All in on Data

We’ve talked quite a bit over the last few days around a number of companies releasing earnings reports/announcements and without wanting to make this a re-hash of what I and others have posted I’d like to share a realisation in my investing decisions and intentions.

As some of you might recall when I very first arrived at this board and the NPI board I was taking a thematic view of growth investing. (I posted long posts of mega themes and potential investment options etc).

Since then I have learnt a lot from the board on the Saul approach to growth stock selection. In particular I have:

  1. Paired down the long tail of my portfolio from well over 100
  2. Grown comfortable allocating much larger absolute $ and % holdings - I think when I first started tracking my portfolio my largest 15 holdings were literally in the 1.5-2% range (now it is 2-10%)
  3. Becoming more ruthless at exiting non-performing holdings
  4. Prioritising higher quality and higher conviction stocks for my allocation

These last few weeks have led me to a conviction in a number of stocks that share an over-riding theme. Yes we are all here looking at growth stocks and yes a lot of our focus has been on cloud based software solution type stocks with subscription models, but for me one particular theme has emerged that unites a number of investment decisions and intentions (whilst still embracing what I have learnt here on this board).

The theme I am seeing reverberating through the stocks that I see having the highest potential, the strongest performance and offering the greatest conviction (and safest risk profile) is data and certain recent announcements have both nudged me towards this as well as better encapsulated the story behind this - so whilst I had always identified Big Data as mega theme, I now have a better vision, stronger conviction and better grasp on connecting the dots that has helped me look at this space and the opportunities therein with a different perspective from TAM potential (forecast growth in data theme markets), to tailwind drivers (digitisation, video file formats, smart phone roll outs, 5G, autonomous driving vehicles etc) to end market use cases (AI, IoT, Edge Computing) to supply and demand conditions (market structure, players & pricing functions).

My interest within the data theme is principally around the storage, transmission and access orchestration of data where I have the highest conviction (high potential certainty with lowest risk), clearest visibility (I’m not a techie) and see the best market dynamics (in terms of potential market leaders yet small enough to offer 10x opportunities with the most manageable competitive landscape).

The realising of this theme has really for me been driven by a number of exceptionally aspirational, extremely compelling, highly corroborating and frankly astonishingly revealing investor communications.

These include:
Micron - https://seekingalpha.com/article/4176022-2018-micron-analyst…
discussed here: http://discussion.fool.com/micron-update-and-upgrades-32977162.a…

Pure Storage - https://seekingalpha.com/article/4177184-pure-storage-pstg-p…
https://seekingalpha.com/article/4176020-micron-technology-m…
discussed here: http://discussion.fool.com/long-pstg-investor-presentation-slide…

Nutanix - https://seekingalpha.com/article/4155979-nutanix-ntnx-invest…
discussed here: http://discussion.fool.com/ntnx-3q2018-33077953.aspx?sort=whole#…
latest results slides here: https://seekingalpha.com/article/4177051-nutanix-2018-q3-res…

Pure and Nutanix have been extensively discussed on NPI and is worth a search there.

For me the key messages are:
Pure Storage - penetration of only 1Bn of a 30Bn addressable market, forecast to grow at a 30% CAGR to 2Bn by 2021, redefining the data storage model with a disruptive vision of shared accelerated storage and leading the industry a data storage ownership business model with ES2
Micron - 62 Trillion GB to be created by 2021, cloud data capex increasing by 2.6x by 2021 & longer term demands of AI, in memory device computing, IOT & Mobile banishing both the pricing cycle and the supply/demand swings in the future for NEW MICRON.
Nutanix - 48% CAGR in hyperconvergence to 2021, targeting 3Bn in billings by 2021 and looking at a TAM of $178Bn in cloud operating/management systems! $178 Billion WOW.

Pure is a 40% growing, turning profitable this year & operating with positive free cash flow.
Nutanix is a 40-60% grower on the verge of positive cash flow & aiming for profitability within 2 years.
Box is a 25-35% grower, with great subscription visibility and aiming for profitability in 2 years.
Micron is a 60-90% grower, massively profitable with a P/E of 6 and for those of you still old school enough to think along PEG lines <0.1

I think there’s a real possibility that either Pure or Nutanix can become the de facto cloud operating system and do what Microsoft did with desktop PC computing O/S in the PC era. The only other search for a de-facto standard that comes close to this would be a search for a de facto operating system standard setting player in AI! (Nvidia?). That’s exciting.

