Actionable insights — Weekly SSR: the reaction-function trade, the revisions tell, the paper-gains strip-out, the concentration audit, the negative-EV screen, the unaffected-price entry

1. Trade the data through the Fed's reaction function, not the economy — with a stated band where "bad is good"

1. Before a labour or inflation print, decide which regime you are in. Write down the level at which weak data stops helping — his is the unemployment rate: below ~5%, soft data suppresses hike risk and lifts equities; above it, soft data becomes a recession, and the sign flips.
2. Convert the print into the only variable that matters — the market-implied probability of the next policy move — and size the trade to the change in that probability, not to the headline number.
3. Act on the hedges first. A collapse in hike odds is a reason to reduce the short/hedge book in a defined percentage, not to add gross length; the longs should already have been added during the fear that preceded it.
4. Confirm the read with the wage line. If average hourly earnings undershoot alongside weak payrolls, the print is disinflationary rather than merely weak — "no major wage inflation risk" — which is what keeps the "bad news is good news" reading valid.

• The unemployment rate approaching 5% (the point the rule inverts); the wage line alongside the jobs line; the implied-probability move rather than the surprise itself; and the next test of the regime — CPI Aug 12 (core 2.6% → 2.5% expected) and PPI Aug 13 (core ex-food-and-energy 4.7% → 4.1%).

2. Read the revisions, not the headline — a two-month restatement is the real print

1. When a labour report lands, immediately net the headline against the revisions to the prior two months. A −23k print with −103k of revisions is a −126k report; a +100k print with −100k of revisions is nothing.
2. Track whether revisions are consistently one-directional across administrations. A persistent pattern of boasting the first print and revising it down later is a reason to discount the initial release as an input rather than treat it as data.
3. Use the revised base, not the reported base, when you extrapolate the trend — and cross-check it against an independent series measured differently (job openings, claims) to make sure both are pointing the same way.

• The size and direction of the two prior months' revisions relative to the headline; whether openings/claims confirm; and the sector composition of the decline (services and professional roles rolling over is a different signal from goods).

3. Strip non-operating mark-ups out of a reported earnings season before you believe it — and date the payback

1. When index earnings growth blows out, decompose it. Find the companies carrying large stakes in private businesses and separate operating EPS from non-operating unrealized gains — value changes on holdings they never sold, which flow through the income statement as profit.
2. Recompute index growth with those gains removed. That number, not the headline, is what the market will eventually be marked against.
3. Ask what has to happen for the gains to repeat. Mark-ups on a private stake require a new funding round or an IPO at a higher price; a listed holding that has already fallen forces a mark-down. If neither is likely, you now have a dated earnings cliff — the same quarters next year.
4. Check whether sell-side estimates for those quarters have adjusted. If forward estimates still embed the one-off gains, the miss is mechanical and pre-scheduled, and you can position around the calendar rather than the news.

• The operating-vs-reported EPS split in the 10-Q; whether the private holding raises again this quarter; the listed holding's price versus its last carrying mark; and whether 1Q/2Q-2027 consensus has been reduced for the non-repeat. The falsification: another up-round large enough to reset the base.

4. Audit who is actually producing the index's growth — then map their dependency chain

1. Rank the index's earnings contribution by company, in dollars of index EPS, not by market-cap weight. Note any single name responsible for a double-digit share of the total growth.
2. Establish whether that contribution is durable or cyclical: check capacity sold forward, allocation percentages customers are receiving, and the next year's expected growth rate against the current one (a deceleration from 850% to 100% is still growth, but the second derivative is what re-rates).
3. Trace the demand back through the chain and write down the percentages: who buys from whom, and what fraction of each company's revenue depends on the next link. A ring — where A funds B, which buys from A, whose revenue is C's customer — is not fraud, but it means the whole ring contracts together.
4. Identify the financing constraint at the top of the chain. If the buyers are spending more than their operating cash flow, the ring's survival is a credit question — so watch credit instruments, not equities, for the first crack.

• Single-name share of index earnings growth; capacity sold forward and customer allocation percentages; each link's revenue-dependency percentage; hyperscaler debt issuance (>$100B raised, excluding non-recourse off-balance-sheet); and CDS spreads on the most levered spender as the leading indicator.

5. Screen for a negative enterprise value — then apply the three gates that decide whether it is a trade

1. Screen for companies where net cash exceeds the market capitalisation. Compute cash per share against the share price: you are being paid to take the operating business.
2. Gate one — the burn. Quantify the quarterly cash loss and divide it into the cash pile. That is your clock. A negative-EV stock "is only cheap if the cash isn't burned before it reaches shareholders."
3. Gate two — the catalyst. There must be a named, dated process that converts cash into shareholder cash: a strategic review, a sale mandate, a buyback authorisation. Sequenced disposals and cost cuts already executed are evidence the process is real, not rhetoric.
4. Gate three — management's history. Ask what this management did with cash the last time it had a lot. A team that historically deployed a cash hoard into expensive acquisitions is the reinvestment risk, and it is the reason the discount persists.
5. Only size it when gates two and three clear. Until then it is a watch — "it only becomes a compelling trade if phase three results in concrete capital returns or a total sale."

