The foreign exchange industry got an early reminder of risk when on the second full day of trading this year the market saw another flash event. What, if anything, does this mean for FX market liquidity and volatility in the year ahead though? Colin Lambert finds out.
The very sharp moves seen in FX markets at the start of the year triggered yet another round of introspection over conditions in the FX market with commentators pointing the finger of blame at one or more of algos, news from Apple, thin markets, Japanese retail and poor execution. Although Profit & Loss understands that industry players have been approached for data logs by certain regulators, the chances of an investigation turning up a convincing catalyst for the moves are thin.
It was actually just 30-40 minutes after New York markets officially closed for the first trading day of the year when FX traders were given their nasty reminder of both the fragile nature of liquidity and the markets’ propensity to panic, when a yenbuying spree took place seeing several other currencies, most notably AUD, GBP and TRL falling sharply. In just seven minutes, the Australian dollar fell by 3.5% against its US equivalent, while the Turkish lira fell over 5.5% and sterling 1.6%. At the same time, the Japanese yen rose by just over 3.7%. Even what many FX traders consider the bastion of boredom – EUR/USD – moved, falling just under 1%.
The moves were quickly reversed to about 50% and then lost ground was steadily made up until just 12 hours late,r the AUD was trading at or above pre-crash levels and the yen was giving up pretty much all of its gains. That the euro and sterling recovered their respective losses much quicker has led some to speculate that what was seen was a clear out in yen crosses.
“The selling definitely started in AUD/JPY,” says a trader at a bank in Australia. “The amounts weren’t huge but the selling was relentless and at that time of day liquidity just isn’t there.”
While agreeing that yen crosses were the trigger for the moves, a senior e-FX trader in Asia also believes the lack of risk appetite in the window as well as the slightly different behaviour of the pricing algos in what is widely termed the “witching hour” after New York closes may have played a role. “No-one trusts the market in that window and the pricing algos tend to react as much to the pace of selling as the volume. Ally that with the lack of risk appetite at that time of day and you have a recipe for some sharp moves.”
It may not all be bad news, however, for several trading sources report, as one puts it, “a great start to the year”, thanks to the opportunity offered by the event. “We had plenty of customers putting bids and offers in the market as the event progressed,” says an FX salesperson in Asia. “Our own traders also got involved and they all had a pretty good day. One active trading client told me they made in 15 minutes what last year they would have expected to make in two months. If you’re alive to the opportunity, the rewards are there.”
Inevitably the quick nature of the move triggered debate about the benefits of human traders over algos and indeed sources at FX trading platforms say the percentage of GUI interaction went up “significantly” while the API traders’ involvement fell. “This is to be expected because the machines don’t guess where the market is and a lot of our automated trading volumes come from market makers, who widen or pull back,” explains a senior source at an FX platform. “Part of the GUI uptick was undoubtedly traders seeking to profit from the reversal, but I also think that a lot of desks sit on customer orders that are a way off the market and when they see the market move they quickly put those resting orders in.”
The combination of opportunistic traders and resting customer orders meant that conditions in the market did not deteriorate to the levels seen in October 2016 when sterling dropped more than 12 big figures in a matter of minutes. “We learn from these events, and one of the things we learned from the sterling crash was how you need people on the ground to exploit the opportunity,” says the head of FX trading at a major bank. “We actively encourage our traders, especially in Asia where liquidity can be patchy, to take risk on during these moves.”
Crucially, the FX trading head observes, however, that taking risk on does not mean assuming it from clients. “Good traders are more proactive than that. They analyse the situation quickly, fill client requests as they come up (although most are handled by the machines), but if they can’t see a fundamental reason for what has happened, they act. It’s why a strong trading business has a healthy balance of man and machine – there’s a time when one is more useful than the other.”
Looking ahead, the FX trading head believes greater focus on the opportunities afforded by flash events at the expense of merely handling customer flow, will mean the scale of the moves themselves will diminish. “Don’t get me wrong, we will still have some big, quick, nasty moves, but unless there is a real reason behind them – the UK leaving the EU being one example – traders will be quicker to trade against it and that in turn will provide stability.”
