Deep Learning Network Portfolio: Building a Minimally Correlated Portfolio Deploying Network Analysis


In this article, we have used hedgecraft‘s approach to portfolio management. However, unlike hedgecraft, we have used a sub-graph centrality approach.This sub-graph centrality approach is what makes our approach unique. Using insights from Network Science, we build a centrality based risk model for generating portfolio asset weights. The model is trained with the daily prices of 31 stocks from 2006-2014 and validated in the years 2015, 2016, 2017, 2018 & 2019. As a benchmark, we compare the model with a portfolio constructed with Modern Portfolio Theory (MPT). Our proposed asset allocation algorithm significantly outperformed both the Sensex30 and Nifty50 indexes in every validation year with an average annual return rate of 26.51%, a 13.54% annual volatility, a 1.59 Sharpe ratio, a -21.22% maximum drawdown, a return over maximum drawdown of 6.56, and a growth-risk-ratio of 1.86. In comparison, the MPT portfolio had a 9.63% average annual return rate, an 18.07% annual standard deviation, a Sharpe ratio of 0.41, a maximum drawdown of -22.59%, a return over maximum drawdown of 2.2, and a growth-risk-ratio of 0.63.


In this series, we play the part of an Investment Data Scientist at Bridgewater Associates performing a go/no go analysis on a new idea for risk-weighted asset allocation. Our aim is to develop a network-based model for generating asset weights such that the probability of losing money in any given year is minimized. We’ve heard down the grapevine that all go-decisions will be presented to Dalio’s inner circle at the end of the week and will likely be subject to intense scrutiny. As such, we work with a few highly correlated assets with strict go/no go criteria. We build the model using the daily prices of each stock in the Sensex. If our recommended portfolio either (1) loses money in any year, (2) does not outperform the market every year, or (3) does not outperform the MPT portfolio—the decision is no go.

Asset Diversification and Allocation

The building blocks of a portfolio are assets (resources with the economic value expected to increase over time). Each asset belongs to one of seven primary asset classes: cash, equity, fixed income, commodities, real estate, alternative assets, and more recently, digital (such as cryptocurrency and blockchain). Within each class are different asset types. For example stocks, index funds, and equity mutual funds all belong to the equity class while gold, oil, and corn belong to the commodities class. An emerging consensus in the financial sector is this: a portfolio containing assets of many classes and types hedges against potential losses by increasing the number of revenue streams. In general the more diverse the portfolio the less likely it is to lose money. Take stocks for example. A diversified stock portfolio contains positions in multiple sectors. We call this asset diversification, or more simply diversification. Below is a table summarizing the asset classes and some of their respective types.

An investor solves the following (asset allocation) problem: given X rupees and N, assets find the best possible way of breaking X into N pieces. By “best possible” we mean maximizing our returns subject to minimizing the risk of our initial investment. In other words, we aim to consistently grow X irrespective of the overall state of the market. In what follows, we explore provocative insights by Ray Dalio and others on portfolio construction.

The above chart depicts the behavior of a portfolio with increasing diversification. Along the x-axis is the number of asset types. Along the y-axis is how “spread out” the annual returns are. A lower annual standard deviation indicates smaller fluctuations in each revenue stream, and in turn a diminished risk exposure. The “Holy Grail” so to speak, is to (1) find the largest number of assets that are the least correlated and (2) allocate X rupees to those assets such that the probability of losing money any given year is minimized. The underlying principle is this: the portfolio most robust against large market fluctuations and economic downturns is a portfolio with assets that are the most independent of each other.

Visualizing How A Portfolio is Correlated with Itself (with Physics)

The following visualizations are rendered with the Kamada-Kawai method which treats each vertex of the graph as a mass and each edge as a spring. The graph is drawn by finding the list of vertex positions that minimize the total energy of the ball-spring system. The method treats the spring lengths as the weights of the graph, which is given by 1 – cor_matrix where cor_matrix is the distance correlation matrix. Nodes separated by large distances reflect smaller correlations between their time-series data, while nodes separated by small distances reflect larger correlations. The minimum energy configuration consists of vertices with few connections experiencing a repulsive force and vertices with many connections feeling an attractive force. As such, nodes with a larger degree (more correlations) fall towards to the center of the visualization where nodes with a smaller degree (fewer correlations) are pushed outwards. For an overview of physics-based graph visualizations see the Force directed graph drawing wiki.

In the above visualization, the sizes of the vertices are proportional to the number of connections they have. The color bar to the right indicates the degree of dissimilarity (the distance) between the stocks. The larger the value (the lighter the color) the less similar the stocks are. Keeping this in mind, several stocks jump out. Bajaj Finance, ITC, HUL, and HeroMotoCorp all lie on the periphery of the network with the fewest number of correlations above Pc = 0.325. On the other hand ICICI Bank, Axis Bank, SBI, and Yes Bank sit in the core of the network with the greatest number connections above Pc = 0.325. It is clear from the closing prices network that our asset allocation algorithm needs to reward vertices on the periphery and punish those nearing the center. In the next code block we build a function to visualize how the edges of the distance correlation network are distributed.


  • The degree distribution is left-skewed.
  • The average node is connected to 86.6% of the network.
  • Very few nodes are connected to less than 66.6% of the network.
  • The kernel density estimation is not a good fit.
  • By eyeballing the plot, the degrees appear to follow an inverse power-law distribution. (This would be consistent with the findings of Tse, et al. (2010)).

Intraportfolio Risk

We read an intraportfolio risk plot like this: ICICI Bank is 0.091/0.084 = 4.94 times riskier than Maruti Suzuki (MSPL). Intuitively, the assets that cluster in the center of the network are most susceptible to impacts, whereas those further from the cluster are the least susceptible. The logic from here is straightforward: take the inverse of the relative risk (which we call the “relative certainty”) and normalize it such that it adds to 1. These are the asset weights. Formally,

Next, Let’s visualize the allocation of 100,000 (INR) in our portfolio

Subgraph Centrality-Based Asset Allocation

Bajaj Finance receives nearly 12.58%, Bajaj Auto gets about 12.58%, HUL 8.15%, Infosys 4.52%, and the remaining assets receive less than 0.5% of our capital. To the traditional investor, this strategy may appear “risky” since 60% of our investment is with 5 of our 31 assets. While it’s true if Bajaj Finance is hit hard, we’ll lose a substantial amount of money, our algorithm predicts Bajaj Finance is the least likely to take a hit if and when our other assets get in trouble. Bajaj Finance is clearly the winning pick in our portfolio.

It’s worth pointing out that the methods we’ve used to generate the asset allocation weights differ dramatically from the contemporary methods of MPT and its extensions. The approach taken in this project makes no assumptions of future outcomes of a portfolio, i.e., the algorithm doesn’t require us to make a prediction of the expected returns (as MPT does). What’s more—we’re not solving an optimization problem—there’s nothing to be minimized or maximized. Instead, we observe the topology (interrelatedness) of our portfolio, predict which assets are the most susceptible to the subgraph centrality of volatile behavior and allocate capital accordingly.

