The blistering rally in AI stocks is on shaky footing and this does come after a disappointing forecast from Broadcom.
Well, we did get earnings out from Oracle and in pre-market trade.
We are looking at shares of Oracle lower now.
The company did spend more than expected on CA last quarter fueling concerns on Wall Street about the profitability of its new AI and for business at the same time, the broader software sector is facing.
An existential crisis now with the agents driving development costs lower, investors are offloading former premium software stocks at a discount, hearing their competitive modes are gone.
But are these beaten down stocks a rare buying opportunity or deadly value trap?
Well joining me live here at the New York Stock Exchange is Kai Wu, Founder and Chief Investment Officer of Sparkline Capital.
Good morning.
Thanks so much for joining me.
Great to be back.
A lot of competing headlines this morning and here in New York.
I do want to mention what a performance for the Knicks yesterday.
So I'm sure that you're a little tired recovering from yesterday evening, but given what we're seeing across markets, there are so many concerns about what's happening in AI and especially on the heels of Oracle's earnings.
So how are you looking at the landscape right now?
Yes, I think Oracle is such an interesting case.
Because they are a kind of legacy software firm that pivoted hard into cloud AI cloud data centers, so they kind of represent both sides of this K-shaped market.
On one hand you have the semiconductor companies, the hyper scales, the folks building all the infrastructure for AI, and then on the download you have the software stocks which for whatever reason a lot of investors in the market believes are in the crosshairs of disruption, that they are.
You know, to follow the popular narrative going the way of the dodo or or Blockbuster or Borders, some of these companies.
Yes, so I do want to get your take on how you're looking at disruption, especially in relation to evaluation.
Yeah, so I think I'm a value investor and that's kind of how I was trained.
The instinctive thing to do when you see a stock like Adobe or like Salesforce that goes down 50 to 80% and is trading at a PE ratio of half the market is to just buy.
That's what you're trained to do.
But I think.
Especially in times of technological change that can oftentimes be a mirage, right, so I mentioned these classic examples of disruption, the Kodaks of the world.
These are companies that, you know, at one point were considered super innovative, like had great brands and seemed on top of the world, and then the technological landscape shifted landscape shifted, and over the next 5, 10 years, these companies basically went bankrupt.
And if you look at the way it played out in all these cases, you have this kind of interesting pattern where the market first materializes fears around disruption for these companies.
Prices start to fall, but it takes a while for fundamentals, think like sales, to really kind of show up with a lag.
So what happens is you look at the price to earnings, price to sales ratios of these companies, they oftentimes look really cheap, which obviously sucks in and lures.
The value investors who say, oh wow, this is a cheap stock right before they kind of go out of business.
So I think it's a really perilous time to be naively applying value metrics into these periods of AI and technological change.
So you're highlighting some of the concerns that we're focused on, including the software sector, that term apocalypse that entered our lexicon in 2026, as well as Some of these concerns about these metrics, especially when it comes to technological disruption, just because it might not be the same as other areas of the market.
So what are you focused on when it comes to the AI trade and what do you make of some of the concerns out there that liquidity is being drawn away from the AI trade just given some of the mega IPOs that are coming down the pike?
Yes, I mean those are all really interesting points.
I say to answer the metrics question. generally investors focus on trailing PE or current earnings or expected next year earnings.
I feel like especially in times where moats can instantly evaporate due to the advent of a new technology, looking at trailing or spot metrics is generally not the way you want to go.
I think that's a recipe for disaster in many cases.
Instead, I think what you should do is look to history and ask the question of which companies actually survived.
And thrived in periods of disruption.
The examples I gave in my research piece recently were The New York Times and Walmart, who respectively survived the newspaper and you know e-commerce like brick and mortar retail decimations.
And so why were they able to survive?
Two reasons.
One was that they leaned into the new technology.
It took them time, it took them years in many cases to pivot and adopt e-commerce and online media.
But they got there.
But more importantly, and this is the second point, they possessed unique intangible assets.
