StockResearch Intelligence Weekly #30

1. Opening

The market finally had a bad week. A real one, not an afternoon wobble that vanished by the close. From August 14 through Friday, August 21, the S&P 500 fell 1.43% to 7,674.37. The Nasdaq Composite lost 2.05% to 26,180.45. The Dow dropped 0.85% to 53,277.01, and the Russell 2000 fell 1.65% to 3,017.87. All four major indexes finished lower. No breadth story to hide behind this time.

The damage was concentrated, though. Semiconductor ETF SOXX fell 5.52%, technology ETF XLK lost 3.53%, and QQQ dropped 2.41%. Intel, Jabil, Seagate, Flex, and ON Semiconductor all took double-digit hits. This was the part of the market that had carried too much confidence and too many clean narratives. The trade got crowded. Then rates pushed back and the floor disappeared for a few days.

Healthcare went the other way. XLV gained 4.33%. Moderna rose 129.20% after its individualized mRNA cancer treatment with Merck cleared a Phase 3 efficacy test in melanoma. Merck gained 12.30%. That was a genuine clinical result, not a social-media rumor or a conference sound bite. It was also one drug program. Anyone treating a 129% weekly move as a normal entry signal needs to put the phone down.

The macro tape was rough. Minutes from the July Federal Reserve meeting showed three officials had preferred a rate increase, while several others said a hike could soon be appropriate if inflation stayed firm. July housing starts fell 12.4% from June and 13.5% from a year earlier. Single-family starts dropped 9.9%. Industrial production rose just 0.2%. Gold gained 5.45%, the USO crude proxy rose 6.35%, and long Treasuries went nowhere. The market was buying hard assets and healthcare while cutting exposure to expensive growth.

That gives us a better screen. Pullbacks are useful when the business underneath them is still working. ON Semiconductor fell 10.22% even after reporting stronger revenue, cash flow, and a better third-quarter guide. Cummins lost 6.90% after a quarter with $932 million of net income and $1.25 billion of free cash flow. Cigna stayed cheap but its operating mix and our forecast model still refuse to give a clean BUY. Good. A portfolio should contain decisions, not six copies of the same opinion.

Key themes

• Rate pressure finally cracked the crowded technology trade.
• Healthcare leadership came from real clinical news, but Moderna’s move was too violent to chase.
• Oil and gold rose while Treasuries stalled. Inflation risk did not leave the building.
• The useful opportunities are damaged stocks with current operating proof, not damaged stories with a large analyst target.

2. Strengths

• Healthcare had the only clean leadership. XLV rose 4.33% while every major index fell. Moderna’s Phase 3 melanoma result changed the value of its cancer platform in one morning, and Merck gained 12.30% alongside it. Moderna finished at $145.13 after rising 129.20% for the week. That is a victory lap, not an entry. Healthcare was strong, but the best-known winner already ran far past its old valuation map.
• Hard assets gave the tape somewhere else to go. Gold rose 5.45%, USO gained 6.35%, and energy ETF XLE added 2.79%. The move fit the week’s higher-rate and inflation anxiety better than the bond market did. TLT was flat, which meant long-duration assets got no rescue from falling yields.
• The prior edition’s restraint held up better than its aggressive calls. YETI’s SPECULATIVE BUY lost 0.09%, beating SPY by 1.26 percentage points. EQT’s HOLD also beat SPY slightly. Fluence fell 13.63%, validating the decision to avoid it. CRH and Microchip did not work in week one. That part goes in the ledger too. The lesson is not that HOLD always wins. The lesson is that a target gap cannot repair bad timing, and a NO TRADE can be a useful call even when nobody gets to post a green return beside it.