Alternative Allocation Strategy: Allocate Capital in the Maximum Independent Set

The maximum independent set (MIS) is the largest set of vertices such that no two are adjacent. Applied to our asset correlation network, the MIS is the greatest number of assets such that every pair has a correlation below Pc = 0.325. The size of the MIS is inversely proportional to the threshold Pc. Larger values of Pc produce a sparse network (more edges are removed) and therefore the MIS tends to be larger. An optimized portfolio would therefore correspond to maximizing the size of the MIS subject to minimizing Pc . The best way to do this is to increase the universe of assets we’re willing to invest in. By further diversifying the portfolio with many asset types and classes, we can isolate the largest number of minimally correlated assets and allocate capital inversely proportional to their relative risk. While generating the asset weights remains a non-optimization problem, generating the asset correlation network becomes one. We’re really solving two separate problems: determing how to build the asset correlation network (there are many) and determining which graph invariants (there are many) extract the asset weights from the network. As such, one can easily imagine a vast landscape of portfolios beyond that of MPT and a metric fuck-tonne of wealth to create. Unfortunately, solving the MIS problem is NP-hard. The best we can do is find an approximation.

Using Expert Knowledge to Approximate the Maximum Independent Set

We have two options: randomly generate a list of maximal indpendent sets (subgraphs of such that no two vertices share an edge) and select the largest one, or use expert knowledge to reduce the number of sets to generate and do the latter. Both methods are imperfect, but the former is far more computationally expensive than the latter. Suppose we do fundamentals research and conclude Bajaj Finance and HUL must be in our portfolio. How could we imbue the algorithm with this knowledge? Can we make the algorithm flexible enough for portfolio managers to fine-tune with goold-ole’ fashioned research, while at the same time keeping it rigged enough to prevent poor decisions from producing terribe portfolios? We confront this problem in the code block below by extracting an approximate MIS by generating 100 random maximal indpendent sets containing Bajaj Finance and HUL.

The generate_mis function generates a maximal independent set that approximates the true maximum independent set. As an option, the user can pick a list of assets they want in their portfolio and generate_mis will return the safest assets to complement the user’s choice. Picking Bajaj Finance and HUL left us with Sun Pharma, Hero Moto Corp amongst others. The weights of these assets will remain directly inversely proportional to the subgraph centrality.

Allocating Shares to the Deep Learning Network Portfolio

In this section we write production (almost) ready code for portfolio analysis and include our own risk-adjusted returns score. The section looks something like this:

We obtain the cumulative returns and returns on investment, extract the end of year returns and annual return rates, calculate the average annual rate of returns and annualized portfolio standard deviation, compute the Sharpe Ratio, Maximum Drawdown, Returns over Maximum Drawdown, and our own unique measure: the Growth-Risk Ratio.

Finally, we visualize the returns, drawdowns, and returns distribution of each model and analyze the results the performance of each portfolio.

Visualizing the Returns

Pictured above are the daily returns for Deep Learning Network MIS (solid green curve), Deep Learning Network (solid blue curve), and the Efficient Frontier portfolio (solid red curve) from 2015 to 2019. The colorcoded dashed curves represent the 50 day rolling averages of the respective portfolios. Several observations pop: (1) Deep Learning Network MIS significantly outperformed Deep Learning Network Portfolio, (2) Deep Learning Network Portfolio substantially outperformed the Efficient Frontier, (3) Deep Learning Network MIS grew 158.4% larger, falling below 0% returns 0 out of all the trading days, (4) Deep Learning Network grew 139.3% larger and (5) the Efficient Frontier grew 49.8% larger. Next, let’s observe the annual returns for each portfolio and compare them with the market.

In comparison, the Nifty50 had a -4.1%, 3%, 28.6%, 3.2%, -1.03% (YTD) annual return rate in 2015, 2016, 2017, 2018, & 2019 (YTD) respectively. Deep Learning Network Portfolio and Deep Learning Network MIS substantially outperformed both the market and the Efficient Frontier. Both Deep Learning Network portfolios grew at an impressive rate. Deep Learning Network MIS grew 19.1% larger than Deep Learning Network Portfolio and 108.6% larger than the Efficient Frontier, while Deep Learning Network Portfolio grew 89.5% larger than the Efficient Frontier. What’s more, Deep Learning Network MIS’s return rates consistently increased about 25% every year, whereas the return rates of Deep Learning Network Portfolio and the Efficient Frontier were less consistent. Deep Learning Network Portfolio MIS clearly has the most consistent rate of growth. We’d expect this rapid growth to be accompanied with a large burden of risk—either manifested as a large degree of volatility, steep and frequent maximum drawdowns, or both. As we explore below, the Deep Learning Network portfolios’ sustained their growth rates with significantly less risk exposure than the Efficient Frontier.

Visualizing Drawdowns

Illustrated above is the daily rolling 252-day drawdown for Deep Learning Network MIS (filled sea green curve), Deep Learning Network (filled royal blue curve), and the Efficient Frontier (filled dark salmon curve) along with the respective rolling maximum drawdowns (solid curves). Several observations stick out: (1) the Deep Learning Network portfolios have significantly smaller drawdowns than the portfolio generated from the Efficient Frontier, (2) both Deep Learning Network portfolios have roughly the same maximum drawdown (about 22%), (3) Deep Learning Network on average lost the least amount of returns, and (4) Deep Learning Network’s rolling maximum drawdowns are, on average, less pronounced than Deep Learning Network MIS. These results suggest the subgraph centrality has predictive power as a measure of relative or intraportfolio risk, and more generally, that network-based portfolio construction is a promising alternative to the more traditional approaches like MPT.

Deep Learning Network and its MIS variant dramatically outperformed the Efficient Froniter on every metric (save Deep Learning Network MIS’s annual volatility). These results give credence to the possibility that we are on to something substantial here as we have passed the criteria of our go/no go test. Outperforming MPT by these margins is no simple feat, but, the real test is whether or not Deep Learning Network Portfolio can consistently beat MPT on many randomly generated portfolios. To wrap up this notebook, let’s take a look at how the returns for each portfolio are distributed and move to the conclusion of Deep Learning Network Portfolio Optimzation.

Visualizing the Distribution of Returns

Above are the returns distribution for each portfolio: Efficient Frontier (in red), Deep Learning Network (in blue), and Deep Learning Network MIS (in green). The Efficient Frontier algorithm somewhat produced a portfolio with a normal distribution of returns; the same can’t be said of the Deep Learning Network portfolios as they’re heavily right-skewed. The right-skewness of the Deep Learning Network portfolios is caused by their strong upward momentum, that is, their consistent growth. In general, we’d expect a strong correlation between the right-skewness of the returns distribution and the growth-risk-ratio.

It’s important to emphasize that deviation-based measures of risk-adjusted performance implicitly assume the distribution of returns follows a normal distribution. As such, the Sharpe ratio isn’t a suitable measure of performance since the standard deviation isn’t a suitable measure of risk for the Deep Learning Network portfolios.

