Audit
Generated from the same release the rest of the site readsThis release
| Version | topics-2026-10-14-signal-following |
|---|---|
| Classifier version | no-timeout-classifier-1 |
| Released | 14 Oct 2026 |
| Rail | Replay |
| Current rule version | R6_reparam |
| Schema version | public-replay-9 |
The contract
- Entry. 8-slice TWAP over 25 minutes.
- Stop loss. A position closes at a loss of 3.50%.
- Trail arms once price moves 5.00% in the call's favour.
- Trail offset. Once armed, the trailing stop sits 2.00% behind the best price reached.
- Terminal timeout. None — a position closes only at its stop or
its trailing stop. The contract's own
terminal_timeout_hoursfield is still carried by this release and is deliberately not shown: a released contract is no-timeout, so printing that field would state a rule no result used.
| Desks | Fees | Slippage | Funding / day |
|---|---|---|---|
| Crypto Futures Signal, Forex Signal, Commodity & Index | 0.07% | 0.10% | 0.03% |
| US Stock Signal, Commodity & Index | 0.05% | 0.10% | 0.00% |
| ID Stock Second Liner | 0.20% | 0.30% | 0.00% |
| ID Stock TOP 90 Days Volume, ID Stock Bullish Trend Swing | 0.20% | 0.25% | 0.00% |
| ID Stock KONGLO ONLY | 0.20% | 0.15% | 0.00% |
| Forex Signal | 0.00% | 0.01% | 0.00% |
| ID Stock FCA & MICIN | 0.20% | 0.40% | 0.00% |
Methods and coverage
The headline column is the ladder: 5-bid ladder (ours, not Solagin's)
(five-bid-ladder-exit-avg-1).
The publisher's own exit rule is computed alongside it over the same alerts, so the
two can be compared rather than taken on trust.
| Method | Published method | 5-bid ladder (ours, not Solagin's)headline |
|---|---|---|
| Win rate | 37.0% | 52.0% |
| EV per trade | -8.72 | -4.58 |
| Profit factor | 0.69 | 0.72 |
| Finished trades | 6,980 | 5,787 |
| Both columns are on a 1,000-unit budget per alert, so EV per trade reads on the same scale on each side. one 1,000-unit budget per alert, five bids weighted 1:1:2:2:4. | ||
Over the 5764 alerts where both methods produced a result — the only alerts the two columns can be compared on — the ladder's win rate is +14.7 points, its EV per trade +5.48 and its profit factor +0.07 against the published method.
| Filled | 6164 | A bid was reached and capital was committed. |
|---|---|---|
| Never touched | 0 | No bid was reached before the bars ran out; no capital was committed. |
| Cancelled | 130 | Price ran in the call's favour first, so the unfilled bids were cancelled with no capital committed. |
| No fill (total) | 1171 | Every unfilled row, whichever reason applies. |
How much of the corpus this is
| Desks | 9 |
|---|---|
| Corpus signals | 8323 |
| Scoreable | 7838 |
| Released | 8238 |
| Scored on the default rule | 1689 |
| Released but unparsed | 1 |
| Withdrawn by Solagin | 4 |
Released trades are a subset of the corpus, so a win rate on this site is not the win rate of everything Solagin has ever published.
Quote flags
This release states quote_executable and rule_applicable on
8238 record(s). A call whose quote could not be transacted at is
excluded from every figure on this site.
Corrections
A published result is corrected when an audit proves it wrong. Nothing is removed silently: the old value stays visible on the trade it belongs to, and the corrected value is the one counted here.
1 correction in this release.
| Date | Correction | Rows | Summary |
|---|---|---|---|
| 24 Sep 2026 | correction-2026-09-24 | 56 | 56 crypto futures results recomputed after we found gaps in the price history they were calculated from: 15 changed between win and loss, 41 kept their result but closed at a different time or price. |
Withdrawals
A withdrawal is the publisher's act, not ours. A withdrawn alert is not counted in any result and is left out of every figure on this site; it stays listed so the deletion is on the record rather than quietly gone.
4 withdrawn alerts in this release.
| Symbol | Desk | Withdrawn | Source |
|---|---|---|---|
| BULL.JK | ID Stock KONGLO ONLY | 12 Oct 2026 11:10 WIB | live_event |
| GTSI.JK | ID Stock Second Liner | 08 Oct 2026 22:42 WIB | live_event |
| CCSI.JK | ID Stock Second Liner | — | manual |
| ACRO.JK | ID Stock Second Liner | — | manual |
Regimes
| Rule version | Regime | Current | Signals | Trades | Win rate |
|---|---|---|---|---|---|
| R0 | Formation (long-only, ML-alert/zones) no published exit rule: scored on our default rule | No | 851 | 0 | — |
| R1 | Vocab expand (LONG/SHORT ZONE) no published exit rule: scored on our default rule | No | 20 | 0 | — |
| R2 | Majority-SHORT (CICIL, shorts) no published exit rule: scored on our default rule | No | 55 | 0 | — |
| R3 | Feed restructure (.P, 1000x, SPLIT5) | No | 14 | 1 | 0.0% |
| R4 | Format shift (bare dir, ENTRY DI HARGA) | No | 228 | 220 | 48.6% |
| R5 | Post-maintenance reparam (inline exit recipe) | No | 1405 | 1339 | 59.7% |
| R6 | Full reparam (CRYPTO PROBABILITY template) | Yes | 1419 | 1355 | 50.7% |
Win rate and trades are read from the headline column (ladder), so this table moves with the rest of the site. A regime with no finished trade in that column shows a dash rather than a zero: nothing was measured, which is not the same as measuring nothing.
What is not here
- This release is a static file with no private plane in it: no ledger, no Telegram archive, no internal endpoints.
- The vendor's raw candle data is never included in the bundle.
- The per-signal pinned input hashes are shown on each trade's own page, not collected here.
- Two populations are not performance claims and live on their own pages: the calls whose quote could not be transacted at (Indices) and the calls on a quoted yield rather than a price (Yields). Both are read from the same release as this page.
This page is generated from the same release the rest of the site reads, so it can be checked against it.