What鈥檚 the background behind Trading Apps?
I have now worked for a total of 23 years in securities finance IT (Lehman Brothers, Credit Suisse, and Deutsche Bank). In 1996, I was the CEO of RTFM Ltd, which developed a securities finance system called Martini, which I then sold to SunGard in 2004. Many institutions still use Martini or its rebranded version, which has been known as Apex since 2008. Martini was an enterprise proven, global multi-asset system that was designed to do what I now consider as the core functions in securities finance鈥攖rade entry, position keeping, and feeding to settlement systems, and it had a very similar functionality to the other major vendor systems: Global One, 4sight, Anvil, and Loanet.
After selling the company, I had three years away from the securities finance business, working, with the intention of taking my technology skills into other industries with similar characteristics, but missed the dynamic pace of the financial services industry, when compared to other traditional industries.
I came back into the market in 2008, working on a part time basis for IT consulting firm Rule Financial, assisting the securities finance practice and also as board director. As a consultant I found it much easier to engage with my previous customers and discuss what was happening in the markets.
Having had three years off, I expected to find the technology and business had progressed in what they were looking for. In hindsight, I probably shouldn鈥檛 have been surprised to find that very little had changed!
When talking to the front office, they were concerned about the same problems that they had eight years prior (in many cases 20 years prior). People who had older systems hadn鈥檛 taken the opportunity to build out or replace them, which struck me as interesting.
In 2009, my ex-colleague at Martini, Jeff Lloyd, who is now the chief technology officer of Trading Apps, contacted me.
Jeff is a true 鈥榞arage engineer鈥�, and always tinkering with software. In his spare time and using his experience installing Martini for 10 years, he had written a set of software to build applications (at Trading Apps, we call it GLASS). I wasn鈥檛 attracted straightaway, but after he went to work with a German investment bank and built the desk a new position screen using his software in such a short time frame, I became interested. At this point, I helped him to commercialise the product.
Three months later, he contacted me again to say that the bank would like to buy another IT solution, this time to create and send daily push lists. It was at this point that I got the idea of creating apps that solve specific problems and that can sit on top of existing systems. The front office can get the tools they need without the usual method of ripping out an old system and installing a new system (which ends up at the same place鈥攅xcept years later and millions poorer).
So what we needed was ideas for apps from real market experts. Fortunately, at precisely this time I was contacted by Jean-Paul Musicco, who had been global head of securities finance at Deutsche Bank and had just left his position as head of desk finance at SAC Capital.
So we formed an equal partnership with Jeff providing the IT skills, Jean-Paul the business ideas, and me translating between IT and business speak in the middle. Roy Zimmerhansl also joined us for a time to provide further exceptional business knowledge.
That was the foundation of Trading Apps, which officially launched in 2011. It is a very different beast to Martini, in which you either bought the whole system or nothing al all. With Trading Apps, on the other hand, you can pick which functional gaps you would like filled by our apps. We have apps for investment banks, agent lenders and hedge funds.
By providing solutions that can sit on top of existing systems we offer a much more cost effective approach to providing the front office with effective tools.
The other vendors and in-house IT staff tend to want to reinvent the wheel. They take out old technology, and put in a new system that will give you some slightly better technology but often pretty much the same functionality. This is always offered as the 鈥榮tepping stone鈥� to providing the tools being asked for by the desk. Unfortunately, years later and millions in the red, the new tools usually never materialise because other priorities take over.
One of the biggest changes I have seen over the last 20 years has been the shift from a small team of developers often sitting on the trading desk and directly accountable to that desk, to a centralised IT structure where the IT department might not report to the investment bank management at all, but instead to a group level operational report. This has led to a focus on operational and compliance issues to the detriment of the revenue producers. It has also led to a massive increase in red tape.
Trading Apps鈥檚 approach is to say, leave the core systems as they are and if necessary, make peripheral changes to hardware to increase performance. There may have been tens of millions of dollars put into the existing system, so there is no point removing it. Instead, we will put the front office functionality you need on top of it.
This is exactly what we did at a major agent lender. They had Global One as their core processing system and various sources of third party data and market information, only accessible via different screens. We wrote a suite of apps to sit on top of Global One, which gave a completely new way of trading. Our solution has proved to be much more automated and more focussed on generating revenue out of their current portfolio with their counterparts. That has proved to be very successful, with the idea of putting Trading Apps on top of existing infrastructures definitely resonating with the front office guys, who ultimately have to foot the bill for all of the IT costs, so they are delighted to see a tangible return on their investment as well as in a much shorter time frame to production.