Not all have such an optimistic view, however, for as one relationship manager who focuses largely on retail accounts says, there is a structural reason why we will see more of these events. “While it’s not fair to say that the brokers squeezed their clients out of positions on January 3, although I am sure some firms on the edge of respectability do that every chance they get, the brokers generally are more sensitive to the clients’ exposures since the Swiss franc event,” the relationship manager says. “The problem the FX market has is that Japanese retail will forever be chasing carry and that means short yen and longer term positions. If you’re long AUD/JPY or TRL/JPY as so many are you are not going to cut the position every night – you lose the carry which is the reason for your trade in the first place.
“This means that the most popular trade of a pretty large market segment is structurally ‘risk on’, which means every time something nasty happens – such as the Apple warning – these positions gets cleared out,” the relationship manager continues. “Until we see Japanese investors change their strategy, the FX market will always be vulnerable to ‘risk off’ in other markets – and don’t forget that corporate advisories such as that from Apple, come after the New York close – right in the most illiquid window in foreign exchange.”
Not everyone is buying the theory of a clear out of yen shorts, however. The research team at RBC Capital Markets told clients that its positioning aggregate suggested USD/JPY positioning may not be as long as the IMM data in isolation portray (it did note the data were old as the last available was from December 18).
RBC’s analysts also noted that currency managers’ returns (using the Parker index) have recently been entirely uncorrelated to moves in spot USD/JPY, which, they say, suggests discretionary positioning is light. They also observed that risk reversals were bid for puts to an unusual degree at the time. “The normalised average of these three measures – our aggregate positioning measure – is close to neutral,” RBC said. “Of course, this won’t stop USD/JPY selling off when markets are as risk averse as they have been this week, but reports that investors are “all the wrong way” are not a convincing explanation for the size of [these] moves.”
Another popular catalyst to emerge in the immediate aftermath of the move was that risk parity strategies, that have become more popular in recent years, could have prompted events. Sources point to the increasingly popular risk-neutral strategies in the market and observe that the Apple warning and subsequent sell off in after hours trading of its stock, would have triggered these.
“Risk parity strategies react to volatility and when equity vol climbed post-Apple these funds had to rebalance and that only made things worse,” observes a senior manager at a US hedge fund. “My only question would be how this filters through to FX markets? I am not sure of the connection there, but there are so many correlations out there now I wouldn’t be surprised if a few people had them.”
Another trading source points to the hedging strategies of the non-bank market makers as a potential link, saying, “Just as these firms hedge their FX risk in non-FX markets, so they are increasingly making markets in these other products and perhaps they are hedging back into FX and that’s why things went sideways on Jan 3.”
A source at a non-bank market making firm doesn’t rule out the connection but doesn’t give it much weight. “This is putting two and two together and coming up with three. I don’t think there was a link between the Apple news and the FX moves beyond it was a trigger for ‘risk off’.”
Although the price action was undoubtedly violent, there are some who are questioning whether it should be accorded “flash” status. Of course, actually defining a “flash” event is no easy task itself, but conventional wisdom is that it involves a quick, multiple standard deviation move in a limited number of assets that is quickly reversed – during the sterling flash event for example, other FX pairs were relatively quiet. A flash event in these circumstances is simply an extreme supply/demand imbalance or a liquidity problem.
What was interesting about the event on January 3 was that even EUR/USD moved. As one market maker tells Profit & Loss, “Everything went crazy, it wasn’t just the yen.”
While this may be true to an extent, other traders point to the crosses again as the culprit. “Yes, EUR/USD dropped sharply but it was people trying to sell EUR/JPY through the legs,” says a trader in Asia. “It was the same for all the majors that moved, people weren’t pricing them normally, they were pricing for a sharp drop in all yen crosses.”
There is also the question of where the positions were. In a note to clients on January 2, market information firm InTouchFX observed that retail traders were long Turkish lira against yen and, with a nice understatement on January 3, pointed out that, “These positions will not have been easy to get out of, especially with Tokyo shut.”