3. Weaknesses

• Semiconductors finally lost the argument. SOXX fell 5.52%, XLK lost 3.53%, and Nvidia declined 4.64% ahead of earnings. Jabil fell 13.74%, Intel 12.13%, and ON Semiconductor 10.22%. This does not prove AI demand is dead. It proves investors had priced too much certainty into a group where spending, inventory, margins, and rates still matter. Wednesday’s Nvidia report now carries a silly amount of weight.
• Rates and housing moved in the wrong direction together. The Fed minutes showed more appetite for tightening than the market wanted. July housing starts fell 12.4%, single-family starts dropped 9.9%, and long Treasuries failed to rally. Permits rose 5.0%, so construction is not dead. The current activity data still weakened. Expensive growth stocks hate this combination because the discount rate rises while cyclical demand gets harder to trust. The market had been pretending those two problems could not arrive on the same day. They can.
• The selloff was broader than one bad technology session. SPY, QQQ, IWM, and DIA all finished lower. Industrials fell 3.36%, utilities lost 3.48%, financials declined 1.17%, and small caps gave back 1.68%. There was no clean flight into defensives outside healthcare and a narrow group of staples. That matters for entry discipline. A stock can be down 10% and still have no buyer. We are buying current cash flow and guidance. Everything else can sit on the watchlist and think about what it did.

4. Weekly Market Recap

5. Top 3 Stock Picks for Long-Term Investment

1. ON Semiconductor (ON) – BUY ON WEAKNESS

$74.21 | $28.89B market cap | 23.2x FY2026 consensus EPS | 50-day / 200-day: $92.77 / $77.10 | Mean target: $104.67 | Target gap: +41.05%

• Thesis: ON Semiconductor just reported the kind of quarter that should survive a bad chart. Second-quarter revenue reached $1.604 billion, up 9% from a year earlier. GAAP gross margin was 38.4%, operating income was $258.6 million, and net income was $226.8 million. Free cash flow reached $425.4 million. That was roughly four times the prior-year level.

Management guided third-quarter revenue to $1.65 billion to $1.75 billion and non-GAAP gross margin to 40% to 42%. The company also expects AI data-center revenue to more than double in 2026. Those numbers do not erase semiconductor cyclicality. They do show that ON entered the selloff with improving revenue, cash conversion, and margin guidance. That is operating proof. We need it because the chart is ugly.

• Why now: Shares fell 10.22% for the week and closed at $74.21. The stock is below its $92.77 50-day average and slightly below its $77.10 200-day average. The mean target is $104.67, leaving a 41.05% gap. The gap matters because the business just improved. A large target gap attached to falling revenue would be decorative garbage. This one has a quarter behind it.

The entry still needs patience. Friday’s weekly low was $73.52. A stock that falls through both moving averages can keep falling while analysts slowly move their targets down. Staged buying is the only sensible response. Start small. Make the company prove that the third-quarter guide survives the next update.

There is also a clean way to judge the call. If revenue reaches the guided range and gross margin moves back above 40%, the selloff was probably an entry. If revenue holds but margin misses again, the stock may deserve the lower multiple. We are not buying the word AI. We are buying the chance that better mix and cash conversion are already arriving in the reported numbers.

• Forecast Model check: The one-day model forecasts UP 0.13% to $74.31 for Monday, August 24 at 57.33% confidence. The three-day model forecasts UP 0.24% to $74.39 for Wednesday, August 26 at 58.78% confidence. Both are weak signals. The seven-day model is different. It forecasts UP 3.30% to $76.66 for Tuesday, September 1 at 78.58% confidence.

We agree with the seven-day direction. We do not agree that a forecast below the 200-day average is enough for a full-size BUY. The model is describing a bounce, not a repaired trend. The business case does the heavy lifting here.

• Risk and verdict: Auto and industrial demand, inventory corrections, customer budgets, margin execution, and another semiconductor de-rating are the main risks. BUY ON WEAKNESS: begin with a small position around the current reset. A recovery through $77.10 would be the first useful signal. The $92.77 50-day average is the harder test. The $104.67 mean target only matters if revenue and gross margin keep moving up.