Whether this is the start of the revolution to re-vitalise the front office, we鈥檒l have to wait and see.
Given that sales is based on maintaining relationships, do you follow IT individuals as they move from bank to bank?
We do get referrals from people who have moved on, but we are not aggressive from a traditional IT salespersons鈥� point of view. We have more than enough interest from people who have heard of us through word of mouth. If we make our customers happy, they spread the word. We currently sell to three types of clients: agent lenders, investment banks and hedge funds. Between those three we have a large potential customer base. In the future, we will also be building solutions for beneficial owners such as asset managers or pension funds.
Especially from the agent lender and the investment banking customer base, the business can be divided many different ways depending on the structure of how a firm manages their clients, regional reporting lines, not too mention their products. International securities finance desks can include both cash and swap products, which can include a European/UK trading location as well as various Asian trading hubs. Then you have the high volume US equity business that has totally different demands on the systems. Fixed income financing desks can also include corporates, treasury, and collateral management, so where there are probably around 200 reasonable size financial institutions active in securities finance around the world, you can then multiple that by five or six times and they can all be different customers of ours. From the hedge fund perspective, again there are hundreds of reasonably sized institutions that consider desk finance in one way or another.
We are talking to a bank at the moment where we are in discussions with the US equity desk, the London fixed income desk and the agent lending desk all independently and for different apps. Unlike Martini, which was often considered as a 鈥榮trategic鈥� purchase requiring many layers of authorisation, our apps have a pay back for the initial investment in less than a year, sometimes in as little as three months, so they can be purchased as a 鈥榯actical鈥� solution that provide an immediate impact to your current fiscal year.
You mentioned that a front office had problems with position screens and triparty. What were other specifics that they wanted to change?
The big thing we鈥檙e moving to now is pre-trade, and this is resonating well with the market. Pre-trade means all the negotiating activity that happens before a trade is booked.
Martini started from trade entry, and in our arrogance had many checks and balances in the trade entry forbidding this and that. It then dawned on me that the trader had already agreed the trade with the other side, so what was the point of all this validation after the trade had been agreed?
What would they do if they found out they couldn鈥檛 do it? They would have to go back to the counterpart, eat humble pie and say they can鈥檛 book this trade, which is not ideal.
One of the more significant changes that has occurred in the industry over the past ten years, is the ability to handle mass transactions via an AutoBorrow process for the general collateral priced transactions. EquiLend鈥檚 AutoBorrow process handles an awful lot of transactions daily that are now booked automatically and require minimal trader interaction. We are actually working with EquiLend to build adaptors into their T2O product for both the borrower and lender. This will give seamless integration of T2O into the banks own systems, which has been the biggest barrier to using T2O since its launch some years ago for negotiating mid-tier and hard to borrow securities.
Outside of AutoBorrow though, the market works on Bloomberg and email. The traders are emailing transaction requests all day; sometimes sending Excel lists attached to them.
If you take, for example, an agent lender such as BNY Mellon, as a lender you are receiving requests to borrow securities all the time, and these are coming in via Bloomberg and email into the desks. These were processed manually.
What we鈥檝e written is a tool that automatically reads these Bloombergs and emails, with no particular format that they have to be in.
It looks for any security ID, quantity, start date and end date, and creates a pre-trade record or a borrow request record. We then take that request record and compare it to their availability and also a set of validation rules鈥攚hich is what the trader would have had to do manually鈥攁nd then we present it to the trader: 鈥淵ou鈥檝e just received this request, this is the availability you have: what action would you like to take?鈥� They can then simply accept or reject it, or better yet, counter with another rate. After this, an automatic message is sent back through the same Bloomberg email channel back to the broker/borrower to say they鈥檝e either accepted or haven鈥檛.
All of that process was manual from an agent/lender and borrower perspective. I see this as a big change in the market. Ultimately the process can become more of a black box, whereby it can automatically respond to the borrower request. It is an exciting area to be working in and I am really enjoying the challenges that our new approach presents.
← Previous interview
State Street
Paul Fleming
Next interview →
Brown Brothers
Keith Haberlin
Image: Shutterstock 