While Deep Learning Network had less pronounced maximum drawdowns it was more frequently bellow 0% returns (1.59% of the time) than its MIS variant (0.53% of the time). These values dwarf that of the Efficient Frontier, which painfully experienced negative returns a third of the time. It’s interesting to note that the maximum loss of the Deep Learning Network portfolio is an order of magnitude smaller than their maximum drawdowns. This relationship is in contrast to the Efficient Frontier’s maximum loss which is on the same order of magnitude as its maximum drawdown. It’s also interesting to point out that Deep Learning Network has a lower probability of falling below its rolling 30, 50, and 90 averages than its MIS variant. Taken together, Deep Learning Network’s smaller average rolling maximum drawdown and smaller probabilities of falling below the above rolling averages indicate its growth is more consistent than its MIS variant. On the one hand, Deep Learning Network’s growth is more consistent than its MIS variant while on the other hand, the MIS variant has a more consistent growth rate. Stated another way: Deep Learning Network’s “velocity of returns” is more consistent than that of the MIS variant’s, whereas Deep Learning Network MIS’s “acceleration of returns” is more consistent than that of the Deep Learning Network portfolio.

Future Portfolio Allocation & Conclusion

A similar analysis as displayed above was repeated for generating the optimal portfolio and the subsequent allocation. Following were the results of the same:

HUL: 11.73% , ITC: 11.73% , Bajaj Auto: 11.73% , Sun Pharma: 11.73% , ONGC: 11.73% , Asian Paints: 11.73% , NTPC: 7.6% , PowerGrid: 7.6% , Tech Mahindra: 3.88% , Infosys: 3.88% TCS: 2.76% HCL Tech: 2.76% HeroMotorCorp: 0.87%

Thus, Sector Allocation proposed by our Deep Learning Network Algorithm is as follows: FMCG: 23.46% , Automobile: 12.6% , Pharma: 11.73% , IT: 13.28% , Energy: 26.93% , Paints & Varnishes: 11.73%


In this article, we built a novel algorithm for generating asset weights of a minimally correlated portfolio with tools from network science. Our approach is twofold: we first construct an asset correlation network with energy statistics (i.e., the distance correlation) and then extract the asset weights with a suitable centrality measure. As an intermediate step we interpret the centrality score (in our case the subgraph centrality) as a measure of relative risk as it quantifies the influence of each asset in the network. Recognizing the need for a human-in-the middle variation of our proposed method, we modified the asset allocation algorithm to allow a user to pick assets subject to the constraints of the maximal independent set.

Both algorithms (Deep Learning Network and Deep Learning Network MIS, including the benchmark Efficient Frontier) were trained on a dataset of thirty-one daily historical stock prices from 2000-2014 and tested from 2015-2019. The portfolios were evaluated by cumulative returns, return rates, volatility, maximum drawdowns, risk-adjusted return metrics, and downside risk-adjusted performance metrics. On all performance metrics, the Deep Learning Network algorithm significantly outperformed both the portfolio generated by the Efficient Frontier and the market–passing our go/no go criteria.

Harsh Shivlani
Team Leader– Fixed Income & Derivatives
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Harsh on LinkedIn
Neil Jha
Team Leader – Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Neil on LinkedIn

China is coming up with cryptocurrency – shock or surprise?

The meeting of finance and technology, widely known as fintech, is changing the landscape of investment management. As the saying goes, it’s tough to make predictions, especially about the future events. But it’s manifestly worth the effort because catching big trends is how fortunes are made and catastrophic losses are avoided.

Blockchain-related topics are extremely hot nowadays and cryptocurrencies are one of those. So, what is a cryptocurrency? From the word itself you can see that it has something to do with cryptography and currency. For its part, cryptography is the process of converting ordinary plain text into unintelligible text and vice-versa. Modern cryptography deals with confidentiality: information cannot be understood by anyone, integrity: information cannot be altered, and authentication: sender and receiver can confirm each other.

Putting all the pieces together, cryptocurrency is a medium of exchange value (just like ordinary money) that exists in the digital world and relies on encryption, which makes transactions secure. A cryptocurrency is an alternative form of payment to cash, credit cards, and cheques. The technology behind it allows you to send it directly to others without going through a third party like a bank. In short, cryptocurrencies are like virtual accounting systems.

As you can find, there are many exciting use cases for this. You can send money back to your family without incurring large international fees if you’re working in a different country. Merchants no longer have to worry about payment fraud because people can only spend what they have. Summing up, Cryptocurrency is a radically new way of paying that makes all the transactions secure and helps to get rid of intermediaries represented by banks, which also contributes to a significant reduction in the commission fee.

The cryptocurrencies can either be based on blockchain technology or can be centrally issued, circulated within a community or geographic location, or tied to fiat currency. Blockchain is a revolutionary ledger technology, with a wide array of potential applications from smart contracts to healthcare systems, but it did not catch the attention of speculators and the media until Bitcoin surged from $0.009 to more than $11,000 per coin. There are more than 869 cryptocurrencies, but without fundamentals, they are little more than “trust machines” and, as such, are nearly unanalyzable. They generate no cash flow, making discounted valuation approaches inapplicable, but this criticism applies to gold as well.

Although it is cheaper to invest in the early stages, during a new cryptocurrency’s initial coin offering, doing so may overlook the network effect that favors older altcoins (alternative cryptocurrencies other than bitcoin).

Cryptocurrencies are going to play a major role in the coming years and China has decided to be part of that future, in a big way. China’s official digital currency is nearly ready. As much as China frowns on cryptocurrency, it’s happy to introduce its cryptocurrency. There is a great deal of confusion and misunderstood facts surrounding the legal status of cryptocurrency in China. Various headlines like China Bans Bitcoin, China Bans Crypto Exchanges, China Bans Bitcoin Mining, and many more make most people unclear on where China stands on cryptocurrency and whether that has any real impact on how its citizens behave.

The People’s Bank of China has revealed that its digital currency, “can now be said to be ready” after five years of research work. Don’t expect it to mimic crypto, however. According to payments Deputy Chief Mu Changchun, it’ll use a more complex algorithm and structure. This project of coming up with own cryptocurrency of China was started by the former governor of China’s central bank, Zhou Xiaochuan, who retired in March. He decided to come up with the digital currency which will protect China from having to adopt a technology standard, like Bitcoin, designed and controlled by others. 

Facebook Inc.’s push to create cryptocurrency Libra has caused concerns among global central banks, including the People’s Bank of China (PBOC), which said the digital asset must be put under central bank’s supervision to prevent potential foreign exchange risks and protect the authority of monetary policy. Sun Tianqi, an official from China’s State Administration of Foreign Exchange, said, “Libra must be seen as a foreign currency and be put under China’s framework of forex management”. Dave Chapman, executive director at BC Technology Group Ltd also said on similar lines that, “It is without a doubt that with the announcement of Libra, governments, regulators and central banks around the world have had to speed up their plans and approach to digital assets. They have to consider the possibility that non-government issued currencies could dramatically disrupt finance and payments.”

How the cryptocurrency issued by China will be different from other cryptocurrencies, might be one of the questions coming to your mind. To begin with, in launching the new cryptocurrency, referred to as DC/EP for Digital Currency/Electronic Payment, the People’s Bank of China (PBOC) has stolen a march on both Facebook and other central bankers who have been discussing the possibility of a cryptocurrency and how it’s the implication. What sets China’s DC/EP apart from libra and Mark Carney’s(Bank of England’s Governor) “synthetic hegemonic currency” (SHC), according to Paul Schulte(The founder and editor of Schulte Research, a company does research on banks, financial technology, bank algorithms, and credit algorithms), is that while libra is little more than early-stage computer code and the SHC doesn’t appear to have gone much further than Carney’s mind, the Chinese cryptocurrency is ready to launch. “China is barreling forward on reforms and rolling out the cryptocurrency,” says Schulte, who now runs a research firm. PBOC will be the first central bank to come up with its cryptocurrency. Unlike the decentralized blockchain-based offerings, this one could give Beijing more control over its entire financial system. It would increase the PBOC’s ability to root out risks and crackdown on money laundering. It could also give the government an unprecedented window into individuals’ private lives.

