reports earnings on Wednesday, and for a lot of traders that is the entire AI story. The stock is up about 41 percent this year, still near all-time highs. The Magnificent Seven have carried the indexes. If you are like many active traders, a third or more of your portfolio is tied up in AI and mega cap tech.
The headline numbers from Nvidia will matter, but what happens on that call will not be the only factor. Three quieter signals in the background are already shifting. Together they tell you whether this is just another strong quarter, or the moment when the AI trade moves from easy mode into something much less forgiving.
This article walks through those three red flags, how this week’s Nvidia call fits into them, and what to do with your AI exposure when they start to move.
Red Flag #1: The VC Down-Round Wall
The first signal is showing up where the AI boom started: in venture-backed startups.
PitchBook data says down rounds have hit 15.9% of VC-backed deals in 2025 so far, a 10-year high. Almost a third of those deals, 29.3%, are in AI and machine learning companies. That’s a lot of repricing in a part of the market that’s supposed to be the future demand engine for GPUs and cloud.
The micro picture looks just as stretched. Series A AI companies are spending about $5 for every $1 of revenue they generate, roughly twice the rate of earlier cohorts. The median Series A AI company now reaches $100 million in cumulative burn in about three years instead of six. Bain & Company estimates the AI sector needs around $2 trillion in annual revenue by 2030 to justify current investment levels. Today, AI and machine learning revenue still lives in the tens of billions.
Put those pieces together and you get a funding wall. As VC “dry powder” gets used up and fundraising slows, late-stage AI startups find it harder to roll their funding. When they can’t raise Series B and C rounds on reasonable terms, they cut back. That shows up as fewer H100 orders and fewer new data center buildouts.
You don’t need much imagination to connect that to NVDA’s guidance.
What to Watch
The main number is the share of VC-backed deals that are down rounds. The baseline right now is 15.9% for 2025. If that climbs into the low- to mid-20s, roughly 22–25% of deals, you are seeing a clear escalation. At 30% or more, funding stress is no longer a background worry, it is right in front of you.
PitchBook and similar VC data providers are often summarized in financial and tech news coverage. TechCrunch and other venture news sites regularly highlight down-round trends, and platforms like Crunchbase and Dealroom publish stories on AI startup funding pressure.
How to React When It Moves
At today’s 15.9%, this is an elevated but manageable level. It is a “monitor” signal, not a “do something drastic” signal. If it climbs into the 22–25% range, funding stress is spreading. At that point, it is reasonable to trim 5–10% of your NVDA, , and exposure on strength rather than on panic days and start redirecting some of that capital toward dividend names such as Atmos Energy (), McDonald’s (), and Waste Management ().
If down rounds push toward or above 30%, the funding wall is in front of you. That is when it makes sense to trim another 10% or so of your AI and mega-cap tech bucket and accelerate the move toward dividends and defensives.
Red Flag #2: Hyperscaler Capex Guidance
The second signal shows up in a place traders already watch: earnings calls from the big cloud and AI infrastructure players.
The AI trade rests on one big assumption: that cloud giants will keep spending heavily on AI infrastructure for years. Right now, the picture is mixed.
Meta is the first big name to sound more cautious. In Q3 2025, it raised its capex guidance range to $70–72 billion for the full year, up from $66–72 billion. The dollar range went up, but the tone shifted. Management stressed being “more disciplined” with AI infrastructure spend and pointed out that near-term returns on AI investment remain uncertain.
Microsoft, by contrast, is still pressing the accelerator. In Q1 FY26, its capex hit a record $30 billion. CFO Amy Hood said they feel “very confident” that spending is tied to contracted business they need to deliver, and Azure AI revenue is growing around 35% year over year. Amazon’s AWS capex remains at historically elevated levels without a clear sign of a pullback, and Alphabet raised its full-year capex guidance to roughly $91–93 billion, up from a prior $75–85 billion range.
