as part of the yagi clean season organize order to explode so it is in fintech so we have sai darmayohan executive director to share his knowledge and ideas with us hello yes you're audible yeah so let me uh just share my screen and then i'll just i can get started right yes so in the meantime let me just give you some background about myself so i was um i'm i was born in colombo and then um just a second and i did my schooling actually in india chennai before actually moving to the us where i did my undergraduate and since my since graduating from university i based i have been in the financial services industry working for the mostly um most of the big banks u.s banks specifically and then i have just recently moved into the fintech space but i focus specifically on fintech is quite a wide area actually i have the slides i'm just trying to get them sorted out so that i can share and just give you a minute so can you see the powerpoint slide now yes can see okay so great so uh yeah as i just mentioned um so i've been in the financial services space for a very long time since graduating from college it's 15 years and i started off um in the banks in securities uh trading and um over time i've evolved more into a client services role and now i moved into fintech where i'm doing something similar then i'll also touch upon why fintech actually is a different and interesting opportunity in the financial services space so fintech is actually quite broad if you look at what forbes defines as spin tag uh it covers personal finance uh wall street which is basically what i work in uh crypto that's relatively new investment is also really is a related field to wall street it's basically how pension funds asset managers uh invest their money and as well as retail investment investors as well that's kind of picking up um significantly especially in the u.s where you have uh robin hood and things like that uh lending is more uh it's peer to peer lending where traditionally people have had to go to the banks to borrow money now they can actually have there are a lot of fintech applications or providers that lend money though at a much higher interest rates and then payments if you look at payments payments is probably the one of the biggest beneficiaries of fintech until recently you needed a credit card um or cash to pay and now you pretty much can pay with most technology devices like from your iphone watches and stuff like that as well as making online payments all of those have become so much easier with fintech so that's probably the area that's seen the biggest growth and uh last but not the least is real estate it's i was quite surprised by it it actually has 14 distribution in fintech um so interesting things about real estate is if you look at the uk market it is actually very transparent and almost digital so back in the day you would have estate agents basically listing properties and taking three to five percent commission for buying and selling houses and uh nowadays the st agent commission is roughly around one percent uh in in the uk so um it's largely driven by uh technology advances in transparency in terms of um buying and selling houses it's made it so much more transparent less need for value add from an estate agent this is quite different each market is quite different i think u.s is also quite relatively transparent i'm not sure how transparent it is but um some opaque markets like canada where state agents still get paid three to five percent for buying and selling houses so there's a lot of wastage in terms of uh um for people in terms of having to pay such high fees to make transactions and fintech has definitely made changes in some countries regulations will play an impact in terms of how much they can actually penetrate each market it applies to most fintechs from personal finance wall street crypto is not highly regulated but still um governments are starting to go regulate them investing is also very highly regulated lending and payments so when i what i'm going to do is since my background is more in the securities trading space i'm going to talk about up what the wall street give you some background about wall street because investment banking i'm not sure how familiar you are about it so what i'll do is i'll give you a high level background about investment banking what are the different units and business units and how specifically i'll then touch upon sales and trading and investment to some extent and how fintech can actually fintech has already made some um significant changes in the landscape and how they can affect the future as well so there are five main areas in investment banking one is underwriting so this is where corporates and governments raise money so if let's say the government wants to build a bridge and it needs money for that then obviously it goes to the um one way to do that is basically sell bonds uh raise money and um so that it can invest in certain projects similarly for corporates as well so corporates can issue stock or corporates can issue bonds so underwriting is that is one business a related business is mergers and acquisitions here what when corporates want to but one company wants to buy another company like uh uh or two companies once want to merge example the ba british highways in liberia corporate merger then each company will have its own mma bank advisory that advises on how to structure it and how how much to pay and how much shares each company gets similarly you have corporate acquisition so recent one is salesforce buying slack so obviously sales for such m a bank and advisory bank and slack will have its own advisory bank where the banks will advise the companies on how to structure the acquisition and how the payments are made and whether you pay in stock cash all of that details and the last one is the leverage buyout so here is when a company usually a private equity firm um takes in a lot of loan basically um leverage to basically buy another company so blocks when blackstone bought hilton it basically was a leverage by art where they actually borrowed money to buy another company and banks usually provide funding or help them arrange funding for these kind of deals the next one is the third one is sales and