UBS Chief Investment Officer now a Video Game AI

File this one under -- "They'll never automate my job, oh wait". We've got three delicious data points for you. The first is Google's voice generation platform called Google Duplex. We're sure you've seen the demos by now (if not see the source below), so we'll merely place this into context. Google's virtual assistant has an experimental new feature that can be your agent by calling restaurants and other small business and booking appointments. Google has the map of all SME data, their hours and phone numbers, can generate and route call, and now makes a robot that sounds eerily human as well. The virtual agent comes with "ehmms", "umms" and lip smacking in its voice generation algorithm, to the point where the clerk really has no idea they are speaking with a machine that's doing busy work. Neural networks are getting really really realistic with speech.

Second, remember Alibaba, the Chinese version of Amazon plus eBay plus all of Facebook and JP Morgan in one, give or take. One of the requirements of the platform is to enable merchants to advertise and sell goods to consumers. But the scale of the selling is beyond human management -- with some days seeing $25 billion in revenue. So the firm has launched an AI written copy generator which can manufacture description of products based on the millions of data points the firm already has on prior commerce. Yep, just casually writing 20,000 lines of proposed description, in styles ranging from "promotional, functional, fun, poetic or heartwarming.” The company claims this tool is now used on average 1,000,000 times a day. 

Last data point, which picks up nicely from several observations we made prior about HSBC using Pepper robots in branches and other physical/digital interactions. UBS is launching something fresh in Switzerland. The first is a cute virtual assistant animated object that will be able to help people do basic account actions in physical branches. It looks to us like a Siri or Cortana attempt, but for finance troubleshooting. The second is an animated 3D rendering of the firm's Chief Investment Officer, to be displayed on a screen while visiting a private banker. This AI CIO will be able to answer more complex questions in real time about markets and investing, and has been developed by IBM and FaceMe. We'll let you connect the dots.


Source: Finextra (UBS), Alizila (Alibaba AI writer), Ars Technica (Google), Autonomous NEXT (Alibaba $25BHSBC)

Bringing together thousands of people with Augmented Reality

Let's keep it light-hearted. Here are things we did not know about Pokemon Go, the augmented reality game. It has seen over 800 million downloads, and generated over $1.2 billion in revenue since inception. Niantic, the firm that makes the game, has raised $200 million to build a Harry Potter version of the same. And the game has just run through a partnership with several cities including Akron, Ohio; Charlotte, North Carolina; Philadelphia, Pennsylvania; and San Jose, California. See the link below in the Sources for more detail -- we found the study fascinating.

Niantic and the cities created joint events that drove people outdoors and on adventures through physical communities by sprinkling scarce digital objects along the way. Like Pacman eating dots or bunnies following a trail of nibbles, people were guided along a video game narrative to participate in battles and events. In San Jose, 35,000 people showed up and spent $450k; in Philly, 10,000 had a more bespoke experience along historic landmarks. In some cases, players could modify the game in progress. What is also notable is that the most effective way to get community engagement was not to send invites to individual players, but to networks, which then self-organized and came together. It's a version of migrating a web-based forum or community to the physical world. 

So what, you say? Ok: (1) 22 million virtual and augmented reality headsets will be sold this year -- still a far cry from Whatsapp adoption, but laying the groundwork for mass adoption nonetheless, (2) software platforms like Metaverse let any person or SME program their own AR apps, which can be epxerienced on the major mobile platforms, (3) blockchain companies have laid down the groundwork for scarce digital goods owned and maintained by a community, (4) bank retail branches and physical shopping are both facing pressure from the migration online. Massively multiplayer augmented reality games could be used to bootstrap crypto economies, financial services engagement, or new types of commerce. Or at least, to watch the World Cup on your tabletop.


Source: Venture Beat (PokemonGo revenueHarry Potter), Mobile AR (City ReportMetaverse), Prosthetic Knowledge (Soccer), Crypto projects (DecantralandBubbled*), Tech Radar (AR market); *disclosure -- Lex is advising Bubbled.

Lessons from Monzo's annual report and £33M losses

We love unfair comparisons, but there's a reason behind the madness. Crypto currency exchange Binance is on track to print $1 billion in profits this year, while neobank Monzo has a £33 million loss to show for its £109 million in venture funding. Here's another one: Coinbase now has about $20 billion in addressable (custodied?) assets, while Monzo has £71 million (<£150 per account). One way to think about these companies is (1) store of value in crypto currency, vs (2) facilitating payments and commerce via fiat. And in this way the comparison evens out. The crypto companies to date have failed in making BTC a medium exchange, instead choosing to take economic rents through capital markets. The neobanks have hit the wall of trying to get profitable at scale, though Monzo's 750,000 users and £2 billion in facilitated payments transactions points the way. Looking at Revolut, we see about 2.2 million users and $18.5 billion of transaction volume. That's a medium of exchange story.

