matplotlib and reads the already-calculated results.

FILES
inputs.json — all fictional exact endpoints and FX assumptions.
time-series.csv — those eight points, UTC, Warsaw offset and fold.
return-comparisons.csv — six explicitly named endpoint comparisons.
fx-normalization.csv — one same-window synthetic currency comparison.
results.json — calculation results and 17 fixed expectation checks.
replay.py.txt — original calculation and read-only verification source.
return-definitions.png / .svg — original plot of three defined return measures.
plot.py.txt — optional plot source; Decimal CSV remains the numerical evidence.
sources.json — primary documentation and pinned timezone source.
api-shape-checks.json — independent public request receipts and response field
names. These are not trade event times or synchronized venue observations, and
no captured numeric quote is used in inputs.json or in the return calculation.
SHA256SUMS.txt — SHA-256 file manifest, excluding the manifest itself.

LIMITS
No claim is made about which convention an unidentified app uses, a typical LTC
move, provider caching policy, executable prices, fills, fees, taxes, portfolio
cash flows or investment merit. A provider-reported opening statistic need not
be an observed trade at the exact wall-clock boundary. Do not relabel this exact
synthetic-endpoint model as an independent audit of an exchange's trade history.