With Amaranth's managers having decided to throw in the towel and sell off the fund's remaining assets, we have seen a couple of interesting new developments. Nick Maounis' letter to his investors, reprinted (among other places) here, announces that the fund will be suspending redemptions until the end of October in order to sell off the rest of the assets at the best price they can get.

We've also received an analysis of the collapse from Hilary Till at the EDHEC business school in Nice - see Risk News for a summary.

IBM and the Economist Intelligence Unit have released a study forecasting that product groups of traders, now averaging forty-strong, will be four-strong by the end of 2015. The culprit, it says, is "electronification" - OTC's new favourite word - in particular the move towards rapid electronic trading and trade processing.

In addition, the report predicts a split between alpha and beta - "In the future, investors... will be less willing to pay Alpha fees for Beta returns, and the way that funds are priced and managed will have to change." Pensions funds in particular will be driving the change be shifting from managed mutual funds to a combination of hedge funds - for alpha - and index-linked products or derivatives for beta, paying close to zero in transaction costs for the latter.

The profit will come from principal rather than agent trading, and from advice and value creation rather than simply intermediating in trades.

Overall, no huge surprises (everybody is expecting a market dominated by electronic broking, after all) but some interesting thoughts.

(Incidentally, readers will be unsurprised to learn that IBM recommends switching to using an external specialist to handle your strategic IT requirements.)

Discuss the future of trading in the Risk Forum.

Following on from yesterday's entry, another interesting paper from IBM appears at Finextra - it covers similar ground but focuses more on the advances in IT which will make the cuts feasible.

The figures are impressive: electronic buy-side order flow will go from 52% in 2005 to 80% of the total in 2007. This represents an increase from 1.2 billion shares a day in 2004 to 3.1 billion a day in 2007 - a revolutionary shift in the way the markets work.

Algorithmic trading will have to be fast enough - in the tens-of-milliseconds range - to catch favourable prices before they shift. At present, the LSE sees over 3,000 quotes for each completed trade, a real waste of resources for the traders. Cutting latency will be vital to bring this down - the report suggests that players will start to co-locate with market data sources to cut transmission times, and may also construct "military-grade networks for assured data services".

The headline of this post, incidentally, is IBM's way of explaining that voice brokers (or "lone gunslingers") are outdated and set to be replaced by computerised trading systems (or "robots with machine guns").

Comment on algorithmic trading, robots or machine guns at the Risk Forum.

A couple of recent stories could mean trouble for the Tokyo Stock Exchange, now in the middle of rebuilding its IT infrastructure to avoid any more embarrassing mishaps (such as this and this.)

The exchange has announced consultation, but little progress, on the refit - we are now approaching the first anniversary of the original problems. Meanwhile, Euronext.liffe has become the latest exchange to extend its opening hours in order to attract the Asian market - others (such as Eurex) have also moved in that direction.

As exchanges gradually become more globalised, TSE risks obsolescence. However, there is hope - the prospect of competition over latency times (mentioned here) will make locality important even as the global market makes it irrelevant. Latencies across the Atlantic or Pacific, produced by the time taken for electronic signals to travel thousands of miles at light speed, can be up to 250ms - a crippling length of time for tomorrow's algorithmic traders. Ironically, the future may be local, rather than global - a "single financial market" is a laudable goal, but we can't change the laws of physics.

Discuss the future of trading in the Risk Forum.

There are a couple of interesting features that are unique to the emissions market.

First of all, it's entirely dependent on the government to provide the product involved.

Second, participants in the market - like SG, which launched its Orbeo emissions trading joint venture with Rhodia today - really believe in their market. They're given to statements like "This has to work". To the cynical - such as some attending the Orbeo launch - this evokes memories of the 1990s dotcom bubble.

Despite the price fluctuations earlier this year, the carbon market is still liquid and reliable. The crunch will come soon - this year or early in 2007 - when the European Commission either manages, or does not manage, to get its troublesome member states to agree to new, lower total emission allowances in their next National Allocation Plans. If the EC gets its way, the market becomes short - and thus nicely busy. If not - if the national governments succeed in giving effective subsidies to their own industries - the market goes under.

And this is a point where the true believers are right. A sudden upset in, say, the credit derivatives market could cost a lot of people a lot of money. But that is all. A failure in the emissions market could lead, indirectly but fairly certainly, to the greatest global disaster in history.

Discuss the emissions market on the Risk Forum.

From Sean Park, head of digital markets, credit trading and debt syndicate at Dresdner Kleinwort, this short film (requires sound; less than 5 minutes) paints an interesting picture of one possible future for the financial industry. Park blogs at www.parkparadigm.com, where he's outlined a few of the ideas behind the film. (For people who'd rather not watch the film, here's an article on the same subject.)

His predictions cover two main themes - an impending sequence of mega-mergers in the financial industry (LSE/Euronext/NYSE, BNP/ABN and Deutsche Bank/SG are the examples he gives) and the consequences of the completion of the IT revolution in trading. Broadly, banks get out of the intermediary business and become "huge, regulated pools of capital", while the markets shift onto the vast electronic platforms of AmazonBay Financial Markets. Meanwhile, President Obama has liberalised electronic betting (or "outcome trading" as it is euphemistically called") and GoogleCorp (CEO: Lachlan Murdoch) is hedging its advertising exposure with World Cup futures...

