➤ Patreon Club Members access to entire S&P 500: tinyurl.com/sp500february2025 ➤ RUclips Club Members access to entire S&P 500: ruclips.net/user/postUgkxKuHYa0ZCuvQgumIb8343yD5vzqXdtCQV ➤ Patreon FREE member access to Dow Jones: tinyurl.com/february2025dow ➤ Member-only livestream on February 10, 2025 at 2 PM ET on Google Meet (check Patreon and RUclips for links!)
The biggest lesson I learned in 2024 about the stock market is that nobody knows what will happen next, so practice some humility and low a strategy with a long-term edge.
Nobody knows anything; You need to create your process, manage risk, and stick to the plan, through thick or thin, While also continuously learning from mistakes and improving.
@@MercuriosBakers Inflation is gradually going to become part of us and due to that fact, any money you keep in cash or a low-interest account declines in value each year. Investing is the only way to make your money grow. Unless you have an exceptionally high income, investing is the only way most people will have enough money.
@@ShepherdsHarleys I've shuffled through investment coaches; they can positively impact an individual's portfolio. But do your due diligence to find a coach with grit that withstood the 2008 crash. MARGARET MOLLI ALVEY was better and smarter than all the advisors I have ever worked with. I’ve never met anyone with as much conviction.
The last time I invested in a stock that did extremely well was during Covid when I invested in Moderna & Zoom and I am so sad that I missed out on Nvidia. I have $360,000 in cash and am looking for new sectors to invest in for the next five years. Any suggestions?
I think AI is entering a new phase, and you should look into top companies investing in AI research. With your budget, it's a good idea to consult a financial advisor for personalized advice.
I agree, working with an advisor helped me build a $1.6m stock portfolio. Her guidance was invaluable, especially as an early investor in NVDA, AVGO, ANSS, and LRCX. Information, insight, and predictability are crucial in the stock market.
I'm cautious about giving specific recommendations since this is an online forum and everyone situation is unique, but I've worked with Jessica Dawn Walters for years and highly recommend her. Look her up to see if she meets your criteria.
She goes by Jessica Dawn Walters I suggest you look her up. To be honest, I almost didn't buy the idea of letting someone handle growing my finance, but so glad I did.
I am a numbers person and I think you can easily get an AVERAGE growth rate over the next 5 years of 30% and over 10 years in the 18-22% range. I have owned it since 2020 and continued to add throughout so I have some with 800% profit with some at 30%. My assumption could be quite low but it's better to run with something conservative that you can live with and be pleasantly surprised over the 10 year period. I tend to look in the 5 year range since technology moves so quickly. I am most impressed with Jensen's forward thinking about where things are headed and preparing a product and service to meet that growing need that others cannot see. Only own a few companies, know them inside and out and don't let the fact they are a stock impact your desire to own them. You own a company! My Buffett approach. Good luck my friends
The best way to find Nvidia’s EPS in 10 years is to call up Microsoft, Meta, Amazon, Alphabet and Tesla and ask them how much they plan to spend on GPU’s 😂
I would say if rigetti or a quantum company invents a gpu that’s far more superior then nvidia. Thats when nvidia will deflate into another intel. Unless nvidia buys that company out. Then nvidia will continue to reign supreme.
Nathan.... you have really found your path! As a longtime NVDA holder (who has taken out every nickel invested and a few more) this was very satisfying. It's very difficult to walk away from NVDA and at times very difficult to stay on board. I did weather the 2020 drawdown by taking consolation that I never got into the red and now with all of my investment taken out I have to remind myself I really can't lose.
That is interesting to hear the simulation results. Duly noted. Predicting the future is still just a roll of the dice. It was a wonderful dip yesterday, Monday Feb 03.
Different inflation rates would be implied within the existing simulations, to some degree. So if inflation was super high, earnings grow faster, but multiples contract, that’s one of the simulations.
@ I think with historical Monte Carlo you end up with a 1000+ useless predictions rather than one. I don’t think markets are random (Monte Carlo). Also doesn’t stdev imply symmetry and bell curve returns distribution? Is that likely? If you’d run a Monte Carlo on Nvidia 3 years ago, would the subsequent returns have fallen within the prediction?
I personally can't imagine a company being worth 20 trillion ($800 price) - 50 trillion ($2000 price) in 10 years unless something insane happens with the company that would probably be extremely monopolistic... but if regulation is stripped out, it could happen if it's allowed to run rampant. Not sure I like the implication of an AI/robotics company being worth that much, as it likely means massive job loss, robots everywhere, AI agents, "paperclip problem" optimization scenarios, AI being so saturated in society that life is claustrophobic (i.e. where it doesn't need to be, societally, or shouldn't be, legally), not to mention anticompetitive practices, warfare... or one could go optimistic and say it's uniquely positioned to solve the world's solvable problems, like fusion power, quantum computing, climate change, simulation, medical applications, space travel/colonization, taking over impossible or dangerous jobs etc... probably all... interesting modeling
Without any simulation, I can tell you it's very likely in the near term it will be worth a hell of a lot less. If the bullish estimate here were true, NVDA would have to be close to 100% of the value of the entire US stock market.
