Category Archives: Lending

Is a European consumer credit model coming to the US via ecommerce?

If you are a retail customer in most European countries, you are quite familiar with this sight:

Selling credit or selling televisions?

If you are from the US or the UK, chances are this will not be a recognizable image. There is a fundamental difference in the way retail consumer credit is built and distributed between the US/UK and Central Europe, which can be summarized as (I am oversimplifying):

  • USA/UK > revolving credit through credit card (financial institution or merchants)
  • Rest of Europe > installment credit in store or through payment card (financial institution or merchants), often linked to a specific purchase. The interest cost is either carried by the customer or the merchant.

In that perspective the launch of companies such as Affirm, Bread in the US, as well as the geographic extension of Klarna are quite interesting. Effectively, they are introducing in the US installment credit at checkout, a very European product dominated in its traditional form by companies often linked to banks such as BNP Paribas Personal Finance, Santander Consumer Finance, Credit Agricole Consumer Finance, except the new players are starting from e-commerce and mobile e-commerce checkout.

The daily life of an ecommerce site

If there is one think that ecommerce website hate, its the Cart Abandonment Rate. It basically means you had everything right from acquisition to product mix, to site/app experience but when comes the time to pay, the user drops out. Making the checkout experience seamless is a constant concern and making any change / adding additional steps is a risky move. This will be a challenge for the emerging players who have to prove offering credit brings more users than it deters in the checkout process.

On a broader scope, the impact of Fintech on the American Credit Landscape is one of the most interesting trends in the recent years. Whether it is the new crop of credit at checkout players or the marketplaces lenders such as Payoff* or Lendingclub, they all participate in Credit Transformation, moving consumers from Revolving Credit to Fixed Term Credit. Compounded by Millenials’ attitude toward Credit Cardshttp://www.bankrate.com/finance/credit-cards/more-millennials-say-no-to-credit-cards-1.aspx , I can’t wait to see what consumer finance will look like in the coming years.

Are we at the beginning of a move beyond Credit as we know it?

2016 and beyond in fintech: a few thoughts

valthorensLooking at 2016 with the experience of the past 7 years in financial services. it will be a pivotal year for financial services. In many ways we are coming to the end of a phase, that started with the world‘s the most important financial crisis since 1929. The FED hike is upon us, after having experienced one of the most destructive slow downs, of which the effects are still very much acute through the world. While banks are still experiencing the effect of the crisis notably through restrictive regulation and a continuing string of financial scandals, their public messaging is one of innovation and change. For the raft of new players in financial services, the funding environment has never been as good and some striking successes are showing the way forward. A new wave of technology, from blockchain to AI and omnipresent sensors is shaping a new world (hopefully not brave).

2016 will see a strong competition in the pure banking space. In Europe especially, a crop of new banks and alternative banks will push strongly into the market. European digital banks such as Fidor are expanding beyond their core market.The UK regulator’s move to lowering the barrier to entry in becoming a bank will come to reality with Atom Bank, Tandem, (…) establishing themselves as a brand to customers. Alternative solutions based on prepaid born in various european countries are also expanding beyond borders, with SEPA as a core foundation to propose bank like services. With PSD2 looming, traditional banks have the opportunity / incentive to more easily expand collaboratively with startups.

What about the alternative lenders, roboadvisors and other digital financial services players? Having spent the last few years building a trusted brand as well as consolidating a customer base, it is highly possible they will start leveraging this to expand horizontally into other markets, whether collaboratively or by launching their own services. For example, blended remittance and multi currencies current accounts are highly complementary services for clients with attachments to multiple countries.

This collaboration between emerging players will most likely extend beyond end customers, towards balance sheet management, especially for emerging banks looking at matching assets to their new found deposits. Looking beyond the pure European and US context, in a developing world that is increasingly interconnected, via diasporas and large economic regions, the winning platform of the coming decades appears to be the messaging platforms: Wechat, Line, Facebook Messenger and WhatsApp are all showing significant growth, engagement and increasingly financial services integration. 2016 may be the year where we will see Facebook becoming more integrated with financial services, leading with seamless blending of messaging, commerce and payment first and potentially P2P transactions next.