Yesterday I took a ruthless look at some of my holdings and exited successful small and large growth holdings I had in Materialise and Appian that were offering 10-20% growth rates and trimmed my Abiomed holdings that is yes still a 30% grower but sat at a P/E of 161 and redirected the proceeds into topping up Pure Storage, Micron and Nutanix and I intend to add more next week.

Before this latest buy-hi round of top ups, my US Port “Data” theme positions amounted to ~12%:
Nutanix - 3.5%
Micron - 2.75%
Pure - 2.5%
Box - 1.75%
Arista - 1.5%
I am looking to raise the combined total to 15-20%. (I’m leaving out Nvidia and AMD - another 1.5%, from this which are related but downstream as well as database software plays including Hortonworks and MongoDB - another 3%). I would also include my Keppel Data Center REIT holding which is probably another equivalent of 1% but sits in my Singapore portfolio.

The only other themes where I have this level of long term “sleep well at night” total conviction for a 10 year plus outlook are in:
The rise of ecommerce (hence my 15%-20% combined holdings in Shopify, Ali Baba, Hubspot, Wix, The Trade Desk & Square)
The cloud (hence my holdings in Twilio, Ali Baba & Tencent (HK) and individual cloud plays and my interest in Pivotal, Red Hat, Microsoft, AWS, Zuora).

The next highest conviction after these for me is possibly in cyber security: (~6% across Okta, Fortinet, CyberArk, Palo Alto Networks)

Ant

76 Likes

Since then I have learnt a lot from the board on the Saul approach to growth stock selection. In particular I have:
1) Paired down the long tail of my portfolio from well over 100
2) Grown comfortable allocating much larger absolute $ and % holdings - I think when I first started tracking my portfolio my largest 15 holdings were literally in the 1.5-2% range (now it is 2-10%)
3) Becoming more ruthless at exiting non-performing holdings
4) Prioritizing higher quality and higher conviction stocks for my allocation

Great Post Ant!

Those 4 points resonate with me as well. I’m so glad I finally discovered this board last fall!

  • Since late October, I’ve gone from nearly 90 holdings down to 16.

  • Had the same initial discomfort with building a position thinking only in terms of $ amount, and then realizing, jeez its really just 3-4% of my WHOLE portfolio

  • I too took a ton of underperformers (Like TMFBreakerRob would say) out back and shot them. :slight_smile: And truthfully, now that they are gone, I feel relieved not even having to think about them anymore.

  • point 4 - seems like a duh no brainer thing, but when I think back to my best performers and realize what a small overall % they made up of my portfolio, I try not to beat my self up because I was doing the best I could at the time. I too, have been working hard to get my favorites to be the highest % weighted companies.

Always enjoy your posts, and thanks again.

Matt

21 Likes

I too took a ton of underperformers (Like TMFBreakerRob would say) out back and shot them. :slight_smile: – Matt

Some companies need shooting. :slight_smile:

Prioritization is a good thing. How can you have your “top ten ideas” (for example)… and justify funding #30 or #40?

I’m on the cusp of violating that view, but a lot of my positions are represented by a handful of call options. Only around ten large positions currently.

Rob
Rule Breaker / Market Pass Home Fool & STMP/MTH Maintenance Coverage Fool
He is no fool who gives what he cannot keep to gain what he cannot lose.

2 Likes

Ant,
Not to put too fine a point on it, but I think it valuable to distinguish between data and information. If you look in almost any dictionary, you will find circular definitions, essentially making the two synonyms. They aren’t. At least not in my way of thinking.

Before I retired from a Fortune 50 firm, I was a member of a relatively small IT group of Enterprise Architects. My area of specialty was Information Management. Although I freely admit that after 8 years in retirement, my knowledge base in the field is growing rapidly stale, I think a few things endure. One of those things is the need to clearly define terms. The organization I worked for maintained the official dictionary of IT terms for the entire enterprise. Reflect for just a moment that the term Information Technology evolved from Data Processing. I wrote the definitions for data and information. So far as I know, they are still in use today.

I could go to great depth on this subject, but I’ll do my best to condense it. Information is consumable and meaningful to humans, data is the electronic representation of the binary gunk interpreted and manipulated by machines (in that this is an IT discussion, I’m mostly ignoring processing and manipulation of analog signals). Some folks will maintain data is all ones and zeros, but even that is far from the truth. It is the presence or absence of magnetic flux, or the voltage differential, or some other quantifiable signal (i. e., laser light, quantum computing, etc.) synchronized to a clock. Data is completely unintelligible and in many of its forms imperceptible to humans.