Currently, the P/E of NVDA is sitting around 40-50. In the next 10 years, the P/E would probably be normalized to about 25-30. If Nvidia maintains 25% annual EPS growth, its 10-year stock price estimate is between $2,500 - $3,000 per share, assuming a normalized P/E of 25-30. Of course there are risks such as AI demand slowdown, regulatory challenges, and competition from other companies. Your Monte Carlo Model is not too far away from my estimates. There are challenges for Nvidia but based on what I see, the demand for new better and faster hardware will continue to accelerate. AI is still in its infancy, and there is growth in the area of computer vision, autonomous systems, and general intelligence (AGI). With that said, I would not worry about the noises around Deepseek on Wall Street.
I love the concept of monte carlo simulations. However, I can see where the parameters can really skew the probability numbers, especially in the chip sector where things are deeply cyclical. IMO, it's predictibility would be much if the optimistic end of the spectrum is tempered. 10 years is an eternity in the chip sector, and we are in a period of peak optimism. Or maybe even better yet, (I say this not knowing how you built your model) you more heavily weight the growth rate probability in the early years than the later years to account for the increasing uncertainty. For NVDA, it's all going to boil down to how well or even if they can maintain their competitive advantage and where we are in the cycle. I could easily see them running hard for 3-5 more years and then hitting a wall.
Also I don't get the assumption of a 50x multiple in 10 years (which is the average you seem to assume) whatsoever. If earnings growth drops from 40% over the last 10 years to 10% as you assume, then there is no way that anyone will still be buying this stock at a 50x multiple.
@@zweidritteleinfach2087 The stock is priced based on hyper scale capex spend. Since deep seek showed that in theory you don’t need as much hardware, bc you can create more efficient running models, this means that hyper scale capex spend will drop and the stock price corrected. What the market hasn’t priced in is all the business use cases for AI that will require nvidia chips to run. There’s two camps here - the camp that says AI is a tech fad and it won’t progress beyond the hyper scalers and the other camp which sees AI evolving into AGI that will be a legitimate human replacement. Which camp you fall into will determine how bullish you are on this stock…
It's because there's massive uncertainty in what they'll be able to do in 10 years. This is a company that literally invented the GPU and has been able to find new uses for it over the past three decades. FSD cars, robots, software, in addition to maturing current AI trends. But notice that the simulation shows a very low probability for those outcomes.
5 years is a more accurate time frame to work with. 30-35 PE avg is more likely over 5 years. 10 years is too long to estimate based on the ever changing industry. After 3 years do another 3-5 year estimate
Nathan, I think you are doing a good thing to help people understand it's about the numbers that help one drive their investment decisions. I managed monies for many years and most people want the return but are unwilling to stay on the ride they jumped on reach their desired destination. While I think 10 years is to long a time frame reassessing the ride over 3-5 year periods of time works better to keep the people focused, you are doing a good thing to get people to NOT look at just the price of a stock but the real underlying value
I don't get the point or the need to use some sort of Monte Carlo simulation to do this. Starting at a 40% annual earnings growth and assuming this drops to 10% over 10 years and that this drop follows an arithmetic sequence you can get the answer on a pocket calculator in a couple of minutes. That result will have very much the same predictive power as running 100 000 MC simulations, since deviations from the mean are bound to cancel out. The key parameter here is the assumption that earnings growth drops from 40% to 10% over 10 years. That assumption will have a vastly higher influence on the resulting fair value estimate than whichever computational model is used after the fact.
For the median result, sure, it’ll be extremely similar. The primary value, in my mind, is the ability to model uncertainly and understand what kind of downside you could have if things go poorly; in effect, build in a margin of safety.
S3 Partners tweeted on 1/28, "Short Interest is now $32.12 billion, 271.3 million shares shorted, 1.15% SI % Float. $NVDA is the largest U.S. short ahead of $AAPL, $MSFT and $TSLA."
There's no right answer but ideally you want no more than 5-20% of your portfolio into any single stock, and only over 10% if you're extremely optimistic on a company. So if you have say $100k to invest you might want to buy 84 shares for 10% of your portfolio ($119 per share x 84 shares = about $10,000). Double that if you're very optimistic, or half if you're not as optimistic.
Why 10 years investment horizon for high growth stocks? Most of the time high growth is not sustainable for long periods.. Looking at history.. Remember Yahoo, Nokia, Blackberry…
Investors should be open to changing their mind when they notice that thesis for buying the company has changed. Buy with intention to hold for 10 years, but sell when company no longer resembles the one you brought
I am not sure how to access new Monte Carlo simulations with my RUclips membership. I saw Coffee Can Scorecards after hitting the link, but no simulations. Help?
@@NathanWinklepleckCFA OK...I was looking for the 100,000 possible futures data for estimated 10 year future valuations as mentioned in the video. Perhaps I misunderstood. Thank you.
In other words, I see the redesigned score cards in the preview under the membership tab, but the Google Doc link downloads the older version score cards with no futures data.
Is there any sort of excel template where we can input the numbers and run the simulations ourself? I would love to do this for other stocks as well! Thanks
AI, robotics and a whole lot more are just getting started, so it would foolish to assume that the defacto industry leader wouldn't continue to benefit significantly.
Let’s make a simple calculation. If the S&P500 has an average return of ~9% per year (historically), then in 10 years it would appreciate: (1.09)^10 times, or 2.37 times. This is not MonteCarlo, but a good approximation. Now, per Nathan’s calculation, the median appreciation for NVDA over the same period would be ~484/125 or 3.87. In other words, 3.87/2.36 = 1.64 times the S&P500. This is a slam dunk. Just keep doing DCA and keep including NVDA in your portfolio. That’s it.
