That was when the Swiss National Bank introduced mayhem to the FX market with its unexpected decision to scrap a policy of capping the Swiss franc against the euro. For three years the market had seen the central bank keep a lid on its currency, only for the franc to move more than 40% and surpass the Russian rouble as the world’s most volatile currency. The move was likened to a nuclear explosion. Market veterans have said they’ve not seen anything like it.
What will the decision mean in terms of the popularity of FX algos? The short answer is that it depends. In the hedge fund community, there are those who argue that reliance on algos is what rescued them. For corporates, at least in Switzerland, there is now an argument for much more active currency management. And despite a long list of market casualties from the SNB reversal, there have been no prominent reports of algo-based mishaps.
“It’s becoming more and more popular in general to use algos.”
Christopher Cruden, the CEO of Swiss-based fund Insch Capital Management SA, said his fund’s systematic approach was vindicated by a decision not to intervene and overrule the trading suggestions of its algorithm.
“If you’ve got a set of rules, the important thing is to follow those rules. If you don’t follow them then you don’t really have rules at all,” Cruden said.
Insch, he said, had been up slightly more than 4% for the month just before the SNB news and soon found itself down about 2-1/2%. “That’s a sizeable move for us, considering that our gearing is very, very low,” he said.
“Our decision at our 2 o’clock conference call was to follow the system, so that’s what we did. And we are currently down I think about half a percent,” Cruden said. “I’ve been doing this for 35 years – that was outside my experience. But our algorithm allowed us to treat it in exactly the same way we treat every other day.”
Andy Flury, CEO of AlgoTrader, which is also based in Switzerland, said it will take time for the trading community to assess the full implications of the central bank move.
One change to be expected will be a shift away from mean reversion strategies which were obviously common under the old Swiss central bank policy regime. “Whereas now I would assume trend-following/momentum-type strategies to become popular again on the EUR/CHF,” Flury said.
AlgoTrader, which started as an in-house solution for a volatility hedge fund in Switzerland, markets software for hedge funds, trading groups, proprietary traders and family offices. The software handles a number of tasks, including backtesting execution, reconciliation and reporting.
“We built our own software because we couldn’t find anything suitable and only about two years ago we started selling the software independently to other hedge funds,” he said.
While most of the firm’s clients are in the US and to a lesser extent Europe and Asia, it does have some customers in Switzerland.
“Generally speaking, if I may compare our US customers with our Swiss customers, what I see is that Switzerland is generally a lot more cautious and conservative with algorithmic trading,” Flury said.
“our algorithm allowed us to treat it in exactly the same way we treat every other day.”
That said, Flury said algos were gaining appeal in Switzerland. “It’s becoming more and more popular in general to use algos. Switzerland is probably a bit behind the US. But I think Switzerland is also catching up.”
Cruden of Insch Capital said that in effectively ending a period of strong coordination among central banks, the SNB is now ushering in a return of interest rate differentials and a period of larger moves.
“And that environment is precisely the sort of environment in which rules-based systems trading will best work,” he said. “In other words, I would say the last five years has not been a quote-unquote normal environment, because of the sort of linked-arms solidarity of central banks. Now that appears to be broken, we will return to what I would say is a more normal environment, which is much, much better for people like us.”
While Switzerland is known for its fund community, the country also boasts a large and active corporate sector.
Sebastian di Paola, a partner at Pricewaterhouse Coopers AG in Geneva, said Switzerland has a diverse group of multinationals of various sizes, many of whom are exporters which will be affected particularly by the downward movement of the euro against the Swiss franc. Some, such as those in the luxury goods sector, may find it easier to pass on shifts in the cost base, including FX, to customers than for instance those in a traditional industrial manufacturing.