Two more thoughts on neobanks. The burn should slow down and economics seem likely to improve. On the revenue side, consider that most of the neobanks (Monzo included) started out as pre-paid cards that you load, with a nice mobile interface. That's pure cost, because the Fintech has to pay a third-party for each card while making no revenue of any kind. So Monzo's conversion from pre-paid card to current account under a banking license matters, because they can actually make spread revenue on deposits. On the cost side, the neobank claims to have reduced the cost of maintaining an account from  £65 to £15 -- pretty good operating leverage, but for the upfront cost of acquiring the customer. Since the market is crowded (Revolut, Starling, N26, B Bank, Finn, etc.), we expect venture funding to continue fueling the turf war.

And second, the implementation of Open Banking may not be going according to plan. As a reminder, European legislation PSD2 is supposed to expose incumbent banking data via structured APIs to third parties that want to build upon banking information and money movement. In theory, this lowers the stickiness of bank accounts, allows data to travel safely into aggregators and apps, and lays the groundwork for financial bots and agents that make shopping decisions. Surely, neobanks would benefit from the ability to see and move these traditional assets. Well, maybe not. According to blog Open Banking Space, major barriers stand in the way erected by incumbents: "(1) lack of rich data or functionality on the account information APIs, (2) a regressive method coupled with very poor authorisation journeys on the banks’ platforms, (3) technical challenges such as that posed by a lack of immutable transaction IDs’, and (4) the absence of any bank-provided, data rich testing environments." Who will get blamed for this end of the day? The apps trying to build experiences, not the incumbents. But if you don't want to deal with incumbent infrastructure, there is always Bitcoin. 


Source: Problems in Open Banking, Monzo (current accountsAnnual Report), Coin Telegraph (Binance), Treasury XL (PSD2)

HSBC's branch robot and BBVA's facial recognition payments

We are always searching for what's exciting about Fintech, what's at the edge of the wave. It's encouraging to see, for example, JP Morgan launching its digital bank Finn out of St. Louis into the world, Goldman's Marcus lending to lots of subprime risks, and Venmo (i.e., PayPal) putting out a debit card to fight over the neobank consumer. But it's also sort of obvious. This is the innovation of 5 years ago, deployed at scale. Of course large finance firms have no choice but to innovate and copy startups, of course startups will diversify products from payments and banking to investments and insurance. We know the direction of travel, there is only one way to go.

The edge in consumer banking, apart from crypto, is figuring out the role of the human in the context of artificial intelligence. About four years ago, we started seeing chatbot companies like and Kasisto building out natural language interfaces into financial data, and virtual agent companies like Digit and Trim perform account actions like savings and planning. Now, robots are spilling out into the physical world to do the emotional labor of human employees. Take a look at HSBC, parking Softbank's Pepper robots into its physical branches. And while yes, this is a gimmick like Saudi Arabia giving citizenship to the Sophia robot, it is another step towards embodied digital agents.


The robot is currently used only for generic queries, like a chat window on a website but in a physical location. Its cartoonish form moves it out of the uncanny valley, to which both Sophia and something like Soul Machines still succumb. It is not integrated into a user's account or actual financial situation, but we can see a future where such automated interfaces are a bridge to bring in new customers -- both those that are not tech savvy and require the emotional labor, and those that want to geek out over a new user interface. And if you want to understand how something like this could become a window to your financial soul, just check out BBVA's new payment system. A camera placed at the checkout counter reads your face, identifies it using machine learning, and charges your account. All it takes is one camera, and one robot.


Source: Reuters (JP Morgan's Finn), Business Insider (Goldman subprime MarcusSoftbank Pepper), Techcrunch (Venmo debit), Independent (Sophia citizenship), FSTech (BBVA facial recognition payments)

Alibaba's New Retail digitizes physical retail stores

Under our augmented commerce theme, we believe that mixed reality, AI and other digital enhancements will leave the browser and become anchored in the physical world through hardware and software pushed by companies like Apple and Amazon. Such a transition will be as transformative to retail commerce as the web and the browser were to e-commerce. The United States is showing glimmers of this through incremental symptoms. For example, Walmart and its competitors are implementing an IBM blockchain solution called Food Trust, which acts as a ledger for each food item across the entire supply chain, and cuts down identification time from 6 days to 2.2 seconds. Or look at DHL employing augmented reality headsets to help employees identify required packages in a warehouse.


But Chinese tech companies are many steps ahead. A great Axios article focused our attention on an Alibaba initiative called "New Retail". A mom and pop store can pay $6,000 for a digital renovation, and a $620 annual membership fee which locks the store into the Alibaba ecosystem. The digital tools include a heat sensor to track foot traffic, an AI-backed app and the Alipay payments system, and the entire Alibaba delivery and fulfillment infrastructure. The store becomes a physical endpoint for Alibaba's e-commerce platform. JD and Tencent are doing the same.


A chain called RT-Mart updated 400 of its stores in this manner. Here are the bells and whistles: (1) one-hour arrival on e-commerce orders in a 3km radius to the physical store location, (2) orders through a branded app are fulfilled by inventory from local physical stores and carried by conveyor belts to packing/delivery areas, or all the way to a customer's home, (3) in-store physical e-commerce kiosks with payments using QR codes, and (4) red envelope coupons that are gifted within Alipay can be redeemed for physical goods. Will customers adopt these solutions? Well, 500 million already have the needed app on their phone and are trained in  digital payments behavior. Maybe China will save the American mall! And as we've been saying about Amazon, technology and finance are mere enabling features of the commerce happening in the system. 


