Profiting from the AI supercycle and then it's eventual possible bust (share price bust not the tech or economic bust)

[DegenerateGambler]

Hi guys,

 

I was thinking back to the dot com bubble and how some traders took both long and short on some names to profit from both directions.   How would you guys measure that the AI mania is near a top?  And flip from long to short to do 2x the profit on it?  Right now I have net long SNDK positions which I expect it to hit 2k, but from strictly valuation criteria, that wouldn't really be a maniacal top - which looks more like the chart of VCX (Fundrise Innovation Fund) which had stakes in Anthropic and SpaceX and jumped to 30 times NET ASSET VALUE.  Then it blew up.

https://investorplace.com/hypergrowthinvesting/2026/04/the-vcx-frenzy-is-a-warning-for-ai-ipo-investors/

 

I wanted to short VCX at 25 times net asset value but the shares were restricted for shorting - so i lost that opportunity.  But if you look at some AI names like NVDA, SNDK - they are not expensive on a forward PE basis given their growth trojectories over the next 2-3 years.   As businesses I wouldn't want to hold them over 10-20 years, but I do believe the AI supercycle will produce a chart similar to VCX.  So the question is, how do we profit from both direction given the velocity of the moves? 

 

For example - Allbirds (BIRD) converted from a shoe-making company to buying chips to rent to AI companies and its shares went up 600% and crashed, I tried to short that one too but couldn't find any shares available.  But those are individual cases with BIRD and VCX.  I was thinking of the actual entire trend, if riding it up then ride it on the way down, could produce generational wealth.  But currently am not yet skilled enough to know when the mania will end and do the "flip".   Would greatly appreciate any insight into this.

[WayWardCloud]

That's obviously the holy grail with any trend : getting in on the upside and getting out (or turning short) right at the top. Kudos for trying and I wish you great success but how many people have been consistently able to do that in history? I worry you might be trying to be too smart for your own good.

 

If you think there is a bubble of irrationality in a certain part of the market, just stay out of that part and find something else that you like at a price that you think is fair or cheap in another unrelated industry. Because the second you decide to "play" it, whether long or short, you make yourself dependent on that very crowd of irrational price setters that send shoe companies to the Moon. I don't personally think we're in an AI bubble but, for example, I've never believed in cryptocurrencies and I've always considered Tesla to be overvalued by a factor of at least 10x. Right or wrong, I still feel the same 10 years later, but boy has the market taken the other view! How could I possibly gauge when the tide will turn around since it's based on nothing but crowd mania? I don't have any control over those people and I have no idea what might make change their minds and when. And most importantly how do I avoid getting doubly smoked if I'm wrong? I remember some famous investor was short Tesla and long GM like a decade ago. He had a whole deck that made a lot of sense on paper.

 

I became a better investor once I decided I should focus on making money and not on feeling smarter than others. I realized there was a lot of misplaced ego involved in my contrarianism.

Edited May 1 by WayWardCloud

[DegenerateGambler]

Well this year my short ideas produced way more gains than longs, but those were some things that were easy to value, for example:

GDXU short at 400

PLTU short at 90

RDDT short at 208

CAR short at 669

 

But I feel the AI bubble is not as easy to measure.  As for TSLA - it is indeed overvalued by 10x, and they have declining revenue and increasing capex to bet on something others are also competing against them on - self driving.  And a human-shaped robot may or may not even be the optimal way to do things since many tasks require specific-shaped robots.  

 

Again, I think it's worth it to consider how to measure "ridiculousness" to profit from the eventual (almost certain) collapse in share price from mania.  I guess am trying to gain the skill George Soros would describe as "reflexivity" which somehow he found a way to have a magic touch on how to identify boom-bust cycles.  But probably you are right, this does sound like trying to be too smart.

 

[Libs]

2 hours ago, WayWardCloud said:

That's obviously the holy grail with any trend : getting in on the upside and getting out (or turning short) right at the top. Kudos for trying and I wish you great success but how many people have been consistently able to do that in history? I worry you might be trying to be too smart for your own good.

 

If you think there is a bubble of irrationality in a certain part of the market, just stay out of that part and find something else that you like at a price that you think is fair or cheap in another unrelated industry. Because the second you decide to "play" it, whether long or short, you make yourself dependent on that very crowd of irrational price setters that send shoe companies to the Moon. I don't personally think we're in an AI bubble but, for example, I've never believed in cryptocurrencies and I've always considered Tesla to be overvalued by a factor of at least 10x. Right or wrong, I still feel the same 10 years later, but boy has the market taken the other view! How could I possibly gauge when the tide will turn around since it's based on nothing but crowd mania? I don't have any control over those people and I have no idea what might make change their minds and when. And most importantly how do I avoid getting doubly smoked if I'm wrong? I remember some famous investor was short Tesla and long GM like a decade ago. He had a whole deck that made a lot of sense on paper.

