Palantir Technologies is a relatively new company. My research questions here is: Is Palantir Technologies likely to outperform all FAANG stocks post 2023-2025? In this particular piece, I will break my research question into two-sub questions.
1) Is Palantir Technologies a "new type Google stock" and is Palantir Technologies a new Google in AI (instead of search engines)?
2) Is Palantir technologies a stock that is likely to outperform META (formally Facebook), Apple, Amazon, Netflix/Nvidia and Google (known as the FAANG stocks)?
This is a slightly different type of article, which is aimed at looking into whether there is an emerging opportunity in a company that is yet to become as famous as Google or Baidu (before it has become that well known and successful). To start helping my own answer on the first sub-question - is Palantir a new type of Google?, I found it funny to "Google" whether Google uses Palantir, this is what I found as ready answers.
The answer is Google Cloud and Palantir are working together on a joined solution
Another simple question to ask was, given their focus on AI, are they the leader in what they claim to be?
To clarify one thing, the first research sub-question is not aimed at finding out whether Palantir is the actual new "Google". No, rather, the question aims to examine whether buying Palantir today is just as a good of an opportunity it was to buy Google back in 2004 or 2009. This was the period I was personally adopting the Internet as a brave 15 year old in 2004. Post 2009, I remember that there was a saying that you open Microsoft's Internet Explorer one time and this is to download Google's Chrome. In a way, this was showing that Microsoft had started to lose its edge in at least one sub-segment of its operations. Only if someone had told me that if I personally love Google so much, maybe I should have bought some of its stock. But 15 year olds care about much more important topics than the potential returns of their future (or at least me, when I was 15). This ideally shows also that when I was 15 and Google was barely beginning to be a well known company (and it was a leader in only one thing - search engines). It barely had introduced its browser somewhere in 2008, so in 2009, it was about 1 year old. Basically, no one knew that Google would still be the leader in this space in 2009, but I imagine some people invested and made a lot of money, betting on this idea that it will. Since, I do not have a time machine, I have missed buying Google before it was the well-known "Google". So I am interested in finding something similarly attractive today.
Today, blockchain, Artificial Intelligence (AI), 3D printing, renewable energies seem to be the important new technologies (that are yet to be massively adopted). Overall, I will eventually have different pieces looking into every one of these. Of those, I already have one on the Bitcoin mining vs Bitcoin stocks and the bitcoin mining industry accessible via this
link.
For now, in this one I focus only on AI. Below I have made two short sub-sections to clarify some beginner's terminology. Just like with computers where there is a difference between software and hardware, the same difference exists between AI players who focus on one of the two - AI hardware products or AI software products (for now, I cannot yet find an AI company which is like Apple/Microsoft and does both).
Infrastructure for AI (hardware)
In another piece, I examined in one piece how I see Nvidia's A100 GPUs as the infrastructure of many AI tools. I'd even say that none of Tesla's products (the self-driving cars or Optimus, the AI robot which builds robots) can exist without having one of Nvidia's GPUs. In short, I believe almost every "smart" Tesla product will contain an Nvidia GPU. You can find this piece
here. This will hold true at least as long as other competitors can come up with similar products to Nvidia's A100 GPU or Tesla decides to make its own (both seem at least 5-10y away as of now), so for now I believe NVIDIA to be the key company in providing the physical infrastructure for data centres and AI-based products (anything which has self-driving, or any automated AI robot).
In other words, all leading AI products are likely to use NVIDIA's GPUs in their AI products. So in my opinion Nvidia will continue to be the leader in the in AI hardware (in terms of anything which uses (GPUS). Currently one of these costs between $10,000 and $70,000, so they can easily be what accounts for a 5-10k expense on each Tesla which costs 50-100k (i.e. 1 GPU is likely to be 1/10 of the entire Tesla product cost). I guess only Elon can tell me if I am correct on my math here.
So my humble opinion is that Nvidia will be the leader in the hardware for all products in the AI space which will need "to see" and "think" and "decide" what to do on their own.
Software AI (software operating system to operate the hardware, solutions, services)
I wanted to write the part on Nvidia to 1) clarify that AI is not one thing, it is at least two - hardware and software and 2) to have a hardware AI company so I can present what I think will be the leader in the Software AI space for the next 5-10 years.
Palantir and software AI at a Glance
If Palantir is already the leader in the AI software space and AI is expected to be one of the fastest-growing and most profitable industries, then it would be very logical to think that Palantir is yet to grow it's revenues and profits from its digital solutions because the AI space is still very small - less than $150 bil industry. I'd say we are in 2009 for Bitcoin, i.e. one year after the 2008 crisis was created, just like Palantir was created 6 months after the COVID-19 crisis. To support this argument I will use these.