trading which is basically what i work in so here it is mostly buying and selling of securities stocks bonds foreign exchange commodities oil wheat everything um and now slowly getting moving into crypto as well where you work as a banks act as a facilitator between institutional clients when i say institutional clients it's basically asset managers pension funds large institutions so banks usually don't deal directly with retail investors like you and me if we want to buy some stock we basically go to a small brokerage firm that'll basically that could potentially send that let's say you want to buy vodafone or google you would basically go on to a brokerage firm log and log into a brokerage firm and place an order let's say 400 shares and that will be routed the the brokerage firm will probably send it to a bank or some kind of a liquidity provider that will provide a price or credit in the market but banks mainly focus on asset managers and institutional clients the next one is asset management so here banks also provide asset management service so in addition to providing services for asset management banks also have an asset management division not all banks have it but most banks tend to have an asset management division but they actually manage the funds uh for pension funds or other asset managers a more complex asset might be managed by banks and then the last one is research because to facilitate sales and trading and asset management you actually need to have in-depth research of companies on in which you're investing or planning to invest in so researchers quite a significant investment for most banks and that's also come to a lot of reason for a lot of regulatory that regulators have been trying or made some inroads into separate research from sales and trading and asset management so there is no banks don't get a head start ahead of the clients to which we sell research so bank is a highly regulated uh industry so um obviously these have a lot of impact for companies as well as people investing in them so regulation is quite key there have been a lot of huge scandals um in the past so um regulation is quite tight especially in the u.s and europe in development in
double emerging markets uh it's the enforcement of regulation is not so it's not very strong so you do tend to have concerns in terms of investing over there transparency and things like that now going into the sales and trading business um it's so what i want to do is give you an overview of the business before we actually go into the technology aspect of it so it's mainly divided into three parts you have the front office which is basically what i would call the revenue generating um unit of the best of sales and trading so here all the strategy of development marketing sales client engagement product development pricing trading deal negotiation all of that comes in there as front office and then you have middle office what you call middle offices here risk management compliance customer onboarding so this is considered the middle office of a let's say a bank and the back office is basically human resources id infrastructure so it is everywhere now so when i say id infrastructure it's basically the networks the pcs their phone infrastructure as well as the servers and the data centers and things like that and data processing accounting clearing and settlement so all of this is considered back office the reason this is quite important is because sales like sales and trading has actually had a lot of technology investment in the past so securities trading was one of the early adopters of digital trans transformation in the 1960s so what i mean by this is even around 60 years ago you could actually place an electronic order through a computer to buy and sell stocks so that's quite unique because not many industries were adapted the use of technology back then and investment banks continued to be a technological innovation leading technological innovation in the 80s and 90s so most banks or some of the major banks actually had their own programming languages so if you look at morgan stanley they had something called a plus which became eventually became kdb right now it's very widely used in the financial industry and it's also being used in other industries it's basically a flat file time series database but it emerged out of a bank in morgan stanley and then goldman sachs also has their own programming language spam so again this was developed in the 90s mainly to meet the needs they didn't feel the existing languages programming languages at that time met the needs of the financial services industry so in terms of processing very large time series data and being able to roll things out quickly so they develop their own programming languages and were kind of techno innovators in that space but what happened over time is let's say in the around 2000s or mid between 2005 2010 uh regulation and when we had when we had the issues with um the financial crash in 2008 the there was a huge um obviously uh banks were banks didn't have a good pr and that's uh around that time so um regulators obviously started focusing a lot more into what the banks do and trying to place more onerous regulations financial capital requirements on the banks so this slowly led to the decrease in investment in technology and it has limited and the focus mainly has been in terms of keeping up with the regulation and making sure banks don't get a bad pr again so but obviously that hasn't always happened banks have had issues even in the last decade but for most part a lot of investment has actually gone into the regulation aspect of it and that itself is a separate area called red tech where there are vendors who actually provide technology that banks can now use to kind of be better smarter in terms of adhering to the regulations so that's a very growing field rig deck i'm not going to cover that in detail but i'll probably just give you and speak about it when i can so what that's happened is over the last 10 to 12 years banks haven't really adapted the latest technology in terms of they haven't embraced open source code and migration to the cloud so most of the technology banks have are probably around 15 years old which large