Source: Alibaba (AxiosAlizila), WSJ (Walmart and Blockchain), UploadVR (AR for business)

EOS liquidation of Ether reserves impacts markets


ICOs were the reason the crypto prices shot to the moon last year, and they are also (part of) the reason why we are now seeing a prolonged weak market. Maybe this is obvious to some of you, but it's worth pointing out the dynamic. When consumers became excited about token crowdfunding last year, they allocated cash and their Bitcoin capital gains to buying Ether, which allowed them to participate in token sales. This year, new consumer demand for Ether has been weak in light of regulatory and execution uncertainty. But the need to sell Ether has skyrocketed because of all the ICOs that have raised crypto assets, and now again need cash to pay staff in the real world.

The $16 billion or so of ICO proceeds, if not more when adjusted for market fluctuations, is a large chunk of Ether's $45 billion market cap. The asset has shed about $90 billion of value since its peak at $135 billion. Assuming ICO liquidations of $5 or so billion led to the continued pressure on the Ether price, the negative impact has been magnified 10-20x via sentiment and illiquidity. Again, this is not the only variable at play as correlations to Bitcoin and other large coins are around 90%, and speculation still prevails. But it is important to note the structural dynamics. If you have any idea how we can test this hypothesis, let us know!

As an example, take EOS and it's $4 billion of crypto currency. The project has hit community issues almost immediately out of the gate, needing to re-write the governance structure and creating negative sentiment in the space. But more damaging is when $100 million worth of Ether is sold by EOS in a single day on Bitfinex. The lack of institutional liquidity matters because the price is far more sensitive to large blocks in response. And when a company is able to raise money in the form of its top competitor's token, and then create ongoing selling of that token across thin markets, we don't think this create good market dynamics. Further, as more large ICOs are done privately -- see Tatatu's $575 million or Telegram's $1.7 billion closed pre-sales -- the demand to buy Ether in order to buy the ICO token is simply not there.

Source: Coinmetrics (chart), EOS (CoindeskCryptoslateTrustnodes), Coindesk (Tatatu)

If Amazon Did It: Entry into Mortgage Finance and Wealth Management

Ah, the constant specter of BigTech gobbling up finance. In this installment, rumors are circulating that Amazon may be interested in acquiring LoanDepot, a non-bank lender founded in 2010 with over 6,000 employees and 180 locations in the United States. The asset in this case is geographic reach, loan underwriting software, and the human presence that is still necessary to deliver financial products to the long tail of the American population. But such rumors are most likely false, especially since the CEO of LoanDepot sarcastically shot them down. Also, imagining such an acquisition betrays a lack of understanding of Amazon's platform strategy, which optimizes around users within the Amazon ecosystem, and not sales of financial product. 

So we want to point you to two resources to help anchor that platform strategy. The first is a comprehensive analysis by CB Insights on the financial services moves that Amazon has made to date. Highlighted below are a few of the key graphics. Amazon's product strategy has been around growing payments and lending -- in order to facilitate commerce in their core offering. Meaning, if underbanked customers can use Amazon's cash products or consumer credit to purchase goods, that's great! If merchants can get an SME loan via Amazon that helps them offer more products on Amazon's platform, that's great! If purchased products are insured via Amazon's partners, that's great! You can see why re-financing a mortgage doesn't quite make sense in this context, until Amazon sells residential real estate that is.


And second, we spent some time with the 2018 World Wealth Report from Capgemini. It surveys the growth in global HNW investors, their preferences and asset allocations. And this year, they also asked whether investors would hire a BigTech firm as a wealth manager. About 40% of the respondents said they would give such a firm 10-50% of their share of wallet, with Asian and Latin American millionaires showing the highest appetite. Further, the report suggests that most likely outcomes for BigTech entry into finance are: (1) partnering with manufacturers of asset management product, (2) unbundling the financial services industry through competitive payments and banking products that are priced below industry levels, (3) providing services and software into incumbents, and (4) partnerships via messaging platforms. Least likely outcomes are (a) no market entry at all, and (b) acquiring a wealth management firm. Google, Apple, Amazon, Alibaba and Tencent are all desired entrants into the wealth space. So if you're a financial incumbent, time to accept your fate and get into a pole position for partnership.


Source: CBInsights (Amazon summary), Inside Mortgage Finance (Loan Depot), Capgemini (World Wealth Report

SEC Fines Betterment $400k, while SigFig and SmartAsset Raise $50mm and $28mm

Growth is expensive and risky. The brightest example is Facebook, with its now outdated mantra -- move fast and break things. Facebook did move extremely fast, and it may have broken some fundamentals, like the concept of privacy for an entire generation, as well as functioning democracy. But we digress. Even in Fintech, companies like Zenefits (and SoFi, culturally) have run into walls by taking shortcuts encouraged by growth. From that perspective, we are not shocked to see a $400,000 fine levied on leading roboadvisor Betterment for accounting practices between 2012 and 2015. This amount is half of the Zenefits fine (about $1 million) for unlicensed insurance brokerage,  and far less than the millions in wire fraud penalties associated with Theranos. But it does throw a wrench into the roboadvisor growth engine. 

