artificial intelligence ETFs

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
chem
Posts: 11
Joined: Sun Apr 12, 2020 7:45 am

artificial intelligence ETFs

Post by chem » Tue May 19, 2020 7:53 am

a new artificial intelligence focused ETF launched this past week, ticker THNQ. Homepage here:
https://roboglobaletfs.com/thnq (holdings also here)
it tracks this index which Robo also created:
https://www.roboglobal.com/robo-global- ... hnq-index/
"Stocks classified within the ROBO Global® Industry Classification as exposed to the Artificial Intelligence theme are given a “THNQ Score,” comprised of factors representing the levels of revenue the company receives from Artificial Intelligence activities, levels of investment the firm makes in Artificial Intelligence, and the market and technology leadership of the firm in the Artificial Intelligence universe."
Unfortunately, I can't find a detailed formula or algorithm for calculation of the THNQ Score (how much of a concern should that be?). But the holdings include companies like iRobot, Twilio, Amazon, NVidia, etc. Pretty interesting set.

What do you guys think of this ETF? I've been looking for something like it for quite a while, but, possibly worth the 0.68% expense ratio for the convenience of not having to manage your own portfolio of stocks as a long-term buy and hold? Could be an interesting alternative to some other general tech or software ETFs. I am personally of the opinion that companies focused in this area will easily outperform the S&P500 during the 2021-2030 time range, and I'm considering putting some money into it after its AUM climbs a bit.

I know that bogleheads in general prefer wide-market indexes, but I'm real bullish on software and tech. I currently have large allocations to VGT and VUG to get some of that exposure, complementing my other Vanguard funds like VIG and VTSAX. There are other specialty software ETFs out there, but I can't find anything else with a strong AI focus (edit: while not being overweight on robotics!).
Last edited by chem on Thu May 21, 2020 10:05 am, edited 1 time in total.

User avatar
David Jay
Posts: 8399
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: THNQ artificial intelligence ETF

Post by David Jay » Tue May 19, 2020 11:28 am

I wouldn't purchase any brand new ETF, especially from a niche player (maybe from a Vanguard or Black Rock). Follow it for a year or two, observing performance and assets-under-management.

Lots of small specialty funds fold up if they don't attract enough investment. Notice that their current holdings are about $2.5M (tiny!).
Last edited by David Jay on Wed May 20, 2020 8:49 am, edited 1 time in total.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

fire_2030
Posts: 10
Joined: Sun May 17, 2020 5:53 pm

Re: THNQ artificial intelligence ETF

Post by fire_2030 » Wed May 20, 2020 12:32 am

When did attlasian (TEAM) become AI focused?

stormcrow
Posts: 73
Joined: Thu Nov 21, 2019 4:12 pm

Re: THNQ artificial intelligence ETF

Post by stormcrow » Wed May 20, 2020 7:43 am

fire_2030 wrote:
Wed May 20, 2020 12:32 am
When did attlasian (TEAM) become AI focused?
Since "AI" became a buzz word picked up by every large corp in the world that has data to manage.

stormcrow
Posts: 73
Joined: Thu Nov 21, 2019 4:12 pm

Re: THNQ artificial intelligence ETF

Post by stormcrow » Wed May 20, 2020 7:49 am

More pertinent to the topic at hand, maybe try BOTZ. It's more focused than just a repackaged QQQ product, and has a fairly large AUM.

Slowtraveler
Posts: 202
Joined: Wed Jan 10, 2018 10:21 am

Re: THNQ artificial intelligence ETF

Post by Slowtraveler » Wed May 20, 2020 9:27 am

At just 39 holdings, seems easy to buy the components directly and avoid the fee. But why bother?

It has a high price to earnings and VTI has huge potential to benefit from AI. Goog and Amzn have bought many and will likely continue to buy many up and coming AI companies. I'm sure Walmart would be happy to have robots stocking shelves and delivering goods. Finance companies will benefit from this. Most companies will implement AI in some way into their business structures.

I bought ICLN (renewable etf) as a play on renewables a few years back. Now, renewables have increased in presence but they're not as profitable or "green" as I once pictured. I sold out years ago but often go back to look and see how much money I gained by selling everything for broad indexes.

AI will definitely help the business world but to say only these few expensive companies will benefit is likely inaccurate. VTI is a solid AI play.

