Correction
Premium morning brief, 14 May to 30 July 2026
Published 30 July 2026.
On 16 dates between 14 May and 30 July 2026, the EHIQ Premium morning brief published the BTC Combined Timer at EMERGENCY, with a recommended position of 0%, when the correct reading of that timer was SELL at 10%.
Every divergence ran in the same direction. On each of those dates the brief advised holding less BTC than the system it reports on actually held. At no point did it advise holding more.
What went wrong
The timer takes two inputs: a price regime, and a macro composite built from Fed policy, inflation, gold, oil and mortgage data. The morning brief was passing a different quantity into the macro slot: the sensor structural-risk index.
Both are integers from 0 to 100, and both are called “score” by the systems that produce them. Nothing raised an error. The brief was internally consistent and wrong.
Because the substituted value was consistently lower than the real macro composite, it landed on the wrong side of a threshold and produced the more severe signal.
The corrected record
Regime was STRESS on all 123 delivery dates in the window. These are the 16 where the substitution changed the published signal.
| Date | Real macro | Value used | Published | Correct | Evidence |
|---|---|---|---|---|---|
| 2026-05-14 | 60 | 20.7 | EMERGENCY / 0% | SELL / 10% | inferred |
| 2026-05-16 | 60 | 20.0 | EMERGENCY / 0% | SELL / 10% | inferred |
| 2026-06-09 | 62 | 27.7 | EMERGENCY / 0% | SELL / 10% | inferred |
| 2026-07-14 | 60 | 24.4 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-16 | 62 | 23.2 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-17 | 62 | 23.3 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-18 | 62 | 21.6 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-21 | 75 | 21.9 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-22 | 75 | 20.9 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-23 | 68 | 22.6 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-24 | 68 | 24.2 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-25 | 68 | 23.5 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-27 | 68 | 22.2 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-28 | 68 | 22.7 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-29 | 68 | 22.7 | EMERGENCY / 0% | SELL / 10% | proven |
| 2026-07-30 | 68 | 23.3 | EMERGENCY / 0% | SELL / 10% | proven |
Proven means the defective line is present in a surviving dated backup of the code for that date. Inferred means the date predates every surviving copy; the value is recomputed from the historical inputs and the divergence follows, but the code as it stood cannot be produced. Three dates are inferred and are marked as such rather than presented as established.
What we cannot tell you
We cannot verify whether any individual brief was delivered to, opened by, or read by any particular subscriber. Our records establish who was paying on each date, not who was reading. Any statement we make about reach carries that limit.
67 paying subscribers were active on at least one affected date, measured from subscription timestamps. Delivery and readership are not verified. We are not claiming that everyone affected saw these briefs, and we are not claiming the reverse.
What changed
Found and fixed on 30 July 2026. Both the BTC and ETH timers now read a single canonical source rather than assembling their own inputs, and that source refuses to return a number at all when the underlying data is stale or unavailable. Where it previously substituted a plausible default, it now publishes nothing and says so.
The same defect existed on a second, internal surface and was fixed the same day. An audit of every other caller of the timer found no third instance.
Questions about this correction, or about your own subscription during this window, go to eric@eventhorizoniq.com. A person reads that address.