It's an interesting film, whether you agree with its predictions or not, and it looks good in a sort of Blade Runner/ Renaissance way.

Discuss the future of trading technology in the Risk Forum. Or, alternatively, discuss the possibility of more large mergers, also in the Forum.

Those of you who came along to the Albannach on Trafalgar Square recently for the 2006 Risk Interdealer Rankings event may see some familiar faces in these photos. Feel free to browse through them, but contact us if you want to use them in your own publications.

And if anyone has any comments on the rankings - should we add more currencies? More product types? Are there any categories that should be split - or even merged? - let us know through the Risk Forum.

An interesting discussion has been going on over the last few days on the tendency in the market, and asset managers in particular, to underestimate the probability of extreme events - or 'tail risk'. One participant is Felix Salmon of the consultancy Roubini Global Economics. The other is Brad DeLong, a professor at the University of California at Berkeley.

Here's a brutally abridged version of the discussion: click the links to read each post in full.

Christopher Fawcett, co-founder of a $4.3 billion London investment group called Fauchier Partners, and also chairman of the Alternative Investment Management Association, told the Times on Friday that he had doubted Amaranth's stability to the point of paying over the odds to extract himself in December last year.

According to the Times, Fauchier spotted 11 red flags in Amaranth's risk management and paid a penalty fee in order to get its $30m investment out fast. Amaranth, Fawcett said, was "a fund with bad risk management and unattractive terms for investors”.

Fawcett's position gives his criticism strength, but he's not the only one to be claiming he saw problems with the collapsed energy hedge fund before the disaster this summer. Blackstone Capital is said to have seen problems earlier this year, and Amaranth was apparently also warned by Nymex in August.

It's looking increasingly embarrassing for the banks that didn't see the collapse coming...

Discuss Amaranth in the Risk Forum.

News of the CME/CBOT merger is now out, with the two exchanges putting cost savings up front as the main reason for combining. According to a study just out from the European IT services company LogicaCMG, European exchanges are going to be feeling the pressure soon as well.

The report, by the veteran industry analyst Graham Bishop, predicts that introduction of the Markets in Financial Instruments Directive (MiFID) in the European Union will open up stock exchanges to competition from other major financial institutions – which could take a lot of business away from the established exchanges.

The report is rather long (and not online: here’s a description), but the crux of it comes in the fourth part, where Bishop outlines three directions that the exchange business could take in the next five years (a short timespan for such a radical set of changes).

A US court has declared Enron chief executive Kenneth Lay innocent on charges of fraud and conspiracy – three months after his death. Lay died on July 6, after his conviction on 10 charges of conspiracy, fraud and making false statements, but before he could be sentenced by the court in Houston. According to the judge in the case, it is standard practice to wipe out convictions if the criminal dies before he has a chance to appeal.

The move is going to make life difficult for Enron creditors and the SEC, who will not now be able to cite Lay's conviction in May in their attempts to recover assets from his estate through the civil courts. Effectively, they will now have to convict him again – although US civil courts require only a "balance of probabilities" standard, rather than the "beyond reasonable doubt" proof necessary in a criminal case.

It could also harm his (still alive and therefore still guilty) colleague Jeffrey Skilling – prosecutors had been going after $43.5 million of Lay's assets when the conviction was vacated, and will now seek to extract the money from Skilling instead, who is already being pursued for another $139.4 million. Skilling was convicted in May and will be sentenced on Monday, October 23.

Back in August, Risk touched on the question of using contracts for difference (CFDs) as an alternative to investing in the underlying equity, in our examination of the spread betting market. We also looked at the Financial Services Authority's attempts to regulate them, for fear of CFD holders using their undeclared holdings to influence corporate strategy.

Gareth Gore, catching up on the story for this month's Risk, has found that the use of CFDs to secretly accumulate stakes in companies under the noses of stakeholders and regulators has led to calls for tighter regulation. Rumours are that the regulator will tighten up the rules on declaring CFD holdings – not immediately, but in the medium to long term. So what impact will this have on the burgeoning CFD market?

Read the full story here. Or discuss CFDs now in the Risk Forum.

Via Lars Toomre’s blog comes a recommendation for a Wall Street Journal piece pointing out the unexpected downside of algorithmic trading. It’s headlined “Algorithmic Trading Inflates Costs”, and the gist of it is that developing, and especially, maintaining algo systems is proving far more expensive than everyone expected – $10 million startup and $5 million annual maintenance is the figure they give. It's an interesting read.

The last couple of years saw a lot of institutions buy constant maturity swap steepener products - and get their fingers burnt. Unexpectedly flat euro and dollar interest rate curves mean that a lot of these products are now well underwater. The payout on these products was based on the spread between two different CMS rates - flat curves mean that this spread is far tighter, and so the coupon is lower.
Naturally, the holders of CMS steepeners are deeply unhappy. But in most cases, there's not a lot the banks can do. Rachel Wolcott has been looking into the problem - and some possible solutions - and her article appears in the next issue of Risk.

Updated on November 6 - the article is now available online here.

More on the structure and pricing of CMS steepeners here. Discuss CMS steepeners in the Risk Forum.