If a company's earnings doesn't affect the actual stock price, it only affects what people might be willing to pay, wouldn't focusing on the willingness of people to pay a certain price be more relevant to identifying future price? So that would be more technical analysis, which identities more of the supply/demand dynamics rather than the hypocritical value of the stock.
Eh, that gets dangerous, imho. Then there is really no limit to what you’ll pay. The only way that works is if others are basing their analysis on something “real” but that doesn’t always turn out to be true. (Ie, houses in 2007, tech in 1999, meme stocks, and a bunch of other examples going back to tulip prices)
@NathanWinklepleckCFA what's the theory behind buying a stock based on the value of the company without regard to what people actually want to pay right now? In a more technical way, why would people change their minds in the future? Why should the company value impact what people want to pay?
I consider this a flawed statistical analysis in that it assumes independent variables. They are not. In billish market conditions earnings growth and P/E ratio are pretty strongly correlated and not independent variables.
Fair points! I could modify the simulation to include some sort of positive correlation, but was relying on the sheer volume to generate reasonable price distributions. Any ideas how you would “fix” it?
@ I use Monte-Carlo simulation together with Bayesian network. In BN, I would model P/E by a conditional probability table linked to earning growth and would give the CPT parameters reasonable sampling distributions.
@@HeranJaskaran I think it will go up. It may hit 110 on the down side first. But earnings out 20th feb. Then 125 will be super cheap. I think it can be closer to 200 by the end 2025. Deepseek won't affect Nvidia much. Microsoft already said they will spend 80bn on data centres and that hasn't changed meta said he sees Deepseek as a positive. Earnings growth won't slow much 60-100% high margins, Fantastic cash generation. Buying back shares and authorized 50bn more. I see this as a temporary dip and a great Buying opportunity
@@quangyung6751 😵💫have been holding in the hopes but it backfired due to this news. Not so sure now i think it will reach 110ish shortly. Long term it will recover but doesn’t look like stabilizing shortly.
Not sure you need to use MCMC. your assumption already led your prediction. Simply based on past average and std to prediction the price is not very useful. First, we need to understand whether AI is the major development tech in next 10 years. 2nd: is NVDA Gpu will dominate the AI hardware technology in next 10 years? Deepseek and AI algorithm improvements do not conflict with AI hardware development but it could shift little bit of NVDA gpu demanding in a short term. For longer term, the computation speed will dominate AI developments. 10 year is too far to predict, but we can assume that nvidia will lead the AI hardware for at least next 5 years. Assuming 40% growth in next 5 years and p/e 50, the price should be at $730 level with plus or minus $150 error window. I would add more nvda shares at today’s price.
lol the problem all of you have is you don’t understand AI or the evolution of it or even the amount of expansion of the tech that still left to grow. It’s gonna blow up next this year many contracts for this year and orders to be filled, soon China will be a buyer too since deep seek is using old Nvidia chips that they bought in a different country, said they bought 20% of Nvidias old chips
LOL...I literally become a billionaire if Nvidia hits 20,000 per share! And I'm only 59! I can't even take it out of my Roth(what's in my Roth)... Even JUST at 1200 per share, I end up the highest paid player in the NFL! I think most likely, in 10 years, Nvidia is in the 300-330 a share. And that's a solid return, but I'm 5 years in and that HUGE dividend that pays me...2200 bucks a year isn't going to get me to stay in. I suppose there could be a rise of the robots and because I'm their shareholder and I imagine the first rule in robotics will be, "protect the shareholder," maybe i should say on and hold at least half my shares!
"The Future" is AI and NVIDIA accounts for 65% of chips for AI. I imagine the impetus for this video is the DeepSeek bombshell. The DeepSeek news is distorted. DeepSeek is a reasoning model and therefore uses more processing power as it gets better. Just as we are taught to read until the 4th grade and then learn by reading after that, DeepSeek may been trained cheaply, but requires more power (more "reading" if you will) to get smarter, and the same chips required for training a reasoning model are required for making it smarter. The massive investment NVIDIA just made in Israel should illuminate "the future" as should the massive investment in Stargate the U.S. govt just made. I would compare it to the semiconductor industry producing chips integral to wireless technologies back in the 90s, say... Qualcomm.
@edwardrichardson8254 lol🤣🤣 deepseek is only used in phones and laptops plus they got nvidia chips illegal plus they truly haven't disclosed how fast it truly is ..remember it's China can't believe everything they say..plus if deepseek is to grow they will need nvidias better newer chips which they can't get ..so deepseek is a bust in my book
Hmmm… Not sure I agree with that, but it definitely could be speculative; I’ll grant you that! High price to pay for lofty expectations, but worthless? Nah…
what people dont realize is that when AI becomes more norm.. things that needs to run AI at optimal level? thats going to be mass production of something. as of now gpu cards runs AI superb so until something is build to overtake gpu cards? nvidia is still king
With ‘Physical AI’ on the cusp of playing out, am buying at every bit of NVIDIA’s dip! .. I believe the opportunity is so big that it’s only limited by our imagination.. once in a generation, life changing ..!
just look at Intel stock if you want to see what will happen to NVDA in a few years. 90% of computers still use Intel chips but yet the stock is 1/3 of what of was 25 years ago.
Intel made significant strategic errors and fell behind competition over the course of a decade. That could happen with NVDA, but leadership is far more forward-thinking, but it’s certainly possible!
@@NathanWinklepleckCFA I interacted with Intel technically decades ago & Intel's commpetitor at the time. The difference between the 2? Intel knew everything, just ask them.
