Do you suspect the S&P might be overvalued, trading as it does on a 23x forward multiple? (For context, that’s 96%ile on a five-decade lookback.)
If the answer to that question’s “yes,” congratulations: You have at least one thing in common with a group of people — professional money managers — whose opinions probably don’t overlap yours very often.
In this month’s installment of BofA’s Global Fund Manager Survey, a net 58% said global equity markets (and remember: “global” stocks are basically synonymous with US shares these days) are overvalued. 165 panelists who between them count more than $425 billion in AUM participated in the September poll.
As the figure shows, that’s up slightly from last month and counts as a new all-time high.
Needless to say, the run-up in equity prices, both in the US and around the world, is in no small part attributable to the AI frenzy. Indeed, “Long Magnificent 7” topped the “most crowded trade” list in the survey again, taking 42% of the vote. “Long gold” overtook “short USD” for the number two spot, after bullion touched a succession of new records.
And yet, when asked if AI stocks constitute a bubble, nearly half of BofA poll participants said no. It was a closer call than it has been previously, though. In July’s survey, for example, 55% said AI stocks weren’t a bubble and just 36% said they were. This month, 48% said they weren’t and 42% conceded things might be getting a little frothy with the chatbots.
Speaking of chatbots, let’s ask one:
Straight from the horse’s mouth.
I asked the exact same thing, identical wording, punctuation and everything, of Gemini 2.5 Pro. The response:
“Yes.”
Here’s the vaguely amusing part. You can click a “Show Thinking” drop-down to expand it’s process. Here’s what was involved in arriving at, “Yes.” (Apparently ethics were involved.)
…
Focusing on User’s Question
I’m currently zeroing in on the user’s core question: “Are AI stocks a bubble?” I’ve identified this as the central point. I’ve also begun to analyze the specific constraints, especially the need for a straightforward “yes” or “no” reply, which adds a layer of complexity to the analysis. The constraints include that it be a yes or no answer.
Assessing Bubble Indicators
I’m now diving into the specifics of a “bubble.” What defines one? It’s when an asset’s price is way over its real value, often because people are getting carried away. I’m noting key signs like super high P/E ratios and crazy hype, much like the dot-com days. I’m also looking at companies slapping “AI” on everything to get investor attention, even if it’s not their main thing. The question of whether this is a bubble still persists.
Examining Bubble Arguments
I’ve just finished a deeper dive into the specific arguments for and against the “bubble” thesis. I’m weighing the extreme valuations against the real, tangible revenues of leading AI companies. I’m also considering the long-term potential of AI, drawing parallels to transformative technologies like the internet. It’s a tough call, trying to separate hype from genuine innovation. I’m focusing on trying to find the one best possible argument that will help choose between yes or no.
Analyzing For a Verdict
I’m now grappling with the core decision: “yes” or “no”? The “yes” side seems stronger, given the high valuations and speculative frenzy, despite real revenues in some firms. It’s difficult to predict when this could shift, but the potential downsides of saying “no” are far greater. I’m focusing on the ethical responsibility of the “forced choice.”
Concluding My Assessment
I’ve reached my final decision, after weighing the arguments. Despite the impressive growth and potential of AI, the market’s current state aligns more closely with bubble indicators. Thus, I’ve concluded on the answer “yes.” The decision was difficult given the constraints, and the implications of the response.
Defining the Answer
I’ve carefully weighed the conflicting arguments and constraints. The core of my current work is solidifying the precise meaning of “bubble” in this specific scenario. The need for a simple “yes” or “no” answer forces me to prioritize the most dominant market characteristics. I’m focusing on why a definitive answer is crucial.
Defining My Final Approach
I’m now fully committed to the “Yes” verdict and constructing the final answer. Despite the ambiguity of the question, my analysis has crystallized around the “bubble” thesis. I’m focusing on crafting a succinct and informative response, keeping in mind the user’s specific constraints. The key is in how best to present this “Yes,” given the request.
Defining My Justification
I’m now striving to craft a brief but compelling defense of my “yes” answer. I’m focusing on providing context that reflects both the risks and potential of AI stocks. The justification should be short enough, but avoid a sense of non-answer.
Summarizing the Current Strategy
The user’s query about AI stocks being a bubble requires careful evaluation. I’ve broken it down to consider arguments for and against, noting market signs and technological impact. The “yes or no” constraint is challenging, but I’ve now resolved that I will say “Yes”. I’ve just now constructed a concise justification to accompany my answer, aiming for clarity while honoring the user’s constraints.
I’m now satisfied that I’ve weighed the evidence and can give a direct and focused answer to the bubble question. I’m choosing a concise answer.
Defining My Response
I’m now confident that my answer must be a simple “Yes,” as requested. After revisiting my instructions and the user’s explicit constraints, I’ve chosen to prioritize their “just yes or no” directive. I’ve focused on ensuring the response aligns with my role as a helpful assistant, while also adhering to this stringent constraint.
Analyzing the Response’s Structure
I’m now fully committed to a single-word answer, prioritizing the user’s “just yes or no” directive. I’ve re-reviewed the conflicting instructions and believe a concise “Yes” response best honors the user’s intent. My justification is a concise “Yes”.
we’re all in a bubble until we’re not i guess.