From an investor's perspective, however, legal and regulatory exposure is just risk. And roboadvisor assets are still attractive. As proof points, SigFig has just raised a $50 million check from growth equity firm General Atlantic, with participation from prior investors like USV, Bain, and UBS. Remember, SigFig powers the UBS roboadvisor. In order for a company to raise growth equity capital, it likely needs to be running at $10-50 million in revenue, which means that the Wall Street contracts that SigFig has signed are probably quite juicy. 

As another example, consumer finance website SmartAsset raised $28 million from Focus Financial and Citi Ventures, among others. Unlike the roboadvisors, SmartAsset is a destination site with massive traffic, claiming to reach 35 million consumers a month. Financially, this is a far more scalable model than trying to gather assets under management. That traffic can then monetize through advertising or lead generation -- for comparison, look to LendingTree (public at $3B) or ClearScore (Experian acquired for £275m). Focus Financial, a conglomerate of independent financial advisor businesses, is looking for its own exit soon, and this investment may give it some runway in a more traditional business. Regardless, there are not that many credible independent roboadvisor assets left, as most have been snapped up by incumbents. Despite profitability questions in the space, the direction of travel is growth.


Source: Business Insider (Theranos), Techcrunch (Zenefits), Investment News (Betterment fine), General Atlantic (SigFig), WealthManagement (SmartAsset), Autonomous NEXT (Clearscore)

Ethereum not a Security Because it is Decentralized

Let's parse today's state of regulation for crypto assets, and the glimmers of what the future will look like. Ethereum and Bitcoin are not securities, said a senior SEC official a week ago. Why? The argument rotates around decentralization -- not because regulators care about decentralized networks, but because you need an entity to lead an offering. The Howey test demands a common enterprise that gives purchasers an expectation of profit solely from the efforts of others. Does that make sense in the world of (1) decentralized networks set up by communities for mutual gain, and (2) changing expectation about tokens as platforms are built? Probably not, but until the courts create a new model, it's what we got, and it is beneficial for Ether.

What's probably not really beneficial for Ether as a development platform are ICOs (vs DApps). Yes, they are still the killer app for crypto, but they have also sapped ETH of its role as a currency for DApps. Instead of a single currency that can power a digital smart contracts economy, we have thousands of disparate tokens of questionable liquidity and value. And while Ethereum itself may have avoided being a security, the tokens launched using it as an offering platform are exposed to continued regulatory risk. The CBOE president, for example, expects to see SEC prosecution of many large token offerings, and potential class action lawsuits against projects that fail to deliver. No amount of disclaimers and structuring will help against an angry mob. 


To reiterate the point, ICOs have been narrowing Ethereum towards a crowdfunding offerings platform. Compare an entrepreneur's choice of Ether vs Stellar, for example, as the choice between the corporate law of Delaware or New York. In choosing a state, you have access to all the common law that has emerged from centuries of litigation. This is like choosing a programming language that has the best code libraries. Regulating this choice for financial disclosure makes no sense. When looking at a particular use case, however, regulatory approval will still be a gate. Square needed to get the BitLicense in New York in order to process crypto payments. The same type of regulation may come to Canada for exchanges and payments companies (KYC/AML for $1k transactions, reporting to regulator for $10k transactions). For a sovereign regulator, the best strategy is to control the choke points.


Source: CNBC (SEC on Ether), CrowdfundInsider (CBOE), Autonomous NEXT (Howey test), Coindesk (Square), Cointelegraph ($10k transactions), Techrunch (Square cash), NY State (Bitlicense), (creating a token)

Public Market Paths for Crypto Billions


Here is a founder's dream. You and a partner own 60% of the company you founded. That company printed $2.5 billion of revenue last year. You have massive political capital to influence the industry in which you are a top player. To cash out, you are looking at an IPO -- the real regulated kind -- of $9 to $12 billion. Welcome to Bitmain, which was founded in 2013 and makes Coinbase and Binance look like amateur town. The company is behind Antpool, the world's largest cryptocurrency mining pool, and is also the largest producer of mining hardware. The next step in the playbook is neural network processing, which also requires specialized graphics cards, a market currently dominated by Nvidia. We still think that mining is something that should be done by traditional banks to support the crypto rails. And so if Bitmain does go public, it will open a fascinating window into the operations and economics of manufacturing global trust machines using hardware and software.


Let's take another potentially meaningful instrument. The saga around a Bitcoin ETF continues to grab headlines, this time with a combined effort from $45 billion AUM asset manager VanEck and crypto company SolidX. Prior efforts were all shot down by the SEC due to an "immature" market -- pointing to everything from liquidity, to execution, to volatility. The lack of an ETF, in turn, created an absolute nightmare in the public equities markets. Instead of buying something efficient at 20 bps, investors flooded into blockchain-pretenders like Long Island Ice Tea. Which caused the SEC to start chasing these opportunists to shut them down. But it looks like VanEck may have an ETF contender that is made up of actual cryptocurrency in cold storage with an insurance layer against hacking. That sounds way better than an ETF made out of CME/CBOE futures to us. Except the planned minimum investment is $200,000.

