BTC indicator alerts
中文An email when one of the published conditions is met. They are computed from the same data as the indicator page, with the same parameters printed on it.
What triggers an email
- Price closes below its 200-week average — the arithmetic mean of the last 1,400 daily closes. Over the 12.3 years to 2026-08-27 that happened on 380 of 4,486 days (8.5%), in 27 stretches; the longest ran 210 days.
- AHR999 falls under 0.4 — price against the 200-day DCA cost, times price against the power-law fit. 169 of 4,486 days (3.8%). The band deepens fast, and the email prints the whole ladder: under 0.45 was 408 days, under 0.40 is 169, under 0.35 is 25, under 0.30 is 3. The lowest reading on record is 0.231, on 2015-01-14.
- Both at once — the rarer case, 108 of 4,486 days (2.4%). The email says so, and says why the two are different measurements rather than one counted twice.
- 1,050 days past the last confirmed cycle low — the one sell-side condition, and it is a calendar count, not a valuation. The three completed cycles ran 1,067, 1,059 and 1,062 days from bear low to bull high: 2015-01-14 → 2017-12-16, 2018-12-15 → 2021-11-08, 2022-11-09 → 2025-10-06. One email a day while the window is open (day 1,050 to 1,075) — so 26 emails per window, 78 in 12.3 years.
- The Fear & Greed index reaches 70, and again at 85 — the alternative.me index, 0 to 100, published once a day. Across its 3,132 daily readings from 2018-02-01 to 2026-09-02 it sat at 70 or above on 645 days (20.6%) and at 85 or above on 95 days (3.0%). One email when the index steps up into a level; nothing on the way back down.
- The Fear & Greed index falls below 25 — 679 of those same 3,132 days (21.7%). One email when it first goes below 25, and the next one only after it has been back at 30 or above.
One email per day at most on the buy-side conditions, whichever hold. The job checks hourly, so a crossing is picked up within the hour rather than the next day. Every email states which signal fired, the readings behind it, and links to the backtest.
Replayed over the whole index history, the two Fear & Greed levels on the greed side come to 48 emails and the fear level to 53 — 101 in 8.6 years, about twelve a year. The four rules (buy-side conditions, cycle window, greed side, fear side) each cap themselves at one email a day, so a single day can carry more than one.
Why the sell side is a calendar and not a valuation
Because every valuation reading that marks a top has fallen each cycle, so a threshold fitted to past tops stops firing. Measured on 5,883 days of daily closes:
| At the cycle top | 2013 | 2017 | 2021 | 2025 |
|---|---|---|---|---|
| Price / 200-week average | — | 15.78x | 3.79x | 2.34x |
| MVRV | 5.06 | 4.25 | 2.72 | 2.29 |
| Price / power-law fair value | 10.81 | 6.46 | 2.89 | 1.26 |
So: AHR999 above 4 fired every cycle up to 2021 and has not fired since. Ranking a reading against its whole history puts 2024–2026 at zero, because the 2013 and 2017 extremes stay in the sample. Ranking against a trailing two years fires in every cycle but also called it in December 2023 at $41,910 and March 2024 at $62,524, before price reached $124,824. A 30% drawdown from the trailing one-year high held in 4 of 8 cases — a coin flip.
The calendar count has its own weakness, stated on the same terms: three observations, one of them out-of-sample, and the low it counts from is only identifiable in hindsight. The email says all of that.
The window itself has contained the actual high in all three cycles. What it does not tell you is which day inside it. In 2017 the window opened at $9,800 and the high came seventeen days later at $19,641 — opening day was the cheapest day of the whole window. In 2021 and 2025 opening day sat about 9% below the window high. So the alert cannot tell you which day inside the window to act; it tells you the window is open. That is why it sends one email a day for as long as the window is open, rather than once — 26 emails per window, 78 in 12.3 years. From day 1,059 the email says so, because that is the earliest day any past cycle topped.
How many emails that actually is
Replayed over 12.3 years of daily closes, through the same rule the service runs: 441 emails in total — but they are not spread evenly.
| Year | Emails | |
|---|---|---|
| 2014 | 0 | from 17 May, when the 200-week average first has 1,400 closes behind it |
| 2015 | 65 | the tail of that bear market |
| 2016 | 0 | |
| 2017 | 0 | |
| 2018 | 13 | |
| 2019 | 8 | |
| 2020 | 8 | the March crash |
| 2021 | 0 | |
| 2022 | 175 | mid-June to end of December, almost daily |
| 2023 | 143 | two long stretches below the 200-week line |
| 2024 | 0 | |
| 2025 | 0 | |
| 2026 | 29 | to 27 August, in six separate stretches |
In a deep bear market this arrives almost every day for months. In a bull market it can be silent for a year or more. If that cadence is not what you want, this is the wrong product — better you know now than after the twentieth email.
Subscribe
Payment is handled by Waffo, which is the merchant of record — it collects and remits tax, and we never see your card. The email address on the receipt is the one alerts go to.
The email language can be changed later by replying to any alert. Every email carries a one-click unsubscribe — your mail client shows it as an Unsubscribe button at the top. It stops the alerts without touching the subscription itself, so cancel the subscription separately if you want the billing to stop.
What this is not
It is not advice, and it is not a signal service. Each email reports a reading, the parameters behind it, and how often that reading has occurred — nothing about what you should do. The thresholds are conventions other people published, not measurements: 0.4 sits inside the band AHR999's author calls deeply undervalued (he draws its top at 0.45), and the 200-week average is a trend line, not a floor. Price has closed below it on 8.5% of days, and once for 210 days running.
The sell-side condition is a day count, not a valuation, and it rests on three observations with one of them out-of-sample. It cannot tell you which day inside its window the high arrives, and if a cycle stretches it is simply wrong. Read it as one fact among several, not as an instruction.
The two greed levels, 70 and 85, are not published bands. alternative.me calls anything from 45 up greed and anything from 75 up extreme greed; 70 and 85 are the levels this site chose to send on, printed here so you can see them before you subscribe. The fear level, below 25, is alternative.me's own extreme-fear boundary.
How much to buy or sell, and whether to act at all, is your decision. The position calculator is free and runs in your browser.