Signet filed second quarter results before the open on September 9 and raised its full year adjusted earnings guidance to 10.45 to 12.15 dollars a share from 9.20 to 11.00, a 13.1 percent increase at the midpoint. It left its sales guidance completely unchanged at 6.7 to 6.9 billion dollars, so every dollar of the raise came from profitability rather than from expected demand. The stock rose 19.13 percent in a week its own sector fell 3.86 percent, and at 97.71 dollars it is 8.6 times the midpoint of the forecast it had just raised.

The catalyst. On September 9, 2026 Signet Jewelers reported second quarter fiscal 2027 total sales of 1,528.1 million dollars with same store sales up 2.2 percent, GAAP diluted earnings of 1.33 dollars a share against negative 0.22 a year earlier, adjusted diluted earnings of 2.19 dollars against 1.61, and adjusted operating income of 107.2 million dollars against 85.4 million. It raised full year fiscal 2027 adjusted earnings guidance to 10.45 to 12.15 dollars a share from 9.20 to 11.00 and adjusted operating income guidance to 535 to 605 million dollars from 480 to 560 million, while leaving total sales guidance unchanged at 6.7 to 6.9 billion dollars. It extended its Bread Financial credit card agreement through December 2035, expecting 30 to 40 million dollars of non comp revenue and gross margin, and repurchased 87 million dollars of stock in the quarter under a 125 million dollar accelerated programme. The stock rose 23.96 percent on September 9 and fell 4.65 percent on September 10.

Source: U.S. Securities and Exchange Commission, Signet Jewelers Reports Second Quarter Fiscal 2027 Results, Exhibit 99.1 to Form 8-K, September 9, 2026.

Why it made the five. It is the largest filed guidance raise in the issue and the cheapest price attached to one. Adjusted earnings came in at 2.19 dollars a share against 1.61 a year earlier and adjusted operating income at 107.2 million dollars against 85.4 million, on same store sales up 2.2 percent. The raise has a named source rather than a vague one: the Bread Financial credit card agreement was extended through December 2035 with a signing bonus and profit sharing. And the catalyst type is the one my own record actually supports. A Guidance Raise has averaged 12.06 percent against the index at one month across eight matured names, the best performing type in this newsletter with a sample bigger than one. The tape agreed for a session: the stock gapped 23.96 percent on September 9 on 6.81 times normal volume.

The case against it. It handed back a fifth of the move the next day. Signet fell 4.65 percent on September 10 on 2.77 times normal volume, which is the market declining to extrapolate a gap it had just paid for. Be careful with the credit card headline too. Secondary reporting described the partnership as worth more than a billion dollars over time; the filing says the company expects 30 to 40 million dollars of non comp revenue and gross margin. I am using the filed number. Then there is what the raise does not include. Sales guidance was left untouched, which means Signet is not telling you it expects to sell more jewellery, only to keep more of what it sells, and a jewellery purchase is the most deferrable thing in this issue if the fuel bill keeps climbing. At about 64 million dollars of trading a day it is also the thinnest name in the five.

Watch next, Sep 11: August consumer price index, 8:30am Eastern. Confirms the thesis: A core reading at or below 0.3 percent leaves the discretionary buyer intact and gives the margin story room to play out. Breaks it: A hot core print with energy passing through means the jewellery buyer is squeezed before the raised forecast is earned.

The takeaway. Read a guidance raise for what it leaves out. A company that raises earnings and freezes sales is telling you where the improvement came from, and margin is the line a cost shock reaches second.

Academy Sports filed second quarter results the same morning as Signet and did the same thing: it raised full year adjusted earnings guidance to 6.50 to 6.90 dollars a share from 6.40 to 6.80 while leaving net sales guidance untouched. The engine was gross margin, which reached 40.4 percent against 36.0 percent a year earlier and carried adjusted earnings up 19.1 percent even though comparable sales fell 0.4 percent. The stock rose 24.12 percent, the largest gain in the issue, and unlike Signet it kept going, adding 5.99 percent on September 10.