A business student's analysis
It is somehow funny to me that in business schools we study that making digital products is actually more profitable for entrepreneurs because costs are lower (and digital products are infinitely scalable). However, there is no theory that suggests that investing in say companies which produce new digital products is a more profitable investment relative to investing in a company which does both or hardware. So as a person who has studied both finance and investing, I find it very interesting that in the investing world, no one actually analyses businesses through the frameworks taught to business students. (This gives me an idea for a unique piece where I will ask is Investment and Business analysis that different? I will eventually link it here).
1. Macro (PESTLE/PEST)
In business studies, we begin with Macro analysis of the PESTLE world. I have already done this since 2020 and I have a summary of my still accurate economic macro views
here. The key takeaway for this piece is that as of 2020 I expected 2021 and 2022 to be crisis years for people and stock markets. I also expected as of January 2023 we will have no new lows for stock markets and that I see us having new highs on stock markets by 2026 or as early as 2024. This is the only piece of macro analysis I need to from there. The other is that the emergence of AI is a significant Technological change (T in PESTLE), which is likely to create political changes (P) - global frameworks needed, economic changes (E) - greater value being created with help of AI, legal changes (L) - about rights of people/robots, and environmental changes (E) - a cleaner and more efficient world, a world dominated by AI or something else). This is to show the emergence of AI (like blockchain) will change the existing PESTLE.
2. Meso (industry + market)
Once the Macro analysis is done, we move to the industry analysis and market analysis. Here I focus only on market analysis and do not do an overview of Porter's five competitive forces (a business model on industry analysis).
Market analysis
Based on Research & Market's industry analysis, the Global AI Market (including all subsegments globally) is expected to grow by 38.1% per year for the next 8 years, growing the size of the industry by 12.3x from $136.6 Billion to $1811.8 Billion (or $1.8 Trillion), see Figure 1 below. If you are confused what CAGR is, it is called compound annual growth rate (which simply means average growth per year for the period which is defined, in this case 2022-2030).
Figure 1. Total Global AI Market size and projected growth
So this research suggests
the industry of AI as a whole is expected to grow on average by 38.1% year.
An earlier an more conservative prediction in terms of the total value is made by Verified Market Research, suggesting that from 2021 to 2030 the value of Global AI software markets alone will grow by 41.30% from $53.54 billion to $850.62 billion (a 14.89x increase), see Figure 2 below. This suggests that while the industry as a whole is expected to grow by 12.3x to 2030, the global software AI industry is forecasted to grow by 14.89x, i.e. faster than the broader AI market (the segment with most growth).
Figure 2. Software AI Market size and projected growth
As of here we can immediately see two things. 1) at least 40% of the total AI market is software AI (if in 2022, total AI was $136.6 billion and as early as 2021 software was $53.54 billion, this suggests that between 40-50% of the total value of the AI market is one of it's segments - the software AI.
The even more interesting number is the difference in CAGR between total AI and Software AI. We can immediately see that Software AI is expected to grow at a rate which is 3.2% higher per year (41.3% for software vs 38.1% for hardware AI). This suggests that business professors know what they are talking about when they explain to us that starting software companies is more profitable (this expectation is also reflected in these two professional industry analysis firms).
In conclusion, this suggests that yes, while Nvidia will likely be the leader in the hardware AI space and Palantir may be the leader in the software AI space, Palantir is likely to grow its revenues faster than Nvidia. This is simply because the market for software AI is expected to grow faster than total AI.
Ok my business student version of me, did a short analysis of the macro and meso. I do not have space to also to the detailed micro, but here is the most important one - the marketing mix.
Ok my business student version of me, did a short analysis of the macro and meso. I do not have space to also to the detailed micro, so let's dive in something short.
3. Micro (company and competitors)
In theory the micro analysis is a long analysis of competitor's and their capabilities and creating matrixes to determine competitiveness of companies (via indexes). It is also the analysis of a particular company's products, promotions, distribution channels (places its available) and pricing strategies (price), these are called the Ps/Ss of Kotler, 2010 (P is for physical products, S is for services).
So let's skip to the most important part the - Product mix
The product mix refers to the company's combination of products or services. In the case of Palantir, the products are rather complex for the novice AI reader.
I guess everybody knows WordPress, it actually created the 2.0. version of Internet with social media. The idea is that WordPress allowed people to create websites - blogs and others, for the first time without the need of Web1 (Microsoft's products) to code in html or java (back then). WordPress also allows developers to create plug-ins, which users of the platform WordPress can pay for. In other words, the company has a fundamental company developed platform, whereby users have the ability to either create on top of it, or purchase from it different additional services.
