I think you're not naming entire nation states, given the way the Chinese are splashing money around there it looks like there is about to be a flood of chipmaking capacity hitting the market and a lot of that will end up competing in matrix multiplying.
Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high.
> Although I hear VCs are working on an algorithm where they burn money directly and the smoke patterns represent the solution to the multiplication. That might keep the margins high.
This has been bothering me for some time now. It’s rather obvious to me, as an engineer, which startups are well positioned and which aren’t… Do VCs really have no standard for due diligence, do they really just not care, or is there something else at play that I’m missing…?
> is there something else at play that I’m missing…?
Depending on the stage, investors don't care about any of the details except for the founders ability to raise the next round and their probability to IPO at a later/advanced stage. The stock is the product and whether you are selling a real time machine or a crystal is irrelevant.
Which is not that "insane" really. If that's how money is being made, and investors are in the business of making money, then that's the path they are going to take.
> The stock is the product and whether you are selling a real time machine or a crystal is irrelevant.
This is the right answer, some people think that the dot-com bubble and mortgage bubble is something for investors to regret, when in fact they made such a ridiculous amount of money they're just searching for the next bubble. Not the next great product.
I agree with you, it is risky, I have no insights into the data but I would imagine that those risks are being calculated with previous W/L taken into account.
We all know there is no 'sure thing' when investing, but if you're a large money manager, you generally want to maximise profits, looking through history, the 'quickest' earners are usually always those from OP, speculation, market share and IPO.. The 'bubble' of these seems to be loosing air but I'm purely speculating based on the tech job market so I could be completely wrong.
There's a point of view that smart investors did, and dumb investors were left holding the bag.
If that's your belief, then thinking it's risky and you may lose your shirt requires believing that you may not be smart. Which is a tough thing to accept for a lot of people, especially ones who think investment success is all about smarts and not luck.
Markets (like most human systems) aren't really 'supposed' to do anything: they've just kind of evolved as various participants in the market have found them useful for different things, and competition between systems has selected for characteristics that allow them to survive. Economists mostly describe systems, ideas like efficient markets are more of a post-hoc hypothesis for why the system survives, not an up-front designed attribute of the system that was desired when it was set up.
(which basically means that if the conditions mean that the optimal strategy in the market is to basically just run pump-and-dump 'greater-fool' scams on everyone else then that's what you'll see occur)
Different to investors are chasing different dreams.
As an exercise, imagine you have a portfolio which will in a typical year lose around 1.2% to the tax man for capital gains. Rather than accepting the guaranteed loss of taxes - you take ludicrous bets on asset classes where you can be guaranteed to book a loss which could be harvested for tax purposes, or will return 100x your investment (at which point you don't care about the tax man).
Viewed in this light, the VC preference for burn bright and burn fast startups makes sense. The worst case scenario for some LPs would be to still have money leftover.
I don't understand how this adds up. If you're losing money to the tax man you're still making money overall. Now, you could say 'screw it' and put the money you're making into long-term risky bets, but either they don't pay off, and you've just lost money, or they do pay off, and you owe even more tax, with even less efficiency. If you wanna gamble, you can gamble if you want, and the supposed idea of VC is that if you take many risky bets you'll be able to win on average, but it's not tax-efficient at all because you tend to win all at once.
(As a general rule of thumb, I'm deeply suspicious of any 'it's a tax write-off' explanation for people losing money: its extremely rare for losing money to make someone better off overall. It may be advantageous for them to move around where and when they say they are losing money, as you'll generally be more tax-efficient if you make money consistently and through activities in areas which are not as highly taxed, but it's not generally a useful strategy to light cash on fire to reduce your tax bill: you tend to be out of more money than you save on tax)
Valuations aren't attached to reality. Look at Crypto, NFTs, Quantum computing etc. You can have a completely non-viable money pit but if it hits the right hype cycle mix you can make huge exponential gains. It's just about what other people can be convinced to invest in. Right now the US tech sector is pretty much the only game in town to dump money and hope for huge returns. Previously it was "emerging markets" but right now there's a lot of regulatory overhead limiting gains in most EM countries.
Okay, tell me, Lets say I want to be an entreprenuer who can roughly learn any skill in tech (pretty young and ambitious) and he just wants to make enough money to live life nicely / travel the world (something like 60-80k USD are fine to me as I live in third world country and plan to travel mostly in cheap countries , I am a pretty frugal person. I want to work the most minimal amount of hours/work hours.
Something like the 4 hour work week. What Idea would you want to suggest me as you are an engineer.
Since, you are an engineer, and it seems obvious to you which startups are well positioned, It seems that you can reason out which trends in startups are well positioned for my specific use case.
I will pay you 100$ of my first income from such a project if it really fulfills what you are saying.
They're not in the business of finding the best ethical fulfilling investment. Their goal is to find nascent niches that can be packed as a trendy future industry and 10x the bag down other investors down the road. Period. It has worked in the past with SaaS, it has worked (sort of) with crypto, it might as well work with AI. The only difference is that now they have to work hard and hope the Chinese don't steal the fire and ruin the long term vision of this revolution being US centred again.
Then I suppose he should first try gaining access if he really knows things "precisely"
It should be on the top of his world priority if anything then.
That’s not actually how professional finance works, at less not on the trading side (I’ve never done vc work).
Usually access and funds is _harder_ to find than alpha. So what’s more important is finding alpha (or whatever metric you are benchmarking against) that you can operationalize even if it’s worse on a risk adjusted basis than some other trade you can’t do.
> Usually access and funds is _harder_ to find than alpha. So what’s more important is finding alpha (or whatever metric you are benchmarking against) that you can operationalize even if it’s worse on a risk adjusted basis than some other trade you can’t do.
