Recession/correction in the next 12 months?
 

I think yes, I don't think the AI bubble is going to 'burst', but it feels a bit bloated to me. Thoughts?

I think this ai thing is going to burst.

But like the guy said: market can be irrational longer than you can remain solvent.

All i can do is hold on and make sure ive got my allocation straight.

Umm. With low interest rates people maybe moving their money from the banks to the markets and AI seems to be the next big thing. Disclaimer: I'm usually wrong.

http://forums.pelicanparts.com/uploa...1760072148.png

The chart above is worth thinking about.

Now, the hyperscalers are not the only ones investing in AI datacenters. There is also ORCL, CRWV, XAi, and many others. And OpenAI is busy doing circular deals so it can ramp up its investment. But I think the hyperscalers currently represent the bulk of it.

Growth dropping from 74% to 15% doesn’t mean investment is going down. It is growing, but growth is rapidly slowing.

Is rapidly slowing growth going to be good enough?

When do the expectations become irrational against the money invested in AI.

That’s what tabs is predicting…

Well, he's predicting more then just that.http://forums.pelicanparts.com/support/smileys/wat5.gif

We’ve had a number of companies pitch their “AI solutions” surrounding DOD contracting, I can tell you at least from my perspective they are 50% hype and hot air. Like many new tech bubbles I think you’ll see a lot of them eventually get found out and fold or be consumed by the ones that survive. I would tread carefully.

DoD+ all agencies has its own AI. Seems like the lower cognitive employees use it to complete their jobs. This might be a bubble when the furloughs and RIFs hit.

I’m not sure what DOD proper has for AI tools, I work for a company that does defense work. Generally speaking it’s a challenge because you can’t just dump all of your data into the LLM because much of it is subject to some variety of export control or classification, and after a certain point you can’t mix those streams. It also has to meet stringent IT security requirements and can’t openly interact with the entire internet like Chat GPT.

When PE's of 30x are the norm for a huge % of the market.

And highly profitable, growing bread and butter companies have PE's under 9.

Something is wrong.

People have been predicting a recession/correction for several years now. I'm sure it's going to happen eventually. It could happen next week, next year, or 5-10 years. The last folks to predict it right before it happens will say "I predicted it!".

The market certainly seems healthy except for the Feb - Apr time period this year when it was down.

My crystal ball is real cloudy. With so much uncertainty about AI, robotics, US capitol investment, tariffs, war, interest rates, government corruption, unemployment, and GDP growth I think it is a fools errand to attempt to predict the future in terms of investment. I simply do what I have always done- Index and chill.

Seems like we're overdue for a recession. I don't count the one that was caused by the pandemic shutdowns. As for a market correction, if it's defined as a drop of 10% or more, we've seen a couple. I don't see highly profitable companies being the cause of a market correction. I'd say it'll be economically driven and likely stem from the lending side again IMO.

http://forums.pelicanparts.com/uploa...1760106834.png

Title of the chart ("Figure 1") is not mine.

http://forums.pelicanparts.com/uploa...1760106945.png

Black curve is left scale. Green curve is inverted, right scale.

I believe a recession is coming in '26. I also believe AI will have nothing to do with why.

Exactly. I'd love to try to time the market, but it's far easier and less stressful to just let it ride for the long term.

Major recession has been impending/overdue since the last one.

It feels like this time (year) its different but it always feels that way. Its gonna happen one of these days but any prediction with enough timing accuracy to be actionable will be a lucky guess.

The one i played with was for cui. Only cui I generated was internal gov estimates for engineering projects. Quicker to use a brain instead of ai on that stuff.