So far, Meta is the cautious one. Microsoft and Alphabet are still pushing hard, and Amazon hasn’t blinked. That keeps this signal in a yellow zone. There is a crack, but not a break.
What to Watch
You can track this through company investor relations pages, major financial news sites, and your broker’s research tools. Read or skim the earnings call transcripts and focus on a few phrases.
Look for how often management mentions AI ROI and whether that language sounds upbeat, neutral, or cautious. Pay attention when CFOs talk about being more “disciplined” or “selective” with capex, or hint at “pullback” or “moderating” infrastructure investment. On Nvidia’s calls, listen for comments about order visibility: how far ahead they can see GPU demand and whether that window is shrinking.
On the calendar, circle Nvidia earnings on Wednesday, November 19, 2025, after the close, and then the next couple of quarters for Microsoft and Amazon. Those are the calls where a guidance change would matter most.
How to React When It Moves
With Meta alone sounding more restrained and the other hyperscalers still leaning in, this is a “watch closely” signal. It does not, by itself, justify big portfolio changes.
If Microsoft or Amazon start guiding capex lower in the next round or two of earnings, that is a different story. At that point you have confirmation, not just a one-off. That is when Red Flag #2 turns red. The playbook then looks similar to Red Flag #1’s escalation: trim 15–20% of your combined NVDA, MSFT, and AMZN exposure on strength. If you are comfortable with options, you can add small 90-day put positions on NVDA, for example a $175 strike sized at roughly 1–2% of your NVDA position, to protect against a sharp drop after a guidance reset. Selling 30-day covered calls on NVDA at strikes around $200 is another way to collect premium while you gradually reduce your exposure.
Red Flag #3: ETF Flows and Sector Rotation
The third signal often moves before the others. You see it in flows and relative performance, not in press releases.
As of mid-November 2025, , which tracks the Nasdaq 100, is up roughly 22–24% year-to-date. , a technology sector ETF, is up about 20–22%. , the Schwab U.S. Dividend Equity ETF, has gained around 13–15%, and , which tracks utilities, is up about 8–10%.
Dividend strategies still yield roughly 3–4%, compared with about 2.3% for the broad market and 10-year Treasury yields that now compete with, or beat, the S&P 500’s earnings yield.
On those numbers alone, you don’t yet see a full-blown rotation. Tech has outrun everything else, yields on dividend strategies look reasonable, and many institutional portfolios are very tech heavy. The setup is there, but the flows haven’t flipped decisively.
Not everyone thinks that’s a problem. Morgan Stanley analyst Joseph Moore, for example, recently kept a $220 price target on NVDA and pointed to strong demand for its Blackwell architecture through 2026. That bull case argues that AI capex has multi-year runway even if some startups stumble.
The point here is not to declare that view wrong, but to give you a way to manage risk if the market stops agreeing with it.
What to Watch
Start with ETF flows. Each week, look at money moving into and out of QQQ and XLK compared with SCHD, NOBL, and XLU. The pattern that matters is not one noisy week; it’s sustained pressure. If QQQ starts seeing $500 million or more in net outflows for three or more weeks in a row, while SCHD and/or XLU are taking in $200 million or more per week, that’s a sign big money is shifting.
Next, look at sector momentum. Simple relative charts of XLU versus XLK, or dividend/value ETFs versus pure growth, will tell you whether defensives are quietly taking the lead. If utilities and dividend strategies beat tech for three or more weeks in a row, you’re likely seeing early rotation.
Finally, watch options sentiment. Put/call ratios on QQQ, NVDA, and MSFT are easy to find through CBOE and many brokers. When the put/call ratio on QQQ moves above 1.0 and stays there, while put/call ratios on dividend ETFs remain relatively low, traders are hedging tech exposure much more aggressively than income exposure.
How to React When It Moves
If flows and momentum are still pointing toward tech and only modest money is moving into dividends, treat this as another signal to keep on your dashboard rather than a reason to overhaul your portfolio.