Let's linger on the differences between these two potential events, and the existential questions they create. One is a currency (or store of value) in which digital economic activity is denominated. The more economic activity shifts into it, the better; speculative trading and labyrinthine financial structures don't count. The second is a traditional oligopolistic champion with regular old equity, selling a claim on profits off $2 billion in revenue. That revenue is essentially a toll-booth on the functioning of the Bitcoin network. If Bitcoin continues to exist at all, Bitmain continues to charge its toll for network maintenance. Which one sounds like more of a sure thing?

Source: Bloomberg (Bitmain), Marketwatch (NVIDIA), WealthManagement (VanEck), Bimain Equipment (BlockonomiHashnest)

10,000 People at Citi Fertile for Machine Processing


Nothing quite makes human people like us perk up more than being described as "most fertile for machine processing". And yet, that's exactly what Jamie Forese, the president of Citigroup told the FT about the investment bank's personnel. Out of 20,000 operating roles, he sees 10,000 potentially going away over the next five years. Now that 50% is a pretty big number, and not everyone agrees. HSBC, for example, see only 5-10% more automation potential over the same time period. So let's cut the pizza at 25%, and still gawk in disbelief.

Yet, this live data point is exactly in line with our analysis of what artificial intelligence will do across the financial services sector. In Augmented Finance, we identified $1 trillion of cost at play across banking, investment management and insurance. Behind that cost are those real human people -- 2.5 million of which are in the United States, and about 160,000 of which sit in the middle office of investment and banking organizations. Looking at the last 10 years only, the FT found 60,000 jobs cut from the top investment banks. The way things are headed, sounds like it's time to take programming courses at General Assembly.

None of us should be glib about the potential impact on the lives of employees of these companies. One of the ethical questions with which we struggle is -- whose responsibility is the welfare of these employees? Does Google and free machine learning software share a responsbility? Do CEOs of too-big-to-fail finance firms share a responsbility? Do investors looking for cost cutting share a responsibility? Does the consumer, wanting to pay nothing for banking, share a responsibility? And if yes to all, how do we come together to make a world where we celebrate not the cost cutting potential, but instead the potential human productivity growth? Is technology a shield or a sword?


Source: Autonomous NEXT (Augmented Finance), FT (Citi), Look and Learn (Lamp Lighter),

Can $250 Million Get Insurtech WeFox Past Lemonade's Litigation


German insurtech startup WeFox -- backed by Ashton Kutcher and banked by Goldman Sachs -- is in the market for $250 million of fresh capital to finance international expansion. That is a meaningful amount of venture for any insurtech company, especially one that just raised its Seed round in late 2014. See the table from Coverager below for the largest raises in their database in the space -- though we would advise you to ignore Theranos. Since 2014, WeFox has changed its name from FinanceFox, acquired ONE Insurance, and intermediated deals with a number of large underwriter incumbents.

So what kind of service do you need provide to deserve a unicorn round? Well, WeFox gives customers the ability to manage all their insurance contracts across products in one place, supported by a personal agent. They act as a mobile-first broker for individuals, and provide an outsourced front office to incumbents that aggregates different insurance use-cases into a single app. The app can be free because large insurance companies pay WeFox to get clients, and then to manage those clients. Can you say B2C2B2C?

Which brings us to ONE. Whereas WeFox is the insurance supermarket, ONE is a proprietary product on that supermarket shelf. And it has just been sued by Lemonade, the radically transparent renters insurance startup, for copyright infringement and reverse engineering. Allegedly, WeFox created fake accounts and made fake claims on the Lemonade app to copy its workflow and process. And Lemonade has hired an expensive law firm -- White & Case -- to litigate. This makes us ask three questions. First, is user interface something that can be protected by copyright? There must be something deeper to this story. Second, are startup ventures now so well funded that they make worthwhile litigation targets? And third, if insurance is ripe for disruption leading to a massive market for new companies, isn't it better to spend cash on acquiring customers rather than lawyers? 


Source: Pitchbook (WeFox Raise), Coverager, (Insurtech Raises), LinkedIn (Lemonade vs One), SPGlobal (Lemonade Growth)

Crypto Funds, RIAs, Regulations in a Box

The cost of launching a startup has fallen from a few million to a few thousand dollars. Why? Amazon and its cloud have collapsed the needed IT infrastructure to a cheap off-the-shelf subscription. Stripe Atlas has made corporate and payments gateway setup a breeze. But what if you're starting a financial entity, and not a software company? What if you're starting a Registered Investment Advisor, and not a wealth tech platform, or if you're starting a hedge fund offering a crypto index, and not a blockchain of blockchains? For that, let's take a look at compliance in a box. 

One of our favorite companies in the independent wealth management space is RIA in a Box. It does what it sounds like -- it sets up a Registered Investment Advisor entity, registers it with the SEC and the appropriate States, and manages ongoing compliance requirements as an affordable service. So if your advisory practice manages $5 million or $500 million, this solution can get you started. In fact, out of 12,000 RIA firms in the US, RIA in a Box has 3,000 as clients. Not surprisingly, Aqualine Capital Partners just acquired the firm through an LBO at an undisclosed price, which tells us that the firm is a cash cow -- an LBO requires a slug of debt that can be serviced by reliable, steady cashflow. And if you control the compliance reporting aspect of a financial entity, it's a very short reach to start selling regulatory, administrative and value-add software.