Topic Author
chem
Posts: 11
Joined: Sun Apr 12, 2020 7:45 am

Re: THNQ artificial intelligence ETF

Post by chem » Wed May 20, 2020 10:15 am

fire_2030 wrote:
Wed May 20, 2020 12:32 am
When did attlasian (TEAM) become AI focused?
yeah, I don't like that THNQ contains TEAM, WIX, and a few other similar stocks. But, the index description does remark that it tracks essentially the AI ecosystem, so, basically you get some companies that make products that are used for AI development or revenue generation by AI companies. Not a fan of that.
stormcrow wrote:
Wed May 20, 2020 7:49 am
More pertinent to the topic at hand, maybe try BOTZ. It's more focused than just a repackaged QQQ product, and has a fairly large AUM.
I looked into that last year, too heavy on robotics for my preference. Also, I contacted the ETF managers, asking why google (for example) wasn't included, and they said google wasn't an AI company. I think that's flat wrong, and google's CEO has since confirmed my view on that.

Similar problem with THNQ is that it's missing Facebook (PyTorch creator...)
Slowtraveler wrote:
Wed May 20, 2020 9:27 am
AI will definitely help the business world but to say only these few expensive companies will benefit is likely inaccurate. VTI is a solid AI play.
I agree that AI/ML will help the total stock market in general, but for a minority part of my asset allocation, I would like to make targeted plays in certain sectors. I realize not everyone agrees with this investment philosophy. That's fine.

randomguy
Posts: 8930
Joined: Wed Sep 17, 2014 9:00 am

Re: THNQ artificial intelligence ETF

Post by randomguy » Wed May 20, 2020 11:12 am

chem wrote:
Wed May 20, 2020 10:15 am
fire_2030 wrote:
Wed May 20, 2020 12:32 am
When did attlasian (TEAM) become AI focused?
yeah, I don't like that THNQ contains TEAM, WIX, and a few other similar stocks. But, the index description does remark that it tracks essentially the AI ecosystem, so, basically you get some companies that make products that are used for AI development or revenue generation by AI companies. Not a fan of that.
if AI really takes off, I am not sure if many of the companies on this list would be winners. Odds are the big money is in some 100m valuation privately held company that will be bought for 10 billion by alphabet. That list to me looks a lot more like a whose whose of current tech companies all of which are pretty much doing AI of some type. Odds are this index will perform about the same as any other tech index.

Seriously how many of these companies do you think will have some AI breakthrough that transforms the world
EA
Twilio
Netflix
grubhub
square
spotify
shopify
intuit
yandex
zendeskt

from the ones I recognize. I have no doubt all of them are doing some AI work. I doubt any of them are building 100 billion dollars of added value off them anytime soon. These are all really weak sector bets. We aren't looking at things like investing in AOL, compuserve and prodigy in the mid 1990s.

handle
Posts: 8
Joined: Tue May 19, 2020 8:26 pm

Re: THNQ artificial intelligence ETF

Post by handle » Wed May 20, 2020 1:57 pm

The importance of artificial intelligence in all sectors of the economy is rapidly becoming obvious, even to people like myself who do not work in the technology sector. I think over the next 10-20 years, applying AI where it is useful for your business will be a requirement to stay competitive, as opposed to an advantage belonging to just a few "savvy" companies. Perhaps there are companies that will profit massively from facilitating this change in business practices, but it does not seem like the ETF you have proposed targets those companies.

Topic Author
chem
Posts: 11
Joined: Sun Apr 12, 2020 7:45 am

Re: artificial intelligence ETFs

Post by chem » Thu May 21, 2020 10:09 am

I renamed the root post, because, I may have found an alternative:
https://www.globalxetfs.com/funds/aiq/

AIQ is a "future analytics" ETF, covering companies that strongly benefit from or develop AI and big data methods. It really hits the highlights, with a little less cruft. Nothing's perfect but perhaps this one is more tightly focused?

Full holdings here: https://www.globalxetfs.com/funds/aiq/? ... dings=true

index info here: https://www.indxx.com/assets/media/Indx ... _Index.pdf
Last edited by chem on Fri May 22, 2020 11:32 am, edited 1 time in total.

DustyCampaign
Posts: 1
Joined: Thu May 21, 2020 6:15 pm

Re: THNQ artificial intelligence ETF

Post by DustyCampaign » Thu May 21, 2020 6:26 pm

randomguy wrote:
Wed May 20, 2020 11:12 am
chem wrote:
Wed May 20, 2020 10:15 am
fire_2030 wrote:
Wed May 20, 2020 12:32 am
When did attlasian (TEAM) become AI focused?
yeah, I don't like that THNQ contains TEAM, WIX, and a few other similar stocks. But, the index description does remark that it tracks essentially the AI ecosystem, so, basically you get some companies that make products that are used for AI development or revenue generation by AI companies. Not a fan of that.
if AI really takes off, I am not sure if many of the companies on this list would be winners. Odds are the big money is in some 100m valuation privately held company that will be bought for 10 billion by alphabet. That list to me looks a lot more like a whose whose of current tech companies all of which are pretty much doing AI of some type. Odds are this index will perform about the same as any other tech index.

