Embracing the algorithmic world around us – With Caution

Humans have worked relentlessly hard in pursuit of a custom-made world that can function as per their needs. Technology is at the cornerstone of all the advances humans have made in this quest. High-end technologies like Artificial Intelligence, Internet of Things, Big Data, Deep Learning etc have become colloquially used words today, thanks to the value they are adding in the life of a common person. For example, Amazon’s Alexa is an artificially intelligent personal assistant that can tell you the latest news or read out a recipe or set reminders for you. It feeds off machine learning algorithms that make it smart enough to have meaningful conversations with humans. Alexa and many such smart machines prove that algorithms are the heart and soul of the brave new world around us. Netflix series “Westworld” and “Black Mirror” , among other shows demonstrate where imagination coupled with technology could take us and hot scary it might become as we become more connected virtually, and less physically with our surroundings.

The Expense of Algorithms – Our Lives are Spun Around Them

An algorithm is a process or a set of rules, to be followed in problem-solving operations by a computing machine. It can perform calculations, data processing and automated reasoning tasks. While staying out of sight, algorithms keep the modern machines running. They make them think and act, as codified by humans. For example, Alexa is a voice-based AI device that is more than just attitude. A small juke box looking device hides a plethora of continuously evolving knowledge aided by machine learning algorithms. Amazon’s AI experts are using big amounts of data to make it understand human language and also be a better speaker. Alexa’s machine learning algorithms learn from the speech modulations of the professional narrators. They fine tune the machine’s synthetic female voice to make it as good as a real human voice. This is one of the many examples of algorithms being used in the world all around us.

Algorithms hide the complexity of machines. They stow away the intellectual property and codes away from human eyes and only let us use the final product. Facebook is a popular case in point. The algorithms on the Facebook website decide what users see in their timeline. It is interesting how users do not get to see all of their friends’ posts. It is the algorithms at work deciding what to show them. Similarly, Google search results are unique for everyone. The recommendations on Netflix or Prime video are also tailored as per unique viewing history. As a user, one does not even need to register these things, but they still are critical to how we live today. This curated world is being driven by algorithms.

The Good and Bad of This World

Google knows when a user has been looking for a particular book and when they turn to the search engine, they display the relevant ads to them. Similarly, Facebook displays content on a user’s timeline as per his interests. This tailoring of results brings a better user experience on the table. It makes internet more useful and contextual to the user. The more the user spends time on the internet, the better his preferences are getting captured every send. The crowdsourced data from different users, applied with appropriate machine learning algorithms, help every benefit by easier google searches (who goes beyond page 1 any more), relevant news on facebook and twitter, amazon suggesting what we may buy, Expedia offering relevant trips, and youtube showing the videos we are likely to watch. US and European capital markets witness over 90% of trading done by algos, on behalf of humans, leading to efficiency.

There are always 2 sides of the coin. While there are a plenty of good use cases of this world, there is a downside too. Is it not too good to be true for facebook to show us the “relevant” news in our feed. Or is it prone to manipulating human minds. Recently held US elections put the bad part in the limelight by showing how Facebook can manipulate user’s exposure to content and try to influence their decision making. Machine learning arguably evolves in any direction without any oversight from humans, and can be very risky and not many understand how it may evolve. Any fake news over the internet can control the minds exposed to it and spur them into a negative action.

And what happens when financial markets are driven by algorithms, and something fails, and all algorithms react in a way that the crash becomes bigger and bigger. This happened during flash crash of 2011.

It is said that there is much bias in algorithms as well. Carnegie Mellon University’s research showed that Google’s online advertisement system was said to display higher paying jobs to more males than females. It is feared that data-based decision making will negatively affect human individuality and identity as well.

Can we trust humans to apply some filtering, their own analysis, on top of the algos curated content that they see? OR would they blindly believe in algos ability to make the right decision for them, without even realizing how that may impact them.

Evolution and The Way Forward

Algorithms define the machines and their processes today. They help devices get smarter and connect to both humans as well as each other effectively. As smart machines will process more big data every second the world over, these algorithms will help machines and artificial intelligence processes become even smarter. There are hypothetical theories in technology like the Singularity theory which states that in the future, intelligence of machines will surpass the intelligence of humans. There cognitive capacity will out run that of the humans. It will be triggered by the runaway technological growth spurring unthinkable changes to human civilization.

However, this is just a theory. Since times immemorial, humans have observed both the positive and negative impacts of any technology that has been embraced by them and the algorithm-based smart technology of today bears no exception to this rule. Maybe regulating the use of algorithms, ensuring users understand what it means and how to take out the biases. There is no denying that algorithms are here to stay and will continue to affect our lives in more ways than we know of today. Saying this, it is also relevant to be prudent while using the technology that simplifies our life.

A next big opportunity for entrepreneurs is to bring the prudence into algorithms to manage their wider adoption diligently. Building police bots for regulation, risk management, compliances, kill switch etc., in the future world driven by machines, shall be critical too.

Busting top 7 myths of Blockchain

Blockchain has become the most “glorified” database of the 21^st century. While there is some proven value of Blockchain (example Bitcoin application) and great future potential; its existing value is being taken out of proportion by the Blockchain “enthusiasts”, “experts” and “advisors”, who do a “lot of conferences”.

Lets look at some of the Blockchain hypes, myths and examine them by “first principle” analysis why they are unlikely to work.

1. Anything on Blockchain means truth

Blockchain achieves truth (transactions validity) in Bitcoin transactions with the help of 1000s of mining nodes, with up to 1000s of servers supporting each node. Miners are incentivized by Bitcoins to perform the mining exercise.

