Gates Are Open

The IPO window is currently wide open. Last week alone, nine IPOs hit the US public market, with the average performance up +10%. Year-to-date, 44 IPOs made a debut, a number already higher than the 39 IPOs from the first half of 2016. What is even more encouraging is the spike in emerging growth (VC-backed tech) IPOs. In the last two month, we saw back-to-back blockbuster listings including Snap, Mulesoft, Alteryx, Okta and Cloudera. (Disclosure: GSV owns shares in Snap).

Looking ahead, we expect to see many more highly attractive, emerging growth IPOs hitting the public market. While we are sceptic of Uber being one of those, a list of other top names are well positioned, in our opinion; Airbnb, Didi, Dropbox, Spotify, Lyft, Palantir, Klarna, or Warby Parker could all be potential IPOs over the next 12 months. (Disclosure: GSV owns shares in Dropbox, Lyft, Spotify, Palantir).

Dropbox’s management had previously talked about achieving key milestones before considering an IPO. After successfully hitting two milestones over the past year — they became free cash flow positive in mid 2016, and hit $1 billion in annual revenue run-rate this January — Dropbox has is now also EBIT profitable, as reported by CEO Drew Houston last week.

Dropbox’s valuation from its last equity financing was at $10 billion. That was back in 2014, in the heyday of the private market “bubble” when “valuations were ahead of themselves.” But with the recent performance and key achievements, we could likely see Dropbox being valued above those levels. Looking at the list of high growth SaaS comps, the median P/S multiple is at 10.5x. What is also interesting is there is only one company with EBIT profitability.

Salesforce, with a 1% EBIT margin and “only” +27% revenue growth, still gets a P/S multiple of 7.3x. Part of it is the premium applied to being profitable. Looking at the comps’ medians, it could be possible to see a profitable Dropbox trading at such levels, assuming its revenue growth rate support the thesis.

Another way to look at a potential public market valuation is to plot the growth rate against the P/S multiple. The scatter plot shows those data points against each other.

Another IPO candidate could be Spotify. The on-demand music leader recently re-negotiated two of its major label contracts — with Universal Music and Merlin — and is still negotiating those with Sony Music and Warner.

Spotify’s fundamental drivers seem to be strong, as evidenced by the accelerating growth in paying users. In March, Spotify announced it passed 50 million paid users, which was up from the 30 million it announced a year ago. Accordingly, we can expect revenue growth to be at similar levels. (Spotify had last reported $2.2 billion in revenue for 2015). With the leverage from the negotiated new deals, we can also expect to see improvements in gross and operating margins in the near-term, which should improve Spotify’s attractiveness to the public market.

Looking at New Media comps, we see a slightly different picture than that of the high-growth SaaS companies. Revenue growth rates are generally lower, profitability is typically higher, and P/S multiples are slightly lower. In our opinion, Spotify could be valued at a multiple near to that of Netflix. Both companies are the leading platforms in their respective fields, both are showing strong users (and revenue) growth, and both have high engagement rates (time spent per user). Also, both have strong management teams and are consistently out-innovating their competitors.

The one major difference is in margins and profitability. Netflix reported a 37% gross margin and a 10% operating margin in Q1’17. Spotify last reported a $194 million in net loss on $2.2 billion of revenue for 2015… So it will be imperative to improve its margins in order to receive a P/S multiple of 5x or above. We believe that the recent label re-negotiations will bring a positive swing in that respect.

Reimagining The Human Brain: Elon Musk Style

CliffsNotes + commentary on Neuralink

LJ: Tim Urban of Wait But Why published a tour de force post on Neuralink recently and I wanted to really understand his ideas without preconceived notions.

To really think about Neuralink, throw out your biases, really try to zoom very far out, and then zoom very close in.

Part 1: The Human Colossus

Humanity has seen a handful of events that has allowed us to get to where we are today. Spoken language, writing, and the printing press were game-changers because they allowed each generation to build upon the knowledge of prior generations.

With books, computers and the Internet, humanity began to communicate as a whole, thus allowing the collaboration that has led to the modern world.

To understand the grand plan for Neuralink, we need to understand the brain, then brain-machine interfaces (BMIs), then get into the “Wizard Era”, then finally Neuralink.

Part 2: The Brain

The brain is the most complex know object in the universe. It has 100 billion neurons.

Neural networks transmit information like transistors in computers — it has 1s and 0s, on and off.

But unlike computer transistors, the brain’s neurons are constantly changing.

This is called neuroplasticity. It’s neurons ability to change and optimize for the evolving world.

We understand the very micro-scale of neurons firing, and the macro-sacle of how the brain overall works as the decision-maker for our bodies but not all the stuff in between. Neuroplasticity make it hard to learn about the brain because the same word or image will not create the exact same neuron firing patterns.

Part 3: Brain-Machine Interfaces

When it comes to our brains, we still interact and talk like we did 50,000 years ago. There’s still the same process of inputting and outputting of information from our brains.

Neurons are small, but we know how to split an atom.

Each neuron has connection with as many as 10,000 other neurons. There are 20 billion neurons in the cortex alone so that means 20 trillion connections. The voltage of neuron could be changing as many as hundreds of times per second.

All of that complexity makes BMIs hard to build.

Current BMIs are limited, but they can do some cool things. For example, motor cortex neuron firing can train computers on actual movements. BMIs allow people to use their brains to play video games, move a motor arm to let them drink coffee, fly a f-35 fighter jet in a simulation, etc.

Also devices exists to help hearing or sight impaired, or stimulation products that can help people with Parkinson’s or other pains.

What exists today are indirect products. Now onto the future.

Part 4: Neuralink’s Challenge

Elon has a formula to not only work on innovation, but catalyze an entire industry.

He believe that in the next 4 years, Neuralink will create BMI products to help people with severe brain injuries.

Major hurdle 1 is bandwidth. There have never been more than a few hundred electrodes in a human brain at once, but we need 1 million simultaneously recorded neurons to really get accuracy to change the world. If we double every 18 months, we’ll get to a million recorded neurons in 2034. 17 years is not bad, but currently it’s only doubling every 7.4 years which means we get to a million recorded neurons by 2100.

Major hurdle 2 is implantation. Need to find a non-invasive way to implant BMIs. There are a ton of other challenges too.

LJ: even for a post of this length, it doesn’t have many answers. Most readers will get a good sense of the overall mission of Neuralink and why the team is pursuing it, but don’t expect to know how all of this will work, not for a few years anyway.

Part 5: The Wizard Era

Writing, typing, reading, hearing, seeing are so slow. In fact, human input and output is 3,000 times slower than an USB.

The BMI era will be gradual and you’ll be used to seeing these devices over time. Just like how today no one bats an eye at Lasik surgery or pacemakers anymore.

The possibilities are incredible. You can think thoughts and direct them to anyone else with the BMI device.

I think we are about 8 to 10 years away from this being usable by people with no disability … It is important to note that this depends heavily on regulatory approval timing and how well our devices work on people with disabilities.
— Elon Musk

Part 6: The Great Merger

AI is likely inevitable.

What humans can do is to be AI.

Elon envisions a world where AI really could be of the people, by the people, for the people.

The only way to do that is to have AI be plugged directly into the human brain and that is the ultimate master plan for Neuralink.

Read the full post HERE.