2. Cummins (CMI) – BUY

$587.68 | $81.09B market cap | 19.9x FY2026 consensus EPS | 50-day / 200-day: $659.59 / $595.04 | Mean target: $787.50 | Target gap: +34.00%

• Thesis: Cummins reported $9.457 billion of second-quarter revenue, $932 million of net income, and $1.25 billion of free cash flow. Adjusted EBITDA was about $1.7 billion, or 17.5% of sales. Diluted EPS was $6.73. Management raised its full-year outlook. The business did not behave like a company that deserved a 6.90% weekly drop.

Power generation is doing real work inside the story. Data-center demand is supporting the segment while the traditional engine and components businesses provide scale and cash. Cummins has several cycles under one roof. That can make the quarterly numbers messy. It also means the company is not dependent on one AI server order or one electric-vehicle launch to justify its valuation.

• Why now: The stock closed at $587.68, below its $659.59 50-day average and just below the $595.04 200-day average. The mean analyst target is $787.50, a 34.00% gap. Even the published low target of $700 sits above Friday’s close. Analysts can be late and targets can be wrong. A complete target range above the market still tells us the selloff moved faster than the earnings framework.

The valuation is reasonable at 19.9x FY2026 consensus EPS. Not cheap enough to ignore a recession. Cheap enough to own a company producing more than a billion dollars of quarterly free cash flow after a sharp weekly reset. We prefer this to chasing a technology name at 40x earnings because somebody said data center six times on a call.

Cummins also gives us a useful portfolio difference. ON is a damaged semiconductor. Cigna is a defensive value stock with a cost problem. Cummins sits between them, tied to industrial demand and power infrastructure with current cash flow. That mix does not remove risk. It keeps the Top 3 from becoming one macro bet wearing three ticker symbols.

• Forecast Model check: The one-day model is DOWN 0.12% to $586.96 for August 24 at 61.08% confidence. The three-day signal flips UP 0.03% to $587.85 at 55.83% confidence. Neither deserves much attention. The seven-day model forecasts UP 0.92% to $593.08 for September 1 at 71.95% confidence.

We agree with the seven-day direction and disagree with its tiny magnitude as a verdict tool. The model expects the stock to move back toward the 200-day average. Fine. Our BUY is based on the quarter, free cash flow, raised outlook, and the size of the valuation reset. This is an investment call with a mild timing tailwind.

• Risk and verdict: Truck demand, industrial orders, tariffs, commodity costs, currency, data-center project timing, and a break in the raised outlook are the main risks. BUY: accumulate in stages below the 200-day average. A close back above $595.04 would improve the setup. The $659.59 50-day average is the first serious recovery level. The $787.50 mean target is reachable only if the raised outlook turns into reported earnings.

3. The Cigna Group (CI) – HOLD

$277.51 | $73.33B market cap | 9.1x FY2026 consensus EPS | 50-day / 200-day: $284.68 / $278.81 | Mean target: $338.33 | Target gap: +21.92%

• Thesis: Cigna looks cheap because parts of the business are working and parts are irritating investors. Second-quarter revenue rose 7% to $71.7 billion. Net income was $1.66 billion, or $6.29 per share. Adjusted EPS reached $7.78, up from $7.20, and management raised its 2026 adjusted EPS outlook to at least $30.45.

The weak spots are not imaginary. Evernorth pharmacy adjusted operating income fell 27%. Cigna Healthcare’s medical-care ratio rose to 84.5% from 83.2%, meaning a larger share of premiums went toward medical costs. Quarterly free cash flow was negative $718 million. Some of that cash movement can be seasonal. It still belongs in the argument. Cheap health insurers can stay cheap for a long time when cost trends move the wrong way.

• Why hold: Cigna closed at $277.51, below its $284.68 50-day average and slightly below its $278.81 200-day average. The mean target of $338.33 implies 21.92% upside. At 9.1x consensus earnings, the valuation already reflects plenty of dislike. That makes the stock interesting. It does not make it urgent.