For any Blockchain use case, worth asking the following,

• Are all transactions going to be public? (example, Banks Blockchain use cases won’t like this)
• You will allow decentralization in the workflow? i.e. Some consensus algorithm (say majority of miner votes), will make final decision in transactions validity (example, Governments or any Central authorities won’t like this)
• Miners understand and can help in validating transactions (example in supply chain finance, how can miners validate any facts about fake invoices, workflows or multiple discount loans?)

“Blockchain is not the magic wand that generates immutable truth”. Its just a means to an end.

2. Private or Enterprise Blockchain makes sense

Among many stakeholders, IBM has been promoting and selling enterprise blockchain use cases. Several Banks have formed use cases for Blockchain consortiums, to improve inter bank data management, operation process improvements.

For any enterprise to use Blockchain internally, or with other enterprises, we have had proven, scalable and efficient technologies for decades, including shared databases (SQL, Oracle, SAP).

Please note that Blockchain is an expensive technology (in terms of resources, as multiple nodes need to be setup, multiple entities need to validate all transactions since the beginning (no concept of end of day final records, so next day’s transactions begin from next day only and not from the first ever transaction like in Blockchain). Decentralization by definition slows down things and introduces security concerns. For a single enterprise with already trusted internal groups, or a hand full of external enterprises, private blockchain is not the best solution. And for Inter-Banks’ transactions, payments, trade settlements; several existing efficient workflows, platforms, protocols like Swift, FIX etc. exist and work fine. Blockchain as a magic wand will not help at all.

– As an example, if some bank says that they can reduce the current trade settlement cycle from 2 days to 1 second, they should ask themselves that why is it 2 days in the first place. It’s not that they were all waiting for the Blockchain technology to be invented for solving this problem.

If you do not believe this, ask any bank who announced any use case, proof of concept (years ago), if they actually scaled on it or not? Do you recall reading any such “scaled up transactions by banks” PR (press release) after the POC (proof of concept) announcement? I guess not. And I rest my case.

3. Transparency / Digitization is best achieved by Blockchain

Several POC’ed or Advocated use cases are nothing but digitization of some data. Examples below,

• Land registry on Blockchain is done by several states, governments etc. Whereas none of them has allowed for independent miner validations (i.e. only states can decide who owns which property and when they approve the transfers), so it just makes online record keeping instead of offline. Such processes by Governments and Companies have been digitized for decades, without blockchain, and more efficiently.
• Invoice discounting If a supplier sends an invoice to a client, that gets accepted, then they can take say 70% loan against that invoice from a bank to develop the products, or get some cash-flow while the products are shipped (say from China to US) which takes weeks. Typically in such loan, supplier, supplier’s bank, client and client’s bank are involved in confirming the transaction and supplier’s bank gives loan after that. This approval process can take days.

The challenge in this loan issuance is that there can be fake invoices, or companies can take bank loan from multiple banks against the same invoice, or that the counterpart client does not exist. Hence banks are cautious in giving such loans.

Now independent miners are bank’s risk departments who can help validate the authenticity of the transactions. And no bank will open their client data to other banks for validation (in fear of losing the client to other bank).

So all that can be achieved is that document transfer between the 4 entities, can be made faster from days to hours, by using digital signing and digital documents transfer. Such solutions exist, and work well without Blockchain.

Most problems advocated here are typical digitization problems that can achieve transparency by making the steps in the workflow public. Think about DHL or any other courier service tracking the package in transit, or Uber’s car driver and trip details. People, who need to know, can know it online very easily, and without needing any Blockchain technology implementation.

4. Blockchain is an efficient way of managing data and transactions

Bitcoin gets around 300,000 transactions per day. To achieve these transactions, there are millions of servers running around the world, consuming more electricity than several countries, to validate these transactions. It takes seconds if not minutes to confirm the transactions. This is of course slow and inefficient.

Blockchain and decentralization by definition slow things, make them inefficient. They should be used only if the “real” benefits of decentralization are likely to be achieved. In most Blockchain use cases, centralized non Blockchain implementation will work best.

5. All crypto-currencies make sense

Bitcoin is owned and run by the community. There is no “known” majority Bitcoin holder. It makes sense for community to adopt it as decentralized currency, as an alternate to fiat currencies.

But different countries like Venezuela launching their crypto currency, or centralized crypto currencies like Ripple make no sense. All they need maybe is a digital currency instead of fiat currency, or digital payment mechanism like Paypal, which has worked well for years.

But if the transaction validation is done by Government or 1 company like Ripple, its not decentralized and is not a crypto currency.

6. Smart contracts are perfect and legally enforceable

Smart contracts on Blockchain maybe useful for achieving conditions based trigger conditions in the Blockchain. Some known challenges with smart contracts must be understood,

• It’s debatable whether such contracts should be coded within the Blockchain, or at the application layer on top of the Blockchain.
• Many smart contracts are poorly coded and hence end up hurting the use cases rather than helping them.
• Very few proven scaled up use cases of smart contracts exist till date. The most successful ones so far are ICO tokens and Crypto-kitties.
• If smart contracts go wrong, good luck. And also, they are not legally enforceable in any jurisdiction so far.

The above challenges maybe overcome over time, as the Etherium alike protocols mature improve and become scalable, making them more accepted widely. And smart contracts shall be more relevant as automation & Iots grow, leading to more “machine interacting with machine” scenarios.

7. All new discounted ICO tokens will rise in value

ICO (Initial coin offering) is the advance sale of a project’s crypto-currencies or tokens, to be used within their platforms or outside, in advance, to fund the development of their platforms. These tokens can be easily sold and traded at anytime, on all crypto-currency exchanges depending on their demand. So, an ICO is when a company raises money in Bitcoin or other crypto-currencies for the technical development of their projects. These tokens are essentially the incentives, for several market participants to use and grow the platform in a decentralized manner. Such incentives are paramount in making a decentralized eco-system operate sustainably. This allows for formation of new economic systems; possibly even capable of transforming / improving wider systems like “capitalism”.