Deputy Chief Mu Changchun described the central bank’s “two-tiered” system, wherein the bank would create the cryptocurrency and a small group of trusted commercial businesses would “pay the central bank 100% in full” to be allowed to distribute it. This dual delivery system is suitable for national conditions of China. It can not only use existing resources to mobilize the enthusiasm of commercial banks but also smoothly improve the acceptance of the digital currency across China. If China’s leaders agree on with this idea of a legal cryptocurrency for the whole country, its introduction will likely be gradual. Early adopters would be barred from using it on investment products, a person familiar with the central bank’s plans says, which would make the impact on monetary policy negligible. 

“China’s strategic plan is to integrate more closely with the rest of the world. Cryptocurrency is just one of the means to have a more internationalized renminbi. It’s all strategic. It’s all long term”, said Charles Liu, chairman of HAO International, a private equity firm investing over $700 million in Chinese growth companies. Finally, the Chinese government said that the cryptocurrency could launch as soon as November 11, China’s busiest shopping day, known as Singles Day.

Pratik Jaju
Team Member– Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2019-21)

Connect with Harsh on LinkedIn
Omkar Pawar
Team Member– Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2019-21)

Connect with Omkar on LinkedIn

Fintech in India- A Global Growth story

FinTech has emerged as a relatively new industry in India. The year 2018 was a big one for the Indian financial technology and financial services ecosystem. With $2.34 billion being raised across 145 deals, fintech finally unseated ecommerce from the top of the list after years of dominance.

So, what really is FinTech?

In short, it is an industry that comprises of companies(such as insurance, asset management, payments) that use technology to offer financial services. Initially, FinTech started its trial by setting its operating base in the banking industry. But over the last five years, it has seen tremendous development and has expanded to insurance and asset management companies as well. 
By leveraging machine learning, FinTech companies are looking to analyze customer expectations and their responses.

In today’s digital economy, a whole new generation of FinTech’s, including nimble new start-ups with cutting edge technology have boomed alongside behemoths and are now valued more than many traditional banks and financial services firms.
We are witnessing specialization in many of the global fintech centres – London emerging as a hub for investments into open banking solutions, while China has become well known for facial recognition associated with biometric technology and Israel, a centre for cybersecurity.
India, however is yet to find a niche to focus on or a specialization, despite the mass availability of smart-technology talent.
As different hubs emerge around the world and technologies become more mature, specialization will be key to further develop India’s fintechs and continue to lure billions of dollars in investments and talent. Being the jack of all trades in different technologies and solutions, from mobile payments to credit scoring, digital banking and peer-to-peer lending, may have worked at the initial stages of fintech development, but the future will be dominated by specialists. 

To answer the many queries we have about fintech, its utilization and growth in the future, I decided to go ahead and take a personalized approach about this. I prepared some questions I had in mind about FinTech in India, and figured the ideal approach to know more would be to ask people who themselves are part of this industry.

I went ahead and interviewed two people, one who works as an Analyst in ‘Advanced Analytics’ division of MasterCard, and the other who is a Head of Sales and Alliances at Airpay. Both had a lot to say about matters relevant to this field, since a lot of work is being done in India to promote FinTech as a rising industry. Here is what they had to say about it.


How is the growth of fintech being supported in India?


India is doing pretty good in terms of fintech and it’s mostly coming from start-ups. Though one of the concerning facts is that most of them are eventually bought by foreign investors so won’t be completely fair to say that India has its in-house fintech innovation or technology. But with Indian government taking great initiatives like launching apps BHIM and UPI interface for P2P payments, it’s a great step towards our contribution to the fintech world.


  • A large unbanked and untapped market in India has been driving in private players willing to bet on it.
  • The Digital India initiative by the GoI has broadly laid down the framework for going digital and you cannot truly be a digital economy without Fintech.
  • The PMJDY led to financial inclusion of the masses, which has been leveraged by Fintechs to grow.
  • The exponential growth of smartphone adoption, along with some of the lowest mobile data rates have brought the next round of Indians online, which will further fuel digitisation of payments, insurance, etc.
  • The NPCI has been actively involved in churning out cutting edge payment methods like the UPI, which has again spurred adoption of digital means.


There are namely 4 spheres of Fintech: In the systems sphere, in the B2B sphere, in the B2C sphere and B2G sphere. How do you think these are getting impacted, whether interdependently or not?


I think we are experiencing a major shift in our focus from P2P to B2B or B2C. The P2P space is quite evolved now unlike B2B or B2C space which has great potential. And, when it comes to B2B space, it’s both Corporates and Small Business where we need to focus because one is much more higher in revenue and the latter one is much more in number. There is still a lot of informal lending when it comes to business space and there is a lot of potential to not only improve payments systems in that space from both cost and speed perspective but also, come up with new technology to encourage small business to start using digital ways of money transfer.

B2C is also evolving everyday where every business is trying to value their customer much more than ever and give as personalised recommendations or offers as possible.


There are definite overlaps in the various spheres and a huge opportunity to create a unified platform. The effort has already begun with the BBPS infrastructure, which brought C2B and C2G payments on a single platform.


FinTech is reshaping categories and disrupting incumbents across a number of financial services sector categories. These include ‘Savings, Personal Finance, Investment & Wealth Management, Insurance, Block chain & Crypto, Lending & Unsecured Credit, and most significantly, Payments. Could you highlight more on the payments system, considering how big it is, not only in India, but all across the globe?


Gone are the days when people used to stand in long queues in banks to deposit/withdraw money for their personal use or to transfer it to somebody else. We are also trying to come out of the cheques world which takes 2-3 days for payment processing. This is the age of fast and quick technology, we want everything simple, fast and secure. You press the button while you’re sitting at home and you can transfer money from 1 part of the world to another. This is already embraced by P2P payments world and companies are coming up with efficient and cheap solutions for Small Business and even large corporates. We are yet to capture the full potential of corporate space but fintech is headed in the right path and card acceptance in Small Business and Corporates is a start of it. So yes, payments system has been modified highly not just for ease but also to make it more secure.


As they say, data is the new oil. Payments are the final layer to any business activity. It is your final signature when doing business with an entity. This payments data generated through digital payments is a rich source for creating patterns and understanding consumer behaviour. We see Fintech as an enabler or a partner for the existing ecosystem, helping them reach out to untapped markets and helping them to create a consumer profile using alternate data points.


38% of the world’s population lack a basic bank account and an even greater proportion lack the simplest of insurance and investment products. Do you think FinTech, particularly in the form of mobile money, is an essential part of the solution for this or it could face barriers?


If we just talk about India, out of 1.3B population, we have 800M mobile users. With such high penetration of mobile in India, fintech in the form of mobile money is an essential part of the solution. Apps like Paytm which cater to small business and small local merchants making them reach out to each and every person in India in turn, facilitating P2P, B2C, C2B, B2B payments. Anyway, the future is mobile or maybe something even smaller where 1 device could serve the purpose of everything so it does make sense that we are trying to fit in every payment form in there.