Healthcare just had a strong week, but Cigna fell 1.79%. Investors were buying clinical upside and selected defensive exposure, not every company with a healthcare label. Cigna now needs to prove that medical costs and pharmacy economics can stabilize while the raised EPS outlook holds. Until then, the target gap is a reason to watch, not a reason to pretend the operating mix is clean.

At the company’s new adjusted EPS floor, Friday’s price is barely above 9x earnings. That is the attraction in one line. The problem is that a low multiple can be the market pricing in lower-quality earnings before management admits it. One more quarter of stable medical costs would change the verdict faster than another analyst target increase.

• Forecast Model check: The one-day model forecasts UP 0.04% to $277.63 at 53.90% confidence. The three-day signal is NEUTRAL at $277.53 with 51.59% confidence. The seven-day model forecasts DOWN 0.60% to $275.84 for September 1 at 69.32% confidence.

We agree with the model’s lack of enthusiasm. The seven-day confidence is just below our 70% conviction line, so it is not a strong bearish call. The combination of neutral short signals, a negative seven-day direction, and an unfinished cost debate supports HOLD. The model does not veto the company. It vetoes the rush.

• Risk and verdict: Medical-cost inflation, pharmacy margins, regulation, customer retention, cash conversion, and execution against the raised outlook are the main risks. HOLD: existing owners can wait for operating confirmation. New buyers should demand a cleaner medical-cost trend or a stronger move back above $284.68. The $338.33 mean target is attractive. It is not a permission slip.

6. Three Low-Cap Stocks to Watch

The low-cap list is Nextpower, Klaviyo, and American Airlines. Each sits between $2 billion and $15 billion in market value.

1. Nextpower (NXT) – SPECULATIVE BUY

$86.18 | $13.07B market cap | 19.7x FY2026 consensus EPS | 50-day / 200-day: $106.54 / $109.26 | Mean target: $148.65 | Target gap: +72.49%

Nextpower is the company formerly known as Nextracker. The new name covers a broader power-technology platform, but the operating engine is still easy to understand. It sells tracking, electrical, and digital systems that improve solar-project output and economics. First-quarter fiscal 2027 revenue reached $935.2 million, up from $864 million a year earlier. GAAP gross margin improved to 35.9%. Net income was $165.4 million, free cash flow was $105.2 million, and adjusted EBITDA margin reached 24.9%.

The stock fell 15.07% for the week and closed below both moving averages. That is why it is here. The $148.65 mean target implies a 72.49% gap, but the real case is the combination of revenue growth, improving margin, positive cash flow, and a brutal price reset. The risk is policy, project timing, solar demand, acquisition integration, and the possibility that investors keep treating every clean-energy stock as the same trade.

Friday’s $86.18 close sat only $0.61 above the weekly low. There is no technical confirmation yet. This is a starter-position setup built around business momentum, not a chart breakout. A rebound through the 50-day average would make the trade easier. It would also make the entry more expensive.

SPECULATIVE BUY: use a small position. The first recovery levels are $106.54 and $109.26. Our risk line is $79, below the current reset and not far enough away to turn a speculative idea into a permanent roommate. The mean target is a long-term reference, not a forecast.

2. Klaviyo (KVYO) – HOLD

$18.33 | $5.49B market cap | 22.2x FY2026 consensus EPS | 50-day / 200-day: $16.72 / $20.86 | Mean target: $27.20 | Target gap: +48.39%

Klaviyo sells marketing automation and customer-data tools to consumer businesses. Second-quarter revenue rose 26% to $370.6 million. The company had more than 205,000 customers, customers above $50,000 in annual recurring revenue grew 36%, and international revenue rose 35%. Free cash flow reached $82.9 million. Management raised full-year revenue guidance to $1.526 billion to $1.534 billion.

That is a good growth quarter. The company still posted an $8.8 million GAAP net loss, and the stock remains below its $20.86 200-day average after falling 37.48% this year. The mean target offers 48.39% upside, but software target gaps can stay wide when investors stop paying for future margins. Klaviyo needs to keep converting growth into GAAP operating leverage.

