This report is going out two days later than usual. I was traveling last week and did not get a chance to publish Sunday's edition, so this edition covers both the week I missed and the week ahead. I am traveling again this coming week and will miss this Sunday's edition too, so the next report will be a bit of a double catch up when I am back on schedule. Without further ado, the Fed's meeting this past week posed questions along with important earnings to catch up on.

Weekly Recap

Fed Chair Kevin Warsh delivered his first keynote as chairman at Jackson Hole on Friday, warning that leaning on forward guidance creates a "hall of mirrors" where the Fed and markets end up reacting to each other rather than the real economy. Markets read his overall tone as hawkish, even as he stopped short of saying that rates would rise. That reading built on Wednesday's data: Core PCE for the quarter was revised higher to 3.6% annualized from an initial 3.4%, and the GDP second estimate confirmed growth at 1.5%, matching the advance estimate but with consumer spending revised up to 3.4%. Both readings pointed to more underlying demand and stickier inflation than the initial numbers suggested, setting the stage for the hawkish read Warsh's speech got two days later. Odds of a near term hike moved up to 66.4% heading into the September 16 FOMC meeting.

On earnings, Nvidia (NASDAQ: NVDA) beat on both revenue and EPS but initially sold off on softer margin guidance, then rallied with the other semiconductor stocks on Thursday. CrowdStrike (NASDAQ: CRWD) and Salesforce (NYSE: CRM) both jumped more than 10% after hours, largely on raised guidance rather than the quarter itself. In each case the market was pricing the path implied by guidance more than the quarter that had already happened, treating the backward-looking numbers as almost secondary. Dollar General (NYSE: DG) also beat on EPS and raised full-year guidance, its sixth straight quarter of comparable-store sales growth, a useful signal that the lower-income consumer is holding up better than feared, even with part of the beat coming from tariff refunds. Traders often invoke the phrase 'buy the rumor, sell the news' for exactly this kind of reaction, and it is worth keeping in mind even with a strong earnings background.

What I'm Watching Next Week

For this this week, JOLTs Job Openings have printed today at 7.359M, ahead of consensus and reinforcing a labor market that is cooling less than expected. That sets up for Friday's Non-Farm Payrolls and Unemployment Rate as the week's key release. Consensus is looking for a rebound at around 58K following last month's shock -23K print. The reaction will hinge less on the number itself than on whether it confirms or undercuts the hawkish read Warsh left the market with last week, since a weak print reopens the door to a rate cut regardless of what he signaled at Jackson Hole.

On earnings, Dell (NASDAQ: DELL) reports Tuesday after the close as a read on enterprise hardware demand. Broadcom (NASDAQ: AVGO) reports Wednesday after the close and is the one I am most focused on, given how directly it extends the AI infrastructure thread from Nvidia's report, including whether its own margin guidance draws the same reaction Nvidia's did.

Research

The simple version of options pricing assumes markets move in a predictable way, where a bigger surprise should produce a bigger reaction. That is not what happened this week. Nvidia (NASDAQ: NVDA) beat both revenue and EPS by a wide margin and still sold off, because investors were not reacting to the quarter that already happened. They were reacting to softer margin guidance and what that implies about the next few quarters. CrowdStrike (NASDAQ: CRWD), Salesforce (NYSE: CRM), and Dollar General (NYSE: DG) moved the other way for the same reason: raised guidance changed what the market expects going forward, and that mattered more than the numbers already in the books.

This is the real-world version of fat tails. Markets do not price the single most likely outcome, they price a whole range of outcomes, including the less likely but more extreme ones. That is why options get more expensive in implied volatility terms the further out of the money you go, even as they get cheaper in dollars, since the market is paying up for the chance of a bigger move than a simple bell curve would predict. The label "black swan" makes events like the 2008 financial crisis sound like a once in a lifetime, one in a million occurrence, but history says otherwise. Large, supposedly rare moves show up far more often than a normal distribution would suggest, which is exactly why traders and risk models started assuming fatter tails in the first place. Guidance is what keeps shifting that range day to day. A quarter that already happened is one fixed data point, but guidance is a claim about the future, which is exactly the part that carries the most uncertainty and the fattest tails.

Warsh's Jackson Hole speech is the same idea applied to the Fed instead of a company. His "hall of mirrors" warning is really an argument that Fed guidance does to rate expectations what earnings guidance does to a stock, narrowing the market's sense of what could happen next rather than just describing the most likely path. Less guidance, in that sense, means leaving more room for the full range of outcomes instead of pretending the future is more certain than it is.

The standard way to put structure on that is the Heston model, one of the more common attempts to price fat tails properly. Black-Scholes assumes volatility is a single fixed number for the life of the option. Heston treats it as something that moves on its own, and two parts of that matter here. The first is how much volatility itself can swing, which lifts both wings of the curve and is the fat tails point above. The second is how volatility moves relative to price. It tends to rise when prices fall, and that relationship, rather than the fat tails on their own, is why downside index options consistently cost more than upside ones.

Heston explains the mechanism, but it is not usually what gets fitted to the screen. That job tends to go to SVI, short for stochastic volatility inspired. SVI describes the shape of a single expiry at a single moment, plotting total implied variance, meaning implied volatility squared multiplied by time to expiry, against log moneyness. It does that with five parameters:

  • a sets the overall level, or how expensive that whole expiry is before any shape is applied.
  • b sets how fast variance grows as you move away from the money, which is the wings, and therefore the tails.
  • ρ tilts the curve, so a negative value makes the put side steeper than the call side and produces the downside skew.
  • m shifts the curve sideways, moving roughly where the low point sits relative to the forward.
  • σ controls how rounded the bottom is, where a small value gives a sharp kink near the money and a large value gives a smooth bowl.

The reason that is worth spelling out in a week like this one is that it separates things which usually get lumped together as volatility going up. Level, tails, and skew are three different parameters, and different news moves them differently. That is my case for far dated, far from the money options getting more expensive, and it is the specific thing I am watching into the September 16 meeting: not whether September volatility rises, but whether the wings six months out do.

Personal

Something I have been thinking about more this week is how much of trading comes down to habits rather than ideas. A good thesis is worth very little if the execution around it is sloppy, and execution is not really a decision you make in the moment. It is the product of whatever process you have built up over hundreds of previous trades.

The reps are what make that process automatic. Doing the proper due diligence, knowing what you will do if the trade moves against you before it does, sizing trades consistently and systematically. None of these are complicated, and all of them are easy to skip when things are going well. The problem is that the habits you build during calm periods are exactly the ones you fall back on during volatile ones, and a volatile week is the worst time to realize the shortcomings in your system.

Reps also give you a record worth looking back on, since small sample sizes flatter everyone and it takes a real number of trades to tell a working strategy from a favorable stretch. I would encourage anyone that is still trying to find what works to stay disciplined and keep going, because the process is what gets built in the meantime.

Disclaimer: Everything here reflects my own opinions and is shared for informational purposes only. It is not financial advice, and nothing in this report is a recommendation to buy or sell any security.