If 2015 was the awakening of the entrepreneurs, investors and incumbents to the potential disruption of insurance by digital first players, 2016 may well shape up to be the coming of more true game changing challengers in the space. In reality, most of the investment and company creation has gone through changing distribution to digital means, focusing mainly on two main trends: the modernization of broker first markets, such as the German or Swiss markets (following the UK market) where digital acquisition is seen as a cost efficient, scalable way to attract customers and the reorganization of health insurance in the US market following the Obamacare reform. What is maybe more interesting is the coming wave of startups looking to challenge the core business model of insurance companies, an early example of which is Oscar. These new players are leveraging lower cost, scalable infrastructures, sometimes through mutual insurance mechanisms.Increasingly, a smart use of the full insurance stack including innovative insurance companies and nimble reinsurance players allows them to succeed at lower scale. Another trend that will shape up 2016 is the change from insuring people for their things to insuring their things for them partly driven by the increasing importance of the internet of things and the facts that objects are more and more blending with the underlying services they provide.

If 2014 was seemingly the year of Bitcoin, with its pricing toping above $1,000. 2015 was the year of transitioning from currency to infrastructure. 2016 will be the year of blockchains as infrastructures for smart contracts. Financial services use cases in clearing and settlement will continue to dominate the headlines with an increasing number of pilots and low scale production releases. Financial services being multi parties by definition, they are a prime market for decentralized trusted software but the use cases go well beyond pure transactional financial services and the coming year will see an increased such projects. Large scale projects are also by definition multi parties and can benefit from a smart contract infrastructure. And lets not forget the elephant in the room, our civil infrastructure which needs to reinvent itself for the digital age.

Finally, taking a longer view, it will be important to focus on two complementary trends. Artificial Intelligence and omnipresent data. Financial services core functions are based on managing capital scarcity and information asymmetry. With the amount of data increasing at high speed, notably through widespread, cheaper and more detailed sensors and the capability to process it progressing in parallel through the use of machine intelligence, the core market of financial services will be affected. The autonomous car is the perfect metaphor for it: what is the insurance market for a sensor full, crash avoiding vehicle? But similarly, what new markets will this create for financial services?

Discussing Machine Learning and Financial Services

It feels like fear around artificial intelligence is slightly receding and that the discussion has evolved toward what it can actually do now and its application in the current crop of technology companies. I, for one, welcome this new step in Machiavellianism from our machine overlords, well played, well played….

Robot pretending not to be able to open a door

Machine learning is about learning, measured as an improvement of an output (From the good read: http://www.androidauthority.com/what-is-machine-learning-621659/)

“ A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.”

Machine learning comes in various shapes but broadly there are important distinctions between:
supervised learning, in which the computer is given example inputs and desirable outputs, with the goal for the machine to learn how to map the two.
unsupervised learning, in which the inputs are not characterized and the machine has to learn what characteristics of the inputs are linked to the desirable outputs.

It is also important to understand that desirable outputs are effectively formulated as algorithms. They are very diverse and not only depend on the problem but also the approach taken to solve it. For example the graph below gives a good overview of potential solutions to a regression analysis problem.

So how does a machine think? Very differently from us as its input are more constrained data sets and its outputs are driven by equations. For example, this is how a computer views a picture of a cat on a carpet in order to be able to classify it as such.

from http://marketingland.stfi.re/how-machine-learning-works-150366?sf=ppelxr

So what can you do with a machine able to understand inputs and find correlations? A lot:

From: https:[email protected]/the-current-state-of-machine-intelligence-f76c20db2fe1

This map is already outdated increasingly, all businesses will leverage machine learning in some form and the infrastructure to do so is increasingly easier to access:
– IBM recently launched APIs to access Watson
– Google open sourced one of its core sofware Library for Machine Learning : Tensorflow
– Microsoft, just after Google announced its own open source release
– Machine learning is an increasing part of Amazon’s computing offer

It makes sense for these large players to open source a big part of their research and benefit from a broader development community. As we discussed above, an essential part of machine learning is gained from data and training, two components that remains firmly proprietary.

In Financial Services, in my mind there are short term two main cases for machine learning: one connected to the interface between humans and finance, the other to optimizing the analysis of vast pools of data.

