Snippets - November 11, 2018

As always, please forward this weekly mailer to friends and colleagues and encourage them to sign up. You can also view this email and past issues in your browser. Enjoy!

This week’s theme: Tether hangs on, we’ll have to wait to see what happens, and ruminating on the “real” price of Bitcoin. Plus introducing Datacoral as a new member of the Social Capital family.
Crypto traders awoke to a strange sight on Monday morning a few weeks ago: a US Dollar was trading for $1.06 US Dollars on one of the largest cryptocurrency exchanges. Huh? 

Well, if you’ve been following Snippets over the past few weeks, you’ll know why: one of those dollars aren’t actually dollars at all; they’re Tethers: the infamous stablecoin we’ve been studying as one of the more intriguing and sophisticated scams of the last long while.

You’ll recall from last week (read it if you haven’t done so already) that Tethers, who purportedly exist to create US dollar liquidity in unbanked exchanges and allow for consistent prices from one exchange to another, can exist on the market because of one critical assumption: that their 1 USDT to 1 USD exchange rate would faithfully hold, in good times and bad. That faith from the market at large allowed the good folks at Tether to print more and more of these promised dollar-pegged tokens and circulate them among the majority of crypto exchanges, juicing them with liquidity that was not actually real, but allowed those exchanges to thrive and make fortunes all around. As it turned out, one of the largest and most fraud-filled exchanges, Bitfinex, is owned by the same people who run Tether - confirming many of the red flags raised by sharp-eyed watchers like Bitfinex’ed on Twitter, and ultimately earning them a subpoena from the US Commodity Futures Trading Commission into whether they had sufficient reserves to back up their claims. Yet as doubt crept in, and a chorus of questions became a genuine assault on Bitfinex/Tether’s credibility, that dollar peg across exchanges stayed more or less intact, for a very long time.

Furthermore, not only was this peg maintained across many different “exchanges”, emanating from its central “hub” at Bitfinex, but for a while it has had to do so without Bitfinex having any banking. That’s because Wells Fargo, who had been Bitfinex’s banker through a Taiwanese partner, withdrew their agreement by which US customers could send money to Tether’s Taiwanese bank account in early 2017. Tether and Bitfinex even sued Wells Fargo as co-plaintiffs (before their common ownership was widely known), but subsequently withdrew the case. Tether was eventually able to secure a new banking relationship with Nobel Bank of Puerto Rico, and then recently a Bahamian bank called Deltec, both of which were already considered to be fronts for money laundering. (They’ve since appeared to secure a banking relationship with HSBC, but we’re not really sure. Customers are supposedly able to withdraw USD from Bitfinex right now, but many people’s experience suggests otherwise. No one really knows!) 

However, I have to admit something very impressive: a lesser stablecoin would’ve for sure broken by now. Tether almost did; a few weeks ago it lots its dollar peg, everybody made a big fuss, but then it gained it back after Tether themselves bought back a large volume of circulating Tether to buoy the price. By the way, this led to a really interesting side consequence whereby the price of Tether falling below $1 caused the price of every other cryptocurrency to jump up by an equal magnitude. Wait, what? Well you, the reader, should be able to understand by now why this happened: because the “USD” prices of all these coins are largely denominated IN TETHERS, not in dollars! So the falling price of Tether created the illusion that Bitcoin, Ethereum and all the other coins rose in value against the US Dollar. Rest assured, most of the mainstream media covering crypto did not pick up on this, and botched their reporting entirely. Still, this was a good reminder of how this scheme is unique among scams: the Dollars versus “Dollars” confusion turns what would otherwise have been uncovered a long time ago into something much more systemically hidden.

Getting back to the point, though, for Tethers to maintain their $1 price across different exchanges while both losing their banking AND seeing a prolonged bear market in cryptocurrency, then one of two things ought to be true, First, the market might be genuinely granting Tether an “illegitimacy discount” of zero; meaning they consider owning a Tether to be equally sound to owning a dollar. But that doesn’t sound right, because there’s been a scramble of activity into the one exchange, Kraken, where people actually can get dollars for their Tether, while Tether themselves appears to have an ongoing, significantly negative equity balance (according to skeptics, anyway. According to Tether themselves, they’re rolling in reserves, obviously.) But the second reason, which is the point I actually want to appreciate as we close out this mini-series on Tether for the time being while we await more news, is really an observation about the crypto market as a whole.

