The S&P 500 Set a Record With More NYSE Lows Than Highs. The Lows Are Where Rising Yields Hurt

Research note. Computed from S&P 500 daily closes, the 10-year Treasury yield and the daily bars of every NYSE-listed issue in our price database, frozen at the October 6, 2026 close. The methods are in How we checked it below; our editorial and AI standards are in How we use AI.

• ·The headline ratio. 69 NYSE issues made 52-week lows and 37 made highs, a ratio of 1.86. Of 470 S&P 500 record closes since March 2013, the previous highest was 1.15, on May 13, 2026.
• ·Who made the lows. Funds, structured notes and REITs made 47 of them, 68%. On a typical record day with ten or more new lows, they make 8%.
• ·The stocks, on the day. NYSE common stocks and ADRs made 35 new highs and 22 new lows, weaker than on most record days (17th of 470) but nothing like the headline ratio.
• ·The stocks, over four weeks. Lows outnumbered highs among NYSE stocks on 18 of the 20 sessions into the record. 300 of 1,635 hit a 52-week low, led by utilities (45% of them) and consumer companies; 150 hit a high, led by energy.
• ·The precedents are mixed. Records on which rate-sensitive issues made most of the lows came in clusters in 2013 and 2018. The S&P 500 kept rising after 2013; both 2018 cases fell more than 10% within months.

The S&P 500 closed at 7,818.93 on Tuesday, a record, and on the same day the New York Stock Exchange produced almost twice as many 52-week lows as highs. On the face of it that is a market making new highs on a shrinking base, the kind of divergence that turns up before tops. On 470 record closes since March 2013, new lows had outnumbered new highs only four times before, in June 2024 and May 2026, and never by more than 1.15 to one.

The count deserves a closer look before it carries a conclusion. The NYSE is not only a stock exchange: 18% of the issues listed there in our database are closed-end funds, exchange-traded funds and structured notes, and another 131 are real estate investment trusts. Most of them exist to pay an income. When bond yields rise, they fall together.

Who made the new lows?

On Tuesday 26 of the 69 new lows were funds, 8 were structured notes and 13 were REITs. Together that is 47, or 68%, from securities that make up 24% of the NYSE issues we track. On a typical record day with ten or more new lows, the same groups make 8% of them. The other 22 lows came from operating companies, against 35 new highs.

That makes the headline ratio mostly a reading of the bond market. The 10-year Treasury yield reached 5.29% on September 30, 2026, its highest since May 2002, and an income fund paying a fixed coupon is worth less when new bonds pay more. On the day itself, the stocks were weaker than on a typical record day, 22 lows to 35 highs, but nowhere near the headline ratio. The day flatters them, though, because the index rallied into it.

The four weeks before the record

From September 9, 2026 to the record, NYSE stocks made more new lows than new highs on 18 of 20 sessions, while the S&P 500 never closed more than 3.2% below its record. 300 of the 1,635 common stocks and ADRs we track hit a 52-week low at least once; 150 hit a high. The lows were not spread evenly. 45% of utilities hit one, and so did 30% of consumer cyclical companies and 26% of consumer staples, which include the packaged-food makers. Utilities and food companies pay steady dividends and trade partly like bonds; consumer companies feel higher borrowing costs and, this year, higher fuel prices. The highs came from the other side: 20% of energy stocks made one, and by count industrial stocks made the most highs (38), ahead of technology (18) and health care (15).

Has this happened before?

Records on which rate-sensitive issues made at least half of ten or more NYSE new lows are not new. They came in clusters: in July and November 2013, as the Federal Reserve prepared to slow its bond buying, and in January and September 2018, as the 10-year yield climbed from about 2.4% to above 3%. Tuesday’s 47 rate-sensitive lows are the most on any record day in our data; the previous high was 35, on September 20, 2018.