At the same time, Switzerland also hosts the headquarters of numerous foreign multinationals, either at a regional or rest-of-the-world sense. “Often that headquarters includes a treasury centre which will be responsible for certainly the European FX hedging, and often those structures also include what are called principal companies, which involve the centralisation of trade flows, including the risks associated with those trade flows. So the FX exposures are actually centralised in the Swiss entity very often from where they can be hedged by the regional treasury department,” di Paola said.
The PwC partner characterised many of these corporate treasurers as sophisticated in how they manage FX risk and he and others suggested that their levels of activity were likely to increase in the wake of the SNB move.
“This move by the national bank was not really expected so I think it has taken both traders and the corporate treasurers by surprise,” di Paola added.
Sebastian di Paola
“This move by the national bank was not really expected so I think it has taken both traders and the corporate treasurers by surprise.”
Thus far, despite the reputation for sophistication, these treasury departments do not appear to have embraced algo trading for execution. “Yes, they’re doing a large share of their hedging through electronic dealing platforms, but I’m not seeing much of what you would term algorithmic amongst corporate treasuries,” he said, quickly adding: “which isn’t to say that they shouldn’t.”
So in that sense, Switzerland offers a new opportunity for algo providers.
Shai Heffetz, MD of FX broker InterTrader, agreed that Swiss corporates will have to be more active in managing their FX exposures. “Absolutely,” he said in response to a question as to whether the policy change will lead to more hedging. “There has been (before now) zero volatility in the EUR/CHF due to the peg.”
The new normal
Still, Heffetz, who called the SNB move a “game-changer” for the industry, said it could have implications for the popularity of algos simply because it would make market participants more anxious.
“Algorithms work within what you call a normal course of action. And then there’s the abnormal. If you look at major currencies, you can expect a1% amplitude volatility intraday,” he said.
One could play it safe and plan for three or four times that volatility. “But I don’t think people have built in the scenario of 20 times that volatility, which is what has happened. A major currency was depreciating by over 20% in minutes. It’s almost unheard of since free-floating currencies have started. Algos are not built for that and a lot of people who are designing them need to rethink how they approach it.”
He said this was the case for both alpha-generating and execution algos.
“The market was completely interrupted, and at that point you needed extremely strong exception handling built into your algorithms in order to have them continue to operate normally,” Heffetz said.
“A lot of algorithms, I believe, got hung. The algorithms did not know how to treat this kind of input. I can tell you from what I know, a lot of systems including algorithms have treated the first few ticks, when the market came back, as what you consider spikes that you need to filter out.”
In the case of the SNB news, many firms had to manually release their systems to allow trading to take place again, he said.
“The market was completely interrupted, and at that point you needed extremely strong exception handling built into your algorithms in order to have them continue to operate normally,”
Among the implications from the Swiss bombshell will be that algo developers will need to rethink how they handle exceptions.
InterTrader includes professional traders and funds among its client base, as well as the retail market. It also hosts algorithms for some clients that want to reduce latency when trading through its platform.
On a day rife with market casualties, InterTrader was one of numerous firms that issued a statement on the day of the Swiss decision to say it was unaffected by the news and continuing to trade.
“What makes this SNB decision so consequential is not so much the depegging of the Swiss franc from the euro, but rather that it was such a massive move – one for which many key players in the market were not well prepared – and the aftermath is like a black hole that can suck massive amounts of credit from currency trading as we have known it,” said Javier Paz, a senior analyst in Wealth Management at Aite Group.
Paz said brokers with an agency model were particularly exposed to the evaporating liquidity and major losses in the immediate aftermath of the SNB announcement. As for major banks and institutional trading firms, he cited analysis by Goldman Sachs and Morgan Stanley as estimating the SNB measure would shave at least 10% off the pre-tax income of Swiss banks UBS and Credit Suisse as they appear to have franc-based loans to close to 700,000 borrowers in Poland and an undisclosed number in Croatia.
And for corporate treasurers, the Aite analyst said the SNB measure “ups the ante” as they now know they must be prepared for central bankers’ sometimes disruptive behavior.