Majority ICOs unfortunately don’t and won’t work because,

– Many ideas don’t need blockchain or ICOs at all. They are centralized, or don’t need community, or have no real business models. As explained above, you cannot “decentralize everything”.

– Team behind ICOs have nothing to lose; Once they raise millions to billions on an idea, basic prototype at best, proven or unproven team, “self-proclaimed experts” or “larger than life celebrities lending their names”, there is no major pressure to execute their idea well. There is no governance on the use of funds. Compare this to typical startups, where entrepreneurs get modest valuation, and have to be “all in” the project to see any returns after 5-10 years or hard work. They are “on the hook”, forever.

And, lack of regulation, accountability and legality of structures; makes it the “wild west“ of finance and reduces the probability of success.

So where does Blockchain work then? Lets take the most successful Blockchain app. How Bitcoin Blockchain REALLY worked, see below the key factors in its success.

1. Shared Beneficial Ownership– Bitcoin Blockchain is owned by the public, striving to build a decentralized financial system where trust is built in a democratic fashion. While Bitcoin was initiated by smart hackers, it was scaled by the tech community and eventually became mainstream, because it seemed to help the whole world and everyone had an equal opportunity to build their businesses around it (like all miners, bitcoin exchanges, etc.). Wikipedia, Linux and several other open and crowd-sourced platforms have scaled in a similar fashion.
2. Independent mining achieves trust– Blockchain does not create trust inherently. Several nodes independently validating all transactions, for some incentive, build trust in a Blockchain. If the Blockchain app did not have enough independent miners, Bitcoin transactions would not have been so trustworthy.
3. True decentralization – Bitcoin is truly decentralized. There is no central point of failure. There is no dependence on 1 person, 1 node, 1 company, 1 Ceo, 1 nation, or 1 leader.
4. No contract among transacting parties– While doing a Bitcoin transfer, payee and payer do not need to know each other. They do not need to have a legal agreement between them defining terms and conditions of their transaction.
5. Some other factors –There are other factors that need consideration, including, multiple parties executing the transactions, transparency of transactions acceptable to all parties, public or private blockchain, etc.

If you can see several / all of the above factors in your Blockchain use case, maybe it can work.

 

In my humble opinion, majority use cases being developed on Blockchain do not qualify for, or need Blockchain at all.

 

PS: I truly believe in Bitcoin and the potential of Blockchain and potential of ICOs to disrupt the corporate structures. I am not even too bothered by inefficient or un-scalable Blockchain protocols yet; those would become better in time, for sure. My concern is that Blockchain is taken out of proportion to solve problems that it’s not meant to and never will.

Quick decision making flowchart.

Branding your Start-up

So you’ve jumped across all the hurdles in your way, put in sweat and blood, and invested countless hours to kick-off your start-up. Now how do you get noticed? For startups, it’s important to “go live” and start generating revenue before running out of initial capital. This definitely drives a sense of urgency to succeed quickly. Many entrepreneurs spend most of their time on the concept and its execution, but should also translate the startup’s big idea into a brand idea. Especially as they cross 3-5 years mark and become a real business. Your brand is your identity; it defines how your customers perceive you. It is the experience you deliver to your customers.

Think about the logos of some of the world’s most admired brands (Apple, Google, Amazon). How do you feel (emotionally) when you see their logos? A brand is so much more than a logo and a slogan, it is reputation and feeling.

We partnered with design start-up Yellow Cursor to help us do rebranding. It did not just mean changing the logo and a color change; the focus was also on the color palette, typography, imagery, tagline, perception, etc. We had to do lot of introspection to really answer some tough questions, which we do not think about it in our day to day lives. They summarized their research to come up with a One-line Pitch for the Brand: ‘Enabling Smarter Transactions’

We then hit the drawing board and came up with a number of possible designs, beginning with dozens of doodles:

We finally settled on the insignia shown below. Because, we are providing the explanation after this paragraph in depth

THE SHAPE

The insignia is built from two arrowheads – an upward and another downward. These arrows represent Trading (Up / Down) and Transactions (Inward / Outward). These arrows also form the half-letters U and T.

Enclosed in these arrowheads, at the core, is a motherboard which symbolizes technology. Two horizontal lines connect the motherboard to the outside world, showing the Open Source nature of the brand. The line also represents currency, as it is used in the symbols of dollars, pounds, euros, rupees, yen, yuan, bitcoins, etc.

The square creates a sense of equality and conformity. It is seen as stable and trusting. It also relates to the earth, with each of the four corners relating to the four points on a compass.

THE COLOURS

Blue stands for positivity, innovation, sophistication, energetic, credibility, integrity, reliability, loyalty and professionalism.

The upper half uses a tint of Blue and the lower half uses a shade of Blue.

THE FONT

The word ‘uTrade’ is written with Esphimere, a freely available, futuristic font.

Is this logo too simple?

Yes it is better. Like any ideas, its better to simplify logos too.

We tested our logo against, Timelessness, Portability, Scalability, Colour and Shape Psychology, Culture-Connect, Applicability and Simplicity.

This logo has passed all of these tests. The entire science converges to this logo. Furthermore, the concept helps us derive the branding for products and child / sister companies in the future.

Is it really necessary for a Growing Company to invest in Branding?

Simple answer maybe,

• No, if you are in early stage and still trying to discover your real sustainable business model.
• Yes, if you have come on the other side of tough startups ecosystem and are profitable now, and planning to scale.

Let’s face it. You’re in a crowded ecosystem. You need a more distinctive position than: “We’re disrupting X” or “We’re like, you know, the Uber of Y.” A meaningful brand strategy can help everyone understand why YOU matter and why they should care. Differentiation is key to survival as a startup, so you need to get this right.