@@NathanWinklepleckCFAas far as volume goes Intel are the # chip maker in the US. It has nothing to do with strategy and has to do with the expectations in the late 90s were too high. The same could happen with NVDA that is all I am saying. In the 90s expectation were too low for AMD.
I wonder how this system can calculate the S&P 500 who's companies are often changing over time? Who knows what future companies will be included or excluded let alone each of their data to enter into this system. A company like NVDA should do better than the S&P 500 as a whole so I would lean to the higher side of the average for NVDA in this case.....For now I have sold my position and will wait for a better time to buy it back. Long term "hodeling" like a helicopter parent as I like to call it, is not my thing because at my age I am running out of time.
@keaka560 I love this overused comment for every stock that go up. Can you please tell what was the CSCO P/E ratio before the dump and tell me todays NVIDIA P/E ratio. Let me help you as of yesterday NVIDIA P/E ratio was 26, it is waaay lower than Tesla, Palantir etc... and for CSCO??... wait let me help you on that too, may be somebody random is reading this post, Cisco (CSCO)'s highest recorded P/E ratio was around 2,128.5, which occurred in the July 2018 quarter, do you get it now? 2128.... 2128 vs 26...
@NathanWinklepleckCFA as a CPA, I look at these companies from a very simple point of view. NET ASSET VALUE per share. So you can say eps all you want, but there are two stories that always exist in a stock. What's the accounting value, and where do the players want the price to be. All I have to do with my player buddies is control enough of the free trading shares and we can wash trade all day. We accept the fact that there will be some longs and some shorts that we have to deal with, but if we can control the float, we can make our options pay us. We love 401k and IRA because it incentives holding and punishes premature sellers. This set of parameters neutralizes all models. Options are the tail that wags the dog.
Nvidia's growth rate will likely be more in line with 24.5% in the next couple of years. The tariffs and restrictions will result in a much lower growth opportunity. The trade war is going to get a lot worse and will not subside anytime soon. The world is sick of Trump and his lies and conspiracies.
➤ Patreon Club Members access to entire S&P 500: tinyurl.com/sp500february2025
➤ RUclips Club Members access to entire S&P 500: ruclips.net/user/postUgkxKuHYa0ZCuvQgumIb8343yD5vzqXdtCQV
➤ Patreon FREE member access to Dow Jones: tinyurl.com/february2025dow
➤ Member-only livestream on February 10, 2025 at 2 PM ET on Google Meet (check Patreon and RUclips for links!)
Your data is profoundly sound. Rational investor you are. - Yoda.
The biggest lesson I learned in 2024 about the stock market is that nobody knows what will happen next, so practice some humility and low a strategy with a long-term edge.
Nobody knows anything; You need to create your process, manage risk, and stick to the plan, through thick or thin, While also continuously learning from mistakes and improving.
@@MercuriosBakers Inflation is gradually going to become part of us and due to that fact, any money you keep in cash or a low-interest account declines in value each year. Investing is the only way to make your money grow. Unless you have an exceptionally high income, investing is the only way most people will have enough money.
@@ChristophersGonzaleza Could you kindly elaborate on the advisor's background and qualifications?
@@ShepherdsHarleys I've shuffled through investment coaches; they can positively impact an individual's portfolio. But do your due diligence to find a coach with grit that withstood the 2008 crash. MARGARET MOLLI ALVEY was better and smarter than all the advisors I have ever worked with. I’ve never met anyone with as much conviction.
@@ChristophersGonzaleza I just ran an online search on her name and came across her website; pretty well educated. thank you for sharing.
The last time I invested in a stock that did extremely well was during Covid when I invested in Moderna & Zoom and I am so sad that I missed out on Nvidia. I have $360,000 in cash and am looking for new sectors to invest in for the next five years. Any suggestions?
I think AI is entering a new phase, and you should look into top companies investing in AI research. With your budget, it's a good idea to consult a financial advisor for personalized advice.
I agree, working with an advisor helped me build a $1.6m stock portfolio. Her guidance was invaluable, especially as an early investor in NVDA, AVGO, ANSS, and LRCX. Information, insight, and predictability are crucial in the stock market.
I'm interested in trying this out. Who is your advisor, and how can I contact this person?
I'm cautious about giving specific recommendations since this is an online forum and everyone situation is unique, but I've worked with Jessica Dawn Walters for years and highly recommend her. Look her up to see if she meets your criteria.
She goes by Jessica Dawn Walters I suggest you look her up. To be honest, I almost didn't buy the idea of letting someone handle growing my finance, but so glad I did.
I am a numbers person and I think you can easily get an AVERAGE growth rate over the next 5 years of 30% and over 10 years in the 18-22% range. I have owned it since 2020 and continued to add throughout so I have some with 800% profit with some at 30%. My assumption could be quite low but it's better to run with something conservative that you can live with and be pleasantly surprised over the 10 year period. I tend to look in the 5 year range since technology moves so quickly. I am most impressed with Jensen's forward thinking about where things are headed and preparing a product and service to meet that growing need that others cannot see. Only own a few companies, know them inside and out and don't let the fact they are a stock impact your desire to own them. You own a company! My Buffett approach. Good luck my friends
The best way to find Nvidia’s EPS in 10 years is to call up Microsoft, Meta, Amazon, Alphabet and Tesla and ask them how much they plan to spend on GPU’s 😂
Answer: a lot
How long before there are cost effective alternatives? Or will AI become more cpu or npu centric
The moat keeps competition from popping up easily…
It will take more than the latest chip technology…
My guess is NVDA has way more buyers (countries , governments etc) of high end chips in the coming years then those 5 companies .