Now think about crypto funds. It's the same problem -- nobody has any idea how to structure them, which regulatory jurisdiction to pick, what bank to use, or how crypto assets are treated. Traditional counsel could easily cost over $100,000 to go through the motions. Enter the Crypto Fund-in-a-box companies! Take Vauban, which provides an interface to select a type of investment fund, its jurisdiction, target size at launch, while a real-time entity structure is built on screen indicating the cost of setup. Other similar plays include Fundplatform, Otonomous, and Bluemeg (note: we don't know or endorse any of these). Could the ease of solving this international puzzle lead to a similar growth outcome for crypto fund entities? Looking at the data, the first 5 months of 2017 saw 40 new crypto funds; there were 45 new entrants over the same period in 2018. Market volatility has not deterred fund formation. That doesn't mean funds won't fail (e.g., Apex Token Fund shutting down after failing to raise $25mm), but it does mean more will keep trying if it's as easy as clicking a button.


Source: RIABiz (RIA in a box), Crypto Fund in a Box (VaubanFundplatformOtonomousBlueMeg), Financial Planning (RIA totals), Autonomous NEXT (Crypto Fund totals), Stripe Atlas

Circular references in Crypto markets as exchanges: Binance and Huobi launch new ecosystem funds


Towards the end of last year, we noted that there was a circularity in the crypto, private and public markets. Large ICO launches were inspiring private companies to follow as well (e.g., Kik), public companies changed their names to blockchain pretenders (e.g, Long Island Ice/Chain), crypto companies pointed to this as progress and Bitcoin went up. It wasn't purposeful market manipulation, but a hype cycle reverberating in a small room. What we're seeing now is, well, kind of worse.

Tezos raised $232 million when the price of Ether was about 50% or less than it is today, so about $500 million now. The result is a lot of lawsuits, no product, and the formation of a $50 million venture fund. Binance raised $15 million through its ICO, which now trades at over $2 billion. The exchange is launching a $1 billion venture fund. Huobi raised $300 million, and though the token trades at a discount, it is also launching a $93 million venture fund. EOS, the largest ICO ever at $4 billion, is committing $1 billion to venture through partners like Galaxy Digital. Another example stuck out at us from a recent Coindesk article, where Meltem Demirors described this cycle -- "[Blockchain Capital invests in Ripple, which owns XRP currency]. Ripple took some of that XRP and gave it back to Blockchain Capital. Blockchain Capital then turns around and invests it in Coinbase ... Coinbase now created a venture fund investing in startups Blockchain Capital is also investing in, who are then turning around and investing in startups with ICOs."

This is billions and billions of capital that were invested for one purpose -- to help the fund-raising team build software products that investors want to use -- that are being re-purposed into another direction entirely. Hey, maybe you hired me to program this website, but I decided manage your retirement portfolio instead. One result is a skill mismatch: lucky coders are not professional investors. Another result is the loss of focus. And the last is the systemic risk to the whole ecosystem. If the investment returns are based on financial engineering and memes, rather than some real economic activity to underpin our excitement, then a regulator pulling hard on one thread will unwind the entire experiment.


Source: ICOs (Binance ICOHuobi), Venture funds (TezosBinanceHuobiEOS), Coindesk (Quote)

For $7.5 Billion, GitHub is now Microsoft, and what that means for Fintech.


Remember when Microsoft got left in the dust by Apple's iPhone? Or when Bing tried to beat Google? If you're a tech firm, missing a platform shift like search, mobile and social is profoundly painful. Well, no more. The enterprise tech giant is in cloud, blockchain, and augmented reality. And it's in enterpise social media -- big time. Microsoft is putting up $7.5 billion to purchase GitHub, which has 28 million developers in the community, and over 60 million code repositories. Think of that as shared documents on a massive cloud drive, but those documents are executable and the people doing the sharing are engineers with razor sharp skillsets.

This is an interesting turn for Fintech and Crypto. To be honest, we always thought of GitHub like a community resource, similar to Wikipedia, and not a corporate entity with founders that want to get monetized. But like LinkedIn, this asset found its way to Redmond. While Facebook is still cleaning off the hangover from Cambridge Analytica and selling customer data, Microsoft has positioned itself as a developer- and open source-friendly ecosystem catalyst. Which is ironic, given its history as a monopolist, open source antagonist, and start-up crusher. 

GitHub has tendrils into every startup team in the world. If you write code, you use GitHub for version control, collaboration, and sharing. Where do you think all the crypto code sits? If you're Ethereum or Bitcoin or anyone else in the world, code commits to GitHub are a fundamental indicator that crypto funds use to evaluate the health of your project. Or, if you're trying to work out crypto governance and decide which proposals to include or exclude in the main codebase, again, you are using GitHub to manage decisions. Or if you're a teenager wunderkind interested in using open source Artificial Intelligence frameworks to build your machine -- e.g., Tensorflow or PyToch -- you use GitHub to learn and get started. Welcome you tired, poor, huddled masses yearning to breathe free; welcome to Microsoft.