In today’s hyper-competitive world, emotion is the secret of many of the world’s most successful brands. It can help your brand become more appealing, differentiated, and loved once you recognize its power. Developing a brand strategy that mixes both the rational and the emotional aspects of your brand will win both the hearts and minds of those people you are trying to capture.

Extraordinary things happen when an organization embraces a purposeful brand strategy. It drives new attitudes, beliefs, and behaviors across your team. It evolves your marketing, communications, and advertising. It inspires new product development. It makes your workplace more productive. It changes how people experience your brand. When companies adapt their behavior to live up to the brand promise, people notice. And that’s exactly the result you want.

Don’t wait too long to discover the true, authentic purpose of your brand, and define why it matters to the people that matter to your business. It should be well worth your time.

Blockchain is likely to be the most hyped database in the current decade. It is the core technology that enabled bitcoin, which has had challenging ups and downs due to business reasons.

Blockchain has the ability to bring lot of efficiencies in how information is created, shared, accessed, secured, and relies upon crowd efforts for validating the correct information. Hence it is compared with the Internet revolution that started in late 90s. Similar to the early days of Internet revolution, Blockchain has lot of useless attempts of me-too players solving a meaningless problem. The important thing to understand is what is the core problem and can block-chain really help at all. 80% efforts should be allocated to validating the business scenario and the solution, blockchain implementation is likely to be the easier technology part (which has possibly scale challenges due to early stage nature of existing blockchain technologies available in the market, discussed further below).

What is Blockchain

Blockchain is basically a database. Blockchain is a distributed database. Blockchain is (arguably) a crowd-managed distributed database. Blockchain is a crowd-managed distributed secure database.

In a blockchain, copies of a ledger are “distributed” and validated by a consensus process, with multiple users independently verifying ledger changes.

Blockchain was invented as the technology to manage and run Bitcoin. As you may know, Bitcoin was the first digital currency, which was not governed by any Central Bank, no notes were printed, and it was largely up to the worldwide users to ensure it works, without any central authority. In Bitcoin, the most well-known blockchain application, tokenized transfers are made directly between payer and payee, effectively eliminating the credit and liquidity risk inherent in the fiat system.

Potential Applications of Blockchain

Decentralize everything is the core theme of blockchain. Since it successfully supported Bitcoin for years, its ready for experimentation in various other industries and use cases, including:

Digitization of Assets

A company or an authority or similar entity typically centralizes all the asset registrations. What if they could move to blockchain to make it more transparent, efficient, faster and secure. This could apply to - Housing / Cars registration - Luxury assets like paintings, diamonds with digital certificates

Digital Legal system

Various legal documents are signed by different parties and stored in physical / scanned format and possibly registered with any central authorities. It could securely move to digital format with standardized / custom contract and digital secure signatures embedded within the blockchain itself. Eventually this maybe used to introduce Smart Contracts which can be actionable events based on conditions like time, etc. For example, the Will of a person maybe to enable inheritance of his wealth upon a certain time or event like his death.

Crowd sourced platforms

Many platforms like Wikipedia could benefit dramatically from Blockchain’s information creation and management system.

Identity

Digital id creation for every individual, combining the decentralized blockchain identity verification can create the largest unique digital ids in the world, which various Governments in different countries have failed to do yet.

Financial Markets

Financial institutions including banks have inefficient processes, and legacy infrastructures to manage transactions. Various post trade transactions for payments (example in banks), clearing / settlement in securities world (stock markets), could be dramatically improved by using Blockchain.

But due to high overheads and delays in validating various blocks before the transaction, blockchain may not be useful in live trading processes as yet. This may change, as blockchain technology gets more mature, scalable and faster in the next few years.

Insurance

Blockchain transforms the way people manage identities and personal information, leading that could change the way insurers review risk and price the premiums. Blockchain applications in insurance are likely to start with digital identity systems and management of personal data. Smart contracts (where the outcome is pre-determined or even based) can help end users self administer their insurance policies.

Technical aspects of Blockchain

Blockchain is a sequential transaction database, which contains a continuously growing list of all the transactions, which have occurred in the system. All the recent valid transactions are bunched into a block (hence the name blockchain), which are then timestamped and stored using strong cryptography. Every new block added to the blockchain database, contains a hash (cryptographic hash) of the previous block, which makes it tamper-resistant. This is because changing any particular transaction in the blockchain database would change the hash of the block being tampered with, which would cascade to all the subsequent blocks added in the blockchain.

The blockchain database may be stored fully or partially by a set of nodes (and/or miners), which can verify any transaction being done on the blockchain and these act in “consensus”. Also, in a bitcoin blockchain, a miner has to solve a hard cryptographic problem to confirm/add a new block on the blockchain. This work done by miners is called proof-of-work (and the miners are adequately rewarded for this computationally expensive work being done by them). The proof-of-work makes the database resilient to attacks, as any attacker will need to have control on more than 51% of the computing power of the network to tamper with the transactional data.

A sample visual representation of blockchain is done below:

Challenges of Blockchain

Scale: Cost of computing, energy consumption in mining for transactions becomes very high as the blockchain increases in size. Also, the growing size of blockchain can be a problem as any need node will need to replicate the blockchain, which may take many days.

Speed Currently, maximum no. of transactions supported in public bitcoin blockchain are 7 per second, with proof of work / validations in place.

Recourse / Modifications: Its virtually impossible to modify a transaction once it is put in place.