I would say if rigetti or a quantum company invents a gpu that’s far more superior then nvidia. Thats when nvidia will deflate into another intel. Unless nvidia buys that company out. Then nvidia will continue to reign supreme.
Where is the Deep Seek got AI hardwares from, Nathan?
Nathan.... you have really found your path! As a longtime NVDA holder (who has taken out every nickel invested and a few more) this was very satisfying. It's very difficult to walk away from NVDA and at times very difficult to stay on board. I did weather the 2020 drawdown by taking consolation that I never got into the red and now with all of my investment taken out I have to remind myself I really can't lose.
great share, Kimm! Good to see you, my friend
Great video, the information you provide along with explanation is amazing. Thank you!
You're very welcome!
New here, good analysis.. i like the wide range you give for probable outcomes. thanks
I bought the dip in Nvidia last week, so I hope at least the median sim happens!
Hopefully a great buy! 👍
Longer term you will be pleased as long as you stay on the ride or the wave.
@@akarpov949 I did the same. Hopefully it pays off. But I can’t imagine it going down. Did you go after PayPal?
That is interesting to hear the simulation results. Duly noted. Predicting the future is still just a roll of the dice. It was a wonderful dip yesterday, Monday Feb 03.
Can you add a factor for inflation?
Different inflation rates would be implied within the existing simulations, to some degree. So if inflation was super high, earnings grow faster, but multiples contract, that’s one of the simulations.
Monte Carlo and randomness isn’t a problem? Standard deviation too?
What?
@ I think with historical Monte Carlo you end up with a 1000+ useless predictions rather than one. I don’t think markets are random (Monte Carlo). Also doesn’t stdev imply symmetry and bell curve returns distribution? Is that likely? If you’d run a Monte Carlo on Nvidia 3 years ago, would the subsequent returns have fallen within the prediction?
I personally can't imagine a company being worth 20 trillion ($800 price) - 50 trillion ($2000 price) in 10 years unless something insane happens with the company that would probably be extremely monopolistic... but if regulation is stripped out, it could happen if it's allowed to run rampant. Not sure I like the implication of an AI/robotics company being worth that much, as it likely means massive job loss, robots everywhere, AI agents, "paperclip problem" optimization scenarios, AI being so saturated in society that life is claustrophobic (i.e. where it doesn't need to be, societally, or shouldn't be, legally), not to mention anticompetitive practices, warfare... or one could go optimistic and say it's uniquely positioned to solve the world's solvable problems, like fusion power, quantum computing, climate change, simulation, medical applications, space travel/colonization, taking over impossible or dangerous jobs etc... probably all... interesting modeling
Without any simulation, I can tell you it's very likely in the near term it will be worth a hell of a lot less. If the bullish estimate here were true, NVDA would have to be close to 100% of the value of the entire US stock market.
Always great to see Nathan video alert! Thank you!!
Thanks! Always good to see a David comment alert 🚨
Caould you do one video on PLTR?
Currently, the P/E of NVDA is sitting around 40-50. In the next 10 years, the P/E would probably be normalized to about 25-30. If Nvidia maintains 25% annual EPS growth, its 10-year stock price estimate is between $2,500 - $3,000 per share, assuming a normalized P/E of 25-30. Of course there are risks such as AI demand slowdown, regulatory challenges, and competition from other companies. Your Monte Carlo Model is not too far away from my estimates. There are challenges for Nvidia but based on what I see, the demand for new better and faster hardware will continue to accelerate. AI is still in its infancy, and there is growth in the area of computer vision, autonomous systems, and general intelligence (AGI). With that said, I would not worry about the noises around Deepseek on Wall Street.
I love the concept of monte carlo simulations. However, I can see where the parameters can really skew the probability numbers, especially in the chip sector where things are deeply cyclical. IMO, it's predictibility would be much if the optimistic end of the spectrum is tempered. 10 years is an eternity in the chip sector, and we are in a period of peak optimism. Or maybe even better yet, (I say this not knowing how you built your model) you more heavily weight the growth rate probability in the early years than the later years to account for the increasing uncertainty.
For NVDA, it's all going to boil down to how well or even if they can maintain their competitive advantage and where we are in the cycle. I could easily see them running hard for 3-5 more years and then hitting a wall.
thanks for your perspectives! and for being critical without offensive... that's a rare trait on the internet these days lol
Great work on the simulations!
Thank you! 😊
Also I don't get the assumption of a 50x multiple in 10 years (which is the average you seem to assume) whatsoever. If earnings growth drops from 40% over the last 10 years to 10% as you assume, then there is no way that anyone will still be buying this stock at a 50x multiple.
@@zweidritteleinfach2087 The stock is priced based on hyper scale capex spend. Since deep seek showed that in theory you don’t need as much hardware, bc you can create more efficient running models, this means that hyper scale capex spend will drop and the stock price corrected. What the market hasn’t priced in is all the business use cases for AI that will require nvidia chips to run. There’s two camps here - the camp that says AI is a tech fad and it won’t progress beyond the hyper scalers and the other camp which sees AI evolving into AGI that will be a legitimate human replacement. Which camp you fall into will determine how bullish you are on this stock…
It's because there's massive uncertainty in what they'll be able to do in 10 years. This is a company that literally invented the GPU and has been able to find new uses for it over the past three decades. FSD cars, robots, software, in addition to maturing current AI trends. But notice that the simulation shows a very low probability for those outcomes.