Source: Techcrunch ($7.5 billion), Wikipedia (infographic), Deloitte, Azeem Azhar (AI frameworks)

Amazon R&D Spend is 2x JP Morgan's Entire Tech Budget


Yet another surprising data point in the 2018 Internet Trends report is the amount of money American companies devote to Research and Development. Tech companies spend 18% of their revenue on R&D and Capex; with Amazon at $34 billion, Google at $30 billion, Apple at $24 billion, and Facebook at $15 billion, with the industry total at approximately $260 billion. More than half of those GAFA figures are flowing into R&D, an investment into the future. We are not sure if that includes the free cash used for acquisitions, but an acquisitive posture in the industry by incumbents is also great at motivating entrepreneurs to start companies in the space and contribute to overall growth.


This clicked for us because we were recently at a Fintech event at a top-5 global custodian, where the audience discussed how historically the financial services industry itself has not had an R&D function, and that Innovation and Corporate Venture arms of financial incumbents targeted at the Fintech industry are now essentially catching up. In fact, this was true for all of venture. In 2010, global venture capital investment was $45 billion, of which less than $2 billion was in Fintech. That's less than 4% and completely out of balance. Remember that financial services as percent of GDP is about 20%. In 2017, global venture capital was $130 billion, with Fintech funding at about $30 billion -- that 23% looks far healthier.


But when we look at financial incumbents, a similar level of future investment is simply not present. We had done a deep dive on tech spending and Fintech readiness by big banks last year, called Bankosaurus. We found that the biggest spender was JP Morgan, with a $9 billion tech budget, of which about $6 billion flowed through the P&L. That's not R&D, but the cost of keeping all tech running, plus investments in technology. In the graphic below, you can see that few banks have more than $1-2 billion to devote to technology, and even less to devote to Fintech. Total tech spend is 5-25% of costs for most banks, which implies 3-12% of revenue, give-or-take. Haircut that by 50% for keeping the lights on, and you have 1-5% devoted to R&D. Thus the leading banks are spending 5% of revenue on "change the bank" initiatives, with the rest quite behind. Not a pretty picture, and one that has to change for finance to remain relevant.


E-Commerce Growing to $500 Billion, Augmented Commerce Coming


There is no greater gift to geeks like us than Mary Meeker's Internet Trends report. If you haven't seen the 2018 version yet, what are you waiting for? Time to read 300 slides in 30 minutes. The key takeaway we remember from last year is the consumerization of the enterprise, and that much of the user interfaces of today's mobile and web apps are directly inspired by 1980s video games. Not surprising, since many of the people building companies today and making design decisions grew up together with the maturation of the video game industry. This year's report has a great section on China and sovereign investment on edge technologies, but lets leave that rabbit hole for another time since we covered that last week.

In this entry, we want to highlight takeaways around payments. The story that comes through the deck is that digital commerce and digital payments are intertwined, and both are rising. E-commerce has grown to $450 billion per year, up from $180 billion in 2010. As share of total retail sales, E-commerce has doubled from 6% to 13%. In terms of everyday transactions, consumers claim to have only paid for items in retail stores 40% of the time, with digital channels comprising the other 60%. This category lumps together everything from P2P payments like Venmo (7%), to money movement through messenger apps (7%! Really!?), to shopping through smart home devices (3%). While this survey sentiment doesn't yet line up with actual retail dollars, it's fascinating to see consumers using so many emerging payment mechanisms.


So we are experiencing this broad digitization of commerce, with E-commerce living in the web and in our mobile apps. But the world is on the verge of another shift, which is augmenting the physical space with embedded digital commerce. Meeker highlights retail stores that use traffic heatmaps to optimize layout, and smart devices in shoes that measure in-store sales conversion (from fitting to purchase). We add the technology of, which is similar to what is likely under the hood at Amazon's stores, and at some point will be at Whole Foods. A machine vision algorithm can watch product stocks, measure sales in real time, and send workers to replenish them. Physical is becoming digital, which we believe have a deep implication on the types of payments companies that succeed. Not point of sale, but point of intent.


2018 ICOs at $9 Billion, But Definitely Slowing


Well, there doesn't seem to be another way to say it -- ICO activity is absolutely and unequivocally slowing down. We were optimistic that token offerings were independent of crypto currency prices -- in part because early stage technology venture activity should be separate from late stage market dynamics. But it was only a matter of time before the slowdown in crypto prices was reflected in a slow down of ICO funding and crypto fund formation. In fact, offerings as a function of crypto market cap, and especially as part of Ethereum's market cap, seem to be fairly stable as a percentage.

The numbers: if we look at all token offerings above $1 million in funds raised, 2017 saw $6.6 billion and 2018 YTD has seen $9.1 billion. So far so good, right! But, if we pull Telegram and EOS out of both numbers, we land at $5.9 billion 2017 YE and $4.1 billion for 2018 YTD. That's still a higher pace than last year, so let's drill down into the monthly figures. In particular, if we pull out Telegram ($1.8 billion) and EOS ($4.1 billion) on a monthly basis, the monthly trend look severely down -- to $560 million from a high of $1.5 billion in December 2017. So unless you believe in the continued presence of mega deals, token offerings have indeed been dragging due to continued regulatory uncertainty, tax overhang, and a lack of tangible progress in software adoption by the mainstream consumer.


