Pre-Market Briefing on Sept 24: Fastly Jumps 13.69% on 2029 Target From Investor Day, and Jobless Claims Are Due Today

All three US indexes closed lower on September 23. The Nasdaq fell 1.13% to 26,936.04, one session after a record close of 27,244.28. The trigger was the September flash composite PMI at 58.4, with input costs rising at their fastest in four years. The same day Fastly (FSLY) rose 13.69% to $29.65, while application software averaged −0.85% and Nasdaq technology −0.82%: a move against its sector, not a sector leading higher; it came from its own Investor Day. The 5-year Treasury yield closed at 4.99% after touching roughly 5.04% intraday. Initial jobless claims come at 12:30 today, before the open, the first claims week covering the September 16 hike. All times are UTC.

 

The unit is the single-day change on September 23 against the prior close; three bars are index moves, three are sector averages of Nasdaq-listed names, and all six share one scale around a zero line. The Nasdaq closed at 26,936.04, down 1.13%; the S&P 500 at 7,708.55, down 0.72%; the Dow at 51,511.59, down 0.68%. On a sector-average basis only materials (+1.11%), consumer staples (+0.22%) and utilities (+0.18%) rose; the largest declines were industrials (−1.85%), real estate (−1.79%) and communication services (−1.69%).

 

The trigger was the September flash composite PMI at 58.4, above 56.0 in August, with services at 58.7 and manufacturing at 57.0 also above prior readings and input costs rising at their fastest in four years. The stronger the survey, the more the market believes the Fed has room to lean on inflation, so the discount rate rises and assets priced off distant cash flow are marked down first. What fell that day was not earnings, it was the discount rate.

 

The five-dimension score runs 0 to 100 against peers and the company's own one-year history, as of the September 23 close; the radar centre is 0, the rim 100, and all five axes share one scale. Trend position scores 80.9, peer strength 100, peer rank 100, sector valuation 99.2 and volatility control 100. The weakest is trend position, and ⚠️ nothing falls below 50.

 

Read the shape with its basis: it measures a one-year window, not this session. The price sits at 80% of the 52-week range, at the upper end, which is why trend position is the most conservative of the five. The stock closed at $29.65 with a market value of $4.6B, up $560M in a day, on turnover of $890M.

 

The unit is again the single-day change on September 23 against the prior close. The six are grouped by hand by what they do (edge cloud / content delivery / cloud security / observability) across two industries, application software and systems software; the chart has a zero line and one shared scale. Fastly is top with +13.69% and Cloudflare (NET), also edge cloud, is bottom with −0.14%, a spread of 13.83 points. Fastly finished 8.69 points above second place, and its excess gain is 13.69 − (−0.85) = 14.54 points.

 

The two closest peers barely moved: Cloudflare (NET) at −0.14% and Akamai (AKAM) at +0.08%. Had the whole edge-cloud group been repriced, those two would not sit on the zero line, so this came from the company's own Investor Day. ⚠️ Still, this is not one stock rising alone: Palo Alto (PANW) at +5.00% and CrowdStrike (CRWD) at +4.97% did rise, on a rotation into cyber-security, which is a different trade.

 

The unit is millions of shares, and the volume multiple is the day's volume divided by average daily volume; both bars are drawn from zero on one scale. The session traded 29.91M shares against a usual 6.16M, a multiple of 4.85x. The multiple carries no direction; it only says how many trades tested that price. At 0.8x a 13.69% gain is usually a few orders lifting it; at 4.85x a lot of stock genuinely changed hands there. ⚠️ It hangs on volume, not turnover.

 

This one is three cards side by side, with no scale and no zero line; all three numbers are target ranges for 2029. At its Investor Day the company set out 2029 revenue of $1.1B–$1.3B, an operating margin of 20%–22% and compound annual growth of 14%–21%. The same event launched three products (AI runtime controls, an AI firewall and API security), and the company said machine-generated traffic already accounts for more than half of its own network. ⚠️ These are long-term 2029 targets, not quarterly guidance; the quarter-by-quarter path only shows up in later earnings reports. An Investor Day lays out the books four years out while the market pays today's price; what matters is whether basis, product support and time frame line up.