 

I became a better investor once I decided I should focus on making money and not on feeling smarter than others. I realized there was a lot of misplaced ego involved in my contrarianism.

 

Good post. I used to wonder why Buffett didn't short the internet bubble, which he so clearly saw. You give the reasons here IMO. We should just stay away, long or short, and stick to our knitting. That's more than good enough. I've burned some capital trying to play this game; not enough to damage long term returns but enough to sting a little. It's just too difficult to get right.

 

And it is about ego, if I'm honest with myself.

 

[DegenerateGambler]

9 minutes ago, Libs said:

 

Good post. I used to wonder why Buffett didn't short the internet bubble, which he so clearly saw. You give the reasons here IMO. We should just stay away, long or short, and stick to our knitting. That's more than good enough. I've burned some capital trying to play this game; not enough to damage long term returns but enough to sting a little. It's just too difficult to get right.

 

And it is about ego, if I'm honest with myself.

 

I found the way to profit from these scenarios is to "shoot fish in the barrel" which means DON'T DO TOO MUCH DUE DILIGENCE like a traditional value investor.  Because the more due diligence you do, the more you become married to an idea.  Wait for an idea to scream at you "this is too easy, why isn't everybody doing this?" type of scenarios.   Buffet is from the Ben Graham school of thought where you have to dig deep into these quarterly reports instead of just waiting for an idea to scream at you - for example shorting CAR when it went from 90 to 800 (I shorted it at 669).  Buffet is very traditional - it would be hard for him to pull the trigger to short internet companies with zero outlook. 

[Lazarus]

My guess is that we're super early in any potential AI bubble. The technology is new, it's the 'next big thing', and we haven't even had significant IPOs yet. Bubbles usually last longer than expected, and they always have big ups and downs along the way as they seesaw between greed and fear.  

 

I think we're just getting started. I bought some AI ETFs and am holding on for the ride.

[DegenerateGambler]

1 minute ago, Lazarus said:

My guess is that we're super early in any potential AI bubble. The technology is new, it's the 'next big thing', and we haven't even had significant IPOs yet. Bubbles usually last longer than expected, and they always have big ups and downs along the way as they seesaw between greed and fear.  

 

I think we're just getting started. I bought some AI ETFs and am holding on for the ride.

I believe we are 2-3 years away from the "mania" stage as well.

[frommi]

When you have a sector ETF like SOXX go up 40% in one month you can be damn sure that we are closer to the top than most people think. I am already thinking about selling puts on 3x short ETF's but chickened out. 20% premium for holding 30 days sounds pretty good.

My way of going short AI is to go long software stocks. It looks like this is a pair trade for most, and it is clear no brainer. Adobe at 10 times earnings growing 10-15% with a clear balance sheet? I mean WTF? Or CSU at 13x this years FCF growing 25% per year? This is to me shooting fish in a barrel.

And there a lot of other high quality businesses on sale right now like BKNG or CPRT.

Edited May 2 by frommi

[valueventures]

17 hours ago, DegenerateGambler said:

Hi guys,

 

I was thinking back to the dot com bubble and how some traders took both long and short on some names to profit from both directions.   How would you guys measure that the AI mania is near a top?  And flip from long to short to do 2x the profit on it?  Right now I have net long SNDK positions which I expect it to hit 2k, but from strictly valuation criteria, that wouldn't really be a maniacal top - which looks more like the chart of VCX (Fundrise Innovation Fund) which had stakes in Anthropic and SpaceX and jumped to 30 times NET ASSET VALUE.  Then it blew up.

https://investorplace.com/hypergrowthinvesting/2026/04/the-vcx-frenzy-is-a-warning-for-ai-ipo-investors/

 

I wanted to short VCX at 25 times net asset value but the shares were restricted for shorting - so i lost that opportunity.  But if you look at some AI names like NVDA, SNDK - they are not expensive on a forward PE basis given their growth trojectories over the next 2-3 years.   As businesses I wouldn't want to hold them over 10-20 years, but I do believe the AI supercycle will produce a chart similar to VCX.  So the question is, how do we profit from both direction given the velocity of the moves? 

 

For example - Allbirds (BIRD) converted from a shoe-making company to buying chips to rent to AI companies and its shares went up 600% and crashed, I tried to short that one too but couldn't find any shares available.  But those are individual cases with BIRD and VCX.  I was thinking of the actual entire trend, if riding it up then ride it on the way down, could produce generational wealth.  But currently am not yet skilled enough to know when the mania will end and do the "flip".   Would greatly appreciate any insight into this.

Curious if there are specific AI names that you view as decent value. I still think SNDK and MU trade at reasonable valuations since their returns have been driven by operating growth rather than multiple expansion, but understand this is a function of perceived cyclicality. I think NVDA is fairly reasonable too (mid-20s forward P/E multiple), as their investor base has started chasing more of the related "picks-and-shovels" plays.