The seven completed episodes went both ways. A quarter later the S&P 500 was higher in four of them, and a year later in five of seven. The cleanest match by mix, July 2013, when rate-sensitive issues made 21 of 22 lows, was followed by a +17.0% year. The closest match by size, September 2018, with 35 rate-sensitive lows, was followed by the fourth-quarter 2018 decline, -19.8% at its lowest close. Three of the seven fell more than 10% within a year. Seven episodes in 13 years are a sample of the rates cycles that happened to occur, and they cannot tell a rate scare that passes from one that does not.

Records during a bond sell-off, since 1963

The NYSE counts only go back to 2011, and they come from today’s listed issues. The rates half of the question can be asked further back without that problem, from the index and the Treasury market alone. Since 1963 the S&P 500 has set 1,183 record closes; on 69 of them the 10-year yield had touched its 52-week high within the prior week. Grouped the same way, that is 11 episodes, 10 with a complete year.

The pattern is a slightly weaker version of an ordinary record. A year later the index was up a median +8.2% after these episodes, against +10.8% after all records, and higher in seven of ten. Over the first quarter the median was +0.9%, against +2.4%. The three that were lower a year later began in September 1965, December 1999 and January 2018. December 1999 is the one that preceded a lasting top. The current episode began on May 26, 2026, with the yield at 4.50%, and Tuesday’s record falls inside it.

What I would watch

I would not sell stocks because the S&P 500 set a record with more NYSE lows than highs. Most of those lows are a statement about bond yields, and the stocks making them are the ones that suffer when yields and fuel prices rise. I would not dismiss it either: an index carried by energy, industrial and technology stocks while utilities and consumer companies slide is narrower than its record suggests. Records set while yields climbed have been followed by a weaker year than usual, and both 2018 episodes in the recent record ended in declines of more than 10%.

• ·The leaders joining the lows. If technology and industrial stocks, which made most of the recent highs, start making new lows, the weakness has moved from the rate-sensitive corners into the stocks carrying the index, and I would treat it as a breadth warning rather than a rotation. The daily counts are on New Highs – New Lows.
• ·The 10-year above 5.29%. A new high in the yield would extend the pressure on everything that pays an income and on the valuations of long-duration stocks. A fall back below 5% would take much of it away. The yield is on Treasury Yields.
• ·Credit joining in. If high-yield spreads widen while yields rise, the move is about default risk as well as rates. Spreads are on Credit Spreads.

How we checked it

Every number here comes from S&P 500 daily closes, the 10-year Treasury yield and our own daily bars for NYSE-listed issues. In plain terms:

• ·A record close is an S&P 500 close above every earlier close, from ^GSPC daily closes since 1927. There are 470 since March 2013 with NYSE counts.
• ·New highs and lows compare each issue’s intraday high and low with the highest high and lowest low of the prior 252 sessions, for the 2,223 NYSE-listed issues in our price database, as the New Highs – New Lows page does.
• ·Groups use each issue’s security type from Massive reference data and, for common stocks, its SEC industry code: funds (closed-end and exchange-traded), structured notes, preferred shares, REITs, ADRs and other common stocks. The first four are the rate-sensitive group. Our database carries few NYSE preferred shares, so an all-issue count from another source, which includes hundreds of them, would show more rate-sensitive lows than ours.
• ·Survivorship. The NYSE counts use issues listed today and today’s classifications, so issues that delisted are missing from past readings, and past lows are understated. The rankings since 2013 belong to that reconstruction. The 1963 study uses only the index and Treasury yields and has no such gap.
• ·Episodes group qualifying record days so that a new one starts after 91 calendar days without one. Returns are S&P 500 price changes without dividends, graded only where the window is complete. “Worst within a year” is the lowest close of the next 252 sessions against the record close; it is not a peak-to-trough drawdown.
• ·Frozen. The exhibits render from a snapshot saved at publication (Oct 6, 2026) with a SHA-256 fingerprint of each of its 4 input files, available below. The box at the top reads current data.