Sun

03

Jul

2011

A Giant Leap Of Speed

In the current arm race towards ever higher speed and ever lower latencies, new ultra-fast machines are just around the corner.

 

The name is FPGA, Field Programmable Gate Arrays. This techno-sounding name is a technology, used by the military for quite a few years already, which allows burning programs directly on a chip.

 

This can make any process hundreds of times faster in some cases. E.g., a FIX message could have all its components parsed at once, instead of through an iterative process.Deutsche Bank has just announced the use of such a technology for its equity trading machines. And as we all know, what happens in equities happens in FX, just a while later. I know of at least 3 banks which are currently looking at these technologies, although sometimes they just buy finished products from specialized (and expensive) suppliers.

 

FPGA techniques can have applications in both matching and routing engines, so we can expect blitz-aggregators any time soon as well.But what kind of speeds are we really talking about? Here are a few numbers:A new company I am in touch with, e.g., is setting up a matching engine which could potentially deal with 4 million trades per second, or a routing system able to deal with 50 million messages per second. A simple message parsing could take 100 nanoseconds. These numbers are a thousand times better than what most of the market deals with today. Yes Ma'am.

Incidentally, the above numbers are also better than most other FPGA speeds seen so far. You read it here first. Call me for more details.

 

Ok this is all good, but do we really need all these goodies? I believe we do, as the quest for good prices will go on no matter what.

 

Already, the probability that a good price issued by a bank in New York is traded on by a fund in London is similar to the probability of a cod in the Atlantic to die of old age.Algo-FX markets are similar to a restaurant, with diners sitting close to or far from the kitchen. When a waiter gets out of the kitchen, good dishes get grabbed first. If you are sitting at the back, you get to eat the leftovers.

 

What I mean by this is that your TCA will increase, and the cost of your TCA must be compared with the cost of these new technologies, which is a quantitative way of deciding whether you want them.

Some banks have already decided. Some funds too, from what I hear.

7 Comments

Sun

03

Jul

2011

Are You Gobbledybooking Me?

Are you gobbledybooking me, Mister?

These last few months, a hitherto unknown word has been making waves, so much so that US authorities themselves have thought it wise or savvy to express their opinion about it.

 

The word is b-book. You saw it here first.

 

For about 10 years, a not so well kept secret was that if a retail FX operation really wanted to make money, it should not or almost never hedge its clients positions. As many retail clients may not have the required professionalism or discipline, so the story says, not hedging them is just the way to go. The stop-losses of the customers become the take-profits of the retail operator. You can even guarantee stop-losses, which makes you look good. Hence the a-book was the book of deals to hedge, because these are perceived to get it right, and the b-book is for everybody else., and where the real gravy comes from.

 

The official line is that positions are not hedged so that buyers hedge sellers and the operator earns the spread whilst still hedging customers. This is partly true. However, most of the time,positions are simply warehoused and everyone hopes for the best.

 

As the Soviet used to say of capitalism : this business is rotten, but what a great smell.

 

At first, this was done without much of systems to control what the operators position was. Alas, retail clients, once in a while, will get it collectively very right, as, e.g., during the financial crisis. This was partly the cause of some of the consolidation we have seen in the sector. Today, b-booking is done in a much more sophisticated way, with algos kicking in to hedge positions or maximise returns.

 

But before going there, let us look at what b-book really means: it just means a client deal is left unhedged for a while at least. First, any trading desk is b-booking its customers at least for a short while. Many of the prices made to customers are narrower than what can be hedged straight away. These deals can only be hedged if customers buy and sell at the same time, which only happens in very active operations and not all the time, or if the trader is savvy enough to bid below the sale price and get given, or already has a position he hedges against.

 

So, narrow spreads are often achieved with at least a little of b-booking.

 

But even the most nefarious rationale behind b-booking, i.e. clients do get it wrong most the time, is useful in this respect. We can say that narrow spreads, which could create losses for the operators, are in a way financed by the less lucky clients. Fortune favours the astute. This is so much true that in the institutional world, clients look for bank having reverse views so they get a better price.

 

So when authorities set up rules that all client deals will need to be hedged in the future, they should perhaps spend some time thinking about the effect on spreads.

 

The saga goes further: eager to say that b-book is not a good word anymore, some operators now hedge all of their clients deals, only pocketing small commissions. Operators can swear that they do not b-book anymore, no no Sir.

 

But... they hedge it with banks providing them with extremely very narrow spreads, which are only for them. We are even hearing of rebates from banks for this business.

 

So look dad, no hands : now you see it, now you dont.

 

The truth of all market is that they are a bridge where willing sellers meet willing buyers. As long as all clients is treated fairly, which is the case in most electronic platforms, the fact that not hedging some positions improves the spread should be seen as a benefit, not a fault, of OTC markets.

1 Comments

Sun

03

Jul

2011

Machines 1 Humans 0

Guess what? Algorithms also have fat fingers, like their human masters.

 

Machines, as we know, compete with humans for records. One of them is the record of the most spectacular market blunder.

