Digital Finance meets Disruptive Technologies
Aus dem CAS Digital Finance bloggt – auf Englisch! – Robert Niggli:
After lunch, we made an exciting journey, diving into “disruptive technologies”, guided by our tutors Michele Forte and René Konrad (both t&im). We started with some clarifications on the terminology.
What’s the meaning of disruptive? (German: unterbrechen, zerreissen)
- Innovation with the potential to replace an existing technology, service, or product or to move it out of the market. Characteristics of a „disruptive“ innovation:
- It satisfies an existing need much cheaper, easier, better or faster
- It creates a new need or market (like the iPhone, iTunes)
- It enters the low end of the market and works itself through continual improvement into higher end of the market (this refers to the classic definition by Harvard Professor Clayton Christensen)
- LCD Displays: entered market first as small displays in watches/instruments, technology evolved into the high end over a series of technology leaps
- Flash memory storage: from USB sticks to the standard memory option in all mobile devices and laptops
- Digital photography: from a low-resolution toy to becoming the standard
- Mobile phones: putting fixed line telephony into a niche
- WhatsApp/Twitter/Snapchat: making SMS redundant
We agree, “disruptive“ is today exponentially used like the term „New Economy“ back in 1999-2001 before the dot-com bubble burst. By the way, who remembers the following companies from the dotcom area? Companies like Gartner fuel the market with their yearly technology surveys (Technology Hype Cycle 2015). For FinTech, in 2015 they declared the hot topic was „Blockchain“, in 2016 it’s „Smart Contracts“.
The hype cycle (see picture on top) follows the sequence of
- Innovation trigger
- Peak of inflated expectations
- Trough of disillusionment
- Slope of enlightment
- Plateau of productivity
It is clear that not all technologies will reach the “plateau of productivity”.
Interestingly: The speed of doubling the integrated circuits per chip every two years (transistors per chip = computer power) has slowed down for the first time (Moores “law”). The limits are reached in the atomic scale related to quantum physics. The further increase will require a new technology leap, today we talk about graphene based chips, optical, quantum or DNA computing.
Artificial Intelligence (AI)
AI (German: künstliche Intelligenz) deals with the automation of „intelligent“ behaviour using computers and software. To jugde this, the Turing-test is an accepted procedure to determine if a machine or application of AI can be regarded as „intelligent“. This is the case once the test person is not able to distinguish the response from a machine from that of a human during the test. Interestingly, one application of Turing-test scenarios are spam-filters: These try to distinguish private mails from mass mailings and advertising.
The context of the use of AI is manifold, evolving and would need a deep dive on its own. We discussed in our lecture the following five applications of AI:
- AI in Data Analytics: Various AI approaches are used to find patterns and connections in data
- Deep Learning (a sub-topic in the field of neuronal networks): Deep learning requires large data sets to train the algorithm, e.g. which patterns it should recognize (e.g. image processing based on satellite data, MRI data for clinical diagnostics …)
- Natural Language Processing (NLP): Is used in the human/machine interface, also uses deep learning approaches to understand speech or written context and respond to it (e.g. Ciril, Chatbots … )
- Big Data: Technology approach enabling the processing of a large scale of fragmented data sources through massive parallelisation
All big tech companies are involved in AI, and many of the previously in-house used technologies like machine learning are now available as services in the cloud:
- Amazon (Amazon Machine Learning)
- Microsoft (Microsoft Azure)
- Google (Google Cloud ML Alpha)
- Facebook (Facebook AI Research (FAIR) using Torch)
One main driver is profiling: Accurate profiling enables targeted and custom advertising, services, consumer choices and the right user experience. Hypothesis: Banks have great potential to improve in this area.
Mobile internet subscriptions will reach 6.4 billion by 2019, with mobile phones being the dominant part. Whereas some companies design the desktop layout of their user interface first, the majority will use mobile for internet usage: This changes priorities (mobile first). Desktop is only relevant in Europe and USA. Especially in countries with no “wiring” to every house, the users enter the mobile space directly, leaving out the desktop (e.g. Thailand, China).
But keep in mind: Screen size and behaviour is linked. Desktop leaves more room for providing more factual information and for the user to analyse. Mobile screens are too small to provide large context, and mobile users have generally less time to analyse context. The characteristics of every digital channel (mobile, tablet, desktop …) and the user behaviour in that channel are different, and have to be considered in the usability and communication strategy.
Internet of Things (IoT)
The number of devices connected to the internet of things will increase massively (e.g. from 8.5 billion in 2012 to 50 billion by 2020). The increase in the number and the variety of sensors and connected devices will drive innovation. Example companies : GE, ABB, Bosch.