Palantir aims to do the same in the AI space, they aim to develop something which we can think of as the software operating system (like Apple's IOS or Microsoft's Windows) or the fundamental infrastructure (like WordPress for websites).
To me this is the most important product because it enables SAAS (software as a service) and while Palantir has put a lot of effort into developing this software, it also has the ability for organizations and enterprises to develop their own business solutions and sell them to other organizations (in other words, Palantir is like the operating system/key infrastructure of AI software).
I compare this company to Google, but it can be compared to Meta too. In a way, these are both companies which started as software companies and relied on building the initial infrastructure - of a search engine for Google and a social media platform for Meta. Both of these relies on users creating content for visits and success of their platforms. However, both of these relied people to do it.
With its business model Palantir products remind me of Web 3.0 products, blockchain solutions, which is because networks are considered open for developers to build apps on them in the same way. To me this indicates that there is a company in the blockchain space that will develop the operating system or the Windows of Blockchains (I will link a publication here). However, Palantir due to its 3.0 thinking and product which is compatible with outside developers building to increase the total usefulness of their solutions, seems well positioned to create strong revenues in the future based on this idea that other companies can help it build even more solutions with no additional cost to the company. In other words Palantir provides WordPress and the developers provide the plug-in software. Or Palantir develops IOS and developers develop different games and apps that we can pay for (but for businesses, not people).
Their other products include more narrowly a software for AI-decision guidance for robots (with usage in military to factory levels), planes and drones (both commercial and military defence), and wind turbines. They also have a product called Gotham (for Batman nerds like me) which they promote as "the operating system of global decision-making" (the one I was comparing to Wordpress, Windows and IOS). The last one is Apollo (see Figure 3 below)
Figure 3. What is Apollo?
An Investment student's analysis based on the business student's analysis
So the business student is telling me that Palantir might actually be a company which is like Google or Microsoft or Apple (10-20 years in the past), interesting. I know the past performance of these stocks will not predict the growth of Palantir, but let me see how Google returns have worked out as an example of what it could look like if the business student's research is correct.
Google grew from about $8 in 2009 (adjusted for stock splits) to about $148 in 2021. This represents an 17.5x (or 1750%) gain from 2009 to 2021 or for about 12 years. Even now, when Google stock is at about $94 at the time of writing this, that is still a 10.75x return (1075%) on the 12 years (or about 80-100% per year).
At the same time, the global AI market is expected to grow 12.3x as a whole and global software AI is expected to be the most profitable segment that can grow 14.89x, would it be reasonable to expect 10-15-20x returns on Palantir stock for 10-12-15 years? Well, if the stock grows linearly to market growth, yes this makes sense. Can it get there faster? Well, if we look at Intel, it's anticipated future growth from 1999 to 2020, was priced in by 2002 when the stock hit an all time high of $70, before the so-called dot.com collapse.
So this tells me that the stock's price may price this anywhere from 1 to 15 years. That is not very useful, but at least it gives me the boundaries of when I can know if this analysis is correct. It also tells me that Palantir stock may have a 15x growth in 15 years as a minimum and a 15x growth in 1-3 years as a maximum. I believe the truth is not always in the middle as the saying goes, but somewhere in-between the statistical range. Therefore, knowing that markets are forward looking, I'd expect that the stock may reach as high as $90, $120 and $180 somewhere in the next 3-5 years.
A comparison between Palantir and FAANG companies
I brought up Tesla and Nvidia together with Google to make the distinction that in my mind, Palantir is much more like Bidu and Google as it offers AI digital services (software). A big difference is that these both started as search engines and are expanding into other segments, while Palantir has started directly with AI and is the leader in the space of software AI sales. Palantir is also very different than Microsoft, Google, Apple, Tesla and Nvidia which all offer (hardware, or a combination of both). Steve Jobs famously said that a company which is "serious about its software should make its own hardware". Therefore it remains to be seen if Palantir actually will make hardware products in the future (and become more like Apple and Microsoft not like Google). If remains only in software, it will be more like Google, Bidu and Meta. The biggest difference with Meta is that Meta is a social network where users create content, Palantir's Gotham operating system and its role of Apollo developers (both products described in the previous section), it relies on providing software as an operating system for enterprises and allows developers to build additional tools on the platform, which other companies can purchase. In that sense, Palantir is like many of the FAANG companies and also a very different one from them.
