One of the oldest rules in technical analysis says that when the MACD line crosses below its signal line, it is time to sell. For some time now Sahmino has been filing setups of exactly this kind across the whole Tehran board every night as "claims": each claim carries a target level and a level that, if touched first, proves it wrong, and on later nights every claim that has answered is settled. As of Wednesday, 16 September 2026 (25 Shahrivar 1405), that ledger holds a number for the famous rule that does not match the common assumption: in market weather we file as rising and low-volatility, only 10 of the 189 settled bearish MACD-cross claims reached their level, or 5.3%.
Background
The rule this number measures is not our claim, and it is quoted from its own textbook source. Investopedia's MACD entry, read on 16 September 2026, states that traders may buy the security when the MACD line crosses above the signal line and sell, or short, the security when the MACD line crosses below the signal line. The same entry notes that Gerald Appel developed the indicator in the 1970s, and it carries its own warning that a moving-average divergence produces "numerous false positives" along the way. So the question is not whether anyone claimed anything strange. The question is how often, in one specific market over one specific stretch, the rule turned out to be right.
That stretch coincides with one of the fastest runs in the Tehran board's history. On our own recorded series, the Tehran Stock Exchange's main index (شاخص کل, TEDPIX), which closed at 6,069,888 points on 23 August 2026 (1 Shahrivar 1405), closed at 7,557,642 points on Wednesday 16 September (recorded 13:24 that day), 24.5% higher in under a month. The same day the Farabourse main index closed at 58,193 points and the equal-weight index (شاخص هموزن) at 2,016,173. This "weather" is precisely what every claim is filed under, and why that filing turns out to be the most important part of the story.
The numbers
The scorecard counts only claims that have settled: they either reached their target first, or hit the level that proves them wrong first, or ran out of sessions. Buckets holding fewer than 30 settled claims publish no rate at all, because the only honest sentence about them is that not enough of them have settled yet. Everything below sits above that floor, all dated 16 September 2026.
In rising, low-volatility weather, bullish claims have 1,688 settlements between them, of which 1,481 reached their level: 87.7%. Bearish claims in the same weather have 721 settlements and 111 hits: 15.4%. Together that is 2,409 settled claims.
The point is that the gap has no exceptions. All seven bullish detectors above the floor sit between 68.1% and 100%: flow turn 90.3% of 601 settlements, range breakout 87.2% of 569, bullish MACD cross 91% of 156, sector rotation 98.7% of 78, liquidity shift 86.1% of 108, news pressure 100% of 32, and asset trend 68.1% of 144. Against them, all five bearish detectors stand between zero and 29.1%: bearish sector rotation zero of 31 settlements, bearish MACD cross 5.3% of 189, bearish flow turn 18.2% of 154, momentum reversal 19.5% of 292, and bearish asset trend 29.1% of 55.
Put the two MACD buckets side by side: same detector, same weather, only the direction of the claim differs. Upward, 91%. Downward, 5.3%.
Drivers
The simplest explanation is the explanation. Over the span of this scorecard, the thing doing more work than any detector was the market itself. When a board rises 24.5% in under a month, any claim whose target sits above today's price has a high chance of being touched, and any claim whose target sits below it has a low one. So that 5.3% says little about MACD in the first instance. It says something about the direction of a claim in a rising market. The distance between 91% and 5.3% across two buckets of one detector is exactly what makes this measurable: if the tool itself were doing the work, both of its directions should have worked.
This is also why the table keeps the weather separate. Take one detector in two different kinds of weather: bullish "asset trend" reached its level in 68.1% of 144 settlements in rising, low-volatility weather, and the same bullish detector reached it in 17.1% of 82 settlements in falling, high-volatility weather. Average those two together and you get a figure that describes neither reality. The rule of thumb that "this pattern works such-and-such a percent of the time" comes from exactly that averaging.
Outlook
This scorecard is a record of the past, not a forecast. No claim in it was built after the fact and every claim was filed before it settled, but everything in it concerns something that has already answered. If the market's weather changes, the right expectation is that these same buckets change, and that the falling, high-volatility buckets, which hold thin samples today, gradually fill. None of the figures above says any share will go up or down, and none should be read that way.
Conclusion
In our own record to 16 September 2026, the bearish MACD cross produced just 10 hits out of 189 settled claims in a rising, low-volatility market. What matters is that the weakness is not specific to MACD: all five bearish families stayed below 30% together and all seven bullish families stayed above 68% together, and that simultaneity shows what was measured was the market's direction rather than any tool's intelligence. If one thing survives from this report, it is this: do not read a signal without its weather. Our reading is that the share of bearish claims reaching their level stays below 30%, as long as the market weather on file stays rising and low-volatility, with a horizon to the end of Mehr 1405 (roughly 22 October 2026); and if that share crosses 30% before the end of Mehr, this reading is wrong.
What to watch
- The filling of the falling, high-volatility buckets: only two of them sit above the 30-settlement floor today.
- The buckets that still carry no rate, among them the moving-average cross and the filing reaction, each holding fewer than 30 settlements.
- The distance between the main index and the equal-weight index, which moved in opposite directions on 16 September and is a plain gauge of that same weather. The latest readings update on Sahmino's prices page.
- The record of symbols reopening after a long halt, such as Mobarakeh Steel after 103 sessions without a trade, whose board claims settle later than most.
- Which side was buying across consecutive sessions, tracked in Sahmino's market coverage.