Key is the data gathering by sensors throughout the value chain, central data collection and management, and turning this data into meaningful output for actions (preventative maintenance schedules, optimizing running level for power plants or manufacturing lines …). One big problem lays in the area of missing API standardisation, there is a lack of industry protocol standards for information exchange. Use cases:
- Houses and buildings with more connected censors (Smart Home, intelligent buildings)
- Remotely handled oil drilling platforms
- Manufacturing lines: optimal maintenance times, prediction of failures (minimizing downtime)
- Traffic/vehicles: cooperative driving, collision avoidance systems, speed management, traffic management (efforts for API standardisation are under way)
- Insurance industry can optimize their risks with more behavioural data from the consumer, these can be acquired through incentive schemes (e.g. monitoring of driving behaviour, exercise tracker via sensors in smart watch …)
- Monitoring of the state of your collateral (Real Estate …)
Blockchain / Distributed Ledger
The disruption by using distributed ledger systems is seen in several areas: It enables peer to peer transactions to everyone, circumventing the need to route it through gatekeepers and intermediaries (like banks, brookers with access to the market, clearing and settlement systems …).
Blockchain is a form of a distributed ledger, where an agreement of validity of a transaction is achieved by untrusted consensus from a network of distributed single ledgers. Example: In case of non-agreement of a Bitcoin transaction, Bitcoin resolves the disagreement by electing one peer as leader, which imposes its changes to other peers by sending a block containing all transactions it accepted since the last block. With this scheme, e.g. double spending of the same transaction is avoided.
Status: In Europe, a consortium of 41 banks is exploring the technology under the lead of the blockchain company R3 CEV. They have recently concluded a trial for solutions for securities trading using distributed ledger solutions. This poses a direct threat to SEPA, SWIFT, SIX, transfer banks, market intermediaries, and to the fee income associated to these transactions.
Where is the greatest impact seen ?
What are smart contracts? Smart contracts are cryptographically captured obligations, represented in a digital distributed system, in this case “asset registry” systems. Todays cryptographic possibilities have enabled the controlled execution and compliance with formulated covenants of a contract. The principles are similar to the distributed ledger network like blockchain. Real applications of smart contracts today are rare. The following list shows potential application areas:
- Structured products, where contract could be executed in a blockchain environment
- OTCC market: peer to peer clearing, private, without clearing house
- Energy utility sector voting systems
- Smart property
Other applications are under discussion and will evolve. Smart contracts are a piece of code, a contract with scriptable clauses in the distributed ledger system. It is not yet clear how such contracts are legally treated and, in the worst case, enforced. These are all discussions which are still in progress. Please check out some demystifying questions about smart contracts.
At the end, as a company you find yourself in a dilemma: Will blockchain (distributed ledger systems) and smart contracts evolve and become the new standard (e.g. like a TCP-IP protocol), so we will assimilate it anyway, or should we be spending millions in the research already today in hope to gain a competitive advantage from the start?
Due to the speed of new developments, the risk for entrepreneurs is that time-lagged regulation might impact their business models afterwards. Politics do not keep up with the pace rate at which new technologies and solutions evolve, so it is the responsibility of the entrepreneur to assess the current situation and monitor the upcoming regulatory developments.
Noteworthy regulatory developments affecting Switzerland are:
- PSD2 (Payment Service Directive 2): The EU rules are expected to apply to Switzerland in 2017/2018. The most discussed feature is the new access to account (XS2A) clause, which requires the payment service provider (e.g. a bank) to provide, on request of the client, unrestricted account access to a third party provider (TPP)
- FINMA rules which facilitate online and video authentication
- Banking licence light for FinTech startups (regulatory sandbox)
- Tax easing for start up companies
FINMA has created a new information site adressing the FinTech community to bundle topics affecting this area. Other areas which will gain momentum:
- Transparency of the market vs. anonymity of transactions in a blockchain ledger
- Role of Swiss banking secrecy in a FinTech and open access market
- Peer to peer lending
Startups and the Situation in Switzerland
We discussed the use case of Ripple, which provides international settlement of payments between banks. The benefits are faster transactions (settlement in seconds compared to days before), reduced number of involved counterparties, and much lower costs for the client (but as well lower fees for the intermediaries). The business model leverages on a standardized transaction protocol, and the service gains attractiveness with every new onboarded bank.
If startups make their business case based on inefficiencies in the existing market, then the financial industry must have a lot of inefficiencies: Venturescanner lists over 1’000 FinTech companies in the area of traditional banking services. And insuretech is gaining momentum.
But how is the situation in Switzerland ?
Ca. 25 of 100 startups are FinTech (2015). As an investor, the situation remains unsatisfactory. On the good side, the overall support and formation of associations and interest groups is increasing: e.g. Digital Zurich, Swiss FinteCH, FinTech Fusion, the presence of Universities and ETH and EPFL which are close to the financial centers and others. Compared to other FinTech-Hubs, the political awareness is regarded as underdeveloped, even though the FinTech area offers a lot of opportunity in the traditional banking areas.
Are FinTechs disrupting Switzerland ?
Fact is: The level of confidence is still very high with traditional Swiss banks. 86 % of clients are satisfied with their current service. TrueWealth as an example: In 2015, it raised 15 millions through crowdfunding, the business ambition is to gather 1 billion assets by 2017. This is not disruption, it is a small evolution.