servers and backend operations which are not the most efficient and also banks value their ips quite a lot so moving to the open source was kind of looked down upon until recently so slowly they're beginning to embrace the move to the open source in cloud uh so what's been driving that recent change is basically the squeeze in profitability and competition from fintech so this is the next stage of transformation which is now happening is basically bands are moving more beginning to embrace open source move to the cloud and adapt to the cloud and compete with some of the more nimble fintech companies in the space so how they're doing that is obviously they need to focus on regulation so what they came with an innovative approach is basically have accelerators or fintech accelerators within each bank so for example citigroup has an accelerator in belfast and um in tel aviv israel so where they have they just focus on innovation you have various teams they come up with ideas and focus on innovation and then when they think it's ready they basically incorporate that into the bank's business so most banks actually now have that accelerators and funding for innovation as a separate standalone unit rather than being within the bank and being subject to funding cuts and things like that so it's uh it is an interesting development in the space so now going into a little bit more detail about what the sales and trading actually is is let's say you have a lot of your pension contribution that you go that you put basically goes into an investment manager and the investment manager basically decides how um they need to invest that money so uh typically a let's say 20 years ago what they would do is they would basically read through all the research reports that the bank's research business uh generates and then decide okay which company i want to invest in and then they would call the bank and say okay i want to buy x shares of this company and then the bank would basically give them a price and usually a reasonable markup in terms of what the market is trading and then um the the portfolio manager will basically buy it and if they need to sell another some seller stock then they would go through a similar process for selling it but what's happened over the time is over the last 20 years especially as the whole thing has basically become digital electronification of investment and trading that's how we call it so electronification of investment is basically the portfolio construction process so in terms of analysis of stocks the risk management and deciding how much which stocks to invest in and which how much to invest in each stock is all basically now you have software that basically helps with this process you have software that can analyze research reports not just research reports but also social media posts and instagrams and facebook and all kind of social media to basically determine which stock to buy based on sentiment and then you have tools that basically help with risk management in terms of how much how to diversify you don't want too much of bank stocks let's say if there's another financial crash then obviously your portfolio will lose a lot of value or if it doesn't so you need to be well diversified that's called risk management and then we have pre-trade analytics that's basically what it means is uh telling how much it's actually costing going to cost you to buy that much amount of stock let's say if you want to buy um 100 million dollars worth of tesla how much would it actually cost it's not going to be 100 million dollars but for 100 million dollars of stock you'll probably end up paying a little bit more in terms of fees because if the market knows that there's a large buyer then obviously there are going to um increase the price so all of that is taken into account in terms of prepaid analytics so you need to take out of that into account when you construct a portfolio so once you construct your portfolio then you need to execute it so that basically you need to go buy some stock or then you might have some stock which you need to sell so you don't want that stocking in your portfolio so you need to sell you basically have buy and sell so traditionally you would pick up on the phone call your relationship manager at the bank and say okay i want to buy x shares of this xyz and then i want to sell x amount of shares and another company or a few lists of companies but all of that is now electronic so you basically create a list and you basically send an electronic order you put it into an order management and you send it to the broker and say okay this is what i need to buy i need to buy x shares of um company abc and sell wire shares of another company or a list of companies and then what this bank actually does is it finds the most efficient way to actually buy and sell that amount of shares so just to give you an um idea of what how much banks deal typically on in london on reasonable uh investment bank would basically do around five to six billion dollars worth of uh execution every single day so all of this is done fully electronically without any involvement for human being so that's where the execution is it's a quite complex process the execution and i'll i think i also have a slide that will later show um the execution stack but what that involves is basically finding out it's something called algorithmic trading so you take an order and then you find out how best to trade it and then sometimes you might actually do principal facilitation so banks through their own money they might actually buy it from the client and then you also have surveillance surveillance is quite new so this is this has grown especially on the back of a lot of front running where let's say some a client wants to buy um 1 million shares of tesla somebody in the bank notices that the client is actually buying tesla and they actually call their relative and say you know what you should buy tesla now because there's another buyer a huge buyer then you can if you buy it now in the evening you can sell it back to him so that those kind of abuse and front running used to happen quite a bit in this industry at least in the late 90s and early 2000s and that's why you have surveillance right now