One of my long standing pet peeve is the lack of innovation in building User Experience for Financial Services. Our representation paradigm is still too close to the accounting standard when only a very few % of the population receives the necessary education to understand those. Natural language interfaces are a rising tide for Financial services, because of their ability to abstract from pure financial data through a more intuitive user interface. This is true both for input and output. Two startups are a good example of this in my mind: Kensho and Narrative Science.

Facebook’s experimentation with M, is a strong indication of the future interface for financial services. Whether existing banks and financial services providers have the skill-set and ability to build and train these agents is an open question. Especially as AI has a strong lock-in effect.

On the data front, one of the key use of machine learning lies first with the ability to analyze at scale and speed data that would require an army of humans. A good example is satellite data. Companies such asDescartes Labs have the ability to determine crop type in satellite pictures of fields across massive sets of information. The other use case will most likely be understanding correlations in the large pool of digital breadcrumbs that people and companies increasingly create. Trading data, Credit scoring, because of their data rich outputs are prime targets. In credit scoring, one of the key concerns will be to make sure that these self learning algorithms comply with non discrimination laws.

Increasingly, if you are working on innovation in financial services and not actively looking into machine learning, you are probably doing it wrong.

Eris Industries Launches Eris 0.10

Late last year, Anthemis invested in Eris Industries. Eris was known in the Bitcoin world for having worked on DOUG, a smart contract based replacement for the Bitcoin foundation (still an interesting concept in these Bitcoin governance days). We were attracted by the quality of the team and their vision. See my blog post at the time.
Eris Industries was one of the early companies to see the potential value of permissioned ledgers and understood the need to make the creation of distributed application accessible to most developers. But at core, Eris Industries’ passion and our enthusiasm for them come from their focus on smart contracts, pieces of code with unique attributes running on a Blockchain. Distributed applications built with smart contracts have the potential to change how corporations are built, how federated organization are run and to make contracts (such as securities in the financial services world) partly autonomous entities. In my view, its one of the most exciting space in technology, because of its potential impact across multiple industries and its longer term its potential impact in redefining how our society is working (a key aspect in Carlota Perez’s theory of technology and economic cycles). The fact that Bitcoin could go from a white paper shared on an obscure forum to what is it now is just a small example of what distributed applications could do.
However most great things don’t come from planned execution but from the creative destruction of innovation. A key driver of that is entrepreneurial endeavor, especially driven by democratized, open sources tools, one of the key fundamentals of why software is eating the world – to take Marc Andreessen’s words. For distributed applications to succeed, we need thousands and thousands of people building stuff this way and for that to happen we need tools to support them. With the Eris 0.10 platform, Eris Industries is building one of those tools, an application framework to build smart contract based, distributed software on Blockchains, any type of Blockchains. Because blockchain development is important, Eris Industries will also continue its contribution to the open source Tendermint project. Researching consensus solutions is another key driver for the success of blockchains in general (permissioned or permissionless).
Some of the key KPIs in an early technology cycle are simple: lines of code and developers:
– The Eris 0.10 is available on Github, the Eris Industries team has dedicated channels to support developers and documentation/tutorials are regularly updated. Read this series of tutorials on Solidity for example.
– Several developers have started to experiment and build on the Eris platform, from large software companies, to banks, insurers and startups. Everledger is one example.
For the blockchain industry to move beyond the newspapers headlines, consultant reports and generic {we could do that} use cases we need more code to be written. Its “interesting” to write about permissioned car access, automated securities settlement, distributed Uber competitors but its way cooler to start building these. We hope Eris Industries will be one of the key partners to do so.

Musings on Full Stack Financial Services startups

(This post has been long in the making). One of the posts that sparked my interest in the last months is a post by Chris Dixon, Full Stacked Startups. In it, Chris highlights several startups such as Nest, Uber, Tesla, Warby Parker as companies that have gone after the market as full-fledged businesses instead stacking on top or in partnership with other players. Notably the …

[…]  full stack approach lets you bypass industry incumbents, completely control the customer experience, and capture a greater portion of the economic benefits you provide.

Recent transactions, such as Twitter’s acquisition of Gnip, are also showing, in my view, the business tension by any tech startups to move vertically upstream or downstream to find the right mix of economic models.

There have also been more conversation around the idea of changing at its core financial services. Marc Andreessen has jumped in the discussion with the idea of building a full new bank: http://qz.com/175512/to-disrupt-banking-do-you-need-to-own-the-bank/.