Private versus public vs mixed: Public blockchains can scale but make process slower, Private ones can be faster but then the use case for blockchain needs to be clear (otherwise a simple shared database may work) and mixed permission based blockchains are in fairly nascent stages yet. Adoption and integration For any disruption, the integration and replacement of current systems and workflows are the biggest challenge. It makes it harder for banks to adopt any change quickly. Legal challenges For any disputes arising from blockchain transactions, there are various loopholes where the existing legal frameworks and courts may not be able to resolve the disputes easily. Future of Blockchain Its possible that blockchain may go through an internet like bubble bust and boom cycle, just compressed into lesser timeframe as shown below.

Businesses run a variety of applications, each containing and holding data vital to the needs of an organization. These applications cannot operate in isolation and eventually; there is a need to integrate them with various other applications. This is primarily driven by complex business needs, which call for information and resources, to be shared across the boundaries of these applications.

This interconnection and overlap of business applications and the information sharing across boundaries of application has led to an emergence of Application Programming Interfaces (APIs), which encapsulate and hide the details of one application from another and enable them to share data and information with the surrounding applications. This leads to better insights into data and showing the linkages with various data sets.

APIs have become more than a tool to connect various applications and their data. But, these APIs have become business enablers. Facebook launched its Facebook Platform in 2007, which led to an immediate opening up of social data, which could be leveraged by the third-party developers. Zynga is a prominent example of this era, which led to its tremendous growth leveraging on the social network’s wide reach and popularity. Within a year of Facebook Platform’s launch, there were at least 33,000 applications, which were running on the Facebook platform, which led users to engage with Facebook in newer ways which would never have been possible before.

Also, apart from having a public API like Facebook, there are a lot of products provide an API, using which developers can extend the functionality of the core product. Atlassian products like JIRA have an API, which extend the core functionality of the product. This makes the users of the product sticky and allows the product to be used in ways, in which the original developers might never have intended. Such usage of API presents a great business case, where you retain customers on the strength of the API and its supporting plugins, which makes the product sticky. There are many examples of businesses, where the core business is to provide an API for the consumers. Twilio is an example, which makes messaging, voice calling available as an API. Google Maps API completely changed the way mapping was done before by incumbents like Mapquest or TomTom etc and made it available to a wide array of applications and in turn generating revenues for Google and empowering businesses like Uber.

Several large major online and technology platforms scale mainly because of the APIs structure. For example in the travel industry, Hotels sell their rooms on their online website and also offer the same through APIs to Travel websites like Expedia, MakemyTrip etc. Traveller feedback is shared through the apis of Tripadvisor. Further level of aggregation is done by likes of Kayak.com which access the expedia, makemytrip apis and get the user best prices. Online Users get a one stop content (aggregated), analytics, and various other features through merging of several APIs on a single platform.

Amazon, one of the largest e-commerce companies, offers its content and products on APIs to Junglee and several other similar price comparison websites. Amazon also sources various products from different offline / online sellers through their APIs.

Fintech cannot be any different. Lot of fragmented information and processing is done across platforms. uTrade provides an API on top of its Algo Trading platform – muTrade. The API is publicly available on Github and allows proprietary traders to develop their own trading algorithms, which they do not need to share with anyone. Whilst we provide several inbuilt algorithms (more than 50 algos currently); we understand that various traders would like to build their own different proprietary algorithm, and we should be able to help them with market access, risk management etc. Within the platform, there is no distinct advantage to an algorithm that runs internally versus the algorithm that runs on an API and they both perform at the ultra low latencies (~10 micros) offered by the muTrade Algo Trading platform. This lets uTrade focus on the best and ultra low latency trading platform and the traders can focus on their strengths and use their prop strategies and logics, for their own benefit.

APIs economy in today’s digital world is the paramount necessity for all technology and Internet based companies. However, the way they are being consumed may change (for better) in the coming years. It would be interesting to see how the APIs shape and change even further, the way business is done in future.

Open Source software has taken the world by storm over the last two decades or so. If you look under the hood of any of the big software applications, you will see a number of open source libraries being used in the product. For example, if we look at the number of libraries being used in the Adobe Reader XI, the list of third party license/notices run into 61 pages, with a variety of open source software being used ranging from graphics libraries to compression and even some libraries from marquee projects like Open Office, Mozilla and the Android projects.

However, we do not see many domain specific application softwares in the open source world. Things are slowly moving into this direction, but there is a lot of progress to be made. And, this may be the next big thing in the open source world, where one would see new companies release mainstream application products which are open source, with an “Open Core” business model and giving the Traditional vendor companies a run for their money with their quality, feature set and rapid innovation.

Benefits of Open Source Software

It is now widely know the immense benefits of open source products:

• Quality - Open Source leads to more high quality-products or libraries. I would mention the example of Boost libraries for C++, which is a very high quality peer-reviewed cross-platform C++ libraries.

  Also, making a product open source leads to more set of eyes looking at the code, thus enabling more peer review of code and probably, leading to faster fixing of defects as well.

• Innovation - When you have open source products, there is a likelihood that more people would be using your product in the way you would never have imagined. This leads to feature requests and often redesign of the applications, making them better in general.

• No Vendor lock-in - Since you have the source code, there is no need to be locked into a particular vendor when looking for upgrades and maintenance of the product. An open source ecosystem leads to more service providers, who would work be more than willing to enhance and maintain upgrades or fulfulling special requirements of the client.

• Transparency - Software cannot get more transparent than this. With an open source product with significant community,

• Security - Obscurity does not help when it comes to information security in software products. Open Source software leads to a healthier and more secure products.

• Stand on the shoulder of giants - Lots of great open source software has been possible because of the large number of open source projects which done been done before. A lot of “big data” applications and frameworks may not had been possible without Hadoop.

OSS being used in Financial Firms

If you look at technology stacks in any of the large institutions (particularly, financial institutions, as we are in the financial technology space), one would see a large number of open source products being used everywhere.