you need to have technology to make sure um like um any sales trader or anybody who actually has information about clients can't actually do that kind of front running so there's a lot of technology investment in that space um your phones are monitored um as well as you're not allowed to use phones inside the bar like do while you're working and things like that so it's a lot of technology investment that again comes under red tech so there are vendors who actually provide surveillance software that banks can actually purchase and use on pcs and phones so once the order has been placed and executed then you have to clear and settle so your clearing and settlement is basically telling um it's all electronic so nobody gets a paper saying you own x shares or why shares it's all centralized so the central authority needs to be notified that the buyer the seller has sold the stock x amount of stock and then the buyer has actually bought it and then they all get notified that it it's actually been accepted and then you have a performance attribution so what we call is transaction cost analysis and market impact which is basically driven out of market impact models so when let's say the portfolio manager wanted to buy a 100 million dollars worth of tesla but it ended up costing 102 million so basically it's basically two percent markup he ended up paying so it's usually not as big the markup usually tends to be around 100 of a percent so it's usually considered in their own basis points so that's basically what uh it costs for the portfolio managed to buy that stock and that's something that needs to be accounted when they want to sell it as well because every time they sell it they also incur that cost as well as finding out if they actually did that better is there a better way to actually trade it did they spook the market uh did someone tell the market that this portfolio manager is actually buying a large amount of shares so the market actually price moved before they could actually buy it so all of that is called performance attribution and transaction cost analysis here again there are a lot of vendors who actually provide these tools so they're quite sophisticated and there is still a lot of room for improvement in the space so which i will cover um in a bit so first i want to go over um the algorithmic trading what it actually means so until let's say 80s and 90s if you actually want if a portfolio manager wanted to buy a stock you would call the relationship manager the manager would basically let's say enter it in the computer and then um basically call someone in the exchange and say okay my client wants to buy x shares of a list of companies and wants to sell another amount of let's say shares in another list of companies so it was a quite an intensive process where it was very manual uh prone to abuse obviously people are having information about somebody else which they can abuse and obviously it was quite inefficient expensive and that's when the algorithmic trading came about as basically an efficient way to actually buy and sell stocks without inefficiencies in the market so you want to take away all the inefficiencies the market impact the front running all of that stuff so they were built on the back of to do to be able to do that you actually need all of you need to have data in electronic form so everything that's happening in the stock market or needs to be available as data as well as for let's say for you to make informed decisions of when to buy what's the right price to buy you need to have historic market data as well as nowadays people use sentiment analysis like in terms of live twitter feeds to kind of figure out what's happening in these companies have they made any announcements if they made any announcement is it positive is it a good announcement is it a bad announcement if it is good then you have to buy it then you want to buy it soon whereas if it is a bad announcement and you're a buyer you probably want to buy it more slowly so that the price can come down so all of that data there's a lot of data processing that needs to be happening in the trading system and then all of this data goes into models so you have market impact models you need to understand the market structure where you're working so the european market structure is very different from the u.s u.s market structure due to regulations so the algorithmic system trading system needs to know about each market how each market is structured uh what are the regulations what are the restrictions in algorithmic trading then obviously you use that with statistical models right now people use decision trees reinforcement learning uh graph networks um and obviously people they're always there's always innovation in terms of finding newer ways to kind of execute and improve your efficiency of trading so that's the models so once you have the models then you actually need to send that order out of market so once a model has decided okay though i want to buy one million shares of this sum right now at this time i want to only buy 1000 and so then it basically creates an order for 1000 that needs to go out the market electronically and to the exchange so again there are several exchanges where you can buy so what we have a technology that says that determines where do you actually go to buy it and once execution is done then there's a feedback loop because then new market data is generated so you you determine how the algorithmic system is actually doing versus your objective function your performance benchmarks that your clients have provided and then you use that again as input data input to the model and then so it's a feedback loop that basically until you fully bought and sold all the shares the client has asked you to do so this kind of process usually trades around five to six billion us dollars every single day at each at some of the big banks so it's quite a lot of automation and risk that's actually happening here and if you look at 2010 there was something called a flash crash when one of these algorithmic system trading system basically went out of control basically and ended up moving the market too much and that brought in a new set of regulations to kind of oversight on the algorithmic