When you see a different pattern, with sustained QQQ outflows, stronger inflows into SCHD and XLU, several weeks of defensives beating tech, and higher put/call ratios on QQQ, it is time to treat Red Flag #3 as a genuine warning. At that point, trimming 15–20% of your AI and mega-cap tech exposure becomes a sensible step. You can start building positions in ATO, MCD, WM, and dividend ETFs like SCHD over four to six weeks, using dollar-cost averaging rather than a single big move. If you are comfortable with options, call positions on dividend ETFs or key dividend names let you participate more directly in a rotation.
A Three-Phase Rebalancing Plan
All three signals point toward the same idea: AI is a hot topic, but your portfolio should consider these warnings.
If you are like many active traders, your current mix might look roughly like this: around 40% in AI and mega-cap tech such as NVDA, MSFT, AMZN, , QQQ and other AI-heavy ETFs; about 15% in other growth names; roughly 30% in dividend, value, or broad index funds; and the remainder in cash or bonds.
You do not have to change that all at once. A more balanced approach might have about 25% in AI and mega-cap tech, around 20% in dividend aristocrats and dividend ETFs, roughly 10% in utilities and infrastructure names such as ATO and WM, and the rest in other sectors and cash. The question is how to get there.
From now through December 2025, Phase 1 is about scouting and nudging. On any strength in your favorite AI names, you trim a total of about 5% of your tech-heavy exposure and send those dollars toward dividend plays such as individual stocks like ATO, MCD, and WM or ETFs like SCHD. At the same time, you start watching the three signals each week so you are not surprised if they move.
Phase 2, running from January through May 2026, is about speed if the signals start firing. If Red Flag #1 moves into the low-20s on down rounds (around 22–25%), you trim another 10% from your AI bucket. If Red Flag #2 flips red, with Microsoft or Amazon guiding capex lower, you tighten again by 5–10%. If Red Flag #3 flashes, because flows and momentum turn against tech, you make a more decisive shift into dividends and utilities, aiming to bring tech down toward 25% of your portfolio and income names up toward 20%.
Phase 3, from June 2026 onward, is about maintaining a steadier mix: roughly a quarter of your portfolio in AI and mega-cap tech, about a fifth in dividend aristocrats and income ETFs, around a tenth in utilities and infrastructure, and the balance in other sectors and cash. At that point, new decisions become incremental. You might move a few percentage points one way or the other based on how the signals look, rather than making wholesale reallocations.
This Week’s Watch (November 15–22, 2025)
If you want to put this framework to work immediately, you can use a simple checklist for the current week.
For Red Flag #1, keep an eye on the next update to PitchBook’s Q3 2025 data later in November. The baseline is 15.9%. Any move above about 18% is worth noting.
For Red Flag #2, focus on Nvidia’s earnings after the close on Wednesday, November 19. Listen for comments on order visibility beyond Q1 2026 and any cautious language about 2026 demand. If you cannot join the live call, the company will post a replay and transcript on its investor relations site, and most brokers will make the key points available.
For Red Flag #3, check ETF flows on Friday. Right now QQQ is still seeing net inflows and SCHD modest inflows. If that relationship reverses and stays that way, take it seriously.
Closing
You do not have to outsmart every analyst on Wall Street to manage AI risk well. What you need is a small set of reliable signals and a plan for what you will do when those signals change.
Most retail traders only rethink their positioning after a big tech drawdown or a scary headline. By then, they are rebalancing into weakness.
You can approach it differently.
Watch the signals. Trim into strength. Build income and defense while you still have room to move. If the AI boom keeps running, you will still participate, just with a portfolio that can handle some turbulence. And if the red flags all start flashing at once, you will not be the last one trying to squeeze out of the same crowded trade.
***
This article is for informational and educational purposes only and is not investment advice or a recommendation to buy or sell any security. The author does not hold a position in any securities or assets mentioned. Options trading involves significant risk and is not suitable for all investors.