Your reports
Value history
What your portfolio was worth at the moments Accuren looked at it — an archive, not a query, and it does not reach back before the recording started.
Accuren records the total value of your positions each time it reads them: hourly on its own, plus whenever you have the dApp open. /value in the dApp draws those readings.
It starts when the recording starts
There is no source for a past price of these tokens. The block explorer publishes the current one and nothing republishes yesterday's, so a value for a date before Accuren was watching cannot be obtained, computed, or bought. A new account sees an empty chart, and that is the honest state rather than a fault.
#Recorded, not reconstructed
The alternative was to replay your balances backwards and multiply them by an archived price. Every step of that is an inference, and an inferred curve is indistinguishable on screen from a measured one. Each point here is instead an observation: this many positions, read at this minute, summed at the prices published then.
#Why the line breaks
A break is a refusal to assert something. Three things cause one, and the page counts each:
| Cause | What the line does | Why |
|---|---|---|
| Two readings priced a different number of positions | Breaks | The total sums only what could be priced. Four of six tokens is a smaller number for a portfolio that did not shrink — joined to a complete reading it draws a crash that never happened. |
| More than six hours with nothing reading the account | Breaks | The sweep runs hourly, so a longer gap means ticks were missed. A straight line through it is a claim that nothing happened in between. |
| A reading could price nothing at all | Point is blank | Recorded as unknown, never as zero. A zero in a value series is a vertical line to the floor. |
The change figure skips incomparable ends
It is measured between the earliest and latest readings that covered the same positions — not simply the first and last in the window. Otherwise a window opening on a partial read and closing on a complete one reports the tokens that were missing as a gain.
#One limit worth knowing
Matching position counts is a necessary condition for comparing two readings, not a sufficient one: swapping one token for another between reads keeps the counts identical and changes what is being summed. The counts are what a reading carries, so that residual is named here rather than papered over.
Every reading is in the CSV export on the page, with its coverage, so any figure can be checked against the two numbers that produced it.