 

The coin ratio (%) is disclosed coin value divided by market value; prices are taken at the September 23, 2026 close and holdings are as last disclosed. The eight bars are drawn from zero on one shared scale, with a dashed line at 100%. The lowest is MARA Holdings (MARA) at 58.4%, the highest DeFi Development (DFDV) at 157.7%; the middle six are left to the chart. Four coins have two entry points each: bitcoin through MSTR and MARA, ether through BMNR and SBET, Solana through FWDI and DFDV, Zcash through CYPH and ZCSH.

 

Six of the eight sit above 100%: the coins each discloses are worth more than the whole company. ⚠️ Above 100% is not free money: it is the price the market puts on share issuance, debt and forced selling at a low. The low end needs reading too: MARA is at 58.4% because it also owns rigs and power assets, which does not make its link looser.

 

 

It started out in product design and is now the largest Solana treasury in US equities by coins held, with roughly 8.16 million SOL as of September 21, 2026, about 1.39% of the circulating supply, staked for yield. ⚠️ The risk sits where the appeal does: a ratio of 157.4% looks like a discount, but the company keeps buying coins by issuing shares, so when the coin price falls the discount and the dilution arrive together. Three steps judge how tightly such a stock is tied to its coin: work out the ratio, ask where the coins came from (cash, issuance or debt), and ask what else it owns.

 

The unit is thousands of seasonally adjusted weekly first-time filings; the bars are drawn from 190, not from zero, and the amber dashed line is the four-week average at 203. The four weeks read 204 to August 22, 206 to August 29, 206 to September 5 and 196 to September 12, the lowest since mid-August; the average fell below 205k to 203k. Today's release at 12:30 covers the week to September 19 and the market looks for 201k; ⚠️ a forecast, not a published number.

 

12:30 falls before the 13:30 open, so this is a pre-market release. It matters because this is the first week of data covering the Fed's September 16 hike to 3.75%–4.00%, decided by a 12–0 vote. ⚠️ Low claims only say there are no mass lay-offs; they do not say anyone is hiring; but for the Fed, no mass lay-offs is enough to keep rates high. Two markers: a print below 200k would confirm employment is holding up, while a jump above 210k would be the first sign tightening expectations are loosening. Barkin speaks at 12:00, Hammack at 12:50 and Paulson at 14:10, with August new home sales at 14:00.

 

The unit is basis points, the change from September 1 to the September 23 close, where one basis point equals 0.01 percentage point; all four bars are drawn from zero on one scale. The 2-year added 46, the 5-year 44, the 10-year 32 and the 30-year 13. With the short end up more than 40 and the 30-year only 13, what is repriced is how many more hikes, not inflation forever.

 

The 5-year is the discount anchor for AI capex, because the payback window for data centres and compute sits in that three-to-five-year stretch. It closed at 4.99% on September 23 after touching roughly 5.04% intraday, the first time above 5% since 2007; the 10-year closed at 5.11% and the 30-year at 5.40%. That is why a strong PMI came with a lower index: the money that got more expensive is the money technology stocks are priced against.

 

One supporting name: IONQ rose 4.42% after fitting an error-correction decoder onto an ordinary CPU, so its machines no longer pause for correction; Superion 256 is also set to join Nvidia's quantum research centre (⚠️ set to, not already there).

 

A: Two things decide it, the sign of the industry average and how many in the group rose. Application software averaged −0.85% and Nasdaq technology −0.82%, both negative, while the stock rose 13.69%. ⚠️ Moving against the group is not rising alone; PANW and CRWD also rose.

 

A: 4.85x is 29.91M shares divided by a usual 6.16M. It only says how many trades tested the price; it carries no direction, and high-multiple down days exist too.

 

A: No. Both are 2029 targets given at Investor Day, alongside compound annual growth of 14%–21%. ⚠️ Long-term targets are not quarterly guidance; the path shows up in later earnings reports.

 

A: The four weeks sit between 196 and 206; drawing from zero would flatten them into a straight line. Truncating makes the change visible, at the cost of reading ratios off bar heights, which is why the chart states the 190 baseline and the 203 average.

 

A: No. ⚠️ Above 100% is the price the market puts on share issuance, debt and forced selling at a low. Read it with two questions: where the coins came from, and what else the company owns.

Disclaimer: This article is compiled and written by the MEXC RealStocks team. The data in this article is based on the closing of the US stock market on September 23, 2026. The content is a compilation of public market information, and individual stocks are publicly discussed targets, which do not represent the recommendation or opinion of MEXC and do not constitute any investment advice. More US stock content: @MEXC | @Alpha_MEXC | @MEXC_Research