Based on your answer, you don’t understand how MonteCarlo simulations work.
50x is the max it will allow a simulation to get to.
5 years is a more accurate time frame to work with. 30-35 PE avg is more likely over 5 years. 10 years is too long to estimate based on the ever changing industry. After 3 years do another 3-5 year estimate
Nathan, I think you are doing a good thing to help people understand it's about the numbers that help one drive their investment decisions. I managed monies for many years and most people want the return but are unwilling to stay on the ride they jumped on reach their desired destination. While I think 10 years is to long a time frame reassessing the ride over 3-5 year periods of time works better to keep the people focused, you are doing a good thing to get people to NOT look at just the price of a stock but the real underlying value
Nobody can predict the future. I will continue to own nvidea because i believe it to be a strong buy for a minimum of 5 to 10 more years.
same.
@@armandoweckmann5699 especially around the current price.
@@armandoweckmann5699 Hwa Wei s GPU against Nvidia.
In 2025
That's a good plan. 3-5 year cycles are the sweet spot in terms of evaluating things
Great video and great information. Thank you for sharing this with us.
Appreciate you!
Oh man if it reached $2,241 🤑 as 60% of my portfolio is Nvidia when I bought it back in Sep for $116.04
If it reaches 1000 in 10 years, i can sell them all and achieve my dream of retiring by 40. Please, let's go!!!
🙌
I don't get the point or the need to use some sort of Monte Carlo simulation to do this. Starting at a 40% annual earnings growth and assuming this drops to 10% over 10 years and that this drop follows an arithmetic sequence you can get the answer on a pocket calculator in a couple of minutes. That result will have very much the same predictive power as running 100 000 MC simulations, since deviations from the mean are bound to cancel out. The key parameter here is the assumption that earnings growth drops from 40% to 10% over 10 years. That assumption will have a vastly higher influence on the resulting fair value estimate than whichever computational model is used after the fact.
For the median result, sure, it’ll be extremely similar. The primary value, in my mind, is the ability to model uncertainly and understand what kind of downside you could have if things go poorly; in effect, build in a margin of safety.
Thank you Nathan, you rock.
Good stuff, curious if would do same analysis rest of MAG7 stocks.
S3 Partners tweeted on 1/28, "Short Interest is now $32.12 billion, 271.3 million shares shorted, 1.15% SI % Float. $NVDA is the largest U.S. short ahead of $AAPL, $MSFT and $TSLA."
Is it going to be worthless? Probably not. But the idea of 40% growth year over year is probably not in the cards.
Exactly once this data center race wears off this will become cyclical business for nvda and then the price will go down or stall.
how many nvidia shares a beginner should buy at minimum ?
@@saifumar3001 depends what you can afford. You can buy 1 share or even fractions like 0.5
There's no right answer but ideally you want no more than 5-20% of your portfolio into any single stock, and only over 10% if you're extremely optimistic on a company. So if you have say $100k to invest you might want to buy 84 shares for 10% of your portfolio ($119 per share x 84 shares = about $10,000). Double that if you're very optimistic, or half if you're not as optimistic.
@@youtuber9991 50% of my portfolio is Nvidia, if it weren't for the Chinese with their chat bot and the tariff war, the shares would be worth $160
You must have been doing this simulations for some time. Whats your success score/ratio in predicting price through this models?
Yep, that is the WHOLE thing. Few notice that...
If you applied the same method of analysis to other stocks years ago, would the expectations meet the actual results?
Great question! I just made the simulation via Python so will work on “going back in time” to see 😉 This can’t be backtested easily, though
Why 10 years investment horizon for high growth stocks? Most of the time high growth is not sustainable for long periods.. Looking at history.. Remember Yahoo, Nokia, Blackberry…
Investors should be open to changing their mind when they notice that thesis for buying the company has changed.
Buy with intention to hold for 10 years, but sell when company no longer resembles the one you brought
What he said!
3-5 years when in technology is the range and then you just continue to evaluate rolling 3- 5 year periods
Nvda to $20000🚀
😂
I am not sure how to access new Monte Carlo simulations with my RUclips membership. I saw Coffee Can Scorecards after hitting the link, but no simulations. Help?
The simulations are inside of the scorecards. :)
@@NathanWinklepleckCFA OK...I was looking for the 100,000 possible futures data for estimated 10 year future valuations as mentioned in the video. Perhaps I misunderstood. Thank you.
In other words, I see the redesigned score cards in the preview under the membership tab, but the Google Doc link downloads the older version score cards with no futures data.
Thank you for this
Thanks!
Wow, thank you very much!! 🙏
Is there any sort of excel template where we can input the numbers and run the simulations ourself? I would love to do this for other stocks as well! Thanks
Great analysis, thanks. Agree 100%!
Will you do simulations for both Jefferies and Ulta Beauty, please?
In conclusion, no one knows anything.
😂
I'll stick to leaps and covered calls for income. :)
You deserve more money from us!! Thank you so much!!
Not at all. I’m happy as a clam 😃 Just appreciate the support!!
Have a Winklepleck convention.
I just look at the exponential trendline LN slope for the last 15 years. It's great.
It is pretty great, but past doesn’t always = future!
10% = too low. 50% too high 10 years. 30% seems baded on DATA reasonable.