Of course, there’s a relationship between data and information. If there weren’t, there’d be no reason and no market for any information technology. But one should not lose sight of how much and how many different technical layers are involved in converting some magnetic flux (data) present on a disc in order to deliver a text character (information) to an eyeball (human perception mechanism). This simple operation involves address/location services, retrieval services, transportation services, interpretation services, conversion services and presentation services. Almost all of which is nearly instantaneous and totally transparent to the human reading the text.

When looking at IT (Information Technology) investments, IMO it’s important to have a rough idea of how the technology resides in the multiple layers between data and information. At a high level, we can break this down into four main domains. OK, I’m ready for a boatload of objections from the techie residents of this board, but bear with me for a moment and remember, I’m viewing this from an investment perspective, not an in-depth technical discussion. But, I digress. The four domains I am referring to are 1) data at rest (storage h/w with attendant s/w), 2) data in motion (network h/w with attendant s/w), 3) data manipulation (CPU, GPU, I/O devices, sensors, transducers and A/D-D/A converters), and 4) data translation/presentation to/from information (applications). I know, an argument could be made to classify some of manipulation h/w I listed as belonging to the data in motion domain or even translation domain. If that makes it easier for you to think about this stuff, go ahead and re-classify it. One might also create a fifth domain to capture all the boundary devices, but I don’t think it’s very useful in most IT investment discussions. Each of the IT oriented companies we discuss on this board resides primarily in one of the four domains.

Big Data in and of itself serves no purpose. It does nothing to serve any human need or desire. What does “big data” mean? If big data is a relatively new phenomena, what does it have in common with “small data” or whatever you want to call the digital stuff that formed the foundation of the IT environment at an earlier date? We can start by saying that big and small data have almost everything in common with respect to how they interact with the four domains I just defined. What sets big data apart are two things: first, just the sheer volume. We used to talk about terabytes (1012) as large data volumes, now petabytes (1015) is common parlance for data volume, and new terms like exabytes (1018), zettabytes (1021) and yottabytes (10**24) are coming into use as shorthand for speaking about ever increasingly large volumes of data (and there’s vocabulary for even larger volumes). So volume is 1/2 of what makes big data big, but the other half is complexity. In the old days (50s, 60s and 70s) of IT, data was simple. Sequential files of EBCDIC or ASCII representations was it. Then came VSAM (I think), random access files of the same stuff. Then IBM introduced IMS (the first large scale database) and a number of not quite competitive database products came out from other vendors, but all data still represented text characters.

Digitized sound first entered the mass market in the mid 70s. The first compact disc was introduced in 1982. The first digital photo products were introduced in 1975, about the same time digitized sound was first introduced. Presently, there are a multitude of file types representing an enormous number of complex native files. While each native file can be rendered for human consumption by specialized applications, big data provides the opportunity to look at these disparate files in combinations and cross sections in order to derive previously hidden information.

Maybe in another post I (or someone else) might look at the IT investments that many of us on this board hold in order to position them in the four domains and gain an understanding of how they complement or compete with one another or in some case determine that a given product is not just best of breed, but maybe the only significant player in an a niche. I think this would be useful in making investment decisions.

For example, Alteryx (clearly in the translation domain) provides statistical analysis of geospatial data. That’s not all that’s offered by AYX, but my understanding is that it is pretty much unique in this area.

Anyway, enough for now. I hope some of the readers here find this post thought provoking and useful.

26 Likes

but all data still represented text characters.

Representing numbers as something other than text characters goes back pretty far … at the very least to the 60s when I started.

1 Like

The sweet spot of transformation is turning information into actionable intelligence.

Never before has so much data been available

Never before has so much data been stored

Never before has so much data been retrievable

Never before could data across systems be assembled into information at a speed that supported decision making

Never before have we moved past “can we relate the data” and into the world of “how should we relate the data”

Hence, we are moving from the age of information into an age of actionable intelligence and we still don’t know how best to do that.

But the companies here are all support that wave, as well as, remove the technical obstacles and make it simple and and turn all the information into cognitive, actionable intelligence.