The catalyst. On September 9, 2026 Academy Sports and Outdoors reported second quarter fiscal 2026 net sales of 1,647.3 million dollars, up 3.0 percent, with comparable sales of negative 0.4 percent, gross margin of 40.4 percent against 36.0 percent a year earlier, GAAP diluted earnings of 2.17 dollars a share against 1.85 and adjusted diluted earnings of 2.31 dollars against 1.94, with ecommerce sales up 12.8 percent and three new stores opened for 327 in total. It raised full year fiscal 2026 adjusted earnings guidance to 6.50 to 6.90 dollars a share from 6.40 to 6.80 and GAAP earnings guidance to 6.05 to 6.45 from 5.95 to 6.35, while leaving net sales guidance unchanged at 6,230 to 6,355 million dollars. It reported a net tariff refund impact of 0.06 dollars a share including reinvestments and 182.1 million dollars of stock repurchased year to date. The stock rose 14.40 percent on September 9 and a further 5.99 percent on September 10.

Source: U.S. Securities and Exchange Commission, Academy Sports and Outdoors Reports Second Quarter Fiscal 2026 Results, Exhibit 99.1 to Form 8-K, September 9, 2026.

Why it made the five. It is the cleanest illustration of what this week rewarded, and the only pick the market kept paying for on the second day. A company whose same store sales went backwards still grew adjusted earnings 19.1 percent, because it held 440 more basis points of every sales dollar than it did a year ago. In a week when diesel rose 24.1 percent, the market decided a retailer that has already proved it can protect margin is worth more than one that is merely selling more, and it paid 24 percent for the proof. At 54.22 dollars the stock is 8.1 times the midpoint of the forecast it just raised, the cheapest in the issue.

The case against it. Measure the raise before you accept the move. Guidance went up ten cents at each end, 1.5 percent at the midpoint, and the stock went up 24.12 percent, so this is a re-rating and not a change in the numbers. It is the same lesson Okta taught in Issue No. 10, and it cost that pick most of its gain. Worse, roughly three fifths of the raise will not repeat: the company reports a net tariff refund impact of 0.06 dollars a share including reinvestments, against a ten cent increase. And the comparable sales figure is the part I keep returning to. Minus 0.4 percent is a customer buying less, and Academy Sports' own chief executive said consumer spending remains pressured, particularly among lower income households. That is the company telling you where the risk is, and I believe it.

Watch next, Sep 16: Federal Reserve decision. Confirms the thesis: A hold leaves credit where it is and the margin story keeps running. Breaks it: A quarter point increase lands directly on the lower income customer the company itself called pressured.

The takeaway. A gross margin gain and a sales decline in the same quarter is a company that got better at keeping money, not at earning it. Both matter, and only one of them compounds.

DigitalOcean's chief executive and chief financial officer raised the company's full year 2027 growth outlook to more than 50 percent at the Goldman Sachs Communacopia conference on September 8, and said roughly 85 percent of its artificial intelligence revenue now comes from higher margin inference work rather than renting out bare machines. The stock rose 12.64 percent that session and 19.79 percent on the week, while the software group it belongs to fell 5.38 percent. Two days later it filed for up to 725 million dollars of committed equipment financing, which is the part of the story most people will miss.

The catalyst. On September 8, 2026, at the Goldman Sachs Communacopia and Technology Conference, DigitalOcean chief executive Paddy Srinivasan and chief financial officer Matt Steinfort raised the company's full year 2027 growth outlook to more than 50 percent and said roughly 85 percent of its artificial intelligence revenue comes from higher margin inference services against about 15 percent from bare metal. The stock rose 12.64 percent that session. On September 10, 2026 the company filed a Form 8-K disclosing equipment financing agreements with MUFG Americas Capital Leasing and Finance providing up to 725 million dollars of committed financing, expandable to 1.025 billion through an accordion feature, funding up to 90 percent of data centre equipment costs on advances requested through September 10, 2027, amortising monthly to September 10, 2030, priced at a term SOFR swap rate plus 2.75 percent a year and secured by the equipment.

