The Financial Services Club's Blog

How far can investment firms innovate?

I’m continually impressed and amazed by the speed of change in the technology of the investment markets.

For example, last year was all talk about low latency and lit versus dark pools. This year, it’s all about private cloud-based services based upon colocation and proximity services. Next year, it will be all about real-time liquidity and settlement.

Equally, the change from old to new trading venues is quite surprising. I was chatting with one of the MTFs yesterday for example. An MTF is a Multilateral Trading Facility, a new class of exchange introduced by MiFID, the Markets in Financial Instruments Directive.

The conversation reminded me that only two years ago, everyone was saying that liquidity doesn’t move and there would be no threat to traditional exchanges from these new trading venues.

Two years later, Chi-X has more trading and market share than the London Stock Exchange (LSE); BATS is in the Top Ten ...

Source: Chi-X, 19th January 2010, 10:30 GMT

... LSE has acquired Turquoise and completely changed strategy and direction; Euronext has launched a major dark pool; Deutsche Bourse is heavily expanding in ultra low latency connections, targeting latencies under 3.3 milliseconds for Frankfurt-Amsterdam, under 4.5 milliseconds for Frankfurt-Paris, under 40 milliseconds for Frankfurt-New York, and under 49 milliseconds for Frankfurt-Chicago ...

... and so it goes on.

Just two years ago, many exchanges thought that seconds were fine. Today it’s milliseconds. Tomorrow it’s microseconds. And then what?

Is it to achieve the ever elusive cross-asset class trading system? The one where a paired equities trade can be combined with an FX hedge across a global arbitrage?

Nah, we’ve already got that.

Is it to innovate new instruments, even when the last set of instruments – Structured Investment Vehicles, Credit Default Swaps and Collateralised Debt Orders – screwed up the world’s economies?

Nah, we’ve already got that too.

Is it to bring on board new exchanges and focus on emerging markets?

Nah, we’ve already got emerging markets, cross-asset class structure exchange products.

So what is the next wave of major change in trading systems and markets, and how will governments manage to derisk the next wave of innovation?

Well, if the move does become one where real-time execution is combined with real-time settlement and real-time risk management, as I’ve stated before, then the challenge for broker-dealers will be to bring added value to clients through arbitraging across venues and instruments.

But hey, that’s what they do now isn’t it?

Of course.

It’s what broker-dealers and market-markers have done for all time and will be what they do for the future.

You see, the challenge is to keep up with the innovations of a Goldman Sachs and it is the reason why Goldmans made $100 million profits for every day of trading last year.

Now there’s a fine border line between making markets and moving markets, and that’s the line Goldman walk.

It will be interesting to see how Goldman and company make markets in the future, between the Obama tax and the new regulatory regime.

But the key will be to continue to innovate with technology ... and they are and will be.

That is why we are moving towards an almost seamless order flow through execution through settlement system, thanks to smart order routing combined with execution management systems.

That’s one of this year’s big buzzes in trading tech, especially as Asia has followed Europe and America’s approach with Japan clearing the way for new Alternative Trading Systems and venues.

So perhaps the big buzz will really be about arbitraging across venues and instruments using global smart order routing, combined with real-time execution and settlement in microseconds.

Hmmm ... sounds risky to me.

Will governments be up to the job of regulating all of these new innovations?

Source: RJMatson, St Louis Post Despatch

What has MiFID done for democracy?

Fascinating meetings with the Russian exchanges RTS and SPIMEX. After my reference to military intelligence yesterday, you may wonder whether SPIMEX is something to do with finding out who started the swine flu pandemic but no, it’s the St. Petersburg International Mercantile Exchange.

The two exchanges have very different start points and focal points, with RTS (the Russian Trading System) starting as a privately-owned exchange back in 1995 whilst SPIMEX is an initiative of the government to create a working commodities exchange.

I had been aware of RTS for a while as, during the MiFID investigations, their name came up many times as a working, highly automated exchange. Originally launched using NASDAQ’s trading platform, they soon developed their own products and services, covering algorithmic trading for equities, forex, future and options and more.

Today, the exchange is one of the best performing, with the RTS Index tripling in value during 2009 to an index high of 1,451 at yesterday’s close, compared to a market low back in February of under 500, although this is still well below the high of 2,487 in May 2008.

