Michael Burry Shorts NBIS & Oracle — NBIS Crashes 13% as 20–40% Correction Warning Lands

By Yogurt · 2026-08-07 · Market Analysis

Michael Burry disclosed short positions on Nebius Group (NBIS) and Oracle — NBIS dropped 13% in a single session on the news. Burry warns of a 20-40% market correction from current highs. Meanwhile a repeating technical pattern across semiconductors, AI cloud, and tech reaches a critical inflection point. Friday's jobs report is the next big catalyst.

Burry Pulls the Trigger — NBIS Pays the Price

On Thursday August 7, Michael Burry — the investor who famously called the 2008 housing collapse — disclosed two new short positions: Nebius Group ($NBIS) and Oracle ($ORCL). The market's reaction was immediate and asymmetric: NBIS crashed 13% in a single session, while Oracle barely flinched.

Why the difference? Because NBIS is a smaller, higher-momentum AI cloud name — the kind of stock where sentiment can do real damage when a credible bear enters the room. Oracle is a $400+ billion enterprise giant with diversified revenue streams that can absorb a short disclosure without structural panic. NBIS, despite its compelling GPU cloud narrative, cannot.

Burry's thesis on both is similar: he believes they are overvalued, stretched relative to their moving averages, and ripe for a mean reversion. He specifically cited a 20-40% correction risk from current market levels — a range he says is justified by the combination of elevated valuations, geopolitical uncertainty around the Strait of Hormuz, and a Fed that is still not done tightening.

NBIS: Was the Chart Already Warning Before Burry?

Here's the question every NBIS holder should sit with: If you didn't know Michael Burry existed — if you never saw his filing — would the chart itself have told you something was wrong?

The answer, according to technical analysis, is probably yes.

Before Burry's disclosure hit, NBIS was trading at a significant distance above its 150-day moving average. That kind of extension — when a stock runs far above its long-term trend line — is not inherently bearish, but it does signal reduced margin of safety. The further a stock trades from its average, the greater the snap-back risk when sentiment shifts.

More concerning is the emerging chart structure. In the coming sessions, analysts will be watching whether NBIS is forming a head-and-shoulders pattern — one of the most reliable bearish reversal signals in technical analysis. A true head-and-shoulders would confirm that the recent 13% drop is not a buying opportunity but the beginning of a larger structural decline. If instead the stock bounces cleanly from here and reclaims its 150-day MA, it would suggest Burry's trade is early and the sell-off was an overreaction.

This is the 48-72 hour test: does NBIS hold, or does it continue to crack?

The Pattern Repeating Across the Entire Tech Sector

Burry's NBIS short didn't emerge in a vacuum. It lands in the middle of a broader technical setup that is appearing, almost identically, across a wide range of technology and semiconductor names.

The pattern looks like this: sharp rise → sharp correction → attempted breakout of descending peak line. Three distinct moves, played out over the last several weeks, now reaching a decision point.

Stocks displaying this exact setup right now:

  • SOX (Philadelphia Semiconductor Index): sharp rise, sharp correction, now testing the downtrend line from above-resistance.
  • SMH (VanEck Semiconductor ETF): identical structure to SOX — nearly overlapping charts.
  • Taiwan Semiconductor (TSM): the same three-act structure, currently in the breakout-attempt phase.
  • Bilibili (BILI): Chinese streaming/tech, following the same choreography.
  • Bloom Energy ($BE): clean energy pivot name, showing the rise-correct-breakout structure.
  • SanDisk ($SNDK): one exception — SanDisk has NOT yet broken upward. It pulled back without the clean breakout attempt. This is the laggard in the group, and either it catches up or it signals that the breakout in the others is false.

When the same pattern appears in this many uncorrelated names simultaneously, it stops being a coincidence and becomes a market-wide signal. The question is not whether the setup exists — it clearly does. The question is what the resolution will be: a genuine breakout to new highs, or a failed breakout that rolls over into a deeper correction.

That answer is likely to come in the next one to two weeks.

NVIDIA: The Stretched Leader

NVIDIA ($NVDA) also fits this pattern, but with an important caveat: it is currently stretched relative to its 200-day and 150-day moving averages. The stock finished Thursday's session negative — not because of any fundamental news, but because it carried too much rubber-band tension relative to its own long-term trend lines.

NVIDIA has technically broken above its descending peaks line — which is constructive — but the extension above its moving averages reduces the risk/reward for new entries. It can go higher from here. The S&P 500 could push toward 7,600 or even 7,800. But the gap between price and moving average is a cushion that's been used up. From a pure discipline standpoint: the time to buy NVIDIA was when it was at its moving averages, not 15% above them.