Answer to sub-question 1:
With these pieces I can already briefly answer my first sub-question - Yes, Palantir is a technology company which is a leader in the software AI space just like Google was back in 2004-2009 when it was dominating every competitor in the search engine space. It was taking overtaking Microsoft's share of the browser usage by double digits percentage every year. Ironically Google is now losing share of its browser market (for instance as of 2019 I have never used Google Chrome, a round 10-15 year usage). Remind me to link here what I think is the best browser technology today and what I use as my new browser (hint, it's browser which is a blockchain-based solution (Web3.0), which I find much better than the traditional layer 2.0 Internet). On Web 1, Microsoft were the gods as people even used Microsoft software to code websites and everything else.
While you might wonder what does this have to do with Palantir, bear with me for a second. The example above shows how two of these companies Microsoft and Google had their respective 10 years of total domination of competitors. For Microsoft, this was Web1, For Google it was Web2. from about 2010 to about 2020. As of 2019-2020 there are emerging competitors even for Google (but not as a whole) but to at least some of their products (and a potential new leader which may dominate the browsing sub-segment in the next 5-10 years).
The relevance to Palantir is that Palantir has not yet had it's 10 years of absolute dominance. The company listed it's stock in September 2021, which was about 6 months before my projected 2021-2022 crisis started. Therefore, it's stock has never yet had the opportunity to grow. It also means, even if it's dominating a space, it's dominating a space in the times of crisis. It is yet to grow as fast or faster than the AI market and yet have it's share of the future 12-14x increased AI market by 2030.
Now, let me turn my focus to my second sub-question: Is Palantir technologies a stock that is likely to outperform META (formally Facebook), Apple, Amazon, Netflix/Nvidia and Google (known as the FAANG stocks)?
To quantity this and have a visual way of understanding what needs to happen, we might look into the graphs below. I look at PLTR/NVDA and PLTR/TSLA below
1. Palantir vs Tesla
As it can be seen from Figure 4 below, the outperformance for Palantir vs Tesla may have started as of October 2022. I we can definitely see that actually Palantir stock grew about from about 0.02 Tesla for each PLTR stock to about 0.06 Tesla for each PLTR stock (i.e. PLTR stock buys 200% more TSLA stock) from October 2022 to February 2023. The value is currently at about 0.043-0.045 Tesla shares for each single PLTR share, or a 100% growth for PLTR/TSLA since the low of October in 2022 to now in March 2023. A technical analyst would say that for this trend to stay valid, we must not break below the 0.618% (0.04 shares)
Figure 4. PLTR/TSLA (the value of each Palantir share/the value of each Tesla share)
This pattern suggests that at minimum it is likely for Palantir stock to grow about 200% relative to Tesla stock (from 1 Palantir share buying 0.04 Tesla shares to one Palantir stock buying at least 0.12 Tesla stocks). In the Figure above, this is the range indicated by the range between two blue lines (from 0.10 to 0.16 Tesla shares for each PLTR share).
The most aggressive case would be something like for each PLTR share to reach a value of 0.36-0.40 Tesla shares from 0.04 Tesla shares for each PLTR share. This would be about 9x vs Tesla shares (i.e. I can potentially buy 9x more Tesla shares if I sell at a ratio of PLTR/TSLA = 0.4, from today's 0.04).
Is this logical or random nonsense?
The EV CAGR is expected to be 10-20x in from 2023 to 2040, which you can see in my piece on Tesla in the market overview section of the piece somewhere in the middle of the post (available
here). At the same AI is expected to grow about 12-14x by 2030 alone for AI as a total and software AI specifically (refer back to Figures 1 and 2 above). Therefore, if software AI is expected to grow 10x by 2030, it makes sense than somewhere as of 2022-2023 we can start seeing outperformance for PLTR over TSLA.
In other words, if the EV market is expected to grow 10x-20x for 17 years, but the Software AI market is expected to grow 12-14x in 8 years, then the software AI industry will grow nearly twice as fast as the EV industry. Therefore, if these analyses are correct, there might be a logical causal explanation to support the reason why Palantir might grow faster than Tesla (and other FAANG companies).
Researcher's Note: I have performed similar calculations on AI vs all FAANG markets (and all suggest Software AI, the focus of PLTR will grow the fastest). As mentioned, the risk of this approach is that even though these expected growth numbers come from professional agencies which charge about $10,000 per report, they might also be wrong. Therefore, if their math on anticipated growth is wrong, it is likely that my own extrapolation of their math could be wrong too (in other words, we return to the basic concept of investing which suggests that to obtain higher return, one must pay a price in bearing a higher risk).
Concluding thoughts on PLTR/TSLA
For now, it is unclear where the top of PLTR/Tesla will be. However, a reasonable starting target for PLTR shares would be to grow by at least 200%-350% from 0.04 Tesla shares for each PLTR share to 0.12- 0.18. It may also be the start of a larger move, given the expected growth of PLTR and its dominant position the software AI space and potentially to grow to buy 0.36-0.40 Tesla shares per 1 PLTR share (this would be a 9-10x or 900%-1000% increase vs Tesla).