What is the reason for such volatility?

Oil.

Chris Weafer, Chief Strategist at Uralsib Bank, puts it in context: “We've recovered very strongly this year mainly because of the recovery in the oil price. Plus, the recovery and optimism in the rest of the world has allowed for the Russian ruble to stabilize."

Or is it because RTS is a keen innovator offering analysis across every trader’s portfolio and position in real-time, as the RTS people I spoke with yesterday said.

RTS now offers highly automated trading across 1,400 stocks, with 90% of the trades using automated systems and 10% Over-the-Counter (OTC). It is also one of the top 40 Global derivatives exchanges and trading is getting far more complicated, with around 15 trades per transaction on average today compared with 10 trades per transaction just two years ago.

Similarly talking with the SPIMEX guys, there is a clear vision that they are trying to avoid being another failed commodities exchange (around 60 have been launched since 1990).

Why did the previous exchanges fail?

Because no-one trusted them by the sound of it.

According to a SPIMEX advisor, it was because of a poor understanding of risk, a word rarely used because Russians do not allow risk to occur. This is why most exchange and exchanges are based upon ‘fundamentals’ – "if I can see it, touch it and trade it, then that’s ok. If you are looking for me to pay now for something that might pay back in the future, no way."

This is why most Russian banks do not provide trade finance, and why oil is an issue.

For example, oil producers currently do not co-operate because they do not trust each other, according to one of the guys at SPIMEX.

Most oil producers are local monopolies with no competitive mechanisms. As a result, when oil prices rose in 2007-2008, Russians were paying more for their oil than America and Europe, even though Russia has more oil reserves and production than most.

So SPIMEX was launched in September 2008 to overcome this, with the Russian anti-monopoly committee creating the right environment to trade on exchange by fining several of the oil producers for anti-competitive practices.

But the real common feature of both SPIMEX and RTS is real-time risk management.

Both exchanges proudly talk about their focus upon real-time analysis of traders’ positions.

In the case of SPIMEX, they offer real-time settlement and straight through processing, so there is no risk for trading.

In the case of RTS, they offer real-time positioning of every trader and every trader’s clients portfolios, not just in real-time for their own trades but also for the knock-on effect of their dealings in derivatives down the line.

I asked RTS about their risk management, and they made clear that for each transaction, the risk is calculated for the trader’s portfolio, including all orders to be filled, in real-time. There are then two clearing sessions during the day. One at 14:00, which takes three minutes to process, and the second is at end of day, and takes fifteen minutes.

If a margin call is made, the broker must cover their position within two hours or, if at end of day, before the start of the next day’s trading.

This discussion got interesting, as RTS and SPIMEX appear to be developing systems that ensure no trader can leverage risk to the levels where the market implodes, and they do this in real-time.

It is the nature of new trading systems and operations that they design things to work in the ideal way, as the outline above is what Europe and USA are trying to develop.

For example, the FSA’s £2 billion technology change program for real-time liquidity reporting is pretty much what RTS has today.

No wonder the gentleman from RTS turned to me towards the end of our chat and asked, straight-faced, “what has the Markets in Financial Instruments Directive (MiFID) done for democracy”.

He probed me about best execution and what it means: “is it just all about price, speed and cost (and likelihood of settlement)?”

I then realised what he was getting at.

Where, in all the developments of MiFID and its best execution, transparency and competitiveness objectives, was the mention of risk?

Where is the focus on real-time risk reporting?

Hmmmm ... it’s obvious that RTS, and the Russian aspirations to build the next major Russian-Asian commodities exchange, is something to watch.

Meanwhile, the major thought that struck me in this dialogue, was that we have two opposites.

In Europe and America, we have an over-leveraged, casino capitalism culture of trading that is now being re-engineered to restrain excessive risk without responsibility.

In Russia, we have a risk avoidance trading environment through real time controls, which needs to increase liquidity and leverage in order to fuel the flow of commerce.

If Russia does not achieve this, then its commodities ambitions cannot be achieved.

So maybe there is a happy medium here between the Russian approach to risk controls of trading, and the Anglo-Saxon approach of using financial instruments to create liquidity.

