EquitiesIn plain sightTuesday 3 November 2026, 10:02
Down Friday, down Monday: the S&P 500 fell on both days, an old warning sign. It has come 11 times a year since 1953, and the month after looks normal.
The S&P 500 fell 1.82% on Friday and another 0.51% on Monday, to close at 7,541.40, CNBC’s quote shows. Traders have a name for that pair: a down Friday, down Monday. The Stock Trader’s Almanac has long read it as a warning, and says the pattern has a “bearish bias sometime during the next 90 calendar days”. With Americans voting in the midterm elections on Tuesday, it has been passed around as a sign of trouble.
The record makes it a weak sign at best. The S&P 500 has given the down Friday, down Monday signal 825 times since 1953, about 11 times a year, by our count from S&P Dow Jones Indices data. Twenty trading days later, the index was higher after 58% of them. After any Monday, it was higher 62% of the time. The gap is small, and since 2000 it has gone the other way.
The idea is simple. A fall on Friday says traders did not want to hold shares over the weekend. A fall again on Monday says the weekend did not change their minds. The Almanac tracks the Dow Jones Industrial Average and counts the first and last trading days of each week. We used the S&P 500, the broader index, and Fridays and Mondays only. Our count starts in 1953.
The Almanac also says the pattern is most common, and does most damage, in bear market years. That part holds. Bad years bring more signals: 24 in 1957, 23 in both 1973 and 1974, and 16 in 2022, all years in which the index fell. Calm years bring few: four in 2010 and four in 2023. But a count that rises in a bad year shows that a bad year is under way. It does not show that one is coming.
We tested the Almanac’s 90-day window on the S&P 500 as 63 trading days, about three months. After a signal, the index fell 5% or more at some point in those three months 34% of the time. After any Monday, it did so 32% of the time. A fall of 10% or more followed 14% of signals, against 11% of all Mondays.
So the warning adds two to three percentage points to the odds of a sell-off, too little to trade on. Three months after a signal, the index was 2.0% higher on average. Three months after any Monday, it was 2.3% higher.
Since 2000 the month after a signal has been better than usual. The index rose 1.45% on average in the 20 trading days after one, against 0.68% after any Monday, and it was higher 65% of the time. In this century, two days of selling have more often marked a dip than a top.
Monday was the sixth signal of 2026 in the S&P 500. The first five came on 30 March, 18 May, 20 July, 17 August and 31 August. After each of them, the index set a new record close within 5 to 35 trading days. The largest rebound followed the March signal, which came with the index 9.1% below its record. Twenty trading days later it was 12.5% higher.
This one starts 3.5% below the record of 7,818.93, set on 6 October, according to FRED. Friday’s fall followed the results rout in big technology shares that we reported on 30 October. Even after two down days, the index is up 10.2% this year.
Midterm years have their own record. Before this one, the S&P 500 gave 32 down Friday, down Monday signals in the October and November of a midterm year since 1954. Fifty trading days later, the index was higher after 24 of them.
Four of the eight losses came in 2018, when the index slid into December. The others came in 1974, 1978, 1994 and 2002. As we wrote on Monday, the market has been higher a year after every midterm since 1946. The signal does not change that record, and 2018 shows the path can still be rough.
A down Friday and a down Monday describe two bad days. On more than 70 years of data, they say little about the next month. What will move the index from here is the news: the election count on Tuesday night, the US jobs report on Friday 6 November and consumer prices on Tuesday 10 November, both at 8:30 in New York, 14:30 in Tirana, as listed in our economic calendar. A rule that fires 11 times a year will fire again before the year is out. Most times, nothing much will follow.
This is analysis and opinion, not investment advice.
ContactWrite to the Notebook
Spotted an error, or a source this piece should cite? Want your own research linked or featured? Tell us. We read every message.
Our research · 52 min read
The lek has not bought so many euros since at least 1995, when Eurostat's records begin. In September 2026 a euro cost 91.85 lek on average, the lowest monthly figure in those records. In 2019 it cost 123.01. Put the other way round, the lek now buys 33.9% more euros than it did in 2019, and 37.7% more dollars.
Continue reading the story