When we think about Bitcoin, or about any cryptocurrency, we think about “price” (in US Dollars, let’s say) as a single, clean number that represents an average of something: all of the buyers’ bids, and all of the sellers’ asks. For something like the stock market, there are in fact many different exchanges in which stocks can trade hands, and it’s easy to move between one exchange and another. Think of different exchanges like storefronts along a common street, where buyers, sellers, and the custodians of their assets can freely walk up and down the sidewalk, making sure prices remain in a nice, healthy equilibrium that reflects genuine underlying sentiment. If we use our analogy from last week, it’s like having many different versions of eBay, all trading Pokemon cards, all fully exchangeable with USD in real time. If I place a buy order for a Charizard on one exchange, I can do so with the confidence that I’m doing so in context with all the other exchanges. After all, the other exchanges are only a few steps away.
But cryptocurrency isn’t like that. The slow, irreversible and pseudonymous nature of using a public proof-of-work blockchain as your transaction ledger makes it much harder to trade with “the whole market”. (Companies like SFOX are working on this, though.) Crypto exchanges are not at all like consecutive storefronts on a street: to borrow David Gerard’s analogy, they’re more like isolated islands. Each island has its own rules, its own integrity standards for activities like spoofing and wash trading, and its own spread of buyers and sellers that’s a lot more isolated from the other islands than you’d want to believe. Many of these islands do not have the ability to move US Dollar on or off the island either. But they do take Tethers, and that’s what allows them to post a “USD denominated price” that has very little bearing to a) actual US dollars, or b) what’s happening on other islands. 
So when we ask the question, “How is Tether maintaining its 1 to 1 peg amidst all of these problems?”, it may appear as though we’re asking, “Is the all-knowing, all-powerful force that is the hand of the market actually still valuing Tethers at par?” But really we’re asking, “How is it that Tether is able to remain at par on each little island, individually. And on any given little island, you can’t really count on the market being able to resolve anything at all. There’s one island in particular, Bitfinex Island, where we’re quite confident that there’s systemic fraud and manipulation going on, and that Bitfinex Island is run by the same people that run Tether. We don’t really know ground truth, because most of these islands are covered in dense jungle; they could be full of scams and grifts and frauds, for all we know, but we can’t know because they’re out of sight to us.

What we do know, however, is that Bitcoin trades for a substantial premium on Bitfinex compared to other exchanges: that’s the market talking, for real. Furthermore, without Tether, many of these islands couldn’t really exist at all, because Tether is one of the only two ways that the islands can equilibrate their prices with other islands. (The other is Bitcoin. Again, that’s why the prices of the entire cryptocurrency market tend to go up and down in lockstep with the price of BTC. It’s also a great revenue source for Bitfinex: with no other way to get their money out, Bitfinexed clients are forced to buy Bitcoin at inflated prices, and Bitfinex makes the spread. Criminal genius!) We also know that on just about every island, the trading volume between Tether and US Dollars is zero: aside from Kraken, Tether holders can’t actually get dollars back. Tether claims that you can, but no one will step forward and actually show receipts

So we’re left with this recursive chicken-egg situation: Tether prices stay consistent between islands because Tether is what creates consistency between islands. It’s a really interesting puzzle: if all of the islands were brought together and Tether’s price were actually put under genuine pressure to break the buck, then my guess is that it probably would, immediately. But on any given island, where there’s no way to actually get USD in or out, Tethers don’t feel the pressure at all because crypto holders can’t cash out. They can only cash out by going to a different island; in our case, Kraken. And the Kraken buy book for Tethers is looking real, real thin...

Not exactly Satoshi’s vision, admittedly. But systems work in funny ways, decentralized ones especially: there’s really no denying that we’re getting what we asked for, whether we realized it or not. 


This week’s must-read isn’t an article, it’s bigger than that: it’s a book.

How the Internet happened: from Netscape to the iPhone | Brian McCullough

As longtime Snippets readers will know, we’re big fans of The Internet History Podcast, which Brian has built and hosted over the last few years. (If you’re looking for some great episodes to start with, try the interviews with Rob Glaser, CEO of Real Networks, Jan Brandt, the mastermind behind the flood of AOL CDs that came in the mail, and Chamath’s episode for some interesting stories about the Winamp / AOL days.) The podcast is fantastic, and the book is too: the story of the Internet is still being written, but there’s a certain undeniable appeal associated with stories about the really early days: when most Americans were getting dial-up internet from tiny local resellers, the earliest web sites found success porting over mainframe terminal applications, and the world had to collectively figure out what exactly the Internet was. Buy the book, listen to the podcast, and send a thank you to Brian for all his great work in putting this together.

“How the Internet Happened” review: building a world online | Jon Gertner, WSJ

Brian McCullough, on “How the Internet Happened”: Triangulation with Megan Morrone

Everyone becomes their parents:

Amazon is mailing a printed holiday toy catalog to millions of customers | Eugene Kim, CNBC

Amazon: retailers gonna retail | Steve Sinofsky

Preschool: I was wrong | Scott Alexander, Slate Star Codex

Biology goes one way…

Sequencing is the new microscope | Laura Deming

How biologists are creating life-like cells from scratch, from the bottom up | Kendall Powell, Nature

While health care seems to be going another:

Why doctors hate their computers | Atul Gawande, The New Yorker

A short history of American medical insurance | John Steele Gordon

The senior bench of tech leadership has some opinions to share:

Elon Musk: the Recode interview | Kara Swisher, Recode

Eric Schmidt on the life-changing magic of systematizing, scaling, and saying Thanks | Conversations with Tyler Cowen