That said, the uncertainty will get resolved. Even if Western regulators constrict the space into a narrow box, there are many legitimate jurisdictions that want to be crypto Delaware. Look at Japan: from Rakuten tokenizing $9 billion of loyalty points, to Mitsubishi bank talking about launching a cryptocurrency. Fear is worse than truth. And, we may indeed be entering the era of mega deals. Many of last year's token projects were built by new teams, like Seed stage venture. This year, more mid-stage companies (e.g., 50-250 employees) are tokenizing some asset of their existing operations. And next year, we may start seeing late stage companies bringing their own DLT projects to the market, and marrying them with public crypto. Just look at the Internet wave: March 2000 was the peak value share a percentage of market capitalizations. Despite the crash, the web has never been more present or important than today. Will crypto follow the same hype cycle curve?


Source: Autonomous NEXT (aggregated ICO data), Kleiner Perkins (IT as %)

Machine Readable Regulations


We started with two difficult entries to highlight how the major platform shift technologies, blockchain and AI, are bringing out some of the worst impulses in human beings to take advantage of each other. And further, these tendencies become enshrined in software -- from decisions learned out of data, to bots endlessly begging to steal your money. From this perspective we pivot to Regtech, and in particular to projects that we think could be antidotes for the malaise.

The first is an effort by the FCA to explore offering regulations in a machine readable format. That means that a regulator would provide standards and perhaps even executable code that could plug into Fintech software stacks. Imagine Python's Django, but with a regulatory module that pre-packages data formats for compliant reporting. Similar ideas have been floated by self-regulatory organizations in Crypto, looking to build into tokens the ability to determine regulatory requirements, like accredited investor status or KYC/AML. But to do this at the level of the regulator is far more meaningful because (a) there is way more law that needs to be translated, which relates to real rather than imagined economic activity, and (b) this regulation is a result of an established governance process, which is still immature in decentralized communities. Imagine putting all of the FCA on Github and satisfying requirements through something like the Digital Asset Modeling Language. Compliance costs would actually become trivial.

But now is a moment of transition. Case in point, last week we attended the fifth London cohort of the Barclays Techstars, where a RegTech startup called Audit XPRT introduced their automated audit and compliance solution that uses machine learning to extract structured rules from unstructured paper documents. The aspiration is to reduce compliance-related costs ($270 billion) by 90% and achieve 5 months of work in 5 minutes. Another example is Governor, which creates dashboards of real-time tracking across Risk, Compliance and Corporate Governance. Or take Suade, which tags existing data with an overlay that maps to regulatory requirements and provides apps out of the box against which the data can be checked, no disruption to the bank’s current architecture. It may feel slow, but the law's getting digital.


Source: Github (Ethereum proposals), FCA (machine readable initiative), Digital Asset (DAML), Thomson Reuters/Tabb Forum (Infographic), SuadeGovernor (infographic), AuditXPRT

Scams in Crypto: 20% of ICOs, 5% of Twitter


Getting a wrap around just how much scamming and fraud there is in the crypto ecosystem is a challenge -- but not impossible. As the industry continues to put up impressive fund-raising figures (with new issues at about 2% of Ethereum market cap per month), just how much of this will become valuable projects? We've written before about how creative destruction is natural for startups, and that failure rates in the mid 90% are a reasonable outcome. We've also pegged hacking of Bitcoin and Ethereum to have been responsible for about 14% of money supply in those pools. But what about outright theft and lies?

Two ideas. First, the WSJ analyzed 1,450 ICOs and found that 271 or 18% of them are just total raw scams. Fake copied white papers, team member photos taken from stock photo websites, nothing behind the project but malfeasance. Yikes. And another version of the same was The North American Securities Administrators Association going after nearly 70 ICO issuers in a coordinated action of regulators across the US and Canada called "Operation Cryptosweep". Which is a totally sweet name, for what is a really regrettable but required clean-up of the crypto ecosystem. A 20% chance to lose your money, for no philosophically meaningful reason, is the wrong price to pay for good financial technology in our opinion.

And second, don't forget the propaganda bot armies. Sure, they can influence elections and spread misinformation, but we didn't expect that they would be used for financial warfare this quickly. The practice in question is copy-cat accounts on Twitter that look like a Twitter influencer claiming to give out free crypto currency, if only you send them money first. This is hacking of the human kind and we monkeys fall for it all the time. As a comparisons: (1) email phishing maxes out at 0.70%, according to Symantec, and (2) bot automation is at approximately 10% of all activity on Twitter. Given that the crypto ecosystem is more prone to Internet memes and bounty programs, we would expect the rate of phishing to skew higher, say up to 5% for crypto-related conversations. So watch where you point that digital wallet.


Source: WSJ (18% scams), NASAA (Operation Cryptosweep), Bloomberg (Bot PhishingHacks at 14%), Autonomous NEXT (Failure rates)