I know at least one person who reads this will know that Palantir was cofounded by Peter Thiel, currently worth about $8 billion and being about the 300th richest person in the world. He is also the original outside investor in Meta (Facebook). Lastly, Thiel is also a co-founder of PayPal.
Here is a photo of young Peter and young Elon, side by side. I mentioned him because Elon is not the richest person the world, while Peter Thiel is respectable at N300. Is Palantir the way in which Thiel catches up to his former colleague's fame and worth? This is just an anecdote of course.
I have looked into the patterns of the remaining FAANGs too. Apple, Amazon, Google and Netflix look similar to PLTR/TSLA. In the case of all of the above, PLTR is likely to outperform them even more than it does TESLA. In other words, their stocks are expected to be weaker than Tesla stock, which itself is likely to be weaker than PLTR stock. The only exception to this is NVDA stock, which I discuss below.
2. Palantir vs Nvidia
Ideally by now it should have become that I did not talk about Nvidia as the hardware leader for AI for no reason. I described this as the company providing the infrastructure for AI tools/robots/cars and all devices which need to "think, decide, see, and make decisions on their own (i.e. process information on their own and decide on one of many programs to execute)". Due to these complex functions that AI cars or AI robots have to perform (and the enormous amount of data they must be able to process in seconds), this will require nearly all these devices to need Nvidia's GPUs, and some of which will need Nvidia's $10-70k A100 GPU, designed for AI. Imagine your self-driving car having a video a card that cannot process new information as fast as needed. This may lead to mistakes which cause death to the passenger. Therefore, I think most companies will opt-in for Nvidia's solutions (until they deliver a cheaper one or competitors deliver a cheaper and equally/better functional alternative).
In short, NVIDIA can be thought of as the company providing the hardware infrastructure, at least in terms of semiconductors and GPUs which are needed for the development of other physical AI products, it therefore provides the infrastructure upon which such products can be built (like roads for cars, or rails for trains).
I also explained that software AI is expected to grow faster than hardware AI. So at one point it should be logical that Palantir will outperform even Nvidia. However, as I explained on my NVDA/Tesla piece (
here), I actually favour owning Nvidia over Tesla. As shown by Figure 5 below, in the near term, I also cannot be sure that Palantir will not register a new low relative to NVIDIA from about 0.035 shares of NVIDIA to about 0.015-0.01 shares of NVIDIA per 1 PLTR share.
While it may seem illogical at first that we are talking about PLTR's potential outperformance and at the same time I am showing that PLTR/NVDA could have a decline of 50 to 80% or a decrease in the purchasing power of 1 PLTR stock relative to NVDA stocks of 50 to 80%. This is why I outlined Palantir as a potential opportunity vs Nvidia in the future, but not yet as of now (the first blue arrow pointing down on Figure 5). One argument here would be that for companies in the AI space to need AI software, its players will first need AI hardware. Another argument to be added here is that while PLTR may be the dominant player 1 market - the software AI, but Nvidia is the dominant and key player for many markets - GPUs for AI, gaming, bitcoin mining + AI-specific GPUs + semiconductors (to simplify). The most important for me in this context is the Bitcoin mining one. This means, that until mid 2023 to mid 2026, it may as well be that Nvidia outperforms PLTR due to the combination of both the growth in Bitcoin and its growth in AI-related GPUs.
Figure 5. PLTR/NVDA (the value of each Palantir share/the value of each Nvidia share)
In other words, while there is an opportunity to use PLTR stock to buy more of any of the FAANGS stocks as of now, the only exception to this is likely to be NVIDIA. In that case we may as well have to wait anywhere between 3 months to 3 years before PLTR stock starts to significantly outperform NVIDIA stock. Therefore, NVIDIA is still the best stock to hold for at least 3 more months and may be the best stock to own to Q1 2026 or until 1 PLTR/NVDA becomes = 0.001-0.015 (i.e. the guideline is not time, the guideline is reaching a new all time low in the ratio of PLTR/NVDA before a potential reversal in this trend takes place. Such a reversal would be indicated by breaking the upper channel line (indicated by the second blue line on Figure 5, the arrow pointing up from the bottom of the channel).