Now there would be a thought ...

*

The Finanser is sponsored by VocaLink and Cisco:

For details of sponsorship email us

The future vision for trading

At the Deutsche Bourse Open Day for IT, I chaired a panel comprising senior figures from Eurex, the International Securities Exchange and IBM.

The focus of the panel was the future of trading technologies, and was a wide ranging discussion covering everything from dark pools to high frequency trading (HFT).

Questions were abounding, such as:

Question:

Is HFT good or bad for liquidity as it gives large players an unfair advantage over small to mid size players and allows market abuses such as flash orders and front running?

Answer:

As long as everyone has equal access to the markets, then it is no issue. The fact that some traders cannot afford the technologies to deal in flash orders is not an unequal market perspective; it’s just that some are better at doing this and can afford the systems to do it. There is no barrier to entry however.

(BTW, I totally disagree with that view)

Question:

At these speeds, is it technically impossible to do compliance? For example, Barclays Capital are just one illustration of the markets becoming more difficult to manage when they are fined for not even being able to get the timing right on a trade execution.

Answer:

Yes, this is more difficult but time stamping trades is an essential core competence of any trading house. The issue lies with low latency and how a broker deals with order flow. If you cannot time stamp effectively when an order is received, filled and executed throughout the trade flow then you are not delivering on that core competence.

This happened recently with TD Securities who were accused of front running trades, and yet how do you prove you weren’t front running when everything takes place in a microsecond? TD Securities fought off a recent investigation to prove they were not front running by being able to demonstrate timings through their order flow, and so yes, keeping time is a basic requirement today.

Question:

What about the whole aspect of the changing regulatory regime with the new rules being developed by the G20, the de Larosiere report, the confirmation that there will be a MiFID II, etc. In fact, MiFID II, according to a speech by Charlie McCreevy in Dublin recently*, will incorporate OTC Derivatives in its next iteration. What does all of this mean for trade execution and technologies?

Answer:

The markets are ready for more regulation although it is interesting that most regulation creates unintended consequences. For example, the law of intended consequences is demonstrated by MiFID where the aim of the regulation was to have more transparency and openness and what did we get? More opaqueness, price fragmentation and a growing off-exchange trading environment using technology to create dark pools. This is the key to the technology used in the trading environment- to create arbitrage, risk and profitability for the client.

Meanwhile, everything will move towards real-time reporting. For example, Eurex is already able to provide real-time clearing at trade execution.  Firms will move towards this model for real-time trade reporting and real-time risk and liquidity management.

Question:

What about Social media & Twitter? Don’t these tools support insider trading? Do we end up with all traders being searched for personal communication devices before they enter the trade floor?

Answer:

No, you don’t need to go to that extent. Facebook and Twitter are electronic communication tools however, so they need to be brought under the controls and disciplines of compliance regulations or be blocked and, if a trader is found in contravention of such in-house policies, then it’s a disciplinary dressing down and possibly termination of employment. But you can do things about these areas. For example, research into this area (download report) clearly shows that the US regulatory rules apply just as much to tweets as they do to email. And there are tools you can implement to monitor such activities and store and forward such information, which is the right approach to take.

Question:

How do you see the development of Cloud Computing in the trading markets?

Answer:

In the US, we have seven major exchanges and I can see the world developing into one where those seven exchanges all place their servers in a massive server farm, co-located with market makers and brokers, so that the systems can deal in real-time. That’s where we are almost at today and certainly will be moving in that direction more and more over time.

The answers above were my own take on the discussions by the way, and not quotes from the panellists, but you get the idea of the flow.

We finished with a vision of the future of trading which I will sum up as being a world where each continent has a massive data centre of co-located servers representing the main trading venues of each economy.

The European data centres will naturally locate in London, Paris and Frankfurt; America’s will be in New York, Chicago and possibly San Francisco; Latin America’s in San Paolo and Buenos Aires; Asia’s in Mumbai, Tokyo, Shanghai, Hong Kong and Australia; and Africa and the Middle East will be based in Johannesburg and Dubai (or Bahrain or Qatar dependent upon how the region’s financial centres compete for dominance).

Fourteen massive data centres with mirroring and backup for each other.

Effectively a massive cloud of trade execution.