Seriously how many of these companies do you think will have some AI breakthrough that transforms the world
EA
Twilio
Netflix
grubhub
square
spotify
shopify
intuit
yandex
zendeskt

from the ones I recognize. I have no doubt all of them are doing some AI work. I doubt any of them are building 100 billion dollars of added value off them anytime soon. These are all really weak sector bets. We aren't looking at things like investing in AOL, compuserve and prodigy in the mid 1990s.
> If AI takes off

It already has. I remember talking to the project manager of Spotify Weekly when it was quite new, and we knew it was a great product, but I had no idea it would be on of the reasons I keep using Spotify every day. The playlist understands me and recommends the perfect tracks week after week, and it's become part of their economic/defensive moat.

Netflix also is a fantastic usecase. Each video preview image is tailor made for you. They use AI to try make the preview as interesting as possible to get you to watch the episode. Things such as have you viewed an actor in a film before will influence the algorithm, and increase the likelihood that you try a new series.

Where I work we use simple AI to detect fraud and reduce the number of human hours required to sift through data - that lets us have a super lean team, many times smaller than our competetors. While people use AI as jargon and as sales tactics, those who really use it to its advantage reap the rewards.

User avatar
David Jay
Posts: 8399
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: THNQ artificial intelligence ETF

Post by David Jay » Thu May 21, 2020 7:13 pm

DustyCampaign wrote:
Thu May 21, 2020 6:26 pm
randomguy wrote:
Wed May 20, 2020 11:12 am
chem wrote:
Wed May 20, 2020 10:15 am
fire_2030 wrote:
Wed May 20, 2020 12:32 am
When did attlasian (TEAM) become AI focused?
yeah, I don't like that THNQ contains TEAM, WIX, and a few other similar stocks. But, the index description does remark that it tracks essentially the AI ecosystem, so, basically you get some companies that make products that are used for AI development or revenue generation by AI companies. Not a fan of that.
if AI really takes off, I am not sure if many of the companies on this list would be winners. Odds are the big money is in some 100m valuation privately held company that will be bought for 10 billion by alphabet. That list to me looks a lot more like a whose whose of current tech companies all of which are pretty much doing AI of some type. Odds are this index will perform about the same as any other tech index.

Seriously how many of these companies do you think will have some AI breakthrough that transforms the world
EA
Twilio
Netflix
grubhub
square
spotify
shopify
intuit
yandex
zendeskt

from the ones I recognize. I have no doubt all of them are doing some AI work. I doubt any of them are building 100 billion dollars of added value off them anytime soon. These are all really weak sector bets. We aren't looking at things like investing in AOL, compuserve and prodigy in the mid 1990s.
> If AI takes off

It already has. I remember talking to the project manager of Spotify Weekly when it was quite new, and we knew it was a great product, but I had no idea it would be on of the reasons I keep using Spotify every day. The playlist understands me and recommends the perfect tracks week after week, and it's become part of their economic/defensive moat.

Netflix also is a fantastic usecase. Each video preview image is tailor made for you. They use AI to try make the preview as interesting as possible to get you to watch the episode. Things such as have you viewed an actor in a film before will influence the algorithm, and increase the likelihood that you try a new series.

Where I work we use simple AI to detect fraud and reduce the number of human hours required to sift through data - that lets us have a super lean team, many times smaller than our competetors. While people use AI as jargon and as sales tactics, those who really use it to its advantage reap the rewards.
Yeah, pretty much all the SP500 companies are using some form of AI. So if one is going to buy an ETF that invests in companies who are users of AI, may I recommend VOO. :happy
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

KyleAAA
Posts: 8281
Joined: Wed Jul 01, 2009 5:35 pm
Contact:

Re: artificial intelligence ETFs

Post by KyleAAA » Fri May 22, 2020 12:36 am

I really question their methodology. Why Are Google, Amazon, and Microsoft not the top 3 holdings at something like 20% each? They are far and away the leaders just based on their scale. I find it hard to take this ETF seriously. This ETF should basically be FAANG plus a few others like NVidia and Microsoft.