12 Likes

…Hence, we are moving from the age of information into an age of actionable intelligence and we still don’t know how best to do that…

Amen, all those names and the only two I recognize and have invested in are Google and Alibaba. I think I have some sense of what goes on and kind of what the plans are at Google. Alibaba, I read the filings and once again just realize I just bet on Jack Ma. Incredibly complex corporate structure. In all the investing talk I rarely hear much about management let alone the folks on the boards. Odd, would be the first thing you would want to know about the little business down the street before you purchased or not.

He is no fool

who gives what he cannot keep

to gain

what he cannot lose

Yep checks out. Rumsfeldian almost. So I see the Motley Fool has about a 400 bagger on Amazon. 40,000% or some crazy thing. As best as I can recall and I think I would remember Google @ about $120 is the best I’ve ever done, ever, shortly after 2005 IPO.

Rule Breaker To Buy AMZN
September 08, 1997

valuation…
…The Motley Fool’s message board on Amazon.com has a few people whose posts consistently cry that this stock is FAR overvalued. From our point of view, these posts have yet to give concrete reasons why that might be so, other than the authors’ admittance of being short the stock. This rather reminds us of similar cries regarding America Online, Iomega, and Yahoo! ever since those stocks entered the public spotlight. What unites those three is their continuing status as market leaders.

…You see why we think Amazon.com is going to be huge?
https://www.fool.com/archive/portfolios/rulebreaker/trades/1…

You know, I’ve read that way more than once and I still don’t really know how they picked Amazon and Jeff Bezos out of what I recall as many wannabes and eyeballs and all that back in 1997. I watched the 100 bagger video and still don’t really get it. Obviously has almost nothing to do (a little bit) with technology and almost everything to do with Bezos. But what did they spot in Bezos.

Anyway, you can go a long time in between posts to see anything about management and/or the board. A bunch of data, usually incoherent and un-actionable.

“Cartoons are like gossamer,” he tells her, “and one doesn’t dissect gossamer.”
https://www.npr.org/sections/monkeysee/2012/07/22/157190213/…

Yeah but. As I recall Kramer’s proposed caption was pretty funny.

3 Likes

Hi Brittlerock

Thanks for the note. I completely agree there’s a difference between data and information and I mentioned that when distinguishing the area of focus I was referring to (as the underlying data space).

Having worked in evidence based consulting for the one time largest database owner on the planet (bigger than NASA) until Google surpassed it and the original Big data company and spent 14 years translating data into information into insights and knowledge I also 100% agree with your point about this value chain.

As an investor I look at this and think a number of points:

  1. Data creation and retention needs is scaling exponentially right now and the 1 to 1 direct benefactors are the memory and storage players. You create a 1000 more GB of data you need 1000 more GB of storage. It’s more of a linear relationship.

On the other hand if you create 1000 more GB of data then you may not need much more in the way of an extraction, cleaning, integration, analysis, visualisetion tool. You might need the pro model or one more license seat but that’s it. Some data storage is needed for pure storage purpose which right now isn’t an information area e.g. the billions of photos that Asian millennials take of their plates of food every mealtime. It just gets stored and accessed and shared not really analysed per se. (Or not yet I guess).

  1. The picks and shovels play on data memory and storage is not changing too much - sure we went from tape to disk to flash but essentially there hasn’t been as much constant churn as you get in the information creation world (with the exception of Oracle).

  2. Then there is the market structure point. There are 4-5 main consolidated market storage guys (EMC/DELL, NTAP, IBM, PURE) and 4 main players in memory (Samsung, Western, Micron and Sandisk) and in data orchestration you have ~3 (Nutanix, DELL/VMWare). In each of these we have an orderly and mature fast growing market with a smaller disrupter overtaking the larger players creating an amazing investment opportunity to enter at a $3bn, 10bn and 60bn player level rather than 30bn, 25bn and 100bn level.

Compare that with the 1000s of data analysis and information software companies that are jockeying for position. Yes I’m invested in Mongo and Alteryx but my point is it is infinitely less certain whether you are backing a future market leader and it is also very unstable part of the value chain where visualisation vs integration vs extraction etc all compete within and with each other for licenses, functionality and $ and also all face the massive risk of AI disruption. AI feeds off base data and disrupts the information creation process potentially. Data is the fuel of AI.

From a basic risk reward perspective and the opportunity to benefit from a linear relationship with the exponential data growth then I’m advocating the base data plays the most. The information space feels more like an individual stock picking exercise with a lot more risk or less safety.

Cheers
Ant

P.s.
A few facts:
90% of all data that is stored has been created since 2016

300 hours of youtube is uploaded every minute.

