Support Board
Date/Time: Mon, 31 Aug 2026 03:57:03 +0000
Post From: Reused ticker symbols showing blended history from prior entity (Denali feed)
| [2026-07-28 15:40:12] |
| TonyCipriani - Posts: 148 |
|
Hi, thanks for trying but just from testing ticker symbol changes that occurred over the past 30 days alone, it looks like the Bar Chart solution missed almost 70% (16/23) of the changes. What it did catch: VSCO > VSXY | BK > BNY | GREE > VIP | FDP > DMC | MWS > PMA | WAI > DPU | UBXG > MF What it didn't catch: LC > HAPN | SCVL > SHOE | SATS > ECHO | LMFA > PWCM | CEPT > SECZ | WTO > FXHO | SKLZ > FIRY | TBH > HODO | SSSS > NSLR | ATLN > CIRC | RAAQ > IQMX | APM > NIKI | ODV > OGG | YOOV > CIGL | PCSC > FRNM | ALBT > CHGA After looking at it carefully over past week or so, I found that I could implement a solution with Denali with the following process: 1. There are some fairly priced corporate-actions data sources that provide ticker symbol changes full history with the effective dates. It's checked on a regular weekly schedule for future purposes too. 2. Automatically detect ticker reuse before merging. Some new symbols were previously used by a completely unrelated stock. I have a Python script that checks whether the "new" symbol's existing local history actually lines up with the company being merged in. It flags suspicious price discontinuities or date gaps at the boundary that indicate leftover data from a prior, unrelated use of that ticker. If flagged, that old legacy data is trimmed out before merging; skipping this check would otherwise produce a fake discontinuity in the combined series. 3. Script the daily (.dly) merge. Once a rename is confirmed, a Python script splices the old symbol's daily records (up to the effective date) with the new symbol's records (from the effective date forward) automatically, across however many pairs need it in one run, rather than doing it per-symbol by hand. 4. Script the intraday (.scid) merge too. The format is simple enough (fixed header + fixed-size records) that I have a Python script to do the same merge Sierra Chart's own "Join Intraday Data" performs, but applied automatically across a whole batch of pairs instead of one at a time through the GUI. This is what makes handling this at scale (hundreds/thousands of symbols) practical instead of a heavy manual burden. 5. Always back up before merging, and run every merge script in a dry-run/report mode first so record counts and flags can be reviewed before anything is applied. 6. Run an automated sanity check after merging: any unexpected multi-day gap in the combined series gets flagged for manual review rather than trusted blindly. 7. Only mark a rename resolved after visually confirming the chart. No artificial spike or gap at the boundary, for both the daily and intraday series. Regarding my deleted post about the splits, I realized that on one of my installations, I wasn't using the Denali US Consolidated feed. I tested it again with the proper feed and the splits look like they work correctly and are in fact automatically done on the server. That looked like an impossible situation but it's fine and I don't think there are any issues with that other than in 0.00001% of cases like with some lower priced stocks. All in all, I think staying on your Denali US Consolidated Feed is better than the Bar Chart feed, as the ticker symbol changes will be dealt accordingly through rename detection/splicing with my automated Python pipeline, driven by a data source that lists the ticker changes. Would be great if a future update could integrate a proper ticker change feed directly, since that's the actual gap and not the price/split data itself. Date Time Of Last Edit: 2026-08-18 05:45:48
|