In other words, while strategic management literature always focuses on finding strategic emerging opportunities for an underlying company (be it a new market, or a new market segment or a new product), the same logic can be applied for finding new investable companies in a strategic way. And while my market overview was extremely useful in providing future market size growth projections, it is in no way helpful for me to decide when to start favouring the ownership of PLTR stock. Figure 5, on the other hand suggests to me that once the ratio of each PLTR/NVDA stock makes a new low, this is likely to be the final low. I will have a
confirmation of this view, which will come only when the PLTR/NVDA ratio breaks above the upper channel line. This gives a much clearer point from which the view that PLTR has stopped underperforming NVDA and has started outperforming it is in effect.
The specific potential opportunity in PLTR/NVDA
Scenario 1: new low PLTR/NVDA
If PLTR/NVDA makes a new low at a ratio of above 0.01, to me this would signal the optimal reward/risk point for "blind entry" (i.e. trying to pick the exact low). This would be worth it in my opinion as it is a potential and amazing opportunity to grow the value of each PLTR share from 0.01 to 0.07 - 0.13 NVDA shares (or a 7-13x growth for PLTR in terms of NVDA). Even if the dominance had a cyclical rotation (meaning it is range bound between 0.01 and 0.35-0.4), this would suggest that somewhere in the future, it might not be unreasonable to expect each PLTR share to buy 0.35-0.40 NVDA shares (this would represent a 35-40x from a low of 0.01).
The key point here is not to guess the exact ratio or the exact timeline, but to show that once the purchasing power of 1 PLTR share may fall by another 50%-80% relative to 1 Nvidia, I am likely to start selling my NVIDIA position into PLTR, as it would become likely that from there PLTR stock grows 7-13x to 30-60x vs NVIDIA. In other words, to me this would be very useful in thinking when to start favouring PLTR stock.
Scenario 2: bottom is not 0.01 it is 0.03
What if no new -50-80% drop for PLTR/NVDA, but say only another -10/-15%? The same, if I were to assume that the bottom is in at about 0,03 (refer back to Figure 5, the middle of the channel range), i.e. the arrow pointing down from 0.035 takes us slightly lower to 0.03 but not to 0.01 (shorter downtrend), I will again need to see a move like the second arrow on Figure 5, which takes the price of PLTR/NVDA above the upper channel line (and once it breaks it up, holds it as support).
In both scenario 1 and 2, the conditions for these to ever be valid and considered is only if PLTR/NVDA price breaks out of the upper channel. Only then, we may have a confirmation of this trend reversal or a visual representation of PLTR stock "talking" to an NVDIA stock.
What remains unclear?
As of now, it is unclear what the company will need to become for it's stock to grow 10-30-60x vs NVIDIA. Therefore, projecting a potential 7-13x to the to 0.07-0.13 from 0.1 is a good start. It is way too early to talk about a re-test of the 0.4 or for each PLTR to buy 0.4 NVDA shares as was the case back in Dec 2021 (see Figure 5 above again).
As of now when 1 PLTR share buys 0.035 Nvidia shares. The highest expectation I can have from here is about 9-10x growth for PLTR stock relative to NVDA from this 0.035 value to about 0.32-0.35 (under scenario 2, no new lows) or about 30x from a new low of 0.01 and 20x from 0.015 (scenario of new lows at 0.01-0.015). So in case I am wrong on expecting new lows here, I already have my initial position, which is also only 1/10 of my NVIDIA position. This is also to force me to track this relationship and not miss the moment which I think is right to convert.
Author's note on cross-charts
The fact that PLTR/NVDA could see a new low, does not mean that Palantir will fall in price (and make a new low in terms of $ value) and Nvidia will grow in price in $ value at the same time (this would be a wrong interpretation of these so-called cross-charts). A cross chart suggests that 1) if both are up, in the short-run Nvidia might grow faster and in the mid term PLTR may grow faster. It also suggests that 2) if both are down, Nvidia is likely to be down by a smaller percentage than PLTR (in the short term).
So these are non-directional charts which do not tell us about whether the prices of these stocks will move up or down in terms of dollars.
The only conclusion we can derive up to here is binary: The outperformance that this cross-chart illustrates will be due to a drop or increase in the price of both (where Nvidia either grows faster or drops less in the short term).
To turn into a 1 or a 0: It is only when I combine this interpretation with the view that AI market will grow by 12-14x in 8y that causes me to interpret that this outperformance for PLTR will happen not when both stocks are down, but when both stocks are up.
Everything until here (ex my author's and researcher's notes) would be the arguments to formulate hypotheses and test whether Palantir stock can actually outgrow the well-known FAANNG stocks + Tesla (i.e. to explain that there might be a reason to 1) expect PLTR to outperform, 2) to test this expected outcome and 3) to comment on the experiment's findings).