And yes, that’s where we move towards – a secure global cloud computing environment, all linked by dark fibre between fourteen massive data centres.

The data centres house the world’s exchanges and their broker dealers.

They can transact in microseconds within exchange and, thanks to terabit lines, can send a trade execution around the world in under a millisecond.

Lightning dealings in order flows through smart order systems that allow billions of offers and bids to be matched in a millisecond.

All of this liquidity flow being watched under a traffic light system of liquidity risk and market monitoring, such that market abuses are visualised and captured during the trade flow in real-time.

The G20 agreements on co-located data centre environments dealt with that one after the financial crisis of 2013, when the dark fibre smart order routers allowed Goldman Sachs to front run ...

Aw shucks, now I’m just getting cynical again aren’t I?

* In his speech to the EU Public Affairs Ireland Conference on 18th September 2009, Charlie McCreevy, European Commissioner for Internal Market and Services, stated:

“There appears to have been a significant migration of share trading transactions from the more regulated MIFID venues to the unregulated over the counter broker dealer venues where substantial unregulated dark pools of liquidity have built up. This gives rise to questions as to whether there are unfair commercial advantages for the operators of these venues and whether the trend undermines price discovery, market integrity and efficiency for the market as a whole. The MIFID review will address this issue.”

*

The Finanser is sponsored by VocaLink and Cisco:
 
For details of sponsorship email us.

Innovations in Payments, Part Three

In the securities settlement markets, we’ve spent years trying to break down barriers to clearing and settlement ... and most of those barriers are still there.

We wonder what to do about it?

And regulation seems the only answer.

Or is it?

As we look at innovations in payments (retail, commercial covered in last two posts), this week is the final thoughts on this as we study payments in the trading room and investment markets.

There are clearly some things afoot which will change things.

The first is that everything is real-time or supra real-time.

Sure, we talk about fast and scalable systems, but in the trading arena we’re talking not just fast but super fast.

What’s super fast?

Well, a human blink of an eye takes about 200 milliseconds or a fifth of one second.

Today’s trading systems trade in under 200 microseconds.

There are 1,000 microseconds in a millisecond, so we can trade 1,000 times faster than the blink of an eye.

That's super fast.

This is the world of latency, which I’ve blogged about often.

But latency really does bring in a new focus on real-time.

For example, I spent last week at a fascinating open day for Deutsche Bourse’s IT and intend to write a detailed update through this week on what came out of that day.

To start with, here’s a chart (double click to enlarge):

This was presented by Colt Telecom and shows linkages between London, Frankfurt and Chicago, with latency of 50 milliseconds or thereabouts. In other words, four messages can move between Frankfurt and Chicago in the time it took you to blink.

Here’s another picture:

This shows how all the exchanges are linked via high speed dark fibre 1 gigabit per second (1G) and up to 10G.  The view is that, by 2011, these lines will be moving data at 6.4 terrabits per second (t/s).

6.4 t/s!

This means we will be moving from a world where it takes about 0.5 seconds to send a message right around the world once (two blinks), to one where that global transaction of data could move around the world in a complete cycle in under one millisecond (blink and you miss it).

This is the real-time world.

Now think about liquidity, risk, clearing and settlement and imagine a world where, as you trade.

Your complete picture of liquidity positions are transparent to all traders internally.

Your liquidity risk exposures, market and credit risk exposures and more are visible to all counterparties in real-time.

Your trade reporting and pricing is visible to all regulatory and supervisory authorities in real-time.

And real-time clearing takes place as part of trade execution.

This world is not a future vision but a reality.

Eurex, for example, thanks to being tightly coupled with a clearing system (Eurex Clearing), are already offering real-time clearing.

Meanwhile, we have regulators moving further and further towards information reporting duties that go far beyond today’s methods.

That is why real-time trade and risk reporting will be a reality in the near-term.

End of day just doesn’t cut it.

Intra-day doesn’t manage it.

Real-time everything.

That’s the future of investment markets.

Real-time visibility and transparency for all counterparties to manage risk.

Mmmm ... now it gets interesting.

$450 trillion in a single place

It’s great being in another timezone as, after my 11-hour flight, I felt really tired so an early dinner and then bed by 10 o’clock ...