Monte Carlo simulations assume efficient markets. Questionable.
They do?
No one can predict Nvidia earning one year away let alone 10 years. The assumption is not reasonable the conclusion is of course baseless.
Hence the reason to simulate it and come up with a range of possible outcomes.
I just trust the intelligence of their CEO.
There will certainly be diminishing returns year after year for NVDA. It would be foolish to think otherwise.
AI, robotics and a whole lot more are just getting started, so it would foolish to assume that the defacto industry leader wouldn't continue to benefit significantly.
I think the only foolish assumption is presuming that you can tell the future.
AI is still in it's infancy, but nvidia create compute capacity. I don't forsee compute capacity value going anywhere but exoponential.
Let’s make a simple calculation. If the S&P500 has an average return of ~9% per year (historically), then in 10 years it would appreciate: (1.09)^10 times, or 2.37 times. This is not MonteCarlo, but a good approximation. Now, per Nathan’s calculation, the median appreciation for NVDA over the same period would be ~484/125 or 3.87. In other words, 3.87/2.36 = 1.64 times the S&P500. This is a slam dunk. Just keep doing DCA and keep including NVDA in your portfolio. That’s it.
If a company's earnings doesn't affect the actual stock price, it only affects what people might be willing to pay, wouldn't focusing on the willingness of people to pay a certain price be more relevant to identifying future price?
So that would be more technical analysis, which identities more of the supply/demand dynamics rather than the hypocritical value of the stock.
Eh, that gets dangerous, imho. Then there is really no limit to what you’ll pay. The only way that works is if others are basing their analysis on something “real” but that doesn’t always turn out to be true. (Ie, houses in 2007, tech in 1999, meme stocks, and a bunch of other examples going back to tulip prices)
@NathanWinklepleckCFA what's the theory behind buying a stock based on the value of the company without regard to what people actually want to pay right now?
In a more technical way, why would people change their minds in the future?
Why should the company value impact what people want to pay?
I consider this a flawed statistical analysis in that it assumes independent variables. They are not. In billish market conditions earnings growth and P/E ratio are pretty strongly correlated and not independent variables.
Fair points! I could modify the simulation to include some sort of positive correlation, but was relying on the sheer volume to generate reasonable price distributions. Any ideas how you would “fix” it?
@ I use Monte-Carlo simulation together with Bayesian network. In BN, I would model P/E by a conditional probability table linked to earning growth and would give the CPT parameters reasonable sampling distributions.
i bought at 125.77 . Am I in trouble seeing all the negative news around this
@@HeranJaskaran I think it will go up. It may hit 110 on the down side first. But earnings out 20th feb. Then 125 will be super cheap. I think it can be closer to 200 by the end 2025. Deepseek won't affect Nvidia much. Microsoft already said they will spend 80bn on data centres and that hasn't changed meta said he sees Deepseek as a positive. Earnings growth won't slow much 60-100% high margins, Fantastic cash generation. Buying back shares and authorized 50bn more. I see this as a temporary dip and a great Buying opportunity
Me too
nope negative news is the best time to buy. You'll be kicking yourself in one year when it's 200. Than you will be why didn't I buy more?
No just dollar cost average add more weekly or monthly when in the red…
@@quangyung6751 😵💫have been holding in the hopes but it backfired due to this news. Not so sure now i think it will reach 110ish shortly. Long term it will recover but doesn’t look like stabilizing shortly.
Fantastic.
Thank you! Cheers!
So many ifs?
Always are…
Not sure you need to use MCMC. your assumption already led your prediction. Simply based on past average and std to prediction the price is not very useful. First, we need to understand whether AI is the major development tech in next 10 years. 2nd: is NVDA Gpu will dominate the AI hardware technology in next 10 years? Deepseek and AI algorithm improvements do not conflict with AI hardware development but it could shift little bit of NVDA gpu demanding in a short term. For longer term, the computation speed will dominate AI developments. 10 year is too far to predict, but we can assume that nvidia will lead the AI hardware for at least next 5 years. Assuming 40% growth in next 5 years and p/e 50, the price should be at $730 level with plus or minus $150 error window. I would add more nvda shares at today’s price.
lol the problem all of you have is you don’t understand AI or the evolution of it or even the amount of expansion of the tech that still left to grow. It’s gonna blow up next this year many contracts for this year and orders to be filled, soon China will be a buyer too since deep seek is using old Nvidia chips that they bought in a different country, said they bought 20% of Nvidias old chips
We can predict the future but we cant predict AI
You mean like the AI robots?!
LOL...I literally become a billionaire if Nvidia hits 20,000 per share! And I'm only 59! I can't even take it out of my Roth(what's in my Roth)...
Even JUST at 1200 per share, I end up the highest paid player in the NFL!
I think most likely, in 10 years, Nvidia is in the 300-330 a share. And that's a solid return, but I'm 5 years in and that HUGE dividend that pays me...2200 bucks a year isn't going to get me to stay in.
I suppose there could be a rise of the robots and because I'm their shareholder and I imagine the first rule in robotics will be, "protect the shareholder," maybe i should say on and hold at least half my shares!
LOL that’ll be great! Remember all of us peasants when you are there 😜
@@NathanWinklepleckCFA Yeah...I'm pretty confident I will never become a billionaire...or worth 9 figures...or anywhere near that.
It is over
What’s over?
im stressed about nvda for feb 2025.... 10y from today, i dont care.
Then you shouldn’t be investing, my friend.