Compilers

GCC and Clang have been leading the way in terms of feature support for C/C++ compilers and they have been way ahead of the commercial competitors, in terms of the new C++11 standard support.

Application servers

<

Open Source Technology Platforms

Given the fact that fundamentally most technology platforms within capital markets do very similar things (access information, manage trades and risk, comply with regulation, etc.), there could be a significant growth potential in the market to collaborate constructively. They can leverage off latest and greatest open source technologies and developing technology solutions for the financial markets in an open-source fashion, leading to lower cost of platform development, deployment and use that is highly desired by various financial services players (especially in the current market conditions where there is a massive slow-down since last few years and the revenue / profit model of these participants has been significantly eroded, and the requirement to spend to technology is only increasing) and let the niche players build their specialty or specific premium solutions on top of the core open source core financial services technology solutions.

Rise of electronic / low touch trading

Investors will need less and less hand holding from brokers and their traders. They will do more self-directed trading with the ease of use of technology. The western and developed markets have already seen this trend in the past decade (around 70% of American and 50% of European trading is driven by electronic tools); the emerging markets will move towards a similar ratio of technology driven trading. The trading workflow will move to more web and hand-device based platforms.

Low latency trading growth will slow down; given the massive investment required to push it to faster speeds (example Chip based / FPGA trading platform cost is very high and the potential returns do not justify the investment)

Self-learning machines leading to virtual and artificially intelligent trading is an area that may witness some reasonable growth.

Virtual simulation of trading environment and virtual trading may meet advances in the gaming industry.

Enhanced User Experience

The user experience for traders would continue to dramatically improve with the ease of information access, simpler and automated trading workflows (like 1 touch or even no touch fully automated trading).

Devices interactivity enhancements

As people become more mobile and the technology through cloud and various computing devices (mobile, tablets, laptops) support the ease and continuity of platform access, the boundaries between all these devices would be blurred and the user would seamlessly pick up any device in any location and be able to access information and trade into any financial markets. The optimization between the location, cloud access, computing power used, based on the proximity and the markets to be traded could lead to interesting business opportunities.

Social driven trading

Social networks would play an important role in the trading life-cycle as the information sharing becomes faster, more trust-able and the investor sentiment analysis becomes more feasible through the social platforms like Twitter, Facebook etc. Social collaboration and interactivity may lead to better trading ideas which can create more value over time.

Blurring of boundaries between capital markets players

There is likely to be more competition across the value chain of financial services players like brokers would continue to match orders like exchanges do, exchanges would offer broker alike trading and algo platforms so as to control the shift of liquidity, institutional investors may also get into broker and exchange roles somewhat to avoid information leakage etc.

There would also be a large consolidation among various capital market participants. And the competition among exchanges, brokers, investors would continue to become more global as world becomes .

Rise of regulation:

The regulatory bodies will try to control the increase in adoption of technologies within the financial services (given that s very hard for them to follow, understand and regulate the innovation in the market) leading to slower future growth in innovative ways of trading. There will be a shift towards more transparent modes of trading (example move from OTC to exchange listed / cleared products).

Tomcat is more or less the standard application container for most Java applications. In the J2EE space, JBoss is as prevalent as WebSphere.

Databases

MySQL and PostgreSQL have led the way for most open source databases. And, in today’s you will not find many applications which do not support one of these two leading open source databases.

Web Browsers

Firefox and Chromium have been the first browsers with support for most of the upcoming standards, be it features in the latest HTML5 standard or CSS standards. Microsoft Internet Explorer has been more of a laggard in this space. Also, commercial browser vendor Opera has ditched their own browser rendering engine Presto in favour of Webkit (Blink, actually).

Open Source Products in Financial Technology

Financial Institutions have been using open source projects throughout their technology stack. Full-suite application software is probably the last mile, where the financial institutions and the industry, in general, can adopt open source produts and bring the associated benefits for everyone in the ecosystem.

Increasingly, there have been products in Financial Technology, which are open source and are being widely adopted. QuickFIX is unarguably the most widely used FIX engine, which is probably being used in all the financial firms (big or small). And, OpenGamma has coming up really well as an open-source Risk and Margining system, which is increasingly being deployed across various financial firms.

Another notable mention is FinTP product by Allevo, which is a very well respected product in the payments and financial transaction processing industry. Allevo has release the core engine of their product FinTP under GPLv3 license and the source code is available on github.

In the capital markets space, there have been products like Marketcetera, Tradelink, Lodestone (Deutsche Bank initiative), which have tried to fill that gap with varying success. We, at uTrade, believe there is an enormous scope here to create a truly world class open source product in capital markets space. uTrade had participated in the Fintech Innovation Labs program last year and presented why Open Source in Capital Markets would work.

We would like to announce that we are opening up the core of our multi-asset trading platform and various other components/modules, which we have created over the past 3+ years. We would offer all this codebase to the community for criticism, adoption, contributions and collaborative development, which we think, will enable us to improve this product with the help of the vast open source community. All this code will be offered under a very liberal Apache License. We’ll be releasing this product on our github page in the next 2-3 months. Watch out !!!

The tremendous growth in financial trading volumes in the last couple of decades has been made possible only with the use of technology. Technology has helped in

• automating the dissemination of market data and other relevant information by exchanges and trading participants

• capturing and consolidation of market data from various exchanges and other sources of relevant information

• making complex calculations on live and historic data, leading to trade decisions

• trading and managing risk

• lowering the latency and cost of trading

• various post-trade processes

Technology is heavily and generally efficiently used by brokers (providing various retail and institutional investors access to financial markets), high frequency traders (who trade in huge volumes aggregating small profits in as many trades as possible, this typically includes arbitrageurs and market makers), quant algo traders (who use quantitative models to predict the future prices etc.), exchanges (who are the liquidity aggregators forming markets), and various other market participants. Each stake holder mentioned above has a significant role in the orderly functioning of the financial markets. Comprehensive Risk Management is the foremost requirement in managing these financial trading businesses.