Note 1: Bitcoin is a strong factor here, not so much from a direct technology perspective, but more from bringing into the public’s mind that the financial services sector can be disrupted / affected in its own core. 
Note 2: API banking has been at the center of what we are building at Anthemis and a personal passion of mine: http://tekfin.com/2012/04/10/the-core-of-the-machine-banking-as-a-utility/  http://tekfin.com/2010/08/16/banking-as-a-platform-coming-soon-with-banksimple/

Jack Gavigan answered with this excellent post on a blueprint for a new disruptive bank: http://jackgavigan.com/2014/04/14/disruptive-bank/.

nextbank_platform

 

A system blueprint is a great way to start but is one of the views of a fractal of perspectives that needs to be taken when considering financial services (I highlighted in red what I think is one of the key area). Another important one is the financial view. A full stack financial services startups is, in my view, a balance sheet driven startup. Balance sheet driven startups are a bit of an exception in the world of technology startups. In the past years, a lot has been made to make these less and less driven by balance sheet. From renting infrastructure to outsourcing functionalities to other companies, most tech startups have been driven at first with little focus on balance sheet. However in the world of financial services whether banking or insurance, balance sheet driven startups are the default structure for full stack startups.

That makes them more difficult to be considered from a venture capital perspective:

– First, they require capital, much more than a typical tech startup. Oscar’s minimum capital requirement for operating as a health insurer in the state of New York is USD 45M : http://www.dfs.ny.gov/insurance/exam_rpt/x9475o13.pdf , most/all of which will need to be kept aside. That’s a $45M raise just for the right to play. Additional funds will be required for development, marketing, …

– Second, they are very difficult to grow hockey stick. Think of balance sheet driven financial services startups as the weird cousin of multi-sided marketplaces startups. Taking the example of a new bank, for every new customer that will subscribe and deposit, a matching capital will need to be added following Basel III or another local capital requirement rule, invested in secure products. In parallel, you will want to deploy your customers’ deposits in money-making investments with risk profiles compatible with your capital requirements. Either you run your own lending / investment business which adds further complexity or you look for partners to deploy. Low risk with relatively good returns investments are chased by investors and your new bank is a small fish in that pond. All of this contributes to make growth more difficult than in a typical startup.

Even for a simpler version of balance sheet driven startups, say a lender with little/no prudential ratio, every growth in customers will need to be matched with an increase in available capital. Kabbage debt raise is a good example of that: ~$53M raised in equity for ~$345M raised in debt.

So why are full stack financial services startups interesting?:

– From an operational point of view, these activities are enormously inefficient in existing banks. The software they are using (Core Banking Software) is old, batch based and difficult to replace – understandably, once you have built a full balance sheet, something that can affect its management is high risk. Anything build on top of this software base is affected, from your customer front end to your risk management software to your lending activities. This leads to more operational margins being taken to ensure you are operating within regulation. A new player will have tremendous opportunities using the flexibility that current software allows. I am playing our book here (Anthemis) but Fidor Bank‘s ability to connect to P2P lending platforms, virtual currency exchanges or to manage multicurrency /commodity accounts is a good example. This is an incredible opportunity space.

– From a business point of view, once you are past the more difficult early stage balance sheet growth phase, you have built a resilient, flexible company. Flexible is not an adjective often used for banks, but with the right infrastructure and API layers I think modern banks will have the opportunity to open themselves to many business models. Built in-house or in partnership with others. This is also the case in terms of their capability to deploy assets. Financial products, liquidity providers, exchanges are evolving at a rapid pace. New platforms appear to access private companies equities, alternative debts (P2P but also factoring, data driven SMB debt). Non banks are becoming investors as well, investing in their own supply chain to guarantee its performance. And these platforms are becoming more and more digital, creating new opportunities for a bank to connect and invest.

Note 3: This is also where the evolution around contracts in the blockchain such as Ethereum, or distributed open ledger such as Ripple (which recently partnered with Fidor Bank) are really important. Making transactions fully electronic and real-time has massive implications for banks in terms of their investments as well as their risk monitoring.

There are a lot of additional perspectives to consider and I will gladly take additional insights, critics, comments. However if you are working on building a full stack financial services startups, whether in banking or insurance, I am really interested in talking with you. There are very few now but I am betting we will see more and more people try in the coming years.