Absolutely! Digital payments give the customer a footprint that banks and insurance companies can leverage. It is redefining the metrics that a traditional bank or an insurance company would use to evaluate a customer and use the alternate data points to curate the best products. Even for existing consumers, Fintechs are helping the large legacy players get down to the brass tacks of profiling and provide products that stay current and relevant. For example, sachet insurance is a great innovation that does away with larger bulky and one-size-fits-all products and gives the consumers exactly what they want and presents the insurer another opportunity for product uptake


Is it a boon for cyber security or are we just unaware of how much of our personal data is actually out in the cloud? Fintrusion? A survey said online fraud could reach $25.6 Billion by 2020. Views on this? So, is fintech actually increasing cyber security?


There is definitely new technology coming up for cyber security and it is surely needed when it is so easy to do an identity fraud or any kind of fraud for that matter because technology will have certain limitations too. Definitely, most of the people are unaware of the quantity/quality of their personal data in the cloud and even, if they are have some idea, they aren’t educated enough on its misuse. And with this whole fintrusion, there is growth of the idea of “privacy” and people are becoming more cautious while sharing their personal data with various websites/apps. But fintech is definitely coming up with increased cyber security. Country like India has 2 step authentication which is a good measure to avoid fraud. Not even that, banks are also coming up with different algorithms to identify fraud and mitigate it in the best possible manner. But there is need for more and at much faster rate especially with new payment methods like Card on File, Contactless, etc. coming in.


Every innovation comes with its own set of risks. We feel the problem here is that the push is mostly on product adoption, rather than consumer education and then eventually adoption. So rather than being reactive, all the stakeholders must be proactive when it comes to data security and consumer awareness.

That said, there is a whole segment of entrepreneurs and technologists working on data security & fraud prevention, so as the industry matures, we should see this challenge being mitigated.


50.2% globally saying they do business with at least one non-traditional firm for banking, insurance, payments or investment management, with the percentage reaching the highest in Asia-Pacific (58.5%). How do you think the consumer response has been to fintech since its inception?


Asia Pacific has a lot of developing countries which is one of the major factors as to why they do business with at least one non-traditional firm. People are still stuck to their old ways and will definitely take time to come up to speed in terms of learning new technology. But overall, fintech has been growing quite fast in these countries and even in better way because fintech isn’t just reaching the consumers or large corporates but also small business or local small merchants which are more prevalent again in developing countries rather than developed countries. China has their own payment system and they are closed to the idea of having the same payment system as the rest of the world but that definitely makes them more secure and India has 2 step authentication which is again unique to it but again, mitigates a lot of fraud already. And because there is a lot of untapped potential and cash is still a king in Asia Pacific, the growth of fintech will be the highest here.


The consumer response till now has been positive, albeit not uniformly. While banks and traditional institutions invoke trust, the generation coming into the workforce (with disposable income in hands and a key drivers of consumption) values speed and customer experience equally, if not more. That said, there is still an entire section of the population that intrinsically thinks of new-age Fintechs as risky. This trust deficit hampers a digital-only strategy. So as a Fintech, our ideal sweet spot is to partner with the traditional institutions to combine the trust that banks inspire with the speed and customer experience that Fintechs offer.

Purnima Dutt
Volunteer- FinTech
(MSc Finance, NMIMS Mumbai. Batch 2018-20)

Connect with Purnima on LinkedIn

Could Libra be the end of the petrodollar?

Let us start with, “What is the petrodollar?”; The petrodollar system is an exchange of oil for USD between countries that buy oil and those that produce it. The origin of the petrodollar can be traced back to the Bretton Woods Agreement, which replaced the gold standard with USD as the reserve currency. Under this agreement, the USD was pegged to gold, while other global currencies were pegged to USD. However, due to massive stagflation, President Nixon in 1971 declared that USD would no longer be the exchange for gold to boost economic growth for the U.S; this led to the petrodollar system, where the USA and Saudi Arabia agreed to set oil prices in USD. This meant that any country that wanted to buy oil from Saudi Arabia, would first have to buy USD. This leads the rest of the OPEC to also follow the same system. The next obvious question that anyone would ask would be, was this petrodollar good for the U.S?

The petrodollar system elevated the USD to the global reserve currency and via this, the USA enjoys a trade deficit and is also a global economic hegemony. As anyone who needs to buy oil needs to first buy USD, this makes USD the most dominant currency in the world. This gives the USA an exorbitant privilege. This essentially means that the USA can fund its current account deficit by issuing dollar-denominated assets at extremely low rates of interest. However, this could come to an end in the near future.

  • One of the reasons for this is that China and Russia have been setting up deals with Oil rich nations, without the use of the USA.
  • The USA historically has been forcing its foreign policy using the petrodollar. However, in the recent past, the USA was unable to do this. Iran, Russia, and Ukraine have signed a 5-year trade deal worth $20 billion. This will not take place in USD and includes the sale on Iran’s oil. Venezuela and Iran also signed a similar oil trade deal.

Could Libra replace the petrodollar?

Unlike most other cryptocurrencies, Libra is backed by real assets. It’s often referred to as ‘stable coin’ because it is backed by real assets and is linked to a basket of diversified global currencies and low-risk bonds. So, via Libra, you could immediately send money all over the world with negligible fees. Facebook is working on Libra along with companies such as Uber, eBay, PayPal, Visa, etc. So, the global acceptability of Libra should not be an issue. Therefore, as Libra is stable and could be accepted globally, we could price Oil in the international market in Libra instead of USD.

Neil Jha
Team Leader – Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Neil on LinkedIn

How Blockchain is changing dynamics for Stock Trading

In the recent past Blockchain has been the most talked about technology, it was originally designed to digitize and decentralize currency (Bitcoin) by using multiple ledger systems. What if there are multiple application of this technology and not just limiting it to the use of currencies. Experts over the world are now validating the use of this technology over the areas such as supply chain management, cross border trades & finance, etc.

Blockchain has gained so much attention and admiration because

  • It is decentralized, no single entity owns it
  • The data is cryptographically stored inside
  • Data tampering cannot be done inside blockchain, it is immutable
  • It is transparent

So the question raised is that can blockchain tech be used in stock trading and if yes then how will it change the dynamics of it.

Recently, SEBI has appointed a Committee on Financial Regulatory Technologies (CFRT) , exploring the possibility of using Blockchain platform in areas of post-trade settlement and fundraising. Japan is already the front runner, as it has already implemented blockchain as its core trading infrastructure at Tokyo Stock Exchange. Japanese brokerages have reportedly initiated an consortium, dedicated to adopt the process of blockchain, the founding member were Rakuten Securities, SBI Securities, Daiwa Securities and Nomura.

Back in 2015 Nasdaq, started the use of its Nasdaq Linq Blockchain ledger technology, this allows private non listed companies on the stock exchange to represent their own shares digitally. With the help of Linq and blockchain it is now possible for private companies to successfully complete and record securities transactions.

Intermediaries Minimization

A single trade between buyer and seller involve stockbrokers, depositories, bank, clearing corporation, etc. The intermediaries are for efficient functioning of the markets but they are not indispensable. Consider brokers, most of them require you to keep a minimum deposit to start an account, forget the stock trading fees, then there are other fees like option fees, etc. In a world where DIY is given so much importance such high fees is unjustified. One survey found that 9 out of 10 millennials would prefer free digital trading platform instead of a regular broker. Budget conscious investors who are unwilling or unable to buy entire shares would be happy to use blockchain powered apps which allows micro-investing.