Source: U.S. Securities and Exchange Commission, DigitalOcean Holdings, Inc. Form 8-K, equipment financing agreements, Items 1.01, 2.03 and 7.01, September 10, 2026.

Why it made the five. It is the one pick here that rose in the half of the market this week was built to punish, and the reason is a change in what the business sells. Inference is compute you charge a margin on. Bare metal is a landlord business. Moving from one to the other is why an 85 percent revenue mix is a more interesting number than the growth target it was announced beside. It is also the only pick still well below its own recent peak, 30.1 percent under a 52 week high of 187.50 dollars, so the re-rating starts from a marked down base rather than from a record.

The case against it. Read the September 10 filing next to the September 8 stage. Two days after raising a growth outlook, the company committed to up to 725 million dollars of secured equipment debt, expandable to 1.025 billion, priced at a term SOFR swap rate plus 2.75 percent a year, with draws available through September 2027. That is the growth plan's funding cost, and it is set by the same swap curve that moved 22.4 basis points at the five year point this week. Every advance the company has not yet drawn gets priced at whatever that curve does next. The growth target was also stated on a stage and not in a filing: a verbal long run number carries none of the liability a guidance range does and can be revised without a document. And this is the second name in this very issue paid for saying 50 percent out loud at the same conference, which is a pattern worth distrusting.

Watch next, Sep 16: Federal Reserve decision. Confirms the thesis: A hold keeps the swap curve where it is and the new facility stays cheap to draw on. Breaks it: An increase raises the price of every advance not yet drawn on a facility that runs to September 2027.

The takeaway. When a company raises a growth outlook and then borrows to fund it in the same week, the interest rate on the borrowing is part of the thesis. Read the filing that follows the presentation.

Cloudflare's chief financial officer raised the company's long run growth target to 50 percent a year from 40 percent at the Goldman Sachs conference on September 9, describing a shift from selling software subscriptions to running an agentic artificial intelligence platform. The stock rose 10.51 percent that session and 9.37 percent on the week, against a software group down 5.38 percent, a gap of almost 15 points in five sessions. The week before, on September 3, it had announced a vulnerability detection and patching service built on OpenAI models.

The catalyst. On September 9, 2026, at the Goldman Sachs Communacopia and Technology Conference, Cloudflare chief financial officer Thomas Seifert raised the company's long run north star growth outlook to 50 percent a year from 40 percent, described the company moving beyond a software as a service model toward an agentic artificial intelligence platform, and named inference, content protection and payments infrastructure as the opportunity set. The stock rose 10.51 percent to 314.18 dollars that session. On September 3, 2026 the company had announced Vulnerability Discovery and Remediation, in early access through Cloudflare Managed Defense and built on OpenAI's Daybreak model family including its dedicated cyber model, available by invitation to selected enterprise customers, citing 60,475 vulnerabilities logged in the National Vulnerability Database by September 2026 against 48,185 across the whole of 2025.

Source: Cloudflare, Cloudflare Partners with OpenAI Daybreak Models to Redefine Vulnerability Management with AI-Powered Edge Defense, September 3, 2026.

Why it made the five. The relative move is the most striking number in this issue. Cloudflare rose 9.37 percent in the same week its own industry group fell 5.38 percent and the 10 year Treasury yield rose 18 basis points, which is exactly the combination that should hurt a company valued on distant cash flows most. It happened anyway, and the reason is that the company reframed what it sells: the announced service searches a customer's code for weaknesses and writes the patch, which is work that customer currently pays people to do. The backdrop it cited is real and checkable, 60,475 software vulnerabilities logged by September against 48,185 across the whole of 2025.