HOLD: watch for a sustained move above $20.86 backed by another quarter of cash flow and margin progress. A break below the week’s $17.41 low would weaken the setup. The company is improving. The stock has not finished its argument with the market.

3. American Airlines (AAL) – NO TRADE

$13.82 | $9.15B market cap | 110.3x FY2026 consensus EPS | 50-day / 200-day: $15.76 / $13.91 | Mean target: $19.90 | Target gap: +43.99%

American reported record quarterly revenue of $16.735 billion. That headline looks great until the rest of the statement arrives. Net income was only $71 million. Free cash flow was negative $351 million. Higher fuel costs squeezed the quarter, and full-year adjusted EPS guidance now runs from negative $0.65 to positive $0.65. That is not guidance. It is a weather map covering half the country.

Shares fell 6.81% for the week and sit just below the $13.91 200-day average. The $19.90 mean target implies 43.99% upside. Ignore it. Airlines carry debt, labor costs, fuel exposure, operational complexity, and brutal pricing competition. A small change in revenue or fuel can erase most of the earnings line. The target gap does not compensate for a business that cannot tell us whether full-year adjusted EPS will be positive.

NO TRADE: there is no entry and no stop because there is no position. The stock can rally. That would not make this process wrong. We are avoiding a weak earnings base, negative free cash flow, and a guidance range with no usable center.

7. Strategic Recap & Forward Outlook

This week broke the easy version of the market. Technology fell, semiconductors fell harder, and every major index closed lower. Healthcare, gold, and energy picked up the money. Bonds did not. That is the part that matters. The market is no longer giving expensive growth a lower discount rate and a clean earnings story at the same time.

Our response is simple. Buy operating proof after the reset. ON Semiconductor gets a BUY ON WEAKNESS because revenue, free cash flow, and guidance improved while the stock fell 10.22%. Cummins gets a BUY because the quarter produced $932 million of net income, $1.25 billion of free cash flow, and a raised outlook. Cigna stays a HOLD because the valuation is cheap but the cost trend and model signal are still messy.

In low caps, Nextpower gets the small speculative slot. The quarter was real and the selloff was severe. Klaviyo is a HOLD until GAAP leverage and the 200-day average improve. American Airlines is a NO TRADE. Record revenue with thin profit, negative free cash flow, and a full-year EPS range crossing zero is not an opportunity. It is an airline doing airline things.

Forward Outlook

• Monday, August 24 – The macro calendar is quiet. PDD reports, but the bigger job is positioning ahead of Wednesday. Watch whether Friday’s technology bounce survives a full session.
• Tuesday, August 25 – July new-home sales arrive at 10:00 a.m. ET. Intuit and Zoom report. Housing will show whether the 12.4% drop in starts was construction timing or a wider demand problem.
• Wednesday, August 26 – Q2 GDP’s second estimate, July personal income and outlays, and July durable goods all arrive at 8:30 a.m. ET. Nvidia reports after the close. This is the week in one day.
• Thursday, August 27 – Weekly jobless claims and July advance economic indicators test the growth story. Marvell reports, along with Dollar General, Dollar Tree, and Best Buy. Chips and the consumer both get another hearing.
• Friday, August 28 – Fed Chair Kevin Warsh delivers the Jackson Hole keynote at 10:00 a.m. ET. Markets will listen for the line between stubborn inflation and weaker housing. There will be one. Traders will argue about three words around it.

Disclaimer: StockResearch Intelligence Weekly is for informational and educational purposes only and does not constitute investment advice, a recommendation, or an offer to buy or sell any security. All data is sourced from third-party providers and believed accurate as of the publication date but is not guaranteed. Price targets and verdicts reflect the author’s analysis and may change without notice. Investing involves risk, including loss of principal. Past performance does not guarantee future results. Conduct your own research and consult a licensed financial advisor before making investment decisions. The author may hold positions in securities mentioned.