trading systems limits on what it can do and what algorithmic trading system shouldn't do and so that it doesn't affect the market market stability so again there was a lot of innovation in the space until around that time and then once the regulations came in uh innovations kind of stalled and it almost kind of took a step back to kind of put limits on what the algorithms can do a lot of machine learning research that was happening had to be pulled back simply because uh the regulators weren't comfortable with machine learning at that time uh to be because it was like a black box if we and if we couldn't predictably say what it would do regulators didn't want them trading in the market and affecting the stability so there was some level of setback in the space about a decade ago but slowly people are gaining comfort and things are starting to move back again in the right direction in terms of innovation so this is something i got from i have the source on the bottom right corner so literally this website does have a good detail a description of the ecosystem so this is basically what it takes to kind of have an algorithmic trading system and this i think is um they've also identified this as pieces of software that can be in aws so these are things that are now in the cloud typically most banks don't have it in the cloud so now as banks are moving and embracing cloud they basically are moving certain pieces of information certain pieces of the infrastructure into the cloud so like market data and this is the algorithmic engine which i described as you can see it's a very small piece in the entire puzzle so here what it means is basically client is sending an order and then order goes into a cloud system which basically sends it into the the internal trading system and that basically uses the algo engine to kind of determine how to trade the algo engine uses market data and then makes a decision and then its decision is sent to the venue gateway which basically goes back outside of the aos into the form infrastructure and then it goes into the exchanges um various different types of exchanges to actually do the trade and once they come back then you go down to the post trade where you actually report to the regulator not regulator finra is basically in the u.s where you report that you've actually made a trade somebody ex client x has bought a certain amount of shares and client y has sold a certain amount of shares so all of that happens in the electronic infrastructure at the bottom so this is an entire ecosystem that is needed to actually do um just the electronic trading and notifying the client that they've actually bought and sold something and there's been a lot of innovation in the space this is considered front office infrastructure and there has in the early 2000s there was a lot of investment uh then the late 2000s like um around 2010 a lot of um investment on the regulatory aspect and now more slowly um this infrastructure is being migrated to the use more modern technologies so what are the challenges and opportunities over here so challenges that we've already touched upon it has very high regulatory cost so the regulators in the u.s and europe
have made it extremely hard for a newcomer or someone to basically just spin up a computer write piece of code connect to the exchange and basically buy and sell stocks so that's made hard for fintech to actually compete in the space there are very few fintechs you can actually compete and the fintechs that are here in the space also need some kind of backing from banks without which you don't have the capital to actually um if something goes wrong if your algorithm goes wrong you could end up being stuck for like hundreds of millions of dollars so you need to have that level of capital uh put into the regulator as well as the exchanges to be able to trade so there needs to be it is quite a capital intensive uh business so what are the implications of that so when you have let's say for every bank you have to put 100 200 million into a regulator and have it stuck there that's money that's actually not making any return so it's just sitting there that you can't actually invest so banks need to be sure that actually that is actually worthwhile so any if you're going to have 100 million dollars sitting at a regulator without any return then um it should be of good use like that you need to be able to get good revenue out of for the for the business that it is supporting so that's what that makes it harder because now the business needs to be a lot more profitable so that you can make up for the loss revenue from capital that's sitting somewhere that you can't actually use and as well as compliance so this is prevention of front-running market abuse so market abuse is basically impacting the market front running is basically somebody basically um taking information about a client and then using that to make money themselves so these there's a huge cost to actually comply with these uh capital and compliance requirements and there are structural costs as well so what you've seen is uh banks have invested a lot in the front office aspect like the trading aspect of it so they have good training infrastructure good mathematical models scientific models data science has been quite intensively used in terms of the trading aspect but what's lacking is the middle and back office so this is basically when you go to the previous slide it's this bottom part of it actually has very little like infrastructure very little investments recently so there are areas in this process which are extremely manual so middle and back office has a lot of manual labor in it so most banks will have their midland back office for let's say supporting the uk uh european markets you let's say some banks have their mid-level back office in birmingham belfast uh berlin that are considered as low-cost areas where you can actually have lots of manpower to kind of do this but obviously that's not scalable and it actually is also what we've seen is there is revenue compression when i say as technology becomes more and more advanced and technology is more democratized you have smaller firms that are slowly creeping in and offering services for cheaper so as so