Great read. I was there to read the comments since they always add a lot of color. All valid points! Honestly you all should do a collaboration post on Nvidia just to takle different perspectives. @thecapitalist @krokerequityresearch @olgausvyatsky
In a bit more optimistic scenario, my fair value was $1,147. However given that we may be at beginning of something that is really big, all these traditional valuation methods may be rendered meaningless.
Jensen Huang says data centers deployed will double in the next 10 years and GPUs in those data centers will have to be replaced in every 4 years. If that scenario materializes, $NVDA can also easily double even from here.
But all these are speculations and investors will be better off sticking with conservative assumptions!
Thanks for your comment, Oguz! You bring up some very good points. I agree with you that there is a lot of speculation out there right now and if Jensen's prediction comes true, $NVDA could thrive even more. However, I also think that the competition will get stronger as they will not give up the market to NVIDIA without a fight. I am on the sidelines and will be watching this with great interest.
GPUs replaced every 4 years would imply depreciable life of four years, yet as of now we see the opposing trend - several companies including Amazon extended the depreciable life of their servers from five to six years. You can always argue that Amazon is using their own custom chips - but that would imply that Nvidia's chips are substantially different. Amazon is not the only example, it's an industry trend. So its an interesting question - either Nvidia's estimate is optimistic, or depreciation expenses across the industry are underestimated. Would be interesting to see which one is true.
Oguz - is there a source for the 4-years replacement rate? I can't find a written source - a recap of a conference presentation, conf call, etc - that would cite this number. But perhaps I am not searching correctly.
No there is no written source. Only the reasoning. Jensen thinks that 4 generations of advancement will be too much to ignore for companies running on GPUs and they will have to replace to stay in the competition.
I guess the main question in this case is the sustainability of the gross margin which is insane (and is reflected in the EBIT margin at the end). But right now the market is not looking at Nvidia through a DCF valuation as there are so many moving parts (beat and raise on the top line for the coming quarters ?). It makes more sense to look at its PEG. In this case, Nvidia valuation is not super rich.
I do agree that if you do a reverse DCF, hypothesis of growth and margin are just phenomenal !
Good thoughts on the potential risks. Tough to bet against this juggernaut growing sales and net income faster than their stock price over the last year though.
Hi, thanks for your comment! I wouldn't call it a "bet". Its an estimate of the intrinsic value of the stock based on fundamentals and future growth expectations. And more importantly, it is not a target price.
Nice write up! Nvidia is certainly an impressive company that's done a good job staying on top of trends.
You did a nice job of pointing out the positives while also pointing out some risks that I think many miss.
I think a lot of tech investors think of every company like a software company, where things are infinitely scalable and the cost of additional sales is essentially zero.
Nvidia is still a hardware company, so like you said, they're reliant on supply chains and partners.
There's also just the physical reality of the expansions they're predicting - as Oguz pointed out, Huang says data centers will double in 10 years. That requires a lot of steel, cement, copper and labor. Where's that coming from? Is there enough excess in those supply chains to meet the predicted demand without driving up costs too much? More data centers will also require a lot of electricity. Will we be able to generate enough to feed them?
In addition to the current valuation, I think there are a lot of risks extrinsic to Nvidia that need to be considered by investors.
Thanks for your comment, TJ! Your point about the external risks is really important. Despite NVIDIA's current success, it doesn't automatically mean that demand will grow indefinitely. NVIDIA is a real company that will face increasing challenges in the future. While it's possible that everything could go perfectly for NVIDIA, I just want to emphasize that this is not guaranteed and will be incredibly difficult to achieve.
Great read. I was there to read the comments since they always add a lot of color. All valid points! Honestly you all should do a collaboration post on Nvidia just to takle different perspectives. @thecapitalist @krokerequityresearch @olgausvyatsky
Thanks for the nice comment! I also appreciate the great comments and insights people give here on substack! I will keep your idea in mind :).
Great post!
In a bit more optimistic scenario, my fair value was $1,147. However given that we may be at beginning of something that is really big, all these traditional valuation methods may be rendered meaningless.
Jensen Huang says data centers deployed will double in the next 10 years and GPUs in those data centers will have to be replaced in every 4 years. If that scenario materializes, $NVDA can also easily double even from here.
But all these are speculations and investors will be better off sticking with conservative assumptions!
Thanks for your comment, Oguz! You bring up some very good points. I agree with you that there is a lot of speculation out there right now and if Jensen's prediction comes true, $NVDA could thrive even more. However, I also think that the competition will get stronger as they will not give up the market to NVIDIA without a fight. I am on the sidelines and will be watching this with great interest.
GPUs replaced every 4 years would imply depreciable life of four years, yet as of now we see the opposing trend - several companies including Amazon extended the depreciable life of their servers from five to six years. You can always argue that Amazon is using their own custom chips - but that would imply that Nvidia's chips are substantially different. Amazon is not the only example, it's an industry trend. So its an interesting question - either Nvidia's estimate is optimistic, or depreciation expenses across the industry are underestimated. Would be interesting to see which one is true.
Thanks for the great comment Olga! That's a really good point.
Oguz - is there a source for the 4-years replacement rate? I can't find a written source - a recap of a conference presentation, conf call, etc - that would cite this number. But perhaps I am not searching correctly.
No there is no written source. Only the reasoning. Jensen thinks that 4 generations of advancement will be too much to ignore for companies running on GPUs and they will have to replace to stay in the competition.
Makes sense, thank you!
I guess the main question in this case is the sustainability of the gross margin which is insane (and is reflected in the EBIT margin at the end). But right now the market is not looking at Nvidia through a DCF valuation as there are so many moving parts (beat and raise on the top line for the coming quarters ?). It makes more sense to look at its PEG. In this case, Nvidia valuation is not super rich.
I do agree that if you do a reverse DCF, hypothesis of growth and margin are just phenomenal !
Thanks for your comment! It seems you are on the bullish side of this discussion :-)
Good thoughts on the potential risks. Tough to bet against this juggernaut growing sales and net income faster than their stock price over the last year though.
Hi, thanks for your comment! I wouldn't call it a "bet". Its an estimate of the intrinsic value of the stock based on fundamentals and future growth expectations. And more importantly, it is not a target price.
Nice write up! Nvidia is certainly an impressive company that's done a good job staying on top of trends.
You did a nice job of pointing out the positives while also pointing out some risks that I think many miss.
I think a lot of tech investors think of every company like a software company, where things are infinitely scalable and the cost of additional sales is essentially zero.
Nvidia is still a hardware company, so like you said, they're reliant on supply chains and partners.
There's also just the physical reality of the expansions they're predicting - as Oguz pointed out, Huang says data centers will double in 10 years. That requires a lot of steel, cement, copper and labor. Where's that coming from? Is there enough excess in those supply chains to meet the predicted demand without driving up costs too much? More data centers will also require a lot of electricity. Will we be able to generate enough to feed them?
In addition to the current valuation, I think there are a lot of risks extrinsic to Nvidia that need to be considered by investors.
Thanks for your comment, TJ! Your point about the external risks is really important. Despite NVIDIA's current success, it doesn't automatically mean that demand will grow indefinitely. NVIDIA is a real company that will face increasing challenges in the future. While it's possible that everything could go perfectly for NVIDIA, I just want to emphasize that this is not guaranteed and will be incredibly difficult to achieve.