@@NathanWinklepleckCFA Investing or trading; fine line of Tom Foolery
"The Future" is AI and NVIDIA accounts for 65% of chips for AI. I imagine the impetus for this video is the DeepSeek bombshell. The DeepSeek news is distorted. DeepSeek is a reasoning model and therefore uses more processing power as it gets better. Just as we are taught to read until the 4th grade and then learn by reading after that, DeepSeek may been trained cheaply, but requires more power (more "reading" if you will) to get smarter, and the same chips required for training a reasoning model are required for making it smarter. The massive investment NVIDIA just made in Israel should illuminate "the future" as should the massive investment in Stargate the U.S. govt just made. I would compare it to the semiconductor industry producing chips integral to wireless technologies back in the 90s, say... Qualcomm.
@edwardrichardson8254 lol🤣🤣 deepseek is only used in phones and laptops plus they got nvidia chips illegal plus they truly haven't disclosed how fast it truly is ..remember it's China can't believe everything they say..plus if deepseek is to grow they will need nvidias better newer chips which they can't get ..so deepseek is a bust in my book
Why are they showing me 3 minute unskippable dementia ads? Do better RUclips.
😂
NVIDIA sucks, Investers tank their stock!
🤯
Nvidia's hype just passed, but this is a real business, not quantum computer stocks that grew 10 times in a month for no reason. just a casino
NVIDIA IS CRYPTO OF NASDAQ. SOECULATIVE BUT USELESS WITH NO FUTURE.
Hmmm… Not sure I agree with that, but it definitely could be speculative; I’ll grant you that! High price to pay for lofty expectations, but worthless? Nah…
what people dont realize is that when AI becomes more norm.. things that needs to run AI at optimal level? thats going to be mass production of something. as of now gpu cards runs AI superb so until something is build to overtake gpu cards? nvidia is still king
With ‘Physical AI’ on the cusp of playing out, am buying at every bit of NVIDIA’s dip! .. I believe the opportunity is so big that it’s only limited by our imagination.. once in a generation, life changing ..!
Is earnings distribution uniform? Is your simulation based on Baysian prior information? Great job. Thank you.
Normal distribution on growth rates 👍 No Baysian
just look at Intel stock if you want to see what will happen to NVDA in a few years. 90% of computers still use Intel chips but yet the stock is 1/3 of what of was 25 years ago.
No comparison. Intel doesn’t have the revenue or fill a needed demand on the same level NVDA. You’re comparing apples to lug nuts. In my opinion.
Nvidia's profit margin: 55%
Intel's profit margin: -29%
Intel made significant strategic errors and fell behind competition over the course of a decade. That could happen with NVDA, but leadership is far more forward-thinking, but it’s certainly possible!
@@NathanWinklepleckCFA
I interacted with Intel technically decades ago & Intel's commpetitor at the time. The difference between the 2? Intel knew everything, just ask them.
@@NathanWinklepleckCFAas far as volume goes Intel are the # chip maker in the US. It has nothing to do with strategy and has to do with the expectations in the late 90s were too high. The same could happen with NVDA that is all I am saying. In the 90s expectation were too low for AMD.
One day of reality destroys an infinite number of simulations!
Where does $500 billion Stargate Project fit in ?
Woohoo. First viewer!
🔥
I wonder how this system can calculate the S&P 500 who's companies are often changing over time? Who knows what future companies will be included or excluded let alone each of their data to enter into this system. A company like NVDA should do better than the S&P 500 as a whole so I would lean to the higher side of the average for NVDA in this case.....For now I have sold my position and will wait for a better time to buy it back. Long term "hodeling" like a helicopter parent as I like to call it, is not my thing because at my age I am running out of time.
Well the easy money or making money from NVDA came and went. This is like a CSCO story back in the day.
@keaka560 I love this overused comment for every stock that go up. Can you please tell what was the CSCO P/E ratio before the dump and tell me todays NVIDIA P/E ratio. Let me help you as of yesterday NVIDIA P/E ratio was 26, it is waaay lower than Tesla, Palantir etc... and for CSCO??... wait let me help you on that too, may be somebody random is reading this post, Cisco (CSCO)'s highest recorded P/E ratio was around 2,128.5, which occurred in the July 2018 quarter, do you get it now? 2128.... 2128 vs 26...
Keep it cordial, please. No need to get upset or offend. We’re just differing in opinions here and that makes us all better.
As long as AI is not able to pretend the future there is still a lot of deep learning to do and chips to sell.
One chart comes to mind...CSCO....
Interesting comparison
@NathanWinklepleckCFA as a CPA, I look at these companies from a very simple point of view. NET ASSET VALUE per share. So you can say eps all you want, but there are two stories that always exist in a stock. What's the accounting value, and where do the players want the price to be. All I have to do with my player buddies is control enough of the free trading shares and we can wash trade all day. We accept the fact that there will be some longs and some shorts that we have to deal with, but if we can control the float, we can make our options pay us. We love 401k and IRA because it incentives holding and punishes premature sellers. This set of parameters neutralizes all models. Options are the tail that wags the dog.
Nvidia's growth rate will likely be more in line with 24.5% in the next couple of years. The tariffs and restrictions will result in a much lower growth opportunity. The trade war is going to get a lot worse and will not subside anytime soon. The world is sick of Trump and his lies and conspiracies.
Hold US Stocks for 10years...silly. Overvalued, monopoly gone, Algor trading...30 years of Mag7 bull run is over?