 |
||
|
Jon Macaskill is one of the leading capital markets and derivatives journalists, with over 20 years’ experience covering financial markets from London and New York. Most recently he worked at one of the biggest global investment banks |
 |
|
The dislocation in US equity trading on May 6 has caused much wringing of hands among market participants and supervisors. It was an unwelcome reminder of how little regulators know about the markets they oversee and how complacent industry insiders become when innovation takes place against a backdrop of steady or rising prices.
Like many changes that cause dislocation the move towards fracturing of the US equity markets took place in plain sight. Hedge funds such as Renaissance do not volunteer details of their algorithmic trading systems but dispersion of stock liquidity to multiple competing venues occurred openly. Most big banks now openly tout their ability to route trades at speed to the exchange or platform with the best price, even if they do not advertise how their own trades or those of their highest-margin clients feature in the pecking order for execution. And limit-order systems are standard even for retail investors.
That made the sudden loss of liquidity on May 6 all the more surprising for market participants.
After markets span out of control soon after 2.40pm, half of all stocks on the S&P 500 suffered an intraday loss of at least 10%, the DJIA dropped almost 1,000 points, nearly $700 billion of value was temporarily erased and several individual stocks and exchange traded funds (ETFs) saw trades print at close to zero.
A report on the crash from the SEC and Commodity Futures Trading Corporation 12 days later managed to be both lengthy – at 151 pages – and surprisingly light on conclusions. The regulators ruled out fat-fingered mistaken orders as a cause then followed the hallowed approach of supervisors everywhere by announcing that further investigation was required.
Some practical steps were announced in the form of a new system of circuit breakers that are triggered by gains or declines of more than 10% within five minutes on S&P 500 companies, however.
Credit Suisse’s AES subsidiary – itself one of the biggest algorithmic traders – conducted an analysis of the impact the new circuit breakers would have had on market movements since the start of 2008.
It found that triggers would have gone into effect fewer than 10 times a day on average, except during the period of greatest markets stress between September and November 2008. In October 2008 triggers would have been breached an average of 40 times a day, but between January 2008 and May 2010 over half of all days would have been trigger-free, while on 80% of days there would have been three or fewer triggers. AES noted that circuit breakers would not prevent big jumps at the beginning and end of the trading day and pointed out that trigger breaches could potentially disrupt natural price discovery whenever news releases such as bid announcements happened to fall within standard trading hours.
The new circuit breakers might help to prevent further flash crashes in the near term but they do nothing to provide clarity on the interaction between players in a market that is both increasingly electronic and fractured.
The flash crash demonstrated that US stock trading is spread across about 50 venues. It also highlighted the extent to which an illusion of liquidity in stocks and options is provided by firms whose main incentive in posting two-way prices is the rebates they are paid by exchanges. This rebate harvesting promotes churning of trades in orderly markets without carrying an obligation to make meaningful two-way prices during times of turmoil.
The downside of trading in ETFs was also made apparent during the crash. Almost 70% of the trades that were busted by exchanges in the wake of the crash were in ETFs, with trades in individual stocks such as Accenture only accounting for 22% of the trades that were cancelled because they took place more than 60% away from their 2.40pm level. The plunge in ETF values relative to the movement in their component stocks suggests that ETF market makers simply pulled their prices when quotes fell.
The disorderly trading in hugely popular and supposedly liquid equity instruments such as ETFs does not inspire confidence in a potential expansion of electronic trading to fixed-income markets.
Electronic trading of credit instruments could be particularly problematic. The credit market is both much bigger than the equity universe – around four times the size even without allowing for all derivatives – and far more complex.
Most of the multiple thousands of individual cash credit obligations barely have any notional liquidity.
Even among benchmark references there are few individual corporate bonds, loans or equivalent single-name credit default swaps that are genuinely liquid. Default swaps retained liquidity better than cash instruments during the 2008 credit crisis, which had the unfortunate side-effect of exacerbating the turmoil in the markets as major dealers and hedge funds such as Citadel saw their assumptions about bond against default swap basis levels overturned, leading to huge losses.
The CDX and iTraxx indices of default swaps were the most liquid credit instruments in the crisis and continue to trade in high volumes at relatively tight bid-offer spreads despite the recent return of volatility to most markets.
Big dealers are accordingly trumpeting their steps to direct CDS index trades to electronic systems and central clearance. This serves to address concern among regulators that banks have been dragging their feet about shifting derivatives to electronic settlement, while concentrating the move among instruments where there is relatively little margin to lose. It also provides helpfully large notional figures to cite when detailing progress towards use of electronic systems.
The liquid rates and foreign exchange markets provide a more encouraging indication of how electronic trading can come to fixed-income markets.
After battling for years to maintain their margins by resisting adoption of electronic systems, the main dealers have finally embraced online systems and reconciled themselves to the inevitability of central clearing of standard derivatives such as interest rate swaps. Dealers are reacting quickly to the stick of regulatory change in part because they now see a carrot in the prospect of substantial volume increases as rates and FX trades go online. Efficient online trading systems for rates and FX instruments can also be used to win or defend market share from rivals, which addresses the paramount goal for most bankers.
Even interdealer brokers, who might be expected to be the last holdouts in the battle to keep trades away from electronic systems, are now embracing change. Seventeen percent of Tullett Prebon’s revenue in 2009 came from trades supported by its electronic platform, and Icap reported that 15.7% of its revenue and almost a third of its profit for the year to March 2010 were from electronic broking. Post-trade risk and information combined with electronic broking to total 48% of Icap’s profit, underscoring the shift in priority from voice trading at the firm, at least for its managers.
The rates and FX markets provide an example of how liquid markets that are dominated by wholesale players can successfully go electronic. A similar shift for credit is likely to be much more problematic.