Facebook and Google sign up to Tim Berners-Lee ‘contract’ | Aliya Ram, Financial Times

Former Instagram leader Kevin Systrom talks about ‘unhealthy’ Internet incentives | Sarah Kuranda, The Information

Other reading from around the Internet:

The CIA’s communications suffered a catastrophic compromise. It started in Iran. | Zach Dorfman & Jenna McLaughlin, Yahoo News

An introduction to approachable threat modeling | Kevin Riggle, Increment

Chrome will soon ad-block an entire website if it shows abusive ads | Jon Porter, The Verge

May you live in epic times | Venkatesh Rao, Ribbonfarm

TenCent’s bruising 2018, as Pony Ma shakes things up in time for its 20th anniversary | The Economist

The cities that never existed: what if the urban visions of famous architects and planners had actually been built? | Darran Anderson, The Atlantic

And just for fun:

All I want for Christmas are these kitschy VC trading cards | Kate Clark, TechCrunch


In this week’s news and notes from Social Capital, we’re delighted to introduce a new member of the Social Capital family: Datacoral.

Introducing Datacoral: a secure, scalable data infrastructure, no management required | Raghu Murthy, Datacoral

Datacoral raises $10 million for its data infrastructure service | Frederic Lardinois, TechCrunch

The Age of Insight | Jay Zaveri, Social Capital

Datacoral has been a part of the family since January of 2016, when its founder and CEO Raghu Murthy joined our Discover Program as its first entrepreneur. The Discover team has a long-term mission: to help solve 40 of the world’s hardest problems by 2045, sometimes by starting companies from scratch if necessary. Data infrastructure, although it doesn’t pull the same heartstrings as other problems on the list like climate change, health care or food security, is a hard problem with long-term consequences. Our ability to create data, compute data, and critically depend on data is multiplying every year, and it’s growing at a much faster rate than we’re able to build architectures that can actually handle its growing volume and complexity. Data we can’t store correctly isn’t just harder to use, it’s also a long-term vulnerability that gets worse the more we have to stretch and bend our existing data infrastructure to handle loads for which it was never designed. When everything becomes made of data, that means that everything is vulnerable. 

Fortunately, there are a handful of people in the world who have already had to build data storage solutions at that gigantic scale, and Raghu is one of them. As Jay explains: Raghu was one of the key leaders who grew Facebook’s infrastructure to handle exabytes of data as they grew from fifty million to one billion users. He’s lived, experienced, and fixed these problems himself, and now with Datacoral he has made that solution available for everyone. 
The Datacoral framework functions as follows:
  1. Break the problem into little blocks: each of these byte-sized operations is called a slice, and does the job of collecting, organizing and enabling access to data. Since these slices are functional tools, they can be deployed anywhere: in the cloud, on devices, or on any machine. 
  2. Make sure the developer retains control: each slice runs inside a developer’s AWS account and on their own servers, allowing users to pull data out as needed into new services without compromising any of the existing architecture or relationship between slices.
  3. Make sure privacy and security are inherently a part of this framework: this is one of Datacoral’s strongest principles: by modularizing data into slices, slices can be designed to be generally robust and secure, avoiding piecewise security approaches that are doomed to buckle under complexity.
  4. Automate data flow and fix broken plumbing: connectivity issues in getting data to play nicely with itself is no longer a bespoke challenge, it’s an aspect of slice design. With robust slices, data can be usefully handled like a continuous flow, rather than as buckets being poured from one place to another.

All of these principles stem from Datacoral’s initial architectural breakthrough and their primitive unit of data storage: the slice. These slices collect and organize data from all kinds of different data sources, and arrange them in a way that makes them more accessible and usable to any downstream user. All together, it delivers something almost unthinkable: and end-to-end data infrastructure stack, ready to go in minutes. 
Designing, engineering and building this architecture did not happen overnight: after two long-fought years of building and working with early customers, including Greenhouse, Front, Fin, Ezetap, Swing Education and mPharma, Datacoral is finally ready to bring on board customers from the outside world, and we’re delighted to partner with Sudip and the Madrona Venture team on Datacoral’s Series A. Congratulations to Raghu and the team, and if you’re interested in using Datacoral for your own data infrastructure needs, head over to to find out more and request a demo. 

Finally, congratulations to the entire team at Cozy for selling their business to CoStar Group! CoStar is very lucky to have you. As always, if you either rent or are a landlord, we strongly recommend you check out Cozy for yourself. They’re in new hands, but still the same fantastic team and product its users love.

Have a great week, 

Alex & the team from Social Capital
Follow us on Twitter @socialcapital
Suggestions for next week's Snippets? Send 'em to @alex_danco
Copyright © 2018 Social Capital, All rights reserved.
You are receiving this email because you are a portfolio company or a friend of Social Capital.

Our mailing address is:
Social Capital
120 Hawthorne Ave
Palo Alto, California 94301

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list