Testable Hypotheses:
For me to answer the second sub-question with evidence (i.e. to collect data and confirm reject my expectations regarding PLTR stock), I will need to formulate at least 6 hypotheses. In other words because I am comparing one thing (PLTR) to 6 (FAANNG), in academia I would have to establish at least 6 hypothesis turning each of these into a testable statement.
Hypothesis is an assumption or an idea proposed for the sake of argument so that it can be tested. It is a precise, testable statement of what the researcher predicts will be outcome of the study. It is an integral part of the scientific method that forms the basis of scientific experiments.
Unlike those in an experiment, these are called - directional hypotheses
They specify the expected direction to be followed to determine the relationship between variables and are derived from theory. Furthermore, they imply the researcher’s intellectual commitment to a particular outcome.
Below I have outlined my 10 directional hypotheses based on this piece. I like hypotheses as a tool to outline statements which can be turned into a testable prediction. After all, we all have our hobbies and mine is to write research and test my own hypotheses.
Hypothesis 1: Palantir stock will outperform Meta stock in terms of return by Q1 2026
Hypothesis 2: Palantir stock will outperform Apple stock in terms of return by Q1 2026
Hypothesis 3: Palantir stock will outperform Amazon stock in terms of return by Q1 2026
Hypothesis 4: Palantir stock will outperform Netflix stock in terms of return by Q1 2026
Hypothesis 5:Palantir stock will UNDERPERFORM Nvidia stock in terms of return by mid 2023 and start to outperform either 1) by Q1 2026 or 2) when the ratio of price for 1 PLTR/NVDA becomes equal 0.01 (from 0.035 today).
Hypothesis 6: Palantir stock will outperform Google stock in terms of return by Q1 2026
Hypothesis 7: Palantir stock will outperform Tesla stock in terms of return by Q1 2026
Hypothesis 8: Palantir stock will UNDERPERFORM crypto mining stocks to Q1 2025 because bitcoin mining stocks and the bitcoin mining industry will grow faster than the AI industry to 2025
Hypothesis 8a: Palantir will underperform RIOT
Hypothesis 8b: Palantir will underperform MARA
Hypothesis 8c: Palantir will underperform BTBT
In this way, I have an amazing set of 10 hypotheses (N1-7+ 8a,8b,8c) that allow us to test in reality whether Palantir stock will bring better returns than FAANNG stocks (hypotheses 1-6), hypothesis 7 will allow us to test whether Palantir will grow faster than Tesla stock
The idea is to find out is Palantir better than all these other 7 stocks which are considered the most attractive tech companies today. In other words, the main research question is aimed at finding out if Palantir is a type of company that resembles and has the growth prospects of a "future Google" and if yes, is it reasonable to anticipate it that it may start having the type of growth that Google had over the last 10 years (both as a company and a stock).
Lastly, hypotheses 8a, 8b, and 8c, will also allow us to further test the hypotheses in my other piece about Bitcoin mining stocks being the most interesting tech sub-segment (read more here). The idea would be simple, if hypotheses 1-7 are accepted, then Palantir is the best tech stock own as of today relative to these other 7 tech examples (FAANNG + TESLA). However, the 3 stocks which might be even better to own in terms of return do not come from AI sub-segment of tech, but from Bitcoin mining. In other words, the last 3 hypotheses allow us to test the hypothesis of the previous piece about Bitcoin mining and the statement that Bitcoin mining stocks are the best tech sub-segment to own to maximize returns.
In short, we will compare Palantir to a bunch of other technology companies which represent different tech sub-segments (e.g. Meta is a social media, Tesla is an EV maker and Amazon is an online retailer). In reality, each of these companies has more than just its core business, but the logic is that they are representative of the types of tech companies out there (I will not bore you with the details, but let's say that for an experiment to be valid we must compare to a wide range of tech companies to be able to conclude anything about it).
From there, I will compare Palantir stock to crypto mining stocks in Dec 2025 to reflect whether all these hypotheses were accepted or rejected.
The purpose of these hypotheses is for me to summarize my opinion in a few statements, test them and write subsequent pieces in which I discuss whether these hypotheses were accepted or rejected, as a requirement of the scientific method to be valid and a continuation of this hobby.
Implications for my personal investing
Given that the company's stock had a low of about $6 from a top of $45 and an IPO price of $11.3, this suggests that buying the stock now would definitely mean buying it at a -50% from IPO price and -87% from it's top. To me this means that the IPO pump and dump part of the trend has already taken place (I have been wrong on these before 1 of 10 times). This simply means as the stock started trading at 11 and was pumped by funds and retail to 45 (about a 3x gain), initial pre-IPO investors who had the stock options at 2-3-5 sold to take a 20-30-100x profit. This also explains why the price is so low, because the supply of shares has increased due to this selling which was never demanded as asset allocators were dropping shares exposure from end of 2021 to end of 2022. In other words, while these pre-IPO investors made money before the Russia/Ukraine war, this stock was never demanded (because who thinks of future AI and growth when there is a war).
