Unpacking The Investment That Doubled In 6 Months
How predicting the future isn’t really that hard
NVIDIA released its latest quarterly earnings last week, during the turmoil that was the American Presidential election. But that didn’t stop the stock from jumping up nearly +25% immediately thereafter, with its price rising from the high $60s to the low $80s.
Looking back on my original April investment at $35 per share, some people have asked me how I was able to pick something that grew 2.5x in 6 months, and is contributing to a portfolio alpha of +25% beating almost everyone on Openfolio.
The answer comes down to two very simple macroeconomic tech trends: AI and Video.
II. Powering Artificial Intelligence
If software is eating the world, then AI is eating software. If you’re not already investing in some form of machine or deep learning in your organization (call it the very enterprisey “Digital Transformation” if you like), then you’re about to get smoked faster than you could ever imagine.
The current state of the art AI being deployed in your favorite technology companies are all using the same approach. It’s mathematical and most of them require two things in order for it to work. Lots and lots of data, and lots and lots of processing power to do all the calculations. In fact, it requires so much processing that if we tried to mimic even a simple roundworm’s neural connections using deep learning, it would test the limits of the world’s fastest supercomputer.
That’s why, in April 2016 I began investing in NVIDIA. They were, I believed, the market leader for GPUs. In other words, the graphics processing chips that were uniquely capable of performing calculations in parallel instead of in series. Said differently, they could do more of the similar tasks, better, than the CPUs in standard computers.
This becomes criticially important for handling the linear algebra calculations necessary for matrix multiplication in deep learning algorithms. Or, what we call ‘mathematic intelligence’.
III. Video as an App
GPUs are also required for delivering high frame rate graphics in video games and video streaming for live, virtual reality and augmented reality. It’s important for advanced rendering, and already being used for architecture, haute couture cars, and for designing physical objects.
We’ve written previously how the future of apps is interactive live video. In order to process all of that data, H.264 and H.265 compression isn’t going to be enough. It’s going to require interactivity over the top of a live stream. We call that Augmented Reality today. We know Apple is working on their own Glasses, even though they’re using AMD for GPUs.
Virtual Reality is an especially big play for NVIDIA, and have shown videos where they’ve stitched together photorealistic 3D environments on the surface of Mars and on Everest that you can explore through your headset of choice. NASA has been using NVIDIA’s tech since 2004 on the current Mars rovers.
IV. Self-Driving Cars & Rovers
GPUs are important for managing the processing necessary for self-driving cars because it’s basically one big robotics system. Sensory input to motor output. Take all the data from radar, sonar, cameras, process it for objects, match that against a database of known objects to understand what it is, then push behavior logic for how to decide what to do.
It’s that Behavior Engine that makes self-driving cars so messy and prone to driving off bridges and crashing in inclement weather.
This becomes critically important on the Martian surface where one mistake means billions of dollars lost to the harsh sands of red.
In addition to NVIDIA, I also invested in MobileEye in April 2016 because they’ve been a leader in automotive computer vision technologies since 1999. They were in the news recently when their marriage with Tesla broke up following a self-driving car crash that lost a life.
I sold the stock, however, a few months back as I began to hear that their CTO believed, and publicly stated, that no further innovation was required for self-driving cars. After being deep into the Biologic Intelligence area and seeing how mathematic self-driving cars are failing, I knew he was flat out wrong.
So, if a key member of the management team has a critical aspect of their business incorrect, I can’t in good faith continue to maintain a financial investment as a fiduciary duty to my own future.
In summary, NVIDIA had spent a billion dollars in R&D on a new AI architecture that enabled them to uniquely address the GPU needs of cloud computing platforms, major social networks and tech companies, and for a big computer in the trunk of self-driving cars. Their Founder and CEO said they needed 5 Miracles to deliver what they did. They definitely pulled it off and we’re planning to use NVIDIA’s Drive PX-2 in our own self-driving car. They’re hardware, we’re software.
There is simply no other company in the world that is uniquely situation to handle the processing power of these three major trends.
V. Other Opportunities
I also successfully predicted the airline industry, choosing Delta as my tour de force. Yesterday, it was announced that Warren Buffett’s fund has invested in every public airline stock, while simultaneously getting out of Wal-Mart.
My guess is due to the fundamental need for ecommerce packages, as well as business travelers, to be shipped by air. Coupled with the low price of oil, it means that an industry that’s traditionally had the lowest ROE of any other is realizing major profits. Each is investing heavily in infrastructure and improvements to the customer experience both in terminals and inside planes.
Delta, for instance, introduced Comfort Plus seating and has segregated pricing even further to make it a new second tier fare, in between Economy and First Class, both as a way to increase profits using assets they already have in place.
Finally, have a look at Coach ($COH), which is having a resurgance in product design. Did you know they initially started as a men’s leather brand and have recently hired creatives from the likes of Gucci and other luxury brands. They’ve managed to marry the two most critical aspects of a successful product business: high quality plus low prices.
So, back to how I predicted that NVIDIA’s stock price would rise? Honestly, it doesn’t seem like that big of a prediction once you understand everything happening in the technology world and the major trends that VCs and tech companies are investing it. They will need something to process all of that data and they will want to choose the best. NVIDIA’s GPU is nothing more than a really great platform.
Much like Facebook is a platform for everyday people to communicate with one another, NVIDIA’s GPUs are a platform for powering other platforms for everyday people. From auto-tagging you in a photo, predicting which emoticon you’re about to type in a text, detecting an object and avoiding it in a self-driving car, and serving you all that video you’ve been watching over the internet, it was only a matter of time before the stock price reflected the real world purchasing behavior of their products.
When you find the new electricity, and find it before the rest of the world knows how valuable it is, you invest and watch as the price goes up, up, and away.
from Stories by Sean Everett on Medium http://ift.tt/2eCys46