16 Likes

nice post Ant gives me a different way of thinking about it

Sure Mauser. Hope it helps. I often worry what Americans on this board must feel seeing a barrage of posts last thing at night and first thing in the morning by some poster who talks weird and spells funny and bangs on about Asia.

Anyhow just thinking about the post - I might have left out NTAP and Citrix from the cloud and data orchestration (virtualisation and hyper convergence) space. But still you get my point hopefully.

Ant

4 Likes

Ant you have a lot more stocks than I do. I concentrate on possible 35% to 40 % growers . With some 25% growers And a small TSLA holding which I will not try to defend.

. At the moment I own no e-commerce stocks but if I did would probably just buy an ETF if it is a thematic investment.

Ant,
As one of many Americans who read and contribute as we are able, I’m confident that I speak for others who have come to appreciate and value your perspective and input on this great board.
How fortunate are we that Saul is willing to be the flux that has brought everyone together in this great community, who together have helped so many who read and/or contribute to this board make a difference in their lives and the lives of their family members.

sjo

10 Likes

The four domains I am referring to are 1) data at rest (storage h/w with attendant s/w), 2) data in motion (network h/w with attendant s/w), 3) data manipulation (CPU, GPU, I/O devices, sensors, transducers and A/D-D/A converters), and 4) data translation/presentation to/from information (applications).

Brittlerock - I like your framework for this and in particular I’d be very interested in your take on Pure Storage’s vision and position in shared accelerated storage - i.e. able to ensure all data is instantly available for (2) Motion and not at rest cold as clearly their mantra has become “NO COLD DATA” in their latest investor piece…

https://seekingalpha.com/article/4177184-pure-storage-pstg-p…

Thanks
Ant

1 Like

Pure is a 6.5% position for me. That makes it my 6th largest position out of 12 (not counting 3 very small, speculative positions).

I’ve reviewed a lot of their presentations and product pages on their website. To be honest, I’m not a hardware guy and I don’t really understand all of their technology offerings. But, h/w technology has a way of becoming quickly commoditized. My sense is that PSTG is bleeding edge right now, maybe they can stay ahead of the pack, or maybe my sense isn’t accurate to begin with. But either way, I don’t think that’s the most important thing.

What I perceive as their moat is their business model, the Evergreen contract I think is essentially recognizing that the h/w technology does get commoditized so they’ve built that into their customer agreement. The customer doesn’t have to plan h/w replacements, data migrations, etc. All the pain in the ass stuff involved with upgrading storage devices; comes with 99.9999% guaranteed up time. What’s the next big breakthrough on speed and/or density and/or power consumption? You don’t need to care, Pure will manage that transition. Why wouldn’t you subscribe to that? And once subscribed, why would you ever leave?

On top of all that, they’re forming marketing and technology partnerships with all the heavy hitters that they have to play nicely with in support of end-to-end data access optimization. What’s not to like?

12 Likes

What’s the next big breakthrough on speed and/or density and/or power consumption? You don’t need to care, Pure will manage that transition. Why wouldn’t you subscribe to that? And once subscribed, why would you ever leave?

That’s exactly it. Once they OWN the customer storage contracts they OWN storage. They can make their business virtual, acquire the next big thing tech wise or in-source/out-source. Whatever they want to do they OWN their future then without having to worry about hardware depreciating or commoditising.

When my previous employer that had the one time world’s largest database encompassing 75% of all prescriptions from over 100 countries around the world, it established “evergreen” contracts. They could manage their panel with switches in and out for their own economic advantage, acquire other data suppliers and sell off existing assets and supply data to customers in paper based reports, disk format or remotely without worrying about migrating customers as the average length of customer retention on their evergreen data contracts were 20-25 years.

It will be interesting how ES2 fits in alongside Nutanix’ Xi cloud service. Either way - data and cloud storage looks like it is going to be owned by these 2 players. They could become another Azure or AWS scale of business in the process.
Ant

6 Likes

Ant,

There is a point you made that is persuasive, and that is you only need so many seats of Talend, as an example, because large or small data sets can be analyzed by the same amount of seats.

You only need so many seats of say an Alteryx (albeit lets not deny the momentum there and the early stage of diffusion of the product) as there are only so many people who need to be citizen data scientists.

However, storage scales 1 to 1, although that is not quite accurate, because the cost of storage per unit of volume does go down every year, but the point is still valid.