• Cash per share vs price; quarterly burn as a fraction of cash; disposals actually closed (not announced); lease and headcount reductions with dollar figures; the strategic-review deadline; and any capital-allocation history of acquisitions over returns.

6. Enter an activist situation at the unaffected price — never on the announcement spike

1. When an activist reveals a stake and the stock gaps, do nothing. Establish the unaffected price — where the shares traded immediately before the press release or the TV interview.
2. Judge the underlying business on its own merits at that unaffected level, ignoring the activist. If it wasn't cheap before, the stake doesn't make it cheap.
3. Wait for the pop to fade back toward that level and buy there, so you own the shares ahead of what the activist actually delivers — a board change, a forced buyback, an asset sale, a liquidity fix, a capital-structure change — rather than paying in advance for the possibility.
4. Use any published activist presentation as the roadmap of what to expect and when, not as a reason to pay up.

• The pre-announcement price as your limit; whether an activist presentation is published; the specific demands (board seats, buyback, asset sale, capital structure); and whether the stock retraces the whole move or holds a premium — a permanent premium means the market has already priced the outcome.

7. Short the mechanical squeeze in a low-conviction name — enter into the extreme, exit on the fade

1. Look for pre-market moves so large they cannot be explained by news — and identify the mechanical cause (a forced liquidation, a short book being closed out). If the driver is covering rather than information, the price has no anchor.
2. Do not short the first print. Let the extreme (here +240%) exhaust and initiate only once the move has already begun to come in but is still far above fair value — the entry is a level relative to the squeeze, not to your valuation.
3. Define the exit as a retracement target on the same day, not a price target on the business. Cover into the fade and be finished within hours.
4. Keep it in the trading sleeve and label it as such — it does not become a position, and it implies nothing about the company.

• A named forced seller or liquidating fund; the pre-market percentage versus any actual news; the level at which the move starts to fade; and whether the name has a real catalyst — if it does, this is not the trade.

8. Separate a timing miss from a guidance miss — buy the first, sell the second

1. When a stock gaps down on results, locate exactly which line missed and by how much. A sub-1% revenue miss is noise; the reaction is positioning, not information.
2. If the scary number is cash flow, find out why. Tax schedules, international cash movements and payment timing reverse next quarter; a genuine deterioration in unit economics does not.
3. Check next quarter's guidance against consensus. Unchanged guidance after a revenue miss means the company is monetising the same customers harder — a quality signal, not a warning.
4. Invert the test for the sell side of the trade: when the beat is enormous but guidance itself is below consensus, the market is right to sell it. The past quarter is sunk; the guide is the information.

• Size of the top-line miss in percent; the stated cause of any cash-flow gap and whether it reverses; next-quarter guidance vs consensus; and any regulatory overhang (an open SEC inquiry, as with AppLovin) that caps the size you should carry.

9. Underwrite a merger arb by counting the risk legs that have been retired

1. List every discrete thing that can block the deal — each antitrust authority, each political veto, each financing or earnings condition, each constituency with lobbying power — and treat them as separate legs, not one blob of "regulatory risk."
2. After each news event, mark which legs are now closed. A cleared phase-one review and a minister formally declining to intervene are two different legs; retiring the political one is often worth more than the antitrust one because it is the least predictable.
3. Look for affected third parties switching sides. When the constituencies who would be harmed by the deal start publicly backing it, the remaining political case erodes — size their weight (share of the market they represent), don't just note the endorsement.
4. Only then set a probability of close, and compare the spread against a modelled break price. Keep the position sized to the break, not to the spread.

• Each authority's stage of review; whether a political veto has been formally waived; endorsements from harmed constituencies weighted by market share; target EBITDA holding even when sales miss; dated regulatory deadlines (DBRG's FERC request by Aug 12); and the state-level gates that outlast federal ones (BHF's NY/MA/DE insurance approvals).

10. Trade the drafted rule — a named excluded supplier is a named domestic beneficiary

1. Read import-ban and tariff drafts for the specific company or product being excluded, not the headline geopolitics. A rule that removes a supplier from a market re-allocates that supplier's volume to whoever is left.
2. Identify the domestic substitutes with the capacity to absorb it, and check the pending decision has a date — a proposal without a timetable is not tradeable.
3. Run the same test on the input side: where a decision would tax an imported raw material, the beneficiary is the manufacturer who has already localised its supply chain, and the tell is a customer paying up for domestic content today.
4. Size to the policy risk. These are option-like positions on a rule that may never be finalised, so express them in names that work on fundamentals too.

• The named excluded supplier and its market share; the domestic substitutes' spare capacity; the rule's expected finalisation date; on the input side, the Section 232 timetable and customer contracts that pay explicitly for domestic content; and whether the beneficiary is investable on fundamentals if the rule never lands.

11. Read an insider sale against the narrative that caused the rally

1. When a stock has run on a macro narrative rather than company results, check whether insiders are selling into it. The relevant comparison is the sale versus the story, not the sale versus the company's fundamentals.
2. Weight it by who sold, how much, and whether options were exercised to do it — a CEO exercising and selling a large block after a defined run is a considered decision, not routine diversification.
3. Treat it as evidence about the narrative's durability, not as a short signal on the business.

• Form 144 filings versus the stock's recent driver; whether the sale followed an option exercise; the size relative to the holder's stake; and whether the disclosure timing allowed the market to react at all.