There has been a major backlash against the financial firms (primarily the ones driven by technology including HFT firms) especially since the Flash Crash - European regulator wanting Half-second delay before execution of a trade or cancelling an order in a bid to temper incentives of trading fast.

It’s like introducing a speed breaker on a high way. People will drive very fast up to the breaker, and speed up again once they cross it. It’s not that hard for HFT firms to adjust their trading algorithms to accommodate for this regulatory change and still do what they intend to do.

• Indian regulator proposed that the exchanges to accept orders alternately from a co-location member (whose servers are located in the same physical location as the exchange) and a non co-location member.

Depending upon the cost-benefit analysis, an HFT firm can setup a co-location server and a server in proximity to the exchange and alternate sending of these orders through those 2 servers by adjusting their strategy.

• Capping the order to trade ratios for all market participants.

Technically the algos can work around this as well. But this can certainly help reduce some over-fishing in the market(i.e. algos sending too many smaller orders at wide limit prices and then cancelling them to search for trading opportunities which are not easily transparent).

• Certain regulators trying to have HFT players notify the exchange or the regulator of the algos used by them

Its part of Germany and s regulatory system, but does not curb the automated trading related over-trading triggers as the regulators or exchanges will not easily be able to understand the detailed functioning of the algos. They can only try and ensure the HFT firms ensure sound risk management and follow the practices strictly at the time of algo approval or audit.

• Retail Investor has a disadvantage and cannot access the same arbitrage opportunities due to lack of access to same technology as HFT firms

Retail Investor should not be trying to do sub second arbitrage the market in the first place. It’s like saying a cyclist cannot ride as fast as a car on a high-way and car-drivers should be stopped or slowed down. Retail investors should look to take slightly longer term view as they had been doing historically.

• Institutional investors can also not compete with the HFT firms’ rise in technology

HFT firms had been driving the use of technology and advancement to low latency trading. In their constant RnD process, they have stepped up in the technology value chain and have enabled the Institutional investors with good and efficient trading tools at affordable prices. For example, execution algorithms like VWAP, Smart Routers were not available to Institutional Investors a decade or two ago, but HFT firms led RnD and trading styles have made execution algorithms available to other Institutional as well as Retail Investors indirectly. Also, the objective of the Institutional Investors is typically to manage portfolios over with longer horizon targets rather than intra-day small price movements.

• HFT firms lead to market failures and crisis like Flash Crash in 2010, Knight Capital’s loss in 2012, etc.

In the absence of HFT, there had been ample crises in the financial industry (and in several other industries too) including the LTCM Crisis in 1998, Sub-Prime crisis in 2007, where the human processes have led to major crisis in the markets. The use of technology if managed well and regulated efficiently can actually reduce the probability and scale of such events.

For example, If an airplane crashes, it does not make the Air Travel a bad business. Air travel brings lots of benefits to our social and economic growth that it’s worth it despite rare unfortunate crashes. The regulators in that industry focus on ensuring the safety of passengers and plans is an utmost priority for the airlines.

In a similar way, HFT firms understand the risk management priority and usually target to mitigate and manage risk first, and then focus on profits.

HFT firms actually help in providing better liquidity in the market, lower cost of trading, tighter spreads, that help institutional investors and retail investors in getting their executions at better prices. Use of technology in the trading value chain has lead to split second order executions

• Imposing extra-taxes on HFT (like a regulatory proposal out there in Germany). Make it expensive for HFT players and reduce their net profits.

Such moves will either move the HFT firms to alternate venues or countries with lower or no taxes or hurt their profits up to a certain point, beyond which it does not make sense for them to be in the business (as some German HFT players are pondering to shut down)

Regulators across the world should embrace the constructive technology developments in financial markets that support growth of businesses and the economies. Their goal should be to understand the market behaviour, advancements in technology, bring transparency to the market, introduce risk management measures like a kill switch (for cancelling all trades when things go wrong), push for stronger stress testing of automation tools, enforce compliance, track patterns that may lead to systemic risks, and lead to a solid and comprehensive risk management across the financial industry.

Financial services plays an important role in the and their functioning. Following the credit crisis (2008), and its various consequences (deleveraging, lesser sophisticated financial products, shift towards increased regulation and transparency, reduced transaction volumes and margins, etc.) have lead to a massive shrinking of the financial services market size. This has led to a large down-sizing amongst the key participants including brokers, investment banks, exchanges, technology vendors, institutional investors (funds), etc.

With the above mentioned business scenario in mind, and the fact that all these participants need technology for automation, growth and even sustenance of their business, do we need every participant (10,000s of brokers, 100s of exchanges, 1000s of technology vendors, 100s of regulators, 100,000s of investment firms worldwide) developing fundamentally very similar technology for their business. For example,

• if a regulator wants brokers to report every trade on behalf of their clients in a certain way, should we have various brokers and various vendors develop the same solution?

• if a broker needs to connect to an exchange, or update its connectivity protocol to the exchange as per the exchange requirements, should we have various brokers and technology vendors spend their limited resources in doing the same work?

• in accordance with the latest regulations like MiFid, Basel, etc., should we have all Investment banks develop technologies to deal with the calculation and management of corporate level risk, liquidity, capital ratios, accounting policies?

• if an exchange or a regulator enhances the risk management policies, or trading workflow (to achieve best execution etc.), should each trading technology vendor and broker develop the same enhancement in their technology?