Maybe this is me being too opinionated, but I try to show that to me, the dynamics behind Palantir's stock price is explained by 3 things - 1) pre-IPO investors selling post IPO at 10-100x profits, 2) the war, which left the sold shares without demand and those who bought at 20-30-40, either patiently waiting (or having sold off their positions at -50-80%) and 3) AI growth has barely begun in 2023 and is yet to gain any traction by 2030. In other words, the worst has passed and I believe the best is yet to come.
One negative thing for me would be that Palantir like Meta and Google is likely to take advantage of the lack of laws on privacy of data in that field and do something similar like Meta and Google have been doing in the past years (to sell data of users) to further increase profitability. This particular last bit I would not take to the bank, but represents my own speculation based on observing US companies' way of doing business. It is to say, I would not be surprised if they cross a line in the name of profitability, which also supports the idea of higher stock price.
Above I also showed, that Nvidia might actually outperform Palantir in the short run. Therefore, the implication is that I can already overweight Palantir relative to other FAANNG stocks (including Netflix), but excluding Nvidia. In other words, Nvidia remains the only stock that I'd keep in a larger percentage allocation relative to all other tech stocks, including Palantir.
I hope I have also managed to show that I like Nvidia for 1) its AI-hardware-centric GPUs which will be needed as the infrastructure for devices that need to "think, see, and decide outcomes or implement processes", 2) GPU usage in Bitcoin mining (making it an indirect and less risky play on the Bitcoin trend).
If you are curious you can see how my opinion is working out for me you can click here.
Reflections on differences with an actual academic piece
In this piece I aimed to offer a few things, 1) an example of how to formulate a research question that can be broken down into sub-questions which provide an easier way to answer the main question. In addition, I showed that answering the first part of my sub-question can easily be done by using traditional business evaluation models (the business student section). In other words, by compiling already published qualitative (information) and quantitative data (like the market size, growth of the AI industry), I was able to answer 50% of the total question (i.e. to show the importance of doing research based on reports from professional research agencies, which already exists). The limitation of answering research questions even partially with other people's data is that it can be wrong. I used 3 graphs, and did not search for another 10 pieces of evidence from other sources (i.e. to be a valid argument in academia) I would have had to link multiple pieces of evidence (on AI growth for example from different sources).
The other 50% (the second sub-question), cannot be answered by any other research because no other research was attempting to answer this question that I am asking. Therefore, to answer this now unique question (or sub-question 2), I had to formulate my own hypotheses. I also showed some simplified arguments behind these hypotheses. Now I have to see if they will get accepted or rejected and then I can answer the second sub-question. For now, for me there is some small evidence to believe Palantir might be a big company and its stock may produce higher gains of FAANNG stocks. One possible explanation is that Palantir's underlying industry is expected to grow faster than that of all FAANG stocks, i.e. software AI is the only industry projected to grow by 41.3% per year for the next 7-8 years). Another one is that Palantir stock never had a bull run as a stock as it was introduced 9 months after COVID-19 and 3-6 months before the Russia/Ukraine war. My expectation due to the war is that if anything, the war helps Palantir sell more defence solutions to the US defence department, homeland security and FBI (I did not look into whether this is true, but if a war doesn't "create" a market need for this, what else will?).
Reminder
If this is the first piece that you read on this blog, I strongly advise you to go to the "purpose of the blog" page to understand what this blog is about. You can do this by clicking
here. At least scroll down to the middle of the page and read the section called - "What does this blog include?" for an explanation on each of my blog series and the type of content that is included in each of the series.
Disclaimer
No one should buy or sell anything because of anything I post inside my blog. This is all that it is, a way for me to track my own thoughts, organize my research and share my unique opinions. Any opinion posted on this blog is just that - an opinion. The purpose of these pieces are to act as a motivational information tool. This is a blog about my life goals and the decisions I use to try to achieve them.
Nothing on this Blog constitutes financial or investment advice, a recommendation that any security/cryptocurrency, portfolio of securities/cryptocurrencies, other investment products, transaction or investment strategy is suitable for any specific person. Every person has unique personal circumstances, financial situation, goals, and risk capacity. No one should use the content of this Blog to make financial decisions.
In short, nothing in this blog is financial advice! Do your own research & Do your own due diligence. If you have investment questions, I highly recommend you seek help from an authorized (regulated) financial planner, which I am not.
