A new paper published by the UK’s Financial Conduct Authority (FCA) claims to throw new light on events surrounding the sterling flash crash of October 2016 by being the first paper to use trade reports to the FCA under EMIR to analyse how different market participants react in times of market stress and their impact on the liquidity dry-up in a flash crash.
The paper has, however, triggered some confusion amongst market participants thanks to ambiguous terminology, mainly the constant reference to “OTC derivatives”, without specifying exactly what products it is talking about. For example, the authors say, “…our paper is also the first to investigate the impact of derivatives on the underlying spot market during a flash crash”, which suggests that spot dealers were observing the sterling “OTC derivatives” market for signals, when it is generally understood that derivatives such as FX options, are priced off the underlying spot market and as such would probably have reacted according to events there, rather than the other way around.
Other OTC derivatives markets mentioned by the authors are the FX outright and swaps markets, the latter is priced off the interest rate differential between the US and UK and therefore less sensitive (but not entirely so) to the spot market.
The report highlights three main behaviours that are found to have contributed to the drying-up of liquidity during flash crash.
The first observation is that dealers withdrew liquidity by widening their spreads and in some cases withdrawing from the market altogether. The authors say their research shows that dealers’ trading activity led to higher transaction costs (the round-trip costs were about 60 times higher during the flash crash compared to normal times) as they charged higher bid-ask spreads when providing liquidity to their clients.
The paper also argues that dealers “impacted trading volume negatively” as their reduced trading activity led to the drying up of liquidity during the flash crash (their trading volume was less than 1% of its average level during normal times). It adds that other financial firms such as hedge funds and asset managers stepped in during the flash crash period and provided liquidity by taking long positions. “However, they offered to buy at less competitive prices,” the paper states with no apparent irony.
Other financial firms were involved in 98% of the traded volume during the flash crash period compared to 35% in normal times, the paper finds, and observes that about 55% of these trades were with non-financial firms who mainly took the short position during the flash crash, while 38% of the trades were with each other. ?
The paper also finds that the inter-dealer part of the Cable market, “which is exclusively used by dealers to hedge? their client trades with each other”, “collapsed almost completely” during the flash crash period. During normal trading, this part of the market accounts for 61% of all transactions, but this share fell to just 2% during the flash crash, the paper states, adding that the absence of this key market during the flash crash meant that dealers could effectively hedge only 31% of their client trades during this episode. This is turn may explain why they withdrew liquidity in the non-interdealer part of the market.
The paper further explains that without the inter-dealer market, dealers have to face the inventory holding risk for every transaction undertaken. Dealers in OTC markets are only willing to accumulate additional inventory during times of stress if there are large price concessions. “This induces a downward pressure on prices and potentially also withdrawal of liquidity,” it states.
This is pretty much in line with other research and anecdotal evidence from dealers present during the flash crash, but where things get confusing is when the authors argue that, “the existence of the FX OTC derivatives market in the spot rate for GBP/USD amplified the initial effects of the flash crash in the underlying spot market”. They say their research shows cross-market effects and bidirectional causalities between liquidity in the OTC derivatives market and its underlying spot market.
“The channel for these bidirectional causalities is that dealers in the derivatives market learn from the underlying spot market (and vice versa) and this can cause a feedback loop in illiquidity between the two markets,” the paper states. “We can confirm this amplification channel via transaction costs, price dispersion and trading volume.”
One sterling dealer who was working during the flash crash, tells Profit & Loss, “All I remember is my options desk screaming at me asking what spot was, I never once asked them what was happening there because, quite frankly, they rarely dealt at that time of day, they wait for mainland Asia to come in – there is no real options market in sterling at that time of day.”
The confusion isn’t helped by the FCA, on its website, offering some rather strange statistics in its introduction, which claims that dealers, “reduced their trading activity in OTC derivatives from normal levels of £32 million per second over the whole day, to just £0.2 million per second during the flash crash.”
Those “normal levels” would equate to average daily turnover (ADV) in sterling alone of £2.7 trillion, or, roughly, $3.7 trillion – or more than 70% of all flow, according to the BIS triennial survey. Even accounting for double counting, sterling OTC derivatives – not spot of course – ADV would be about $1.9 billion, far in excess of the total sterling ADV, across all products, reported by the BIS of $650 million.
In the body of the paper, the authors state that average OTC derivatives turnover is £5.3 million per second, which presents a much more comparable average per day of £477 billion, or $630 billion, but again, they appear to be talking about the ambiguous derivatives and not spot, or both.
An underlying confusion with the paper is the constant references to “OTC derivatives” without specifying exactly what product is being spoken about (the authors regularly refer to “spot markets” thereby highlighting it is treated differently). One example is where, when discussing spreads and average prices, the report states, “At 11:07pm on 6 October 2016, the average OTC derivatives price for GBP/USD decreases from 1.27 to 1.21. After about 10 minutes it recovers to 1.24, and then continues to rise slightly for the rest of the hour.”
This could be observations on the spot price used to price FX options during the event, however, it is not immediately clear, for as noted by the FCA, the vast majority of sterling FX OTC derivatives are in forwards and swaps. It could be that the price used is the average three or six month outright, but again the pricing does not match that in the markets at the time, when six month Cable swaps were trading around 45 points.
It is hard not to suggest that the authors have an agenda, which is to advance the argument for shifting FX trading onto an exchange. A section of the last paragraph would seem to support this supposition, stating as it does, “A number of steps are already being taken by authorities to limit the impact of flash crashes. These include mandating that certain derivatives trade on more transparent exchange-traded venues rather than primarily on OTC markets, and requiring those venues to have appropriate systems and controls in place to manage excessive volatility, such as circuit breakers where applicable.”