Two weeks in one episode, and there is a reason for it. These weeks turned out to be a single story, and reading either one alone gives you the wrong conclusion. Week one: Nvidia guided to roughly 70% revenue growth against Wall Street's 45%, rose 8.7% the next day, and dragged the chip complex with it, while the Fed chair used his first Jackson Hole keynote to say there is more work to do on inflation. Week two: August payrolls came in at 162,000 against expectations near 55,000, and July was revised from a loss of 23,000 jobs to a gain of 21,000. That negative print, the one that anchored the entire August rate-cut rally, never actually happened. Atlas and Michael walk through what a revision that size does to a market pricing central bank decisions off every release, and what a rising cost of capital means for companies that have to raise money to keep going. Commentary and education only, never advice.

Almost nothing in a small company filing comes with a date attached. Going concern language tells you a company may not survive, but not when. A shelf tells you shares may be sold, but not which day. There is one exception, and it is the exchange listing rules. Atlas and Michael cover the clock: the thirty consecutive days below a dollar that trigger it, the letter the company must disclose within four business days, the one hundred eighty days to cure and what curing actually requires, the second window and the reverse split it usually implies, and the hearing that can stretch the whole thing to roughly a year. Then the part that matters most in this segment. Losing an exchange listing can cost a company its fast path to raising money, at exactly the moment it needs to raise fastest.

A small company files a document on a Tuesday afternoon. It is under a page, there is no press release attached, and no headline anywhere. It says the company has a new accounting firm. That filing can be the loudest thing a company publishes all year. Atlas and Michael read it properly: why resigned and dismissed mean completely different things, why the reassuring line about no disagreements is nearly meaningless, and the one exhibit almost nobody opens, where the departing accountant goes on the record about whether the company's version is true.

A drug company posted results on an experimental cancer treatment and gained 177% in a single session. The next day it handed back a quarter of that. No new information in between. Atlas and Michael use that pattern to explain the binary catalyst: one event, one date, and a price that reprices in seconds rather than days. Then the part almost nobody watches for. After a spike, supply arrives from three directions, and the third is the company itself. A stock that triples is a financing window, and windows close, which is why the raise often lands within days rather than months. What to check while the excitement is still running: whether an effective shelf is already on file, how much cash the business holds against what it burns, and why reading the actual disclosure instead of the press release explains most of the next-day giveback. Commentary and education only, never advice.

Two weeks ago the market was pricing a rate cut. That is over. The 30-year Treasury yield hit its highest level in nearly two decades, oil rose about 5% after a US-Iran diplomatic understanding expired, and the Fed minutes revealed that several officials — not just the three who dissented — believe rate increases will be necessary if inflation stays elevated. Then Walmart fell more than 9% on its weakest quarterly sales in over six years. Atlas and Michael walk through the first losing week since July, why energy prices and the hawkish case feed each other, and why manufacturing is running hot while the consumer bends. Plus what a 20-year high in long yields does to the cost of capital for companies that have no profit and must raise to continue. Then the week ahead: inflation data and Nvidia both Wednesday, and the Fed chair's first Jackson Hole keynote Friday. Commentary and education only, never advice.

Inflation cooled, small caps hit records three times, and consumer spending fell off a cliff. All in the same five days.

A small company signs a lease for a warehouse it needs. Thirty thousand a month, ordinary business, except the landlord is a company owned by the chief executive. Nothing about that is illegal and nothing about it is hidden. It sits in the filings, described plainly, in a section most people scroll straight past. Atlas and Michael read that section. Who qualifies as a related party (broader than most people expect), the three places these arrangements get disclosed and which one actually gives you dollar amounts by year, and the catalog of what shows up in real micro-cap filings: officer loans running in both directions, shared services deals, consulting agreements with a director's own firm, and the board member whose fund buys into the company's offering. Then the four questions that separate a normal arrangement from a problem, and the trend test that reveals more than any single filing can. Commentary and education only, never advice.

A company raises $12 million. The headline says the offering closed, and then coverage stops. Almost nobody asks the obvious question: to do what? There is one paragraph in every offering document that answers it, and it ranges from a real commitment down to pure legal wallpaper. Atlas and Michael take apart the Use of Proceeds section: why "general corporate purposes and working capital" is the weakest language a small company can write, why debt repayment is an exit rather than a use, and how to spot the raise that exists only to pay off the last raise. Then the practical half. The three sections of the cash flow statement that reveal whether the money actually went where management said, the runway calculation that tells you if another offering is already coming, and what reading four offering documents in a row will show you that no single filing can. Commentary and education only, never advice.