Built in Regulation

October 24th 1929, also known as “Black Thursday”, as 13 million shares were sold in panic by the investors on NYSE. Nearly 90 years later there is no way to shape up financial sector as such incidents have had happened since then (Black Monday, Black Tuesday, etc) Perhaps blockchain will take the regulatory imperative from traders hands for goods. Users of blockchain with the help of a passkey can access the ledger remotely. Traders could save money by permitting the regulators some oversight into blockchain powered trading. National Stock Exchange (India) is also piloting on such a blockchain system for automatic KYC process.


Some investors buy shares which gives dividends on regular basis and also who wont like some additional money coming in form of dividends. A significant portion of returns is attributed to dividends. “Cash on hand is music to an investors ears.” Automation in the dividend payment process would help the companies in saving a lot of time and money. Blockchain smart contracts help in creating process of self-executing payments. This payment will release the dividends to the customers on behalf of the companies. TMX Global and Natural Gas Exchange (NGX) are testing automatic dividend payments.

Ease in post trade events/ settlement

Over a million securities in India change hand on daily basis. The size of global investing marketplace cannot be comprehended by humans. Efficient trades settle on T+0 or T+1 day. There is need for automation in the settlement of trades happening. The way to do it maybe is with the help of blockchain powered smart contracts. Smart contracts can replace human oversight which happens in settlements of trades which by the way is also very costly. As soon as some pre-requisite is fulfilled these contracts execute. For eg. If a buyer and a seller agree on a price point the trade will be executed, resulting in shorter time lag. Shorter time lag means more money is available ko keep wheeling and dealing. Nasdaq Linq Blockchain Ledger, Australian Stock Exchange, Deutsche Borse are already using such technology for after trade settlement.

Asset Management and Fund Raising

A lot can be said about a company through its fundraising. Tesla’s management has admitted that it often runs negative cash flow. Even if the public weren’t privy to such an admission, Tesla’s relatively frequent return to the fundraising table would have clued investors in. Elon Musk’s electric car company raised $270 million in capital in 2010, $451 million in 2012, and over $18 billion total between 2010 and 2018. Experts don’t expect Tesla to be cash flow positive for quite some time. But perhaps Musk and the heads of other publicly traded companies who require capital could execute fundraising even more efficiently and cost-effectively by adopting blockchain technology next time. Companies will be able to sell stock directly to public without time constraints. Blockchain’s value comes from its ability to conduct fundraising sales and agreements without any middlemen involved. It dispatches cost-effective and immediate smart contracts to execute the transactions, saving money and time without sacrificing quality. aXpire is a blockchain powered company trying to solve this problem.

Tracking Securities Lending

Security lending is not understood by a mass of population. Traders lend ETFs or a commodity within ETF to other parties in exchange of a collateral. Nearly 3.5 trillion USD worth of ETF as of December 2018 was the market share. ETFs are considered as safe return instruments but are not so safe after all, using technology to hedge the ETFs could be critical as the market worsens. Leveraged loans that led to last financial crisis, most of the ETFs are heavily invested in such leveraged loans. For ETF manager lending securities to short seller is low stress and high upside proposition. Tracking the price of ETF is important. They can be tracked through blockchain powered ETFs, and triggers the issuance of collateral if the short seller becomes over-leveraged. Nasdaq Linq Blockchain ledger is tracking the purchase of securities.

Sagar Vikmani
Team Member – Equity Research and Valuation
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Sagar on LinkedIn

Libra – Facebook’s cryptocurrency

Facebook has revealed plans for a new global digital currency, claiming it will enable billions of people around the world without a bank account to make money transfers. The digital currency is called Libra and will allow its billions of users to make financial transactions across the globe, in a move that could potentially shake up the world’s banking system.

Facebook revealed the details of its crypto currency, Libra which will let you buy things or send money to people with nearly zero fees. It released its white paper explaining Libra and the technicalities of its blockchain system before a public launch in the first half of 2020.

The effort announced with 27 partners right now ranging from Master Card to Uber and should launch sometime next year with 100 partners, as it hopes. It is a stable coin backed by a basket of actual currencies and marketable securities. Facebook will only get a single vote in its governance of the crypto currency along with its partners.

The currency will be run by the Libra association as Facebook is distancing itself from the direct management. Facebook’s involvement will be run via a new subsidiary called Calibra that handles its crypto dealings and protects users’ privacy by not mingling an individual’s payments with his/her facebook data. By this an individual’s real identity won’t be tied to his/her publically visible transactions. Calibra will also be launching a digital wallet for Libra, as a standalone IOS, android application and also as a functionality within whatsapp and messenger. Libra is the underlying technology but Calibra is likely how most people will interact with the currency. It will be the first crypto currency wallet that millions of people will have access as it takes advantage of facebook’s massive ecosystem with billions of potential users.

One of the biggest problems that the regulators will have to tackle is drug dealers and money launderers from getting their hands on Libra and using it to move money from the eyes of the law enforcement like with any crypto currency.

“The issue is that once you apply traditional regulation to tokens that are backed by money in the bank then those tokens start to look a lot like normal fiat money, after all most money we use today – credit card, apple pay, PayPal etc is just the digital representation of money that the banks promise to ultimately backup. This is the exact same thing except on a blockchain”- Techcrunch

Libra Whitepaper states that unlike previous blockchains which view the blockchain as a collection of transactions, the Libra blockchain is a single data structure that records the history of transactions and states over time. Facebook has created a whole language for writing commands on its protocol called MOVE (programming language), which is an open source prototype in anticipation of a global collaborative effort to advance this new ecosystem.  The facebook has done its homework to cherry pick the best bits and pieces of other crypto project to create Libra.

Like bitcoin there is no real identity on the blockchain; from the perspective of the block chain itself you don’t exist, only public private key pairs exist. Like hyperledger it’s permissioned (at least to start); initially the consensus structure of Libra will be dozens of organisation that will run nodes on the network, validating transactions. Like tezos it comes with on-chain governance; the only entities that can vote at the outset are founding members. Like ethereum, it makes currency programmable and in a number of ways the whitepaper defines interesting ways in which its users can interact with the core software and data structure. For example anyone can make a non-voting replica of the blockchain or run various read comments associated with objects such as smart contracts or a set of wallets defined on Libra. Crucially, Libra’s designers seem to agree with ethereum that running code should have a cause so as to all operations require payment of Libra as gas for it to run. Also like ethereum, it thinks proof of stake is the future but it is also not ready yet. Like binance’s coin it does a lot of burning. Like coda, users don’t need to hold on to the whole transaction history – states Coindesk.

Now needless to say, this is pulling a lot from the latest and greatest crypto ideas and collaborating it.

Facebook launched 2 crypto currencies, addition to Libra the project will also have a Libra investment token, which is how the stake holders (100 or so partners facebook hopes to have lined up on launch) will make money on this, as Libra itself is not supposed to fluctuate in value.