Topic Author
chem
Posts: 11
Joined: Sun Apr 12, 2020 7:45 am

Re: artificial intelligence ETF

Post by chem » Fri May 22, 2020 7:06 am

David Jay wrote:
Thu May 21, 2020 7:13 pm
Yeah, pretty much all the SP500 companies are using some form of AI. So if one is going to buy an ETF that invests in companies who are users of AI, may I recommend VOO. :happy
The point of this thread, though, is identifying the best ETF that contains a set of companies who which fall into one of the following:

(a) disproportionately benefit from breakthroughs in AI/ML or data science methods
(b) disproportionately profit from the greater adoption of those methods by the S&P 500 at large (or as compared to other tech companies)
(c) AI/ML is absolutely key to their competitive advantage in revenue generation

I realize AI/ML will help the S&P 500 at large, but some companies will benefit much more than others, as users, developers, or as a picks-and-shovel type of play.

User avatar
David Jay
Posts: 8399
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: artificial intelligence ETF

Post by David Jay » Fri May 22, 2020 8:34 am

chem wrote:
Fri May 22, 2020 7:06 am
The point of this thread, though, is identifying the best ETF that contains a set of companies who which fall into one of the following:

(a) disproportionately benefit from breakthroughs in AI/ML or data science methods
(b) disproportionately profit from the greater adoption of those methods by the S&P 500 at large (or as compared to other tech companies)
(c) AI/ML is absolutely key to their competitive advantage in revenue generation
And that is the difficulty with narrow sector funds: defining the criteria in such a way as to capture the performance of the desired sector while remaining investable. THNQ has tried to be investable and thus has compromised theIr AI exposure.

Reminds me of this thread looking for a TaaS fund: viewtopic.php?f=1&t=312353
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

Valuethinker
Posts: 40009
Joined: Fri May 11, 2007 11:07 am

Re: artificial intelligence ETFs

Post by Valuethinker » Fri May 22, 2020 10:05 am

chem wrote:
Tue May 19, 2020 7:53 am
a new artificial intelligence focused ETF launched this past week, ticker THNQ. Homepage here:
https://roboglobaletfs.com/thnq (holdings also here)
it tracks this index which Robo also created:
https://www.roboglobal.com/robo-global- ... hnq-index/
"Stocks classified within the ROBO Global® Industry Classification as exposed to the Artificial Intelligence theme are given a “THNQ Score,” comprised of factors representing the levels of revenue the company receives from Artificial Intelligence activities, levels of investment the firm makes in Artificial Intelligence, and the market and technology leadership of the firm in the Artificial Intelligence universe."
Unfortunately, I can't find a detailed formula or algorithm for calculation of the THNQ Score (how much of a concern should that be?). But the holdings include companies like iRobot, Twilio, Amazon, NVidia, etc. Pretty interesting set.

What do you guys think of this ETF? I've been looking for something like it for quite a while, but, possibly worth the 0.68% expense ratio for the convenience of not having to manage your own portfolio of stocks as a long-term buy and hold? Could be an interesting alternative to some other general tech or software ETFs. I am personally of the opinion that companies focused in this area will easily outperform the S&P500 during the 2021-2030 time range, and I'm considering putting some money into it after its AUM climbs a bit.

I know that bogleheads in general prefer wide-market indexes, but I'm real bullish on software and tech. I currently have large allocations to VGT and VUG to get some of that exposure, complementing my other Vanguard funds like VIG and VTSAX. There are other specialty software ETFs out there, but I can't find anything else with a strong AI focus (edit: while not being overweight on robotics!).
Didn't see Nvidia in that list? A major beneficiary of AI?

Whilst a technology can indeed radically change a whole slew of industries and activities, Warren Buffett's hitherto held view of airlines should be remembered (or was in Charlie Munger) "the Wright Brothers should have been strangled in their cribs" [from the point of view of airline investors].

Aviation has changed the world and with the exception of wars and recessions, world airline passenger miles has probably grown by something like 4% a year every year since 1929 (and faster in the first couple of decades).

But it hasn't been a great sector for investors.

The really interesting AI companies will be Venture Capital backed, and you won't actually get onboard until the IPO. My own view is that if they are good enough, that will never happen - Google or Apple or Amazon will snaffle them up by buying them for $1, $2 billion. This is very similar to what happens in biotech. Most companies fail, the ones with truly promising drugs either do joint ventures with big drugs companies and take a royalty, or get bought out.

Topic Author
chem
Posts: 11
Joined: Sun Apr 12, 2020 7:45 am

Re: artificial intelligence ETFs

Post by chem » Fri May 22, 2020 11:30 am

Valuethinker wrote:
Fri May 22, 2020 10:05 am
Didn't see Nvidia in that list? A major beneficiary of AI?

Whilst a technology can indeed radically change a whole slew of industries and activities, Warren Buffett's hitherto held view of airlines should be remembered (or was in Charlie Munger) "the Wright Brothers should have been strangled in their cribs" [from the point of view of airline investors].

