Seven - 005

Week of August 3–7, 2026. The S&P 500 gained 3.6% to a record 7,757.64, the Nasdaq jumped 5.2%, and the Dow added 3.0% to a record $54,036.93.

Friday's jobs report told two stories at once. Payrolls fell 23,000, missing the 80,000-gain estimate. But unemployment ticked down to 4.1% from 4.2%, not because more people found work, but because labor force participation dropped to 61.4%, its lowest point this year, roughly 1.4 million people exiting the workforce in 2026. Markets read the weak payrolls number as easing pressure on the Fed, sending yields lower (10-year ~4.64%, 2-year ~4.19%).

Earnings drove gains too. Nearly 90% of the S&P 500 has reported, with profit growth tracking around 50% year over year, the fastest since 2021. Atlassian (+35.3%, its best day ever) and Airbnb (+17.4%) led; The Trade Desk (-21.9%) lagged on weak guidance.

Oil slid toward the mid-$70s as Iran and Oman edge toward an agreement managing the Strait of Hormuz, though officials say it won't fully reopen the strait. CPI and PPI land next week and will likely decide the Fed's next move.

Tempus IPO'd in June 2024, so the usual 8-pillar framework (built for 5-year track records) doesn't fit cleanly here. More useful this week: the quarter itself and the deal activity around it.

The business runs on two engines: Diagnostics (lab testing, $289.3M in Q2, +20% YoY, ~75% of revenue) and Data & Applications (the AI/insights engine, $93.2M, +28% YoY, with data-licensing up 36%). Total Q2 revenue was $382.5M, up 22% YoY.

Q2 was Tempus's first GAAP-profitable quarter, $5.6M net income versus a $42.8M loss a year ago, but $98.5M of that came from unrealized securities gains, a non-operating item. Strip it out and GAAP operating loss actually widened to $75.9M from $61.8M, more investment, not yet operating leverage. Guidance was raised to $1.595–1.605B in 2026 revenue, and gross margin expanded to 64.4%.

Tempus is also acquiring Personalis for $1.5B in an all-stock deal (some headlines cite $1.7B in equity value; $1.5B is the company's own enterprise-value figure), adding minimal residual disease testing. Dilution is a real risk given how far TEM has traded from the price the deal was struck at. Expected to close late 2026 or early 2027.

Most 8-pillar metrics come back red or "insufficient data," mainly because Tempus lacks five years of public history, not because the business is deteriorating. Revenue growth is the one clean green pillar (+$1.28B). More telling: forward P/S of 4.56x (in line with the 4.77x industry median), $820.7M in cash (up from $643.8M in Q1), and FDA approval for tumor-only xT CDx worth an estimated $85M in incremental annual revenue starting 2027. Analyst targets average $62.64 (Buy consensus), but several firms have been cutting targets post-earnings, BofA to $52, Stifel to $50, a more honest signal than the label alone.

Bull case: a real profitability inflection, guidance raised twice this year, a genuine 2027 pricing catalyst, a strong balance sheet, and a data segment growing 36%. Bear case: the profit beat leans on non-operating gains while the operating loss widened, growth has decelerated from ~80% to 25%, Personalis adds dilution risk, and a 2.98 beta means outsized swings in both directions.

TEM has been in a downtrend since its $104.32 high last October, trading below every key moving average for weeks. That changed Friday, August 7, when the stock gapped 12.86% ($46.12 to $52.05) on heavy volume, tied to Natera's MRD data validating the Personalis thesis. Real catalyst, not just a technical bounce, but one gap day doesn't undo a year-long downtrend, and gaps like this can fill back down.

The moving averages still tell the bearish story underneath: 20-day $48.24, 50-day $50.08, 100-day $51.98, 200-day $55.54, stacked in inverse order, the classic downtrend signature, even with price now above three of the four. MACD is showing a buy signal and volume confirmed the move. ATR near $3.50 means 6–7% daily swings are normal here. Support sits $50.47–$51.73, with $55 (the 200-day) as the level that needs to break for something more durable than a one-day reaction.

IV Rank ~25, IV Percentile ~32%, cheap premium relative to TEM's own history, the opposite setup from the GOOG credit spread in Seven-004. That favors buying volatility over selling it, hence a debit spread.

Trade analyzed on paper: Sep 11 $52/$55 bull call spread. Net debit ~$0.97, max profit ~$2.03, breakeven ~$52.97, max loss capped at the debit, ~209% return on risk. The $55 short strike lines up with the 200-day moving average, the technical ceiling from The Setup, so the trade caps upside right where the stock likely faces real resistance anyway. The $52 long leg avoids the unusually wide $50-strike market.

Alternatives set aside: a naked $50 call offers uncapped upside but costs roughly 5x more for a stock that just gapped into resistance, a better fit for strong conviction of a clean breakout. Nearer-dated spreads (Aug 28, Sep 4) priced worse than Sep 11 despite less time value, due to wider, thinner markets, a signal to trade the more liquid expiration.

All pricing uses bid-ask midpoints; the $50 strike's wide spread means real fills may differ.

1.4 million. Roughly how many people left the U.S. labor force in 2026 so far, the number underneath Friday's jobs report. Payrolls fell 23,000 in July, missing estimates. Yet unemployment fell to 4.1%, not a contradiction, since the unemployment rate only counts people actively looking for work. Participation dropped to 61.4%, its lowest point this year, as people exited the workforce faster than jobs were lost. A falling unemployment rate isn't automatically good news. The lesson: headline numbers, an unemployment rate, a P/E ratio, an options delta, almost always have a mechanism underneath that determines what they actually mean.

Max loss on the Sep 11 $52/$55 spread is the full ~$0.97 debit if TEM's gap fails and the stock drifts back toward $46–48. Unlike a credit spread, time works against this position even if the stock doesn't fall. The moving-average structure is still technically bearish, and price sits just $0.07 above the 100-day, so an ordinary pullback could put it back below multiple averages quickly. Friday's catalyst is a read-through from Natera's data, not direct confirmation, and the Personalis deal hasn't closed and could still be repriced. Some strikes carry wide bid-ask spreads, so real fills may differ from the midpoint pricing used here. CPI and PPI next week add macro risk for a 2.98-beta name.

Seven-004 built a credit spread on GOOG because elevated IV made selling premium attractive. This week's TEM trade is the mirror image: low IV rank meant buying premium, via a debit spread, made more sense. Building both back to back made the real lesson click: credit spreads and debit spreads aren't different risk tolerances, they're different answers to the same question, is premium cheap or expensive relative to its own history right now. Get that backwards, sell cheap premium or buy expensive premium, and the structure fights the market regardless of whether the direction call was right.

Seven has always meant completeness. In the oldest stories, the world was finished in seven days — and the seventh was the day to step back and take in the whole of it. That's the promise of this letter: seven sections, one complete and careful look at a company, every single week. No shortcuts, no half-finished thinking — the whole picture, with precision, laid out plainly enough that anyone can follow it.

This week, completeness meant sitting with a chart that gave two answers at once, short-term averages saying reversal, the long-term structure still saying downtrend, and building a trade that didn't need to pick a side to be right.

See you next Sunday. — 7even