But storage is not the only thing. MongoDB, as an example, for its Atlas offering, charges per gigabit, thus the larger data stores become, the more MongoDB makes in Atlas.

Thus both scale with data.

The other two companies I am highest on are Okta (that I have not talked much about) and obviously Pivotal. Okta does not scale with data per se, but it has a market of so many people, that maybe that will not matter. Whereas Pivotal also does not scale with data, but it does scale with apps. And the number of apps necessary to replace legacy apps is incredible to behold, and it seems much of this will end up on MongoDB databases, perhaps Atlas.

At this point in time these are my 4 favorite of the next generation companies, and 2 of them scale with data, one scales with apps that creates more and more data, and the other just has an awful lot of people to work with.

On NPI I did raise the question, given the billing structure that Talend uses, which is about the same per seat as Alteryx, with no server products, and none of the expansion packs, like Alteryx just put out Promote and Connect, and dropping the cost per query from $29k to $333 as per the CEO presentation of just how disruptive they are, I cannot figure out how Talend is going to make oodles of cash.

Clearly, at least it seems to me, that Alteryx will have more seats than Talend, which is used by the data jocks, but not by the data jocks plus the citizen data scientists.

But your logic about storage, and I think the other three discussed here, is quite a valid way to look at this as long as these companies are singular so that they will collect most of the tolls in their field unless they really screw up.

Okay, yes, Nvidia. But no need to rehash Nvidia. More of Nvidia’s future value may be built in to the stock already, but this said, as we discussed elsewhere, it predicted that the data center will go from less than 1% GPUs to more than 10% in the next five years or so. That is a 10x or more increase in GPU concentration in the data center, and Nvidia will be the primary beneficiary of that. So I cannot throw out the bath water without also keeping Nvidia in this loop as well.

Nvidia scales with data, because Nvidia is the enabling technology that will create most of the data to begin with.

Anyways, no, I am not in Asia, but I seem to keep Asian hours these days. Now back to my movie. I am slowly moving myself back to not going to bed at the break of dawn, as is my want, as I have court Tuesday morning and 2 hours of sleep just will not do.

Tinker

12 Likes

More of Nvidia’s future value may be built in to the stock already, but this said, as we discussed elsewhere, it predicted that the data center will go from less than 1% GPUs to more than 10% in the next five years or so. That is a 10x or more increase in GPU concentration in the data center, and Nvidia will be the primary beneficiary of that. So I cannot throw out the bath water without also keeping Nvidia in this loop as well. Nvidia scales with data, because Nvidia is the enabling technology that will create most of the data to begin with.

Hi Tinker,
I certainly can understand someone deciding to stay in Nvidia. I was strongly tempted to do so myself. I still am! But here is a simplified version of my thoughts on it.

The way I see it, to increase revenue 50% Nvidia has to go out and sell 150% as many new GPU’s next year as they sold this year, from scratch. Actually maybe 180% as many, as the price continually falls. For one of our SaaS companies to increase revenue 50%, they only have to sell 50% new software programs, as all the current 100% will renew. In fact, they probably need only 25% in new sales, as they will have dollar-based net renewal rates of 125% or so. Now Nvidia may indeed go out and sell twice as many GPU’s next year as this year, and more power to them if they do, but I’ll probably (but not certainly) be in something else.

Best, and thanks for all your contributions,

Saul

25 Likes

By way of keeping it honest and true to my word, here’s a quick update on what I did with these positions this week once the public holidays were through…

Before this latest buy-hi round of top ups, my US Port “Data” theme positions amounted to ~12%:
Nutanix - 3.5%
Micron - 2.75%
Pure - 2.5%
Box - 1.75%
Arista - 1.5%
I am looking to raise the combined total to 15-20%. (I’m leaving out Nvidia and AMD - another 1.5%, from this which are related but downstream as well as database software plays including Hortonworks and MongoDB - another 3%). I would also include my Keppel Data Center REIT holding which is probably another equivalent of 1% but sits in my Singapore portfolio.

After the top ups on Friday and Tuesday this is where I ended up…
Nutanix - 4.5% added 1%
Micron - 3.5% added 0.75%
Pure - 4.5% - added 2%
Box - 1.75%
Arista - 1.5% no change

How did I fund this? I top sliced my Jupai after its pop, sold my failed experiment in Impinj and sold my Five Below which whilst a 100% return for me has run into traffic and faces uncertainty over the impact of Toysrus liquidation plus potential economic slow down.

Ant

2 Likes