 

Let us just mention a recent example : a classic case of fat fingers was expertly turned by heartless programs into a 1000 point freefall of the Dow Jones. Machines : 1. Humans : 0.

 

Comments about our big trek towards machine trading are not always kind. But this is not the first time humankind has bitter afterthoughts about technology. Maybe humans should have stayed in their caves after all. It was even probably more comfortable than we can remember. Back then, women, not computers, remembered where we stored our meagre possessions. The reader can appreciate which is to be trusted more.

 

But no matter how much we would like it, a return to Titans days, when masters of the universe ruled the markets, is impossible. We are condemned to constantly improve our technologies, till the day when they become perfect, which means invisible. Would anyone want to do away with the electric kettle?

 

Machines, for another while yet, are just an extension of ourselves, not unlike hammers, that make fat fingers more dangerous too. And we, their human owners remain to be blames, if only for not installing a seatbelt system on front-end GUIs operated by junior dealers.

5 Comments

Sun

03

Jul

2011

Strap-on algorithmics

Strap-on algorithmics

Not all blogs need to be serious. A few years ago, I took part in a phone conference with an American team, brainstorming about helping clients execute on the ECN I was working for at the time. We had already provided our customers with all sorts of orders, and we were looking to introduce some algo-tools to help them further improve their success ratio.

There are quite a few algorithms that would be helpful in that case, and the equity industry had produced a number of them well before FX markets learned how to spell algorithm without a y...

The idea surfaced to have our own execution tools, available to help customers, somewhat like a dealer helping a client out, albeit an automatic one. I was pushing hard for this, and had a few exotic names for several of them. To describe what I had in mind, I suggested describing these as strap-on algorithms, that would be selected at will by the customers. The first time I used this term, I must confess.

An uneasy silence followed, which I interpreted as tacit approval. I continued referring to strap-on algorithms for the whole phone conference, and enthusiastically described how and why they would perform better than traditional algorithms on the client side, detailing the various attributes the clients would be able to choose from... unfortunately not hearing the barely repressed giggles on the other side of the pond. Someone strongly encouraged me to describe all these in an email...

We did not go all the way, if I can say so, to making brochures with these, although this would have guaranteed instantaneous recognition - nothing better than a good laugh to market your products -.

A week later, a visitor from the US told me how famous strap-on algorithms had become in and around the company. Fortunately, the buzz stopped there, and I am not greeted with a knowing smile every time I speak about how to improve success ratios, and my level of English has improved since.

What linguistic mines are still lurking to explode under my words, I am sure we shall find out

6 Comments

Sun

03

Jul

2011

A Giant Leap Of Speed

In the current arm race towards ever higher speed and ever lower latencies, new ultra-fast machines are just around the corner.

 

The name is FPGA, Field Programmable Gate Arrays. This techno-sounding name is a technology, used by the military for quite a few years already, which allows burning programs directly on a chip.

 

This can make any process hundreds of times faster in some cases. E.g., a FIX message could have all its components parsed at once, instead of through an iterative process.Deutsche Bank has just announced the use of such a technology for its equity trading machines. And as we all know, what happens in equities happens in FX, just a while later. I know of at least 3 banks which are currently looking at these technologies, although sometimes they just buy finished products from specialized (and expensive) suppliers.

 

FPGA techniques can have applications in both matching and routing engines, so we can expect blitz-aggregators any time soon as well.But what kind of speeds are we really talking about? Here are a few numbers:A new company I am in touch with, e.g., is setting up a matching engine which could potentially deal with 4 million trades per second, or a routing system able to deal with 50 million messages per second. A simple message parsing could take 100 nanoseconds. These numbers are a thousand times better than what most of the market deals with today. Yes Ma'am.

Incidentally, the above numbers are also better than most other FPGA speeds seen so far. You read it here first. Call me for more details.

 

Ok this is all good, but do we really need all these goodies? I believe we do, as the quest for good prices will go on no matter what.

 

Already, the probability that a good price issued by a bank in New York is traded on by a fund in London is similar to the probability of a cod in the Atlantic to die of old age.Algo-FX markets are similar to a restaurant, with diners sitting close to or far from the kitchen. When a waiter gets out of the kitchen, good dishes get grabbed first. If you are sitting at the back, you get to eat the leftovers.

 

What I mean by this is that your TCA will increase, and the cost of your TCA must be compared with the cost of these new technologies, which is a quantitative way of deciding whether you want them.

Some banks have already decided. Some funds too, from what I hear.

7 Comments

Sun

03

Jul

2011

Are You Gobbledybooking Me?

Are you gobbledybooking me, Mister?

These last few months, a hitherto unknown word has been making waves, so much so that US authorities themselves have thought it wise or savvy to express their opinion about it.

 

The word is b-book. You saw it here first.

 

For about 10 years, a not so well kept secret was that if a retail FX operation really wanted to make money, it should not or almost never hedge its clients positions. As many retail clients may not have the required professionalism or discipline, so the story says, not hedging them is just the way to go. The stop-losses of the customers become the take-profits of the retail operator. You can even guarantee stop-losses, which makes you look good. Hence the a-book was the book of deals to hedge, because these are perceived to get it right, and the b-book is for everybody else., and where the real gravy comes from.