These are modes that were outside of just the technology itself but were able to allow them to stave off the disruption.
In the case of The New York Times and Walmart, it was like their brand.
It was their logistics network, the human capital, the newsroom, of course, and to some extent network effects too.
So these are, I think, really important things.
So as you Look to sift through the wreckage of the apocalypse as an investor, I think it's important to consider not just what the financials look like because that could look very different in a few years, but also think what are the true motives of these companies.
You don't want stocks whose only moat is the software code because that's basically been eroded.
You want stocks whose moat is their brand, the distribution, their networks.
And then your other question was around the liquidity.
I think one of the interesting things over the past few years has been the dearth of IPOs.
These really hot companies have gone longer and longer staying private and not entering the public markets.
You know, now suddenly in a short period of time we're seeing three mega IPOs, SpaceX, OpenAI, and Anthropic likely come into play.
And yeah, I mean, to some extent the capital markets are not infinite.
There is always a crowding out effect and to the extent investor demands.
To SpaceX or to the AI labs that will almost certainly create a damper on other places where that capital can flow who might be competing for that same sort of investor base.
Yes, and finally, before I let you go, of course we're counting down to that SpaceX IPO tomorrow.
So there are so many questions surrounding this initial public offering.
So how are you looking at SpaceX as we head into Friday?
Yes, well, I think first of all I would say I tend to not participate in day one.
Just because all the data show that historically if you can get in before that's great, but usually there's a pop and then on average across all the data we've seen you don't do well trying to get in on day one.
You kind of wait for things to settle down and then think if you want to be a long term holder of a stock, and we're long term investors of course.
SpaceX is an interesting company, obviously they're known for rockets and for Starlink, but it turns out that he's been Musk has been merging in all these other businesses.
They obviously have now.
Data center business, you know, the Grok model itself hasn't really been so successful, although we'll see where that ends up.
I think you don't want to count them out quite yet, but where Musk has a significant advantage is in hardware.
I mean, he's really, really good in the physical world and hence you know his ability to put together these data centers and then now he's selling off some of the compute to the more cutting edge labs who have the advantage of the software side, which just makes a lot of sense what he's doing there.
I mean, obviously the evaluations are astronomical.
So I think you have to be a little bit cautious, cautious there, of course.
Yes, and finally, before I let you go, I do want to ask you about anthropic as well as open AI.
I am sure that you are paying attention to the headlines that are coming in hard and fast regarding those two expected mega mega IPOs as well.
But how are you looking at their business models?
Yes, look, I mean, I.
Been spending a lot of time in the past two days playing around with the fable model, the new anthropic model.
I mean, obviously there's been a lot of backlash you've seen online against the kind of perceived their perceived kind of gatekeeping of the model and potentially the regulatory capture through the lobbying efforts and fear mongering around AI, which I think is concerning.
That being said, the model itself is quite good as far as I can tell, and you mentioned the headline around OpenAI potentially considering cuts to their costs, and I think that's really concerning because we talked about this, you know, on our last segment that historically, the builders of infrastructure have not been the long term winners because their products have become commoditized and so when you see one of the two major labs say we're going to cut, cut.
Cut our prices to compete and gain market share.
I mean, immediately you ask the question, is there actually a moat here, because you have now these two companies potentially engaged in a price war and then you have the Chinese models coming up from the bottom as well, open source, very, very low cost, and you have the customers, the enterprises which you mentioned as well, pushing back on the cost of tokens.
Hey, this was an interesting experiment, but token maxing isn't working because we're now spending.
More money on the tokens than the employees that we initially fired to make room for this in our budget.
So I think there's a lot of questions around the long term sustainability of those models, even though I will be the first to admit that technology itself is incredible.
Well, Kai, we will have to leave it there, but next time you join us, we will know how the SpaceX IPO went and hopefully we can extend this conversation surrounding tokenomics.
So thank you so much, Kai.
I appreciate your time as always.
Thank you.
Thank you.