Smart Commerce will only succeed with Smart Banking

Bill Ready had a great post at PandoDaily on the growing importance of smart mobile driven commerce. In my view this is one side of the equation of the future of commerce, the other side being the creation of smart banking services.

Using Bill’s example: I book a flight to San Francisco, my financial service app warns me that my travel budget will most likely be exceeded this month and has pushed back the budget allocation for new electronics by 1 month. My extra rental revenue from Airbnb should help cover cash flow needs for the month so that new MacBook is still a go. I take an Uber upon landing and check a restaurant for the group. Bills is split between us automatically, referring back to our positions in a global distributed ledger including interests owed (built on the Bitcoin protocol foundation). After the lunch, I check recent communications from my Angellist portfolio. My portfolio allocation to startups is split across various syndicates. Through tasks performed to help these startups, I have also earned additional exposure to a few. A good way to not only increase my upside potential but build my skills and experience.

Is this future far away? With the increase in sensors in mobile, shops, objects and the digitisation of money, the capabilities of financial services are changing quickly. A lot of this effort in calculation is currently focused on market activities (high frequency trading being probably the most discussed) but I am convinced we will see the same push start in consumer finance. The current push to integrate more data sources, including social data sources, in online lending is a good example.

Mobile is becoming an integral part of people’s financial life. Starbucks success with its mobile app proves that people, when given a good use case for mobile (increase in convenience and additional services) are more than ready to use their mobile. Payments on mobile are increasing at an amazing pace: Paypal’s total payment volume increased to $27 billions in 2013. But the increase in payments on mobile also highlights the gap between how easy it has become to spend online and how little has been done in helping people manage their spending.

Cash was the base budget management tool for a lot of people. A wallet is probably one of the best UX for money. Visually checking how much is left in a wallet is one of the most used and simple budget management tool. The rise of prepaid card with underbanked and neobanked is in some ways following the same trend, as closed cards, especially with easy to access mobile balance reminders, are the modern equivalent of counting the number of $10 left.

However, as highlighted above, as more and more of our purchase experience will not only shift to mobile online or offline but also to 1-click / no click payment, having a single credit card or debit card as a default payment mean can potentially increase the tensions in budget management and understanding of personal finance. In a world where payment is bound to disappear, the pressure for financial understanding will increase further.This is a vast opportunity for financial services startups.

The same technology that let applications recommend you what to purchase, how good a restaurant is or how to manage a fleet of cars / pricing to match demand can be used to optimize your personal financial management. As the age of mobile concierge is coming, the age of mobile financial advisors is coming as well. I am biased, professionally and personally on Simple but they are, in my view, a good example. Smart balance and goals are the beginning of a payment experience based around managing and optimising personal finance. And while this effort begins with spending, it will soon integrate as well with saving and borrowing. Paying overdraft fees with a saving account or other type of liquid assets is an incoherence in a time where a simple excel spreadsheet can compare borrowing and savings rates.

If we push this idea a little further, there is a potential for algorithmic finance becoming even more intelligent. We are on the verge of being able to record how people feel at any point in time. What about a financial algorithm that would help people maximize their happiness over time? What about a mobile agent that prevents you from buying stuff at checkout by automatically reminding of the other activities you would like to do that will be more rewarding?

Welcome to Banks’ new competitors

Founded in 2006 by a single repeat entrepreneur. IPOed the next year. Raised a total of USD 232 M, including a last round in 2013 of USD 150M that puts it firmly in the dollar billion valuation club (aka the unicorn club) …… If you have not figured out which company it is, I will just add 3 words:

– All Blacks

– Sailing

– The Lord of the Rings

If you still have no figured out which company it is or why I am starting to speak about Xero on a financial services / banking blog, then you are pretty much in the same position as most Banks.

Xero

 

The same can be said of Amazon. Founded in 1995 by a former Wall Street Hedge Funder, starting as on online bookstore and now the biggest ecommerce seller and platform online.

How Big is Amazon

 

What do these two companies have in common? They have both started to distribute financial services products via their platform, whether it being working capital loans on stocks or data on small and medium business financial performance.

One way to look at Banking is that it is a data arbitrage business, whether by exclusivity on data itself or control over the aggregated value of data. That data to simplify enormously is used to arbitrage interest rates between deposit and credit. As software is hacking the world, the ownership of financial data is moving from the existing financial players to the new global platforms. 