 

The official line is that positions are not hedged so that buyers hedge sellers and the operator earns the spread whilst still hedging customers. This is partly true. However, most of the time,positions are simply warehoused and everyone hopes for the best.

 

As the Soviet used to say of capitalism : this business is rotten, but what a great smell.

 

At first, this was done without much of systems to control what the operators position was. Alas, retail clients, once in a while, will get it collectively very right, as, e.g., during the financial crisis. This was partly the cause of some of the consolidation we have seen in the sector. Today, b-booking is done in a much more sophisticated way, with algos kicking in to hedge positions or maximise returns.

 

But before going there, let us look at what b-book really means: it just means a client deal is left unhedged for a while at least. First, any trading desk is b-booking its customers at least for a short while. Many of the prices made to customers are narrower than what can be hedged straight away. These deals can only be hedged if customers buy and sell at the same time, which only happens in very active operations and not all the time, or if the trader is savvy enough to bid below the sale price and get given, or already has a position he hedges against.

 

So, narrow spreads are often achieved with at least a little of b-booking.

 

But even the most nefarious rationale behind b-booking, i.e. clients do get it wrong most the time, is useful in this respect. We can say that narrow spreads, which could create losses for the operators, are in a way financed by the less lucky clients. Fortune favours the astute. This is so much true that in the institutional world, clients look for bank having reverse views so they get a better price.

 

So when authorities set up rules that all client deals will need to be hedged in the future, they should perhaps spend some time thinking about the effect on spreads.

 

The saga goes further: eager to say that b-book is not a good word anymore, some operators now hedge all of their clients deals, only pocketing small commissions. Operators can swear that they do not b-book anymore, no no Sir.

 

But... they hedge it with banks providing them with extremely very narrow spreads, which are only for them. We are even hearing of rebates from banks for this business.

 

So look dad, no hands : now you see it, now you dont.

 

The truth of all market is that they are a bridge where willing sellers meet willing buyers. As long as all clients is treated fairly, which is the case in most electronic platforms, the fact that not hedging some positions improves the spread should be seen as a benefit, not a fault, of OTC markets.

1 Comments

Sun

03

Jul

2011

Machines 1 Humans 0

Guess what? Algorithms also have fat fingers, like their human masters.

 

Machines, as we know, compete with humans for records. One of them is the record of the most spectacular market blunder.

 

Let us just mention a recent example : a classic case of fat fingers was expertly turned by heartless programs into a 1000 point freefall of the Dow Jones. Machines : 1. Humans : 0.

 

Comments about our big trek towards machine trading are not always kind. But this is not the first time humankind has bitter afterthoughts about technology. Maybe humans should have stayed in their caves after all. It was even probably more comfortable than we can remember. Back then, women, not computers, remembered where we stored our meagre possessions. The reader can appreciate which is to be trusted more.

 

But no matter how much we would like it, a return to Titans days, when masters of the universe ruled the markets, is impossible. We are condemned to constantly improve our technologies, till the day when they become perfect, which means invisible. Would anyone want to do away with the electric kettle?

 

Machines, for another while yet, are just an extension of ourselves, not unlike hammers, that make fat fingers more dangerous too. And we, their human owners remain to be blames, if only for not installing a seatbelt system on front-end GUIs operated by junior dealers.

5 Comments

Sun

03

Jul

2011

Strap-on algorithmics

Strap-on algorithmics

Not all blogs need to be serious. A few years ago, I took part in a phone conference with an American team, brainstorming about helping clients execute on the ECN I was working for at the time. We had already provided our customers with all sorts of orders, and we were looking to introduce some algo-tools to help them further improve their success ratio.

There are quite a few algorithms that would be helpful in that case, and the equity industry had produced a number of them well before FX markets learned how to spell algorithm without a y...

The idea surfaced to have our own execution tools, available to help customers, somewhat like a dealer helping a client out, albeit an automatic one. I was pushing hard for this, and had a few exotic names for several of them. To describe what I had in mind, I suggested describing these as strap-on algorithms, that would be selected at will by the customers. The first time I used this term, I must confess.

An uneasy silence followed, which I interpreted as tacit approval. I continued referring to strap-on algorithms for the whole phone conference, and enthusiastically described how and why they would perform better than traditional algorithms on the client side, detailing the various attributes the clients would be able to choose from... unfortunately not hearing the barely repressed giggles on the other side of the pond. Someone strongly encouraged me to describe all these in an email...

We did not go all the way, if I can say so, to making brochures with these, although this would have guaranteed instantaneous recognition - nothing better than a good laugh to market your products -.

A week later, a visitor from the US told me how famous strap-on algorithms had become in and around the company. Fortunately, the buzz stopped there, and I am not greeted with a knowing smile every time I speak about how to improve success ratios, and my level of English has improved since.

What linguistic mines are still lurking to explode under my words, I am sure we shall find out

6 Comments