The case against it. This is the most expensive name in the issue by a wide margin, on a multiple of sales in the low forties, and it is the pick whose story rests least on a number anyone filed. A growth target raised from 40 percent to 50 percent was said out loud at a conference. There is no guidance range, no document and no liability attached to it. The product itself is in early access by invitation to selected enterprise customers, which is a long way from revenue. And the valuation is the whole risk: at a multiple like that the price is a promise about the next several years, and the one thing this week proved is that the market has started charging more for promises. The stock closed at 93.7 percent of its 52 week high.

Watch next, Sep 18: September quarterly options expiration and index rebalance, effective at the close. Confirms the thesis: Holding the gain through a rebalance that mechanically reprices the whole software group says the move was company specific. Breaks it: Giving it back into the rebalance says the gain was positioning rather than a change in the business.

The takeaway. A long run growth target is a promise about years you cannot see. The week the discount rate moves is the week the market starts charging more for promises, so notice when one gets paid for anyway.

HCA told the Wells Fargo healthcare conference on September 9 that 99 percent of its business performed as expected or better than in 2025 and that the core business is running at the high side of its long term 4 to 6 percent EBITDA growth range. The stock rose 2.93 percent, the smallest gain in this issue, in a week its own sector fell 4.39 percent and was the worst of the eleven. That is a 7.3 point gap earned by a company saying nothing had broken.

The catalyst. On September 9, 2026 HCA Healthcare presented at the Wells Fargo 21st Annual Healthcare Conference and said that 99 percent of its business performed in line with or better than 2025 expectations and that the core business is running at the high side of its long term 4 to 6 percent EBITDA growth range. It reported 2.1 million adjusted admissions across the first six months of 2026, said commercial volume excluding the insurance exchanges rebounded in the second quarter, and disclosed roughly 22,000 adjusted admissions that shifted from exchange coverage to uninsured status. In its January 27, 2026 fourth quarter release the company had set full year 2026 guidance of 76.500 to 80.000 billion dollars of revenue, 15.550 to 16.450 billion of adjusted EBITDA and 29.10 to 31.50 dollars of diluted earnings a share, and named the expiration of the enhanced premium tax credits among its guidance assumptions. It has since revised that outlook lower. The stock rose 4.93 percent on September 9.

Source: Yahoo Finance, Why Is HCA Healthcare (HCA) Stock Soaring Today, September 9, 2026.

Why it made the five. It is here because of what it is not. Health Care was the worst sector of the week and biotech fell 4.60 percent inside it, and HCA rose, on no new product, no raised forecast and no deal, purely on a confirmation that the business is tracking. In a week when the market repriced the cost of money, a hospital operator running at the top of its own growth range became one of the plainest things in the market to own. It is also the most liquid pick here by a wide margin at roughly 574 million dollars of trading a day, and the furthest below its own peak at 75.7 percent of a 52 week high.

The case against it. I will say the obvious first: a 2.93 percent move on conference remarks is the weakest catalyst in this issue, and the best source I have for what was said is trade reporting rather than a filing. The company filed nothing at all with the regulator between September 1 and September 10. The substance of the bear case is policy, and HCA disclosed it itself. Roughly 22,000 adjusted admissions shifted from exchange coverage to uninsured as enhanced premium tax credits expired, and the company named that expiry in its own guidance assumptions back in January. It then revised its 2026 outlook lower during the year, from an original 29.10 to 31.50 dollars a share, which is why a business running at the high side of its growth range trades between 13 and 15 times earnings instead of higher. A company telling you nothing broke is not the same as a company telling you something improved.

Watch next, Sep 11: August consumer price index, 8:30am Eastern. Confirms the thesis: Medical care services inflation holding up says HCA can still price its own services. Breaks it: Medical care services cooling while everything else heats up means the cost side rises and the revenue side does not.

The takeaway. A company confirming nothing broke is not a company reporting improvement. In a week when everything else was repriced, that distinction was worth seven points of relative performance, and it is still a confirmation rather than a catalyst.