Home / Uncategorized / Causes of the Credit Crash, Part One: The Machines

Causes of the Credit Crash, Part One: The Machines

Technology is blamed as a major cause of the current market meltdown according to an in-depth article in today’s Financial Times entitled: "Algo trading: the dog that bit its master".

These themes build upon the review of capital markets I put forward in December, as well as the discussion of Goldman Sachs’ $1.5 billion algo hiccup last August.

The
gist of the FT article is that capital market firms spend a shedload of
cash on technology: about $41.8 billion per annum, growing at almost
10% a year according to AITE Group.  As a result, what used to be
simple trading and dealing strategies in single asset classes between
humans is now complex, multi-threaded, multi-asset class, globalised,
millisecond strategic plays, which are all automated and cannot be
unravelled easily.

The article likens traders to Rambo’s,
selecting algo weapons to bombard the markets, and to airline pilots
who select destinations and the onboard computers calculate how to get
there.  Funnily enough, Anthony Hilton, the Financial Editor of the Evening Standard, was saying yesterday
that "central bankers are like toddlers driving toy cars.  They think
they are steering but their controls are not attached to anything."

The
issues created by a rogue Rambo are magnified monstrously by the
technology however, as the volume of trades has increased exponentially
such that the smallest human errors are turned into catastrophic losses
before you know it.

And the trouble with all of this is that we have
turned markets from ones that can check themselves into runaway
engines.  For example, we used to mark equities to markets, whilst
complex futures derivatives based upon credit and debt are far more
esoteric.  This is why risk exposures and losses have become so great,
because they are ghosts in the machine.  Ghosts that have returned to
haunt and terrify us months later.

The challenge is therefore how to
use the machines to lower latency, increase margins and find spread
whilst introducing some form of brake.  There needs to be an emergency
stop button somewhere. 

Strangely enough, in another article in today’s FT, they quote the opposite view. 

This one focuses upon latency and has three quotes that were interesting:

“You
are starting to have machines competing against machines. The one that
gets there first is able to exploit the opportunity. Traders are
developing models to look for arbitrage opportunities that may only
exist for a moment. That is why speed matters.”
Kirsti Suutari, Global Business Manager for Algorithmic Trading at Reuters

"Since data travels at the speed of light, the difference in speed
between a co-located server and one 200 miles away is about a
millisecond.”
Jeremy Badman, Partner with Oliver Wyman 

“Reducing latency by 1 millisecond can be worth up to $100m a year to a leading trading house.”
Varghese Thomas, Vice-President at SAVVIS

This
is why the LSE’s new TradElect systems, introduced to the markets last
summer, now claims to be able to execute trades in 6 milliseconds and
process 4,200 orders a second. 

So what’s the point, I ask?

There are no solutions being offered here.

We
know that the markets are getting faster and faster.  I often write and
blog about latency and milliseconds.  Just last week, in a series of
three articles on technologies import to banking (parts one, two and three) I quoted some comments recently heard from exchanges and heads of execution:

“If it takes more than 500 milliseconds to process then it is of no value because it is out-of-date”; and

“We
moved servers to Moscow because it takes 60 milliseconds to route a
trade from London to Tokyo via Moscow compared to 240 milliseconds if
we process the trade via New York”.

The markets are
technologically turbo-charged, connected globally, programmed by quant
rocket scientists and dealing in microvolumes and milliseconds.

This will not change.

So the question is, what to do when we are building ghosts in the machine that will come back to haunt us later?

I think it’s called insurance, although Ambac is an example that even this is not guaranteed.

We
live in a world where algo explosions will hit on a more and more
regular basis.  What we really need is the ability to react to market
movements in real-time in order to unravel an exposed position in
milliseconds, in the same way as we can trade in milliseconds. 

Maybe that is something that humans do best?

Otherwise, the old joke may be coming true.

The joke?

The City is run by one man and his dog.
The dog is there to stop the man touching the computers.
The man is there to feed the dog.

About Chris M Skinner

Chris M Skinner
Chris Skinner is best known as an independent commentator on the financial markets through his blog, the Finanser.com, as author of the bestselling book Digital Bank, and Chair of the European networking forum the Financial Services Club. He has been voted one of the most influential people in banking by The Financial Brand (as well as one of the best blogs), a FinTech Titan (Next Bank), one of the Fintech Leaders you need to follow (City AM, Deluxe and Jax Finance), as well as one of the Top 40 most influential people in financial technology by the Wall Street Journal’s Financial News. To learn more click here...

Check Also

The digital transformation journey

I find more and more people are starting to understand that digital is a transformation …