Unlike Libra a currency that will be broadly available to the public, the investment token is a security according to facebook that will be sold to a much more exclusive audience – the funding corporate members of the projects governing consortium known as the Libra association and accredited investors. While Libra will be backed by a basket of fiat currencies and government securities, interest earned on that collateral will go to holders of the investment token. As previously reported ahead of the official announcement, each of the 27 companies that facebook recruited to run validating nodes as founding members of the consortium, invested at least 10 million dollars for the privilege. The investment token is what they received as a financial reward, but that reward will only be meaningful if the network takes off – states Coindesk.

The assets in the reserve are low risk and low yield for early investors which will only materialise if the network is successful and the reserve grows to a substantial size, facebook said in one of the series of documents that supplement the Libra white paper.

This sound a lot like how an Initial Coin Offering – (ICO) has worked over the past of years, except without the expectation of price appreciation as the reward to early investors.

 We will have plenty of time and a lot of information to dig into in the coming months, but my bottom line and initial take is that the money we have today has not worked very well for all of us, furthering the gap between the rich and the poor. Libra (crypto currency) has the potential to bridge this gap but it has to bypass too many regulatory complications.

If facebook succeeds and receives cash for Libra, it and the other founding members of the Libra association could earn big dividends on the interest. If Libra gets hacked or proves unreliable lots of people around the world could lose their personal information and money. But it is clear that facebook has tried to reinvent money, we will have to wait and see if they can pull it off.

Indrajith Aditya
Team Member – Equity Research and Valuation
(M.Sc. Finance, NMIMS – Mumbai 2018-20)

Connect with Indrajith on LinkedIn

Illuminating the dark side of valuation using a Vanilla LSTM Recurrent Neural Network

One of the major problems faced by investors is to try and distinguish between “Price” and “value”, This problem arises because there are several behavioral biases as well as unethical practices that play a role while doing valuation. Sometimes, the valuation company is given a figure by their client and they do a backward calculation to keep up the pretense. The root cause for there even being a “Dark Side” of valuation is that in the DCF model, we make projected cash flows; These, projected cash flows are based on assumptions made by the valuation company. Now, based on their assumptions the valuation company could grow the cash flows at 100% or 1%, it’s perfectly acceptable as long as they have a story to back them up. Due to this ambiguity in projections, the valuation company provides their client with any valuation they want.

In this article, I have solved this problem by implementing an Artificial Recurrent Neural Network (Vanilla LSTM) to predict future cash flows. This gives us an unbiased prediction and takes away the ambiguity from a valuation that caused “The Dark Side”. The Artificial Recurrent Neural Network is able to detect interrelationships between thousands of diverse market variables and therefore is the perfect analytical tool for the financial forecast. In this article, I have used the Vanilla LSTM to forecast cash flows and arrived at a valuation of Maruti Suzuki Ltd and avoided any biases that usually pay a role while valuating a company.

What is a Recurrent Neural Network?

Recurrent Neural Network is a type of Neural Network where the output of the previous step is fed as an input of the current step. Usually in neural networks (feed-forward), the outputs and inputs are independent; But, in situations where you must predict the next word of a sentence, the previous word is essential. Recurrent Neural Networks solved this issue using hidden layers. The hidden layer remembers some of the information about a sequence. A Recurrent Neural Network remembers all the information over a period. It is considered as a powerful tool because of its ability to remember previous inputs.  An RNN remembers each information through time. It is useful in time series prediction only because of the feature to remember previous inputs as well. This is called Long Short-Term Memory (LSTM). A Vanilla LSTM is an LSTM model that has only one hidden layer of LSTM units, and an output layer used to make a prediction.

  • For forecasting Time series data Via a Vanilla LSTM Recurrent Neural Network, we are taking 10 data points or historical data. These data points are nothing but Net cash flows we arrive at while doing FCFE. Therefore, we start with getting Income statement and Profit and loss from FY 2010 to FY 2019.
  • For each of these 10 years, we take the Profit after tax add the depreciation, Changes in Debt (Current year Debt – Previous years debt) and subtract the Capital expenditure and changes in working capital (Excluding cash). Using this, we arrive at Net Cash Flows of 10 years.
  • These Net Cash Flows are used as Inputs for the Vanilla LSTM Recurrent Neural Network and are used to predict the future cash flows for the next 5 years. The code for this is written on Python by Dr. Jason Brownlee. Each time the neural network is used to predict the future net cash flows, it gives only one output (the net cash flow of the next year). This output is included in the input for the next year. This is how even though the neural network is capable of forecasting only 1 year, I have forecasted for the next 5 years. The Vanilla LSTM is as follows:
  • The terminal growth rate has been assumed at 10% because that is the rate at which the automobile sector is expected to grow as per Research Cosmo. However, to avoid any biases I have done a sensitivity analysis with terminal growth rates varying from 9% to 11%. To find the terminal value we increase the Net Cash flow value of the 5th year by the terminal growth rate and divide this by the difference between the Cost of Equity (Ke) and the terminal Growth rate.
  • These discounted cash flows are added to arrive at Equity Value of the company. This equity value is divided to arrive at Expected market price of the company.


The objective of this article was to remove the assumptions made while valuation because these assumptions create a window for ambiguity which can be used to unethically inflate the valuation of the company. In this article no assumptions were made except the terminal growth rate (taken from a Research Cosmo Report). To tackle this, I have done a sensitivity analysis with terminal growth rates varying from 9% to 11%. Using this new model, we got a valuation at 10% terminal growth rate of Rs 8,417.22. Therefore, using the method I suggested in this article, you can find the “Value” of the company and not “price”. This way you save up on the money you would have paid the valuation company and also get a unbiased valuation.

Neil Jha
Team Leader – Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Neil on LinkedIn

Neural Networks for Finance

Let’s start with, what is an Artificial Neural Network?

An Artificial Neural Network or ANN is a software program that mimics the human brain’s ability to classify patterns or to make predictions or decisions based on past experiences. The human brain relies on inputs from the five senses, while the artificial neural network uses inputs from data sets. ANN have three or more layers of neurons. The first layer of neurons, called the input layer, has one neuron for each input to the network. Each neuron in the input layer is connected to every neuron in a hidden area. The hidden area occasionally consists of more than one layer, in which case each neuron in the first hidden layer will be connected to every neuron in the second hidden layer. The last layer in the hidden layer is connected to the outer layer. The strength of the connection between the various neurons varies with the weights allocated to each of the inputs. To start with a neural network, you would typically define a function that would map any value to a value between 0 and 1. This is called a “Sigmoid” function.

This function will be run at every neuron of our network, it is useful in converting probability out of numbers. Then after you add the input, you would like to seed them to make them deterministic as you would be adding random weights. Then you would add the synapses (synapses are the connection between each neuron in one layer to another neuron in another layer). Each synapse has a random weight attached to it. As you train this neural network, the error rate will go down as the allocation of weights which were initially random would autocorrect to a more accurate value.

Selection of appropriate inputs is one of the biggest challenges while designing a neural network. Selecting which inputs will impact NIFTY 50 is a relatively tougher job than selecting inputs for which mortgage application is most likely to default. Because both inputs and target outputs used in time series forecasting are very noisy (the data has a lot of random movement that has nothing to do with the trend). The data used in rating mortgage application are usually more generalized. If there is no income in that month, the person will default. However, in a time series forecast, the NIFTY 50 could go down today and come back up tomorrow and remain up for two months. Is it possible to claim that this fall in NIFTY 50 just noise? Credit Card customer search is another domain where neural networks are used because they require very specific customers to sustain. This ideal customer should spend heavily and not use revolving credit line. Hence, per card revenue will be below per card cost which will result in a lower breakeven point. This is crucial for a bank as the incidental & incremental exceed the revenue resulting in a non-profitable business. Therefore, implementing neural networks to distribute credit cards will maximize profits.