Precisely, it’s about playing the hand that you have. My curiosity is why VCs are so indiscriminate with their funding. That said, if I had access to billions in capital to invest, maybe my perspective would be different.
You are looking at it the wrong way imo. It's like when developers build a product and only focus on the technical side and neglect/underestimate the sales side or when they buzz about the language used to build something.
VC don't really care about the startup's product. Your pixel machine. That's a side effect.
Growth. Exit. IPO. Those are the 3 words they want to hear.
> I'm not sure if you are taking into account that VCs are remunerated largely on the amount of money they invest rather than the results?
Sure, but is it really worth just spraying a firehose of money across a variety of entities without trying to investigate them at all? It just seems that they could still realize outsize returns and save hundreds of millions as well…
Those Chinese companies will have to fight our western champions, TSMC in a green trenchcoat, TSMC in a red trenchcoat, TSMC in a blue trenchcoat, TSMC in a purple trenchcoat, and TSMC in a yellow trenchcoat.
> And I can't name several customers with very deep pockets working on their own chips to squeeze/compete with NVDA.
Google, Amazon, and Microsoft all have custom ASIC and TPU projects in their pipeline, and what holds true today might not hold true in 5 years.
A major reason Nvidia was able to do so well was because of technical outreach by donating their GPUs to programs all over the US, building a strong albeit self serving OSS relationship, the CUDA ecosystem, and the acquisition of Infiniband.
Much of these advantages can be nullified by competitive margins and pricing for hyperscalers designed and owned hardware.
Cisco was hit by this same situation as server vendors like Dell began integrating their own in-house networking functionality within their servers, and Dell itself was outcompeted by cloud vendors.
Agreed, the only thing I'd add would be AMD's self-sabotage. Pretending to have compute capability (see also "right around the corner") for over a decade when it was actually so broken as to be sub-viable was an enormous own-goal that they doubled down on when they exited desperation mode without rehydrating the teams and tripled down on with the RDNA/CDNA split.
By rights they should have been fast followers, but they stacked the critical mistakes so high that frankly I think NVIDIA's subsidized GPGPU classes are the smaller part of this story. If the people struggling to get OpenCL to work had been able to get it to work (on other than NVIDIA GPUs lol) the situation would look very different today.
Microsoft, Meta and Oracle already spend billions on supposedly completely inferior(according to the Nvidia cultists) AMD chips. Nvidia is the biggest bubble in human history.
PE isn't a cheat code to rational stock valuation, it's the "nothing ever changes" assumption dressed up in a formula. PE looks at past earnings, while the price of a stock buys its future earnings. A modest PE on NVDA says "I expect datacenter revenue to continue" while a modest PE on TSLA would say "I expect robotaxi to fail." These are fair opinions to have, as are their opposites, but if datacenter revenue collapses then today's modest PE won't save NVDA and if TSLA can pull off robotaxi then today's high PE will be vindicated. Stocks are all about Future Earnings, but you can't put that in a column and sort by it so we have PE.
You can certainly have success while avoiding high PE stocks, but you are on a site about startups and your name suggests you work in the tech sector, which are both places where high PE often does make sense and it pays to be familiar with the reasons.
> a modest PE on TSLA would say "I expect robotaxi to fail."
No. Just having competition is enough to destroy the expectations on TSLA.
And realistically, they are abut last on that race, being thrown out of track about a decade ago and never managing to make anything work since then. So, yeah, betting on no competition is a very weird option.
But the market of luxury cars isn't enough to sustain a company with the valuation of Tesla. If interest rates ever stay non-zero, they will need to take almost the entire cars market worldwide, or something else with similar size.
Even if they can get robotaxi to work on a level better than Waymo currently is shouldn’t they just be valued as if they were Uber with $0 paid out to contractors? The labor is not that significant a cost to what’s fundamentally a taxi business
Full FSD would theoretically allow vehicle owners to rent out spare capacity with no hassle.
I saw the low-tech version of this in Honiara: Western expat buys a cheap car. Driver is hired to take kids to/from school, but in lieu of payment, driver is free to run a taxi service as long as he makes his school commitments.
When you don’t have a driver anymore you quickly run into an issue with keeping the vehicle clean.
One time having a drunken stranger puke in your car while you’re not using it will be enough for a lot of people to determine it isn’t worth the hassle.
It’s pretty easy to imagine a number of scenarios that disabuse the nothing of ever renting your car out “no hassle”. Low hassle maybe, but your framing implies you’ve never worked in a job that required engaging the general public.
The more interesting dynamic with Nvidia though is that most of its revenue is actually fueled by bubbles as well. Teslas earnings are much more grounded and natural.
I think you misunderstand that comment. Tesla may be hopelessly overvalues, but their revenue may grow in future. OTOH Nvidia's revenue may have peaked.
Maybe so. Their valuation coming down to earth would certainly ripple. Whatever precedes that fall, it seems unlikely that revenues would remain unaffected.
In all, it’s unfortunate that the US’s most prominent electric vehicle manufacturer is wrapped up in so much noise. Competition is only going to stiffen.
how is their revenue gonna grow in the future when they don't have an edge in their main core product. Batteries are by panasonic. while BYD makes its own batteries.
Tesla has no moat. At least Nvidia has CUDA
I don't why this comment was downvoted. You raise an important point. I also noticed that you (carefully?) made no comment about Telsa's stock price. Instead, you only focused on their earnings -- which are excellent for a car company.
It got downvoted for not parroting “Tesla bad” even if it didn’t claim anything to the contrary and is simply observing one fact — Tesla’s earnings being great. That’s not acceptable apparently.