Interestingly, businesses are more and more leveraging several of these platforms at the same time. For example, online retailers may use both Amazon and Ebay to distribute their products or local competitors. Several online accounting platforms are competing for medium and small businesses, with the aggregated accounting data across companies distributed across them.

Therefore  lot of the early competitive pressure we are seeing is whether each of these platforms have a critical size (and the business appetite) to be an exclusive channel for financial services? Or whether innovative cross platforms financial services providers, such as Kabbage or Fundbox will prove that most of the value lies in cross platform companies? One thing is sure, the market for non-bank financial data, whether productized internally or distributed via APIs, will boom on the next years. 

 

Braintree takes Venmo Touch international with AMEX > what it means for credit

With Simple and now AMEX on board as preferred marketing partners in the US and UK, expect Braintree to follow a similar playbook in future markets. There also could be some significant competition in these initial two markets from other payment card companies (Visa, Mastercard, Discover, etc.) seeking to get their cards installed into Venmo’s valuable default payment card real estate.

via One small step for Braintree, one large step for mobile payments: Braintree takes Venmo Touch international | PandoDaily.

Spot on, multiple card selection on mobile within 1 click payments / facilitated payment (à la Uber) is not a UX problem to solve. It just won’t happen. 

This is why it is so smart from Simple (disclosure, Anthemis is an investor) and AMEX to partner with Braintree to become the default card. But the implications are much more important. With a single default card, the position of credit card is put at risk. Credit Cards are tools made for a card selection environment, with people doing arbitrage while looking at their wallet between debit, credit, credit limits and points. This behaviour is not possible in a 1 click environment, even less in a seamless environment.

Additionally, studies show Gen Y is moving away from credit cards (I am definitely part of that population). According to a recent FICO study (http://www.learnvest.com/2013/06/gen-y-shuns-credit-cards/). “16% of people aged 18 to 29 had no credit cards in 2012, up from 9% in 2005. As a result of lower credit usage, Generation Y’s average outstanding credit card debt was $2,087 last year, down 32% from a $3,073 average for young people in 2007.” According to Frederic Huynh at FICO “it stands to reason that the Great Recession has influenced, to a certain degree, consumer credit behavior as well.”

There is also pressure to move some or all of payments off the card network to direct debit. A potential in Paypal’s acquisition of Braintree is for Paypal to export its arbitrage business model to Braintree. And Dwolla’s effort in building an alternate network is the ultimate push in the direction of direct debit.

This is a great opportunity for credit innovation. What we are seeing in B2B online lending with Kabbage, Paypal and Amazon will very soon spill over in the consumer world. B2B is the low hanging fruit as the market places (Ebay, Amazon), einvoicing networks (Tradeshift), online accounting tools (Xero) act as booth data providers to support credit risk scoring and aggregators to improve cost of origination. But Consumer Credit will be next and the mobile payment providers have an amazing opportunity to act as the future credit platforms. New Banks, such as Simple will also be the winners of this world. The card is only a tool for payment, credit is better managed within the budgeting, goal setting, savings experience of a smart bank.

Credit Card is an obfuscation, the credit and the payment mean are two disconnected products, the digital unbundling machine will soon make it a reality.

Tradeshift’s Christian Lanng on Capital8, Financing, and Supplier Adoption « Spend Matters

Finally, the last parameter is funding cost, for supplier financing. This is important because Tradeshift already has very advanced semantic analyses of content (this is in fact how we build CloudScan). We use this to give CapitalAid and other financing parties access to a proprietary risk model to gauge factoring risk.

A lot of material is being considered in this model, including all past transactional data. This has led us to a model that has low funding costs with much less than 3% bad debt. We can guarantee a pretty high yield, and the rates of return are in the range where hedge funds are buying into the model.

via Tradeshift’s Christian Lanng on Capital8, Financing, and Supplier Adoption « Spend Matters.

< Interesting that it is Tradeshift that appears to supply the risk model instead of each financing company? Or is it because of Capital8 close relationship with Tradeshift? In a sense it could be a model that is close to what Lending Club has put in place with its investment arm. Standard grading and the level of exposure is defined by the portfolio strategy.