MJ futures claims that they achieved a return of 199.2% over a 2-year period using neural networks. Neural networks can identify trends in data that humans might not notice. For example, Dean Barr and Walter Loick at LBS Capital Management using a neural network with 6 inputs. One of these inputs is the ADX, which indicates the directional movement over the previous 18 days. Two more of these inputs are the current value of the S&P 500, and the net change in the S&P 500 value from 5 days prior. It has also been observed that with a network with 3 hidden layers and 20-40-20 neurons in hidden layers was the optimized network with an accuracy of 94.08% for validation dataset.

One of the most common mistakes that most traders make is taking Standard Deviation as their volatility in all their calculations. However, this is a backwards-looking figure and not an accurate anticipation of the future. To solve this figure an implied volatility could be used or volatility using an Artificial Neural Network. This was done in Shaikh A. Hamid and Abraham Habib’s paper on “Financial Forecasting with Neural Networks”, where they discovered that volatility forecasts as per ANN are more accurate than Barone-Adesi and Whaley (BAW) for pricing American options on futures.

  1. They started by selecting appropriate inputs. Technical price data on Treasury bonds was fed along with fundamental data that also effects the market could also be added like fed rates, GDP, money supply, Inflation rate, CPI& Inter-market inputs.
  2. Then they moved on to processing the input data by scaling it between 0 to 1 and normalizing the data.
  3. Specifying a network is the next step. Therefore, they used a feed forward back propagation network. For the input, they took 11 neurons and twice as many in the second layer (In total there are 3 layers).
  4. Then they trained the network with enough historical data.
Neil Jha
Team Leader – Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Neil on LinkedIn

Artificial Intelligence in Finance: A Genesis

Artificial Intelligence (AI) started as the branch of computer science that aimed to make the computers follow logical steps to do some of the basic functions that the humans perform using common sense and has evolved from imitation, extension, augmentation and finally aims to reach human-level AI. In 1956 John McCarthy used the coinage ‘Artificial Intelligence’ and proposed a summer research project on the subject at the Dartmouth college involving scientists of psychology, mathematics, computer science and information theory; marking the beginning of AI as a science and engineering of making intelligent machines and recognizing it as a field of research. Technologies driving AI have since evolved from conventional programming using logical steps, heuristics using neural networks, machine learning using big data analytics to the ability of self-evolution. As a part of AI, machines simulate human intelligence processes like learning (acquire information and the rules for using that information), reasoning (using rules to reach approximate or definite conclusions) and self-correction to find application in expert systems, speech recognition and machine vision.

Kai Fu Lee classifies AI revolution in four waves. The first wave was the internet AI which began about 15 years ago but matured around 2012 and was mostly about using AI algorithms as recommendation engines. Some of the common applications of these are in recommending streaming videos in YouTube and likely next purchase on Amazon. Many users find internet experience getting better and becoming addictive often because of a successful AI recommending algorithm working in the background. Companies like Google, YouTube, Baidu, Amazon and Alibaba made substantial financial gains using internet AI. Another major application of internet AI is in using algorithms as editors giving real time news that is customised to a user, digital reporter and a virtual robot cop that reports fake news.

The use of neural networks, Deep learning and Machine learning in financial services has recently exponentially gone up, with there being “Robo-advisor’s”, such as “Wealthfront”. Traditional financial advisors have high fees, minimum account balances etc. However, in “Wealthfront”, they use risk assessment algorithms to ascertain risk and create highly individualized plans. The created portfolio is also continuously monitored and periodically churned to give the highest return.

The use of neural networks is also used to find cointegrated pairs for pairs trading. As after implementing a rolling Beta, the pair which was once cointegrated for a time period might not be so in the future and might not mean revert. And using machine learning on previously cointegrated pairs that stopped mean reverting.

Even Bank Of America Merril Lynch is implanting enterprise software fintech HighRadius’s Artificial Intelligence solution to speed up stock receivables reconciliation for the banks big business clients. While, some might look at Artificial Intelligence as a path to a brighter future with greater efficiency. There are others like Elon Musk, who refer to the future of Artificial Intelligence as an “immortal dictator”. This might sound hypocritical of him as he created an AI that defeated some of the best DOTA 2 players in the world. This is one of the greatest milestones that AI has ever crossed. IBM’s Deep blue computer defeated one of the best chess players in the world Garry Kasparov in 1997. And In 2016, AlphaGo defeated Lee Se-dol at the board game Go. This further pushed China to peruse AI as the board game GO which was the pride of China.These fears regarding AI are not applicable to the use in Finance; As they are not true while using “narrow AI”, such as in Financial Services.

The second AI wave was the business AI, that makes use of legacy systems data that was already being labelled and stored by traditional companies like the insurance companies covering data on accident claims and frauds, banks on loans and repayment rates, hospitals on diagnosis and survival rates etc. and structured corporate data like historic stock prices, credit card usage, mortgage defaults etc. Early instances of business AI have clustered heavily in the financial sector because it naturally lends itself to data analysis since it runs on structured information and has a clear metric that needs to be optimized. AI, therefore, is ideally applied for optimization for maximisation of the bottom line.

AI finds application in computational finance where an automated intelligent agent is applied for pattern recognition and use it to discover patterns in the stock prices for accurate prediction. Although major companies like Palantir and IBM offered big-data consultancy since 2004, major capabilities in the field emerged after the adoption of deep learning in 2013. Companies like Element AI of Canada and 4th Paradigm of China entered the competition offering algorithms that could mine the data of traditional companies and organizations to improve fraud detection, make smarter trades and uncover inefficiencies in supply chains to use AI for cost savings and profit maximisation.

Whilst there are several credit cards and mobile payment applications that are popular and prevalent, their core services limit spending. An AI powered micro finance app called Smart Finance in China relies exclusively on an algorithm to make millions of small loans. The deep learning algorithm uses a wide range of information including offbeat information like the speed at which the user types his date of birth or the amount of battery power left on the phone etc. of the user from his mobile to predict the repayment potential and authorise loans. In late 2017 the company was making more than 2 million loans per month with default rates in low single digits which outperform many traditional brick-and-mortar banks because it targeted a large user base of potential micro-finance seekers that was ignored by the traditional banking sector – the young and the migrant workers.

The other major applications in the second wave of AI are the use of algorithms for expert medical diagnostics and for legal advice to the judges based on deep learning. The 3rd AI wave is the perception AI where the distinction between the real and the virtual world are being increasingly blurred and digital assistants and Augmented as well as Virtual Reality applications are being introduced. The 4th AI wave is about autonomous vehicles like self-driving cars and autonomous robots to be employed in workplaces. The 3rd and the 4th waves are still in emerging phases but are fast expanding.

Neil Jha
Team Leader – Fintech
(M.Sc. Finance, NMIMS – Mumbai. Batch 2018-20)

Connect with Neil on LinkedIn