[frommi]

1 hour ago, valueventures said:

Curious if there are specific AI names that you view as decent value. I still think SNDK and MU trade at reasonable valuations since their returns have been driven by operating growth rather than multiple expansion, but understand this is a function of perceived cyclicality. I think NVDA is fairly reasonable too (mid-20s forward P/E multiple), as their investor base has started chasing more of the related "picks-and-shovels" plays.

In both cases the bubble is in the E 

Edited May 2 by frommi

[dipod]

18 hours ago, DegenerateGambler said:

Hi guys,

 

I was thinking back to the dot com bubble and how some traders took both long and short on some names to profit from both directions.   How would you guys measure that the AI mania is near a top?  And flip from long to short to do 2x the profit on it?  Right now I have net long SNDK positions which I expect it to hit 2k, but from strictly valuation criteria, that wouldn't really be a maniacal top - which looks more like the chart of VCX (Fundrise Innovation Fund) which had stakes in Anthropic and SpaceX and jumped to 30 times NET ASSET VALUE.  Then it blew up.

https://investorplace.com/hypergrowthinvesting/2026/04/the-vcx-frenzy-is-a-warning-for-ai-ipo-investors/

 

I wanted to short VCX at 25 times net asset value but the shares were restricted for shorting - so i lost that opportunity.  But if you look at some AI names like NVDA, SNDK - they are not expensive on a forward PE basis given their growth trojectories over the next 2-3 years.   As businesses I wouldn't want to hold them over 10-20 years, but I do believe the AI supercycle will produce a chart similar to VCX.  So the question is, how do we profit from both direction given the velocity of the moves? 

 

For example - Allbirds (BIRD) converted from a shoe-making company to buying chips to rent to AI companies and its shares went up 600% and crashed, I tried to short that one too but couldn't find any shares available.  But those are individual cases with BIRD and VCX.  I was thinking of the actual entire trend, if riding it up then ride it on the way down, could produce generational wealth.  But currently am not yet skilled enough to know when the mania will end and do the "flip".   Would greatly appreciate any insight into this.

One lesson I am grateful to have learnt on this website is to largely stay away from bubbles (especially in terms of money I do not feel comfortable losing). I used to try puts on TSLA since I've always had thoughts on it's completely irrational valuation. But after losing money a few times, I've learnt to stop making that trade. Nothing like the pain of loss to teach certain life lessons. Better to find undervalued stuff, buy when the market valuation makes no sense, and then do the time arbitrage till Mr. Market takes his meds. Seems to be a more consistent path to making money than riding waves.

[DegenerateGambler]

1 hour ago, valueventures said:

Curious if there are specific AI names that you view as decent value. I still think SNDK and MU trade at reasonable valuations since their returns have been driven by operating growth rather than multiple expansion, but understand this is a function of perceived cyclicality. I think NVDA is fairly reasonable too (mid-20s forward P/E multiple), as their investor base has started chasing more of the related "picks-and-shovels" plays.

SNDK MU NVDA are all fairly reasonable.  The name I have my eyes on is ZM (Zoom) who owns a 1% stake in Anthropic.  I believe Anthropic will be worth 2 trillion in 2 years, meaning the stake will be worth 20 billion (and ZM market cap is 30 billion with 7 billion net cash).  So you are paying something like 3 billion for 2 billion in free cash flow each year.

[DegenerateGambler]

22 minutes ago, dipod said:

One lesson I am grateful to have learnt on this website is to largely stay away from bubbles (especially in terms of money I do not feel comfortable losing). I used to try puts on TSLA since I've always had thoughts on it's completely irrational valuation. But after losing money a few times, I've learnt to stop making that trade. Nothing like the pain of loss to teach certain life lessons. Better to find undervalued stuff, buy when the market valuation makes no sense, and then do the time arbitrage till Mr. Market takes his meds. Seems to be a more consistent path to making money than riding waves.

I'd agree with this take mostly.  The problem with puts is the time constraint on it.  If you want to short Tesla, just short TSLL - it went from a high of 22.5 to 13 now.  It has no time limit you can keep on waiting.  I do believe TSLA is a short anywhere from 400-450 - their revenue is stalling and they need way more capital to compete on AI/Robotics against the competition.

[dipod]

6 minutes ago, DegenerateGambler said:

SNDK MU NVDA are all fairly reasonable.  The name I have my eyes on is ZM (Zoom) who owns a 1% stake in Anthropic.  I believe Anthropic will be worth 2 trillion in 2 years, meaning the stake will be worth 20 billion (and ZM market cap is 30 billion with 7 billion net cash).  So you are paying something like 3 billion for 2 billion in free cash flow each year.

The ZM idea makes sense as there'll probably be a large bump in price when Anthropic IPOs. I look at it more as an event driven play because the core business does not have many differentiating characteristics.