banks have to reduce their profit margins which basically means they need to be a lot more efficient in terms of overhead so which means a lot of the middle and back office now have been identified as areas of where cost reduction is essential rather than an option and the way they are going about it is using vendors so this opening opportunities for vendors to kind of step in and provide technology and services that banks can actually use so why is it still very attractive to be in the financial services industry it's simply because it's expected to grow so much right now most of the markets that's actually generating money is basically the u.s europe to a smaller extent and hong kong even china hasn't actually scratched the surface and india has is nowhere near it so the revenue potential in terms of expanding this infrastructure globally and being able to trade globally is around roughly three times it's expected to grow around three times the next 15 years but obviously that also means investment you need to invest in each of these different markets you need to understand the regulations of these markets you need to understand the overhead cost overhead of these markets and moving into emerging markets is not very straightforward because the same things that i mentioned like front running market abuse they're quite widespread in less developed markets and it's a huge risk for global banks being exposed to that if because if they are found if one of their employees in a developing market is actually found to have front on a client's order then it's a huge pr disaster for global banks especially after what they've been through so they trend to outsource use third party vendors in areas markets where there are higher risk so high risk of these um or where there is no strong enforcement of regulation there is opportunities for technology providers or local providers what they call brokerages to kind of provide services to the big banks and as these markets mature they're also becoming more homogeneous like uh let's say 10 years ago the hong kong market was very different to how uh it was how very different to how europe and the u.s was regulated but now they're quite similar including singapore so they've all as they mature they you have you can have basically a single system or a single process that you can deploy in these different markets and uh basically then it's base it's just servicing the clients and um why is there more and more money to invest because there's also the rise of middle class globally so as the middle class increases people have more money to invest they're contributing more into their pensions they're contributing into let's say investment portfolios or investment managers where they think they can get more return so there is a lot more money to invest especially in these times where interest rates in developed like in uk the interest rate is almost 0.1 percent so there is no point keeping that money in the bank uh you probably are better off trying to invest it somewhere um where you can get good returns traditionally people go into real estate but now obviously real estate has great so uh investing in equities and cryptocurrency as well as um startups so people are as the middle class grows they have more money to actually invest so that's definitely an opportunity and then disruptive innovation obviously as we said banks are not only looking uh banks are not only looking to big build technology themselves but also looking for vendors who can actually provide innovative solutions the problems they have and so what i'll do is in this slide i'll just go through some of the areas where i think there's most potential for innovation like providing solutions in the near term so this chart on the right is basically something by ernst young about capital markets landscape what are the technologies that actually can have the greatest impact in the financial services industry so what some of the big ones are advanced analytics artificial intelligence artificial intelligence is more on the regulatory side so this is reg tech again red tech if you're interested in rectech artificial intelligence to identify uh erroneous trades market abuse front running finding intelligent ways to efficient ways to identify that so that's actually very popular it's gaining in terms of the direct tech space but it's not an easy space to actually get into because you need to understand the market you actually need to understand um the regulations so you need to some extent you almost need to have like a legal background to understand the rules interpret it and then build the technology to actually solve it so let's say if you want to look at what are the opportunities in that space in sri lanka you would basically want to look at the columbus stock exchange look at the regulatory manual and see basically read through that understand that fully and then speak to compliance regulators to see where the opportunities are and then develop the technologies to kind of solve it but once you solve it in one space has as the markets converge and become more homogeneous it's an opportunity to basically sell that to other markets so that's definitely an opportunity in red tech but it is not as simple as basically building software and selling it but where you can actually add values is analytics so advanced analytics is something where there's always need to be much smarter in terms of how you trade how you invest so one of the areas is sentiment analysis so sentiment analysis is basically reading through live news that's coming in from [Music] instagram or any other apis and basically figuring out how much of a positive or negative you use it is for a particular company so there are a lot of research projects academics who are actually working on it in terms of sentiment analysis and that again ties in with predictive analysis here you're basically trying to take in the sentiment analysis you've done or any other kind of mathematical statistical model you might have and then say predict in terms of how is the stock price going to move in the next five minutes 10 minutes 30 minutes and or even 2-3 hours and these are quite crucial in terms of being able to determine how you price your stock so that's how there are currently a lot of liquidity providers in the market so what they do is they have very strong predictive analytics in terms of short-term models that can predict the stock price in between the next 30 seconds to let's say two hours and they use these predictive models to kind of provide liquidity to both clients and the banks as well so that's another area where there is there's been a lot of progress but i think areas there can be a lot more progress especially in uh emerging markets because predictive analytics work uh works quite well in the u.s in u.s and europe but
