Tickfossil / Observation study 01

A difference is only the beginning.

Two venues. One apparent gap.

Delay changes what you could know.

Two fees reduce what remains.

Tickfossil

Keep the evidence. Replay the difference.

Quotes → observation → cost → evidence10-second illustrated study
Tickfossil

Historical quote laboratory / 01

What survived
the spread?

The price gap you saw is only half the story.
Rewind what was known. Subtract what it cost.

THE EXPERIMENT

Two quote streams.
One shared clock.
No future information.

Local simulation · No trading
01 / Replay bench

A frame worth examining.

Empty experiment · import your quotes or load an exampleAwaiting two streams
Venue ANo quotes
Bid
Ask
Import a quote stream to begin.
SYNC
0ms
t
Venue BNo quotes
Bid
Ask
Only quotes received by the cursor appear.
Gross difference
Buy-side fee
Sell-side fee
=
Simulated net · quote currency

No calculation until both venues have a usable quote.

Observation & costs

Changes apply to the entire replay.

Your data stays in this browser.

02 / Input desk

Bring both sides of the story.

Use the same asset, quote currency and clock origin. Timestamps are integer milliseconds from your experiment start. CSV: timestamp,bid,ask. Up to 5,000 rows per venue; 500 KB per file. No depth or fill data is assumed.

03 / Evidence notebook

Find the moment it vanished.

Positive net changes are logged at quote arrival and expiry events. A stale quote is a missing basis for a calculation, not a zero return.

  1. No event evidence yet. Start a replay.

Sessions save automatically in this browser.

04 / Model notes

The clock is part
of the evidence.

What is visible at time t?

For each venue, take the latest source timestamp s satisfying s + venue delay ≤ t. No interpolation or future data. Quote age is t − s; it is stale only when age is greater than the maximum. Quotes arriving already stale remain unusable.

What exactly gets subtracted?

For A → B, buy at A’s ask and sell at B’s bid. Gross = quantity × (B bid − A ask). Buy fee = quantity × A ask × A fee / 10,000. Sell fee = quantity × B bid × B fee / 10,000. Net = gross − both fees. Reversing direction swaps the venues and their fee rates. Values use your shared quote currency; 1 basis point is 0.01%.

Does a positive net mean an executable trade?

No. This is a historical quoted-price simulation. It does not model order-book depth, fills, slippage, inventory, funding, transfers or exchange permissions. Observation delay is not an execution guarantee. Local calculations use JavaScript floating-point arithmetic and are not settlement accounting.

Why an application token?

Tickfossil proposes portable service access across compatible replay and time-alignment tools, plus contribution records for shared experiments. Such a token would provide a common access unit only if independent services adopt it. The local replay needs no token. No token has been issued and the external record registry is not connected.

Portable experiment record

SOON

A shared attestation registry is not connected. Your experiment can already be saved as JSON and replayed locally. No wallet, signature, payment or token issuance is requested.