Aviation has changed the world and with the exception of wars and recessions, world airline passenger miles has probably grown by something like 4% a year every year since 1929 (and faster in the first couple of decades).

But it hasn't been a great sector for investors.

The really interesting AI companies will be Venture Capital backed, and you won't actually get onboard until the IPO. My own view is that if they are good enough, that will never happen - Google or Apple or Amazon will snaffle them up by buying them for $1, $2 billion. This is very similar to what happens in biotech. Most companies fail, the ones with truly promising drugs either do joint ventures with big drugs companies and take a royalty, or get bought out.
Look at the full holdings, NVDA is in there.

You may very well be right that some of the most interesting AI/ML companies are privately held and will be bought out before they're public. I agree. But those aren't investable for the purposes of ETF or mutual fund discussion.

Anyway, I spent some time digging deeper into THNQ's holdings at names that are at first glance questionable (e.g. TEAM, WIX, INTU) -- and it turns out that they've made major internal investments into AI/ML methods (WIX for ADI web design, for example, and INTU in document processing) and they are significantly hiring data scientists (look at TEAM's careers page), etc. So, those companies do have much more internal exposure to the benefits of ML than a randomly picked S&P 500 or tech company. I imagine the THNQ holdings will have a fair amount of turnover over time, with companies dropping out and new ones being added.

THNQ seems like a good ETF as an AI play. I'm agreed with posts in this thread in that "pure" AI stock plays are really hard to come by. Hopefully over time, THNQ will get more "pure" as public companies focused in this space grow in number and revenue.

edgeagg
Posts: 189
Joined: Tue Jan 23, 2018 1:27 pm
Location: WA-US

Re: artificial intelligence ETFs

Post by edgeagg » Fri May 22, 2020 12:58 pm

Well, as a practitioner in the field, I'm still happy with VFIAX and see absolutely no need to buy (say) Nvidia.

The fact of AI is that you need gpus for training mostly but inference doesn't really need expensive nvidia gpus since Intel is busy commoditizing it. Consequently, regression to the mean and buying the market makes more sense to me. You need way more inference than you do training - once the model has been built, it needs to be deployed widely on cheap hardware, which means intel pretty much. Also given recent partnerships between msft and intel on this, the need for and cost of training models will come down drastically.

AI will be used to accelerate a business problem and those companies that can sell into a specific niche will do well because they address the business problem, which IMO further negates the need of an AI focused fund.

jajlrajrf
Posts: 80
Joined: Sun Feb 09, 2020 6:15 pm

Re: artificial intelligence ETFs

Post by jajlrajrf » Sat May 23, 2020 12:39 pm

I think the reality is that most of the growth in AI is not going to come from publicly traded companies but from startups, and this ETF is not an effective way to invest in those. Certainly giving up nearly 1% to fees to buy pieces of these companies seems like a poor use of money.

edgeagg
Posts: 189
Joined: Tue Jan 23, 2018 1:27 pm
Location: WA-US

Re: artificial intelligence ETFs

Post by edgeagg » Sun May 24, 2020 10:35 am

jajlrajrf wrote:
Sat May 23, 2020 12:39 pm
I think the reality is that most of the growth in AI is not going to come from publicly traded companies but from startups, ....
It depends on what you mean, especially in the current environment. AI infrastructure is being largely commoditized so I don't think that an AI infra startup will do all that well. Business verticals will do well since the big guys will focus more on infra and don't want to take the risk of diving deeply into specific verticals (other than amzn). So AI startups with a clear business focus will do fine. But the ETF isn't as you correctly point out, a good way to expose yourself to that risk and reward. Successful startups get acquired at decent multiples by the behemoths - (which allows people like me not to work for large companies) and hence regression back to the mean.

No matter how you look at it, Bogle is hard to beat (except in tails, if you believe Taleb).

User avatar
JonnyDVM
Posts: 2136
Joined: Wed Feb 12, 2014 6:51 pm
Location: Atlanta, GA

Re: artificial intelligence ETFs

Post by JonnyDVM » Sun May 24, 2020 11:39 am

jajlrajrf wrote:
Sat May 23, 2020 12:39 pm
I think the reality is that most of the growth in AI is not going to come from publicly traded companies but from startups, and this ETF is not an effective way to invest in those. Certainly giving up nearly 1% to fees to buy pieces of these companies seems like a poor use of money.
This is true and I discovered as much when I was poking around previously. Most of the AI work is done my privately owned companies. The successful ones then get scooped up by a big boy like google. That being said, BOTZ isn’t that expensive and I do own some. I figure owning some might help keep me in the good graces of our future robot overlords.
I’d trade it all for a little more | -C Montgomery Burns

Post Reply