If this was the case then could an argument be made that this was not actually a flash event, rather it was a reasonable market reaction to a fundamental change in circumstances – with just a little bit of old-fashioned panic thrown in for good measure. “Given the hour I think the market handled it well,” says a source at an FX platform. “There were good volumes, plenty of resting orders all the way down, the only thing that wasn’t great was the second-to-second pricing. The market found pockets of liquidity as it hit the resting orders but then gapped after working through them.
“Market makers were pretty quick to widen out or shut down – several were actually in the middle of their daily re-boot after the New York close – and until they came in again liquidity was fragile, but you could get stuff done,” the source adds.
The market making source accepts that they “took a P&L hit” during the event, but adds that in the bigger picture it was nothing out of the ordinary and was a few percentage points of a good P&L day. “And the day was still young,” they add.
There is also the question of, as one source asks, ‘can you actually have a flash crash at a time when there is rarely any real liquidity?’ Certainly one can argue that trying to execute anything other than the smallest amounts in the hour or two after New York closes is foolhardy – especially given so many firms now provide intra-day liquidity analysis. “We can pretty accurately judge how much can go through the market at any time of day without unduly impacting it, and in early Asia that number is typically not high,” says a liquidity analyst at a bank in London. “We advise our clients to execute in this window only in the most desperate of circumstances if they have any size of note to do. The problem is many of our systematic accounts don’t have that flexibility and that means they are tiptoeing through a minefield just about every day at that time.”
Notwithstanding the question of whether or not executing trades in this window is prudent, there seems little doubt that at some level this was a liquidity event rather than economic. A brief look at markets some three weeks after shows that not only were the original moves reversed fairly quickly, but things have stablilised since then.
In a note to clients on the day of the flash moves, NAB’s head of FX strategy Ray Attrill wrote, that while the moves did have an “initial catalyst” in the Apple guidance, the 6% drop in the firm’s shares was only enough to see AUD/USD extend its overnight losses from the 0.6982 New York session low down to about 0.6975. “The subsequent rapid fall was to around 0.6730,” he continued. “The earlier move though 0.70 (for the first time since February 9th 2016) could be reasonably attributed to a combination of a renewed bout of risk aversion (lower US stocks and higher ‘VIX’) and sharply lower base metals prices. Aluminium and copper were both off by more than 2%. But this can’t be viewed as justification for the scale of the subsequent price action in both AUD/USD and AUD/JPY.
“The fact that over half the move down in both these pairs has since been retraced is testimony to today’s moves being first and foremost a ‘liquidity event’,” he added.
If that is indeed the case, then the FX industry has something of a quandary in that it has a two-hour window during which it is exposed to liquidity events, but can seemingly do little about it. For all the hope, as expressed earlier, that traders will react to the moves and therefore provide an automatic dampener, there will still be an initial sharp reaction that is likely to hurt those positioned the wrong way.
Some believe that central banks – as the custodians of market stability – should heighten surveillance of FX markets in the witching hour and be prepared to act to provide liquidity as and when necessary. The FX trading head points out, however, that central banks will not react quicker than traders at banks and trading firms “probably slower” and therefore it is best left to the market to self-regulate itself. “If someone is silly enough to chase the market down then they deserve what they get,” the trading head says a little derisively. “The problem is when people are knocking out client orders in these moves. Of course they have to execute orders as the market gets to the level but there is more than a suspicion that occasionally someone is targeting this window to trigger stops or barriers. I think the central banks need to be more focused on this than actually trying to stabilise markets.”
Whichever way one views it, the FX market could see more of these types of moves in the so-called witching hour. For whatever reason, markets are particularly sensitive at the moment and viewing the geopolitical picture in particular suggests that things are not going to calm down any time soon.
This may require a shift in thinking amongst some market participants who have grown over the last decade thanks to strategies borne out of a low volatility environment. As volatility recovers and occasionally spikes, there may well be a shift on the part of some players towards active risk taking rather than passive market making. The change will not be dramatic but it could happen. “Banks especially like to move as a herd, so as one or two do well in these events so the others will want to grab their chance,” observes the FX trading head. “I don’t think the machines will lose share – they’ll continue to gain it as long as markets remain sensible – but we will probably see a few more human resources thrown at the issue, especially in Asia.”