This doesn't make NVIDIA a sell — it makes it a hold with tight stops for those who are in, and a wait-for-a-better-entry for those who are not.

Software (IGV) in a Converging Triangle — Watch $104

The software sector, represented by the iShares Expanded Tech-Software Sector ETF ($IGV), is in a converging triangle — a technical formation where the range between higher lows and lower highs is compressing toward a breakout point.

The level to watch: $104. A clean close above that level would be a genuine breakout signal for the software sector, potentially signaling the next leg of the rotation from hardware (semiconductors) to software. The sector showed strength Thursday in after-hours trading, with several software names — including Clearfire — posting beats on revenue, earnings, and forward guidance.

Airbnb also delivered a strong beat, confirming that the consumer remains healthy. That consumer strength is consistent with what Costco showed earlier in the week. The contrast with TradeDesk — which continues to disappoint — reflects a company whose chart has been warning of structural issues for months. A stock that keeps making lower highs is the market's way of saying it knows something is broken, well before the earnings headlines confirm it.

Friday's Jobs Report: The Fed's Inflection Point

All of the above plays out against a specific macro backdrop: Friday morning's jobs report (Non-Farm Payrolls) is being released as market participants head into the weekend, and its implications extend well beyond today's price action.

The stakes are high because of what comes next. Mid-August brings the Jackson Hole Economic Symposium — the annual gathering of global central bank governors in Wyoming where monetary policy signals are telegraphed, rate paths are hinted at, and markets frequently pivot. The new Fed chair, who has publicly prioritized employment data, will be watching Friday's numbers closely.

The key scenarios:

  • Strong payrolls + rising inflation: September rate hike is confirmed. Markets that have already priced in ~63% probability of a September hike will spike the probability further. Risk assets face headwinds.
  • Weak payrolls + rising inflation (stagflation signal): The Fed faces its most difficult dilemma. Rate hikes would hurt an already-weakening labor market; rate cuts would fan the inflationary fire. This is the scenario that historically creates the most volatility.
  • Weak payrolls + stable/falling inflation: The most market-friendly outcome — gives the Fed room to pause or cut, supporting equities. This is the scenario bulls need.

Complicating the picture: the Strait of Hormuz situation remains unresolved. Energy was the only sector with meaningful gains Thursday (up 0.48%), which tells you the market is not fully confident that oil supply risks are off the table. If Hormuz tensions escalate and energy prices spike, that feeds directly into inflation — the Fed's most watched variable.

The S&P 500: Near All-Time Highs, But Stretched

The S&P 500 closed Thursday down 0.16% — not dramatic, but telling. The Nasdaq fell 0.3%, and the Dow slipped modestly after four to five consecutive days of gains. The VIX weakened further, which in normal conditions signals complacency. Bitcoin traded flat near its support level; Ethereum was also near zero momentum.

The index is near all-time highs — which is constructive — but it is also stretched relative to its moving averages, similar to the individual stocks discussed above. A move to 7,600 or even 7,800 is possible. A 3-5% correction back toward moving average support would be equally normal and healthy. The setup does not scream crash — but it does demand that you know where your exits are.

Burry's 20-40% correction call sounds dramatic. And to be fair, Burry has been wrong on timing before — he's been bearish at various points during this bull market without the crash materializing immediately. But his framework — looking at distance from moving averages, valuation stretch, and macro risk — is the same framework that any disciplined technician would apply. His conclusion may be extreme; his inputs are not.

The Takeaway: Patience, Not Panic

Here's how to hold all of this together:

  • NBIS's 13% drop is a warning, not necessarily a death sentence. Watch the 150-day MA. Watch for head-and-shoulders confirmation or rejection over the next 72 hours. Do not buy the dip until the pattern resolves.
  • The repeating breakout pattern is real — and it resolves in the next 1-2 weeks. SanDisk is the canary: if SNDK breaks up and joins the group, the pattern is likely real. If SNDK stays down, it may signal that the others' breakouts are failing.
  • NVIDIA and the S&P 500 are extended, not broken. That's a different risk posture than a trend reversal. Hold your positions; do not chase new entries at 15%+ above moving averages.
  • Jobs report Friday morning is the single most important near-term data point. Wait for the number before making significant portfolio moves. Volatility in either direction is expected.
  • Michael Burry's short disclosure is information — not a trade signal. He is right about NBIS's valuation stretch. He may be early on Oracle. And his 20-40% correction call may take 6-18 months to play out, if it plays out at all. Use it to sharpen your risk awareness, not to panic-sell.

The next one to two weeks will answer many of the questions that Thursday's session left open. Stay tuned, stay positioned, and — above all — know your plan before the jobs report hits.