... waking up at 1:30 a.m.for a nice five hours of watching movies whilst trying to get back to sleep.

I get to my first meeting and say, “y’know, I’ve been awake since half-one and need coffee”. I then find that everyone else from Europe and America is the same way and up since one or two this morning.

This may be a bleary-eyed SIBOS therefore and I get to my first session, which focuses upon clearing and settlement, to find the audience asks no questions ... yep, they need coffee.

The session opens with a statement from the SWIFT hosts that this is a record-breaking SIBOS with over 5,300 people registered which is more than SIBOS in Sydney three years ago.

So much for a market that has meant to have imploded.

In fact, as I go around the exhibit hall, I see that many of the booths are bigger, brighter and better than ever ... including those of Citi (“we’ve spent a billion dollars on infrastructure in 2009”) and Royal Bank of Scotland (“making it happen”).

Anyways, back to the first session which was a keynote from Michael Bodson, Executive Managing Director of the DTCC.

The DTCC is the large US Central Counterparty Clearing systems (CCP) with the aim of moving into being the largest clearing system globally, as demonstrated by their OTC Derivatives information warehouse which, according to Michael, now tracks $450 trillion worth of credit derivatives in 35 countries.

He emphasised that this is key: a single data repository, a single warehouse, a single hub. The idea of multiple hubs for each region, which is being mooted by the EU as they do not want to rely on a single US-based hub, is clearly wrong in his view.

This is because right now, it is clear to all parties how to register a risk because it is on a single database. If there were multiple hubs, then which risk gets registered on which hub for which counterparty becomes confused.

Or that’s the DTCC’s view.

On a wider note, he did say that post-trade has suddenly become a hot topic because we stared into the abyss last year and survived thanks to collective resilience of the infrastructures.

“A year ago, nothing worked. Stocks failed, bonds failed, everything failed ... except market infrastructures. Infrastructures worked.”

He outlined that the DTCC handled $500 billion in open Lehman Brothers trades during the six weeks after their collapse on September 15th 2008. In fact, during one day, they liquidated $300 billion in open positions.

These are shocking numbers and the fact that the infrastructures worked and were resilient was the key.

This is why the clearing and settlement plumbing of the world’s financial markets has gone from being “about as exciting as goat’s cheese” to being sexy.

Clearing and settlement has moved from being post-trade back office boredom to being critical to everyone’s management of risk and exposure in these global markets of liquidity.

Interestingly, the DTCC had actually run a simulation of a major top ten investment bank failure based upon Lehman Brothers as the case in point, as a training exercise just weeks before their collapse.

A prescient moment.

Michael also referenced that regulators should have three areas of concern in the markets – an asset bubble, liquidity risk or major fraud. The fact that you had Lehman Brothers and Bernie Madoff within months of each other shows what a good job the regulators are doing.

Yeah, baby, yeah.

This presentation made for a good grounding for the panel that followed which was all about competition and value-add between global custodians, central securities depositories (CSDs) and central counterparty clearing systems (CCPs).

To be honest, there’s a challenge there in that you don’t really want competition between these groups if you are trying to manage risk effectively. The ideal risk manager you see, is a risk management of one, where all risk is centralised. That is the point of a CCP or CSD, to have it centralised.

As a result, you really have a natural monopoly when it comes to clearing and settlement. This was the grounding of the panel which comprised:

• Philippe Metoudi of Clearstream representing Asia;
• Monica Singer of Strate representing Africa;
• Ted Rothschild pf JP Morgan representing America; and
• Alan Cameron of BNP Paribas representing Europe;

chaired by yours truly.

It turned out to be a really good panel, debating the nature of ‘value-add’ and ‘competition’ in the post trade clearing markets.

As I chaired it, I didn't take many notes or quotes but am sure it will get a write-up somewhere and will link to this later.  The one comment I did note was the one comment from the floor from a not-for-profit clearing infrastructure:

“CSDs get accused of being fat, dumb, lazy and bureaucratic as state-run enterprises and are then told that it is not right if they try to innovate, compete and reduce prices and costs. We can’t win.”

Yeah, baby, yeah.

*

The Finanser is sponsored by Vocalink and Cisco:
 
For details of sponsorship email us.