The above examples highlight a massive duplication of efforts that go on in an in-efficient financial services sector. When the resources are limited, times are tough (since last 5 years and for foreseeable future), the business model of the sector is being redefined; how could we afford such a duplication in efforts? Should we not try and find common solutions which work for most participants, and have other participants develop the premium technology and service solutions?

Similar initiatives led by open source technology platforms in other domains like Operating systems (Linux), FIX trading engine (QuickFIX), Web browsers (Mozilla Firefox), Web server (Apache), Web scripting language (PERL) have proven to be very productive for different technology eco-systems.

Open source technology primarily means that source code of an application is included in its compiled version, freely distributed (possibly with some re-selling / licensing constraints) and it leads to various developers across the world using that code, enhancing it to their custom requirements, and contributing back to the overall open source code repository. In medium to long term, it leads to enhanced and corrected technologies, and creates a deep platform which is widely used by various developers and users around.

We can broadly divide the financial services technologies into a few categories like Trading Platforms (including order management systems, algorithms, over-the-counter trading platforms, client connectivity system, exchange adapters for orders / executions and market data), Risk Management Systems (pre-trade and post-trade), Corporate level systems (for accounting, back office, etc.), Regulatory and Compliance reporting systems, Data management systems (including analytics), Exchange technologies, etc. Experts into these systems may contribute for a wider productive cause and help create the Financial-services Open Source Technology (lets term this as FOST). Such an initiative, while disruptive and leading to some winners and losers (hence it will be widely criticised first), can be the right collaborative and productive initiative for the future.

Potential criticisms (picking top 3 that come to my mind):

My software system is very complex and sophisticated, no open source can replace it : If you can build it, most likely others can as well. Please try and be realistic. If your system is really that sophisticated and complex, consider sharing it with others through FOST if you feel the initiative is valuable for eco-system over time, else you may continue to work independently.

What about stability, security of the modules offered on FOST? : Is a collaborative effort. We should look up to everyone to selflessly help drive this initiative. In general, open source technologies offer better performance, better security, and more flexibility at competitive cost.

How will such a product be maintained, enhanced, customized, and serviced? my business depends on it : Linux is a very good example. You can use the basic products directly at no cost. You may test it and customize it to your needs. For its premium versions there are products like Red Hat that provide customization and support (FYI, Red Hat is a multi-billion dollar company, so money can be made through the open source technology initiatives). We expect various other companies to come up and offer premium service models on top of FOST.

The key benefit remains a lower cost technology solution for financial services (this cost is typically very high ranging from 10%-50% of the participants overall cost) and the creation of a more efficient, productive, educative, innovative and collaborative eco-system.

Please help spread the word and lets debate this constructively.

Financial trading has grown from manual to electronic to automated with the use of algorithms in the last few decades. The increased adoption of technology will lead to disruptive trends in the current decade and drive exponential growth in newer business segments created within the capital markets (like additional liquidity that was generated by Algo/HFT players in the last few years) . We will possibly witness some of the following trends:

Open Source Technology Platforms

Given the fact that fundamentally most technology platforms within capital markets do very similar things (access information, manage trades and risk, comply with regulation, etc.), there could be a significant growth potential in the market to collaborate constructively. They can leverage off latest and greatest open source technologies and developing technology solutions for the financial markets in an open-source fashion, leading to lower cost of platform development, deployment and use that is highly desired by various financial services players (especially in the current market conditions where there is a massive slow-down since last few years and the revenue / profit model of these participants has been significantly eroded, and the requirement to spend to technology is only increasing) and let the niche players build their specialty or specific premium solutions on top of the core open source core financial services technology solutions.

Rise of electronic / low touch trading

Investors will need less and less hand holding from brokers and their traders. They will do more self-directed trading with the ease of use of technology. The western and developed markets have already seen this trend in the past decade (around 70% of American and 50% of European trading is driven by electronic tools); the emerging markets will move towards a similar ratio of technology driven trading. The trading workflow will move to more web and hand-device based platforms.

Low latency trading growth will slow down; given the massive investment required to push it to faster speeds (example Chip based / FPGA trading platform cost is very high and the potential returns do not justify the investment)

Self-learning machines leading to virtual and artificially intelligent trading is an area that may witness some reasonable growth.

Virtual simulation of trading environment and virtual trading may meet advances in the gaming industry.

Enhanced User Experience

The user experience for traders would continue to dramatically improve with the ease of information access, simpler and automated trading workflows (like 1 touch or even no touch fully automated trading).

Devices interactivity enhancements

As people become more mobile and the technology through cloud and various computing devices (mobile, tablets, laptops) support the ease and continuity of platform access, the boundaries between all these devices would be blurred and the user would seamlessly pick up any device in any location and be able to access information and trade into any financial markets. The optimization between the location, cloud access, computing power used, based on the proximity and the markets to be traded could lead to interesting business opportunities.

Social driven trading

Social networks would play an important role in the trading life-cycle as the information sharing becomes faster, more trust-able and the investor sentiment analysis becomes more feasible through the social platforms like Twitter, Facebook etc. Social collaboration and interactivity may lead to better trading ideas which can create more value over time.

Blurring of boundaries between capital markets players

There is likely to be more competition across the value chain of financial services players like brokers would continue to match orders like exchanges do, exchanges would offer broker alike trading and algo platforms so as to control the shift of liquidity, institutional investors may also get into broker and exchange roles somewhat to avoid information leakage etc.

There would also be a large consolidation among various capital market participants. And the competition among exchanges, brokers, investors would continue to become more global as world becomes .

Rise of regulation:

The regulatory bodies will try to control the increase in adoption of technologies within the financial services (given that s very hard for them to follow, understand and regulate the innovation in the market) leading to slower future growth in innovative ways of trading. There will be a shift towards more transparent modes of trading (example move from OTC to exchange listed / cleared products).