once you start going to emerging markets like middle east even asia asia pacific then these predictive analytics is not really very strong people don't understand the marketplace the complexity of it so um they haven't really invested time and effort into those markets so that's definitely an opportunity where for those fringe markets you can actually develop forecasting mode price forecasting models and it's not too difficult because it's usually you can use available market data that should be available for free of for relatively low cost you can basically ask the exchange uh to provide you a flat file with let's say two years worth of market data or one you start with even few months worth of market data do some analysis and write some models test it out and then you sell those once you have a product then you can basically sell it by the banks or even asset managers and then understanding the cost of trading again this is about um how much is it actually costing you to trade in these different markets so it takes it it involves a good deal understanding of the market structure as well as being able to tell visualize what you actually see so it requires visual data visualization and being able to tell a story of what's happening in terms of your cost of trading this requires both um transaction costs models so price impact cost models which basically specify how much is the expect theoretical expected impact and how much is your realized impact so at a very high level it tells you are you doing better than expectation or you actually need to do a better job in how you actually execute your investment strategy and then you're in the middle office side you have data visualization for compliance and regulatory requirements so this is basically red tech which i basically touched upon so here uh data visualization as well as artificial intelligence are going to be quite key in terms of driving this red tech being it's not a revenue generating field so banks don't tend to spend a lot of time and money on it they tend to just use vendor products because there is no ip in it it's just like a check box exercise that banks need to do so that the regulator is happy with them so since it's not revenue generating it's there there isn't a lot of investments of in terms of proprietary investment internet so again here understanding the local regulations and providing software that banks can actually use to ensure they are compliant with the local regulations of different countries and markets is definitely an opportunity and then you have the digital transformation of the business so what here i mean is the trading aspect of it has actually been um quite efficient in terms of being digital being electronification of the trading infrastructure but um client engagement is actually not very intelligent or very sophisticated so you still have sales people using salesforce which is actually a more generic sales tool rather than something specifically for the financial services industry so banks as well as there are now software companies that actually focusing on um basically sales was equivalent for the sales and trading business where you can actually analyze the different stocks what kind of stocks is a client actually buying and what are the related stocks let's say you could potentially sell to the particular class to that particular client if the client bought a stock a few months ago and the price has actually gone up then basically maybe you just make a call to the client and say hey you actually want to get out of that position you've actually made a good amount of money and our research suggests that that stock might not be moving up anymore so it might be a good time for you to get out of that position so those kind of intelligent client engagement driven by data as well as [Music] transaction history is actually quite key and there's a lot of investment both internally as well as from vendors in terms of the product development in the space and back office is basically crime profitability so this is um because your clients tend to trade a lot in hundreds of millions of dollars in each trade so how much are you actually making because if you're charging them one hundredth of a percent in terms of commission um does that cover your cost in terms of resources they're utilizing do they take up a lot of research and uh hand-holding to actually get that revenue so that's the client p l is something that banks haven't really focused on lately largely simply because the margins were quite large but as the margins tighten in this business um each client needs to be analyzed for the profitability and making sure we don't banks don't put in more resources into that particular client than they should or than they're able to actually make so that is actually quite key and here again ai and data-driven software is being developed by a few companies and um it's not totally um effective but there is progress in the space in terms of understanding uh crime profit profitability so this is a very high level of over a high level overview of the financial services or sales and trading and the fintech association in this particular space um obviously you can go into a lot more details in each of these components like red tech itself is um can be a separate session so but i hope this is giving you a high level overview and then that you can actually use to kind of do your research and figure out if there is anything that's of interest to you then um yeah you can kind of research and investigate more into it so if you have any questions please post it yes available in the fintech thank you so much thank you
2022-08-16