Six Standard Deviation Tech Event Analysis: The Compound Scorecard
The Compound episode was not a generic semis-versus-software debate. It was a market scorecard: six-sigma tech outperformance, a four-sigma semi/software split, rates as the new pressure point, and SpaceX as the public test of whether this is 1997, 1999, or just a blip.
The question in The Compound’s June 2026 What Did We Learn episode was not whether semiconductors are better than software. It was simpler and more urgent: was Friday’s technology selloff the start of something bigger, or just another blip in an AI-led market?
Josh Brown set up the discussion with Nick Colas and Jessica Rabe of DataTrek Research. The episode’s first claim, from The Compound, was that large-cap technology had just outperformed the S&P 500 by a statistically extreme degree over a 50-day window. Jessica then turned that into a framework: separate the tech move from the broad market, separate semiconductors from software, add rates, then use SpaceX and the 1999 comparison as sentiment checks.
That is the article. Not “buy chips, sell software.” The real takeaway is a scorecard for what has to happen next.
Signal One: Tech Got Too Extended
The first chart in the episode compared 50-day relative returns for the S&P 500 technology sector against the S&P 500. Jessica said the average tech outperformance since 2015 was 1.6 percentage points over any 50-day holding period, with a standard deviation around four points. By June 2, tech had beaten the S&P by about 29 percentage points, above the six-standard-deviation line she put near 25.6 points.
That number matters because it changes the market’s risk posture. A normal momentum rally invites dip-buying. A six-sigma move makes traders and market makers ask whether the next downtick is the start of forced risk reduction.
Jessica’s conclusion was appropriately unsatisfying: whether Friday was a pause or the start of a major rotation was unknowable. But the setup was not normal. When a move is statistically that stretched, the correct first response is attention, not confidence.
Signal Two: Semis Were The Epicenter
The second chart narrowed the question. The episode compared semiconductors and software using SMH and IGV as proxies. Jessica said semiconductors typically outperform software over time, but as of June 3, semis had beaten software by 44 percentage points over the prior 50 days, four standard deviations above the long-run mean.
The official fund pages support why those proxies make sense. VanEck describes SMH as tracking a semiconductor index tied to production and equipment. BlackRock describes IGV as tracking North American software companies, with selected interactive media and services exposure.
The episode’s point was sharper than the ETF definitions. Semiconductors had become the concentrated expression of AI infrastructure. Software, which had dragged earlier, began to participate late in the rally. Josh pushed on that point: the final accelerant was not simply money rotating from semis into software, but software and semis rising together. Oracle, cybersecurity, and other software names stopped weighing on tech, and suddenly the rally looked broader.
That is why the selloff mattered. If the last leg up came from everything in tech rallying together, then a reversal is not just about one hot semiconductor basket cooling off. It is about whether broad tech participation arrived too late.
Signal Three: The S&P Was Not As Extreme
The episode did not turn bearish on the entire market. Jessica separated tech from the S&P 500. She said the S&P had risen 15.3% over the prior 50 days, a two-standard-deviation move. In prior cases since 2015 when the S&P first crossed that threshold, the market was higher 50 trading days later, with an average forward return of 7.3%.
That distinction keeps the analysis grounded. The episode was not “everything is a bubble.” It was “technology is statistically stretched, semis are even more stretched versus software, but the broad market’s momentum is not automatically toxic.”
For portfolio work, that argues against all-or-nothing thinking. A tech rotation can happen inside an otherwise healthy market. A semiconductor pause can happen while the S&P remains resilient. The risk is assuming that the index headline and the sector internals are saying the same thing.
Signal Four: Rates Became The New Problem
The conversation then moved from charts to macro. The jobs report and revisions changed the read on the labor market. Two-year yields hit a 12-month high. The rate market started pricing a less friendly Fed path.
Nick’s framing was useful: the selloff might have happened anyway because tech had run too far, but macro did not help. Jessica added that higher rates become a problem for technology when they rise fast enough to raise recession risk. The issue was not that investors had lost belief in AI capex. It was that the rate backdrop had become a new dimension investors were not worrying about a quarter or two earlier.
That is the difference between a clean momentum reset and something more dangerous. If earnings remain strong and rates stabilize, the market can treat the selloff as a pressure release. If rates keep rising and recession risk comes back into the conversation, the same stretched tech chart becomes more fragile.
Signal Five: SpaceX Is The Sentiment Test
The SpaceX section was not a random detour. It was the episode’s public-market sentiment test.
Nick gave the clearest scorecard. He cited historical IPO first-day returns: average gains of 14% and 22% in 1997 and 1998, then 71% in 1999. Applied to the expected SpaceX deal, he said a first-day close in a normal 1997 or 1998 range would be one signal; a 1999-style pop would be something else entirely.
The episode also tied SpaceX to market structure. Nick expected index demand into QQQ after fast-entry inclusion and discussed lockups, float, and potential demand. He also warned that a deal with heavy retail interest can stress trading systems, using Facebook’s 2012 IPO and the 2015 ETF flash crash as reminders that volume itself can become a market event.
That makes SpaceX a better signal than another AI earnings call. Earnings tell investors whether AI capex still has support. A giant SpaceX IPO would tell investors how much speculative demand is still willing to pay up when a once-in-a-cycle asset appears.
Signal Six: 1999 Is The Wrong Shortcut
The end of the episode handled the inevitable comparison: is this 1999?
Nick’s answer was no, at least not yet. The Nasdaq rose 37% in Q4 1999 and another 24% from January to the March 2000 high. Compounded, that was an 84% gain in less than half a year. Then the Nasdaq fell 34% from March 10 to April 14. His point was that the modern tech tape may have echoes of that era, but the magnitude is not the same.
The best line in the discussion was the rule of thumb: a double is a bubble. The Nasdaq was up roughly 34% over the last year, not close to the late-1999 burst. That does not make the market safe. It makes the comparison more precise.
The episode’s 1999 lesson was not “ignore the risk.” It was “do not use the bubble label lazily.” The numbers that would make the analogy stronger are much more extreme than what the market has shown so far.
The Actual Scorecard
The useful output from The Compound episode is a checklist:
| Question | Why It Matters |
|---|---|
| Does tech recover quickly after a six-sigma relative move? | A fast recovery would make Friday look more like a reset than a regime change. |
| Does software strength continue without semis leading every rally? | That would show healthier breadth inside technology. |
| Do two-year yields keep rising? | A rate shock can turn a stretched chart into a macro problem. |
| How does SpaceX trade on day one? | A moderate pop says healthy demand; a 1999-style pop says speculative pressure is still alive. |
| Does volatility stay contained? | The 1999 comparison only becomes serious when upside momentum is followed by violent downside price discovery. |
That is why the article should be read as a market-structure and sentiment piece, not a sector explainer. The semi/software split is one input. The full episode is about whether a statistically extreme tech rally can digest rates, concentration, and a giant IPO without turning into something more unstable.
Bottom Line
The Compound episode did not say the AI trade is over. It did not say the market is in 1999. It did not reduce the whole discussion to semiconductors versus software.
It said tech had just produced a six-standard-deviation relative move, semiconductors had stretched far beyond software, the broad S&P setup was less alarming, rates had become a fresh complication, and SpaceX could provide a public-market read on speculative appetite.
That is a much better framework than the usual bubble debate. The next answer is not in a single chart. It is in the interaction between breadth, rates, IPO demand, and whether buyers treat Friday as a scare or a warning.
