Dutch Speed-Trader Turns to Currencies After Conquering ETFs
Flow Traders wants to colonize markets with its approach to trading, but its core business is coming under attack
The bottom shelf of the fridge is laden with Heineken and Corona. The Corona is on rotation, but the Heineken is a permanent fixture: This is Amsterdam. A few strides away there’s a dark, well-stocked in-house pub.
Up one flight of stairs, the atmosphere is very different. Behind a door that can only be opened with a security pass is by far the largest trading floor for exchange-traded funds in Europe. The 110 traders here, along with 30 colleagues in offices elsewhere, traded €640 billion ($719 billion) in ETFs last year and at least that much in futures, commodities, bonds, stocks, and foreign exchange.
The trading volumes are those of a major Wall Street bank, but the refrigerator—and especially the pub, with its arcade games, pool table, and giant television—is pure startup. This isn’t an investment bank; this is Flow Traders NV, one of the world’s most successful algorithmic trading firms.
Flow, run by co-Chief Executive Officers Dennis Dijkstra and Sjoerd Rietberg, is also pretty unusual for a speed trader. While most firms allow some risk in their trades, Flow uses a more mathematically intensive (and expensive) approach to eliminate as much risk as possible. The pair make this clear during a joint interview in Flow’s sleek, glass-enclosed boardroom. “We do not like to take risk,” Dijkstra says. “When Flow Traders was formed,” adds Rietberg, “the idea was to have a good night’s sleep.” Dijkstra even admits the company name is misleading. “We are not traders,” he says. “We are operators.”
This year the firm will undergo its biggest transformation since it opened for business in 2004. Flow, which handles about a third of all ETF trades in Europe, is seeking to do to currency markets what it’s already done to its core business. The firm’s strategic calculation is that high-speed foreign exchange traders should be able to offer better prices than banks, which typically adjust their bids and offers based on their customers’ creditworthiness and the amount of business they do with the lender.
Rietberg and Dijkstra say the move into currency trading is a natural evolution of the business. That may be, but Flow also badly needs to find a new way to grow: Its shares languish below €32—the price when it went public almost two years ago. In the first three months of 2017, its profit dropped 41 percent as quiet markets reduced its ability to earn money from trading. What’s more, Flow needs to adjust to a changing landscape. In some markets, the pool of income available to algo trading firms is shrinking as competition increases from established trading companies looking to expand into new asset classes.
Flow is at a crossroads. Its distinctive approach to algorithmic trading could enable it to colonize other financial markets—or it could shrivel as rivals attack its core ETFs business. “Flow could be multiple times the size it currently is in 20 years,” says Joost de Rijk, an analyst who covers the company for Amsterdam-based merchant bank Kempen & Co. “But it could also be gone. I think they will reevaluate every year whether this is working. That’s the DNA of the company.”
The market flow does battle in is wildly different from the one that its founders faced 13 years ago. Roger Hodenius was working as an options trader at Optiver Holding BV, a computerized trading firm, when he realized that ETFs, then in their early days, would become a major asset class in their own right. He joined forces with Jan van Kuijk, who’d also worked at Optiver, to set up Flow.
Their timing was fantastic. Amsterdam’s Financial Markets Authority didn’t even begin to discuss high-frequency trading until 2006. Optiver didn’t set up its own ETF desk until a year later. Meanwhile, investors had begun shoveling money into ETFs, with assets under management rising from $319 billion in 2004 to $2.4 trillion in 2013, according to BlackRock Inc.
In 2014, after a decade leading Flow, Hodenius and Van Kuijk handed over control to Rietberg, a former head of trading, and Dijkstra, then chief financial officer. Van Kuijk, who’s 50, stayed on as vice chairman. Hodenius became an adviser; now 45, he’s moved to Ibiza, Spain, where he owns a hillside villa.
Rietberg, who’s been with the firm since its second year, worked his way up from the trading floor. Sharp and intense, the 37-year-old still stalks the trading desks, studying screens and quizzing traders. Dijkstra, a 45-year-old veteran of Dutch bank NIBC and a career CFO, handles the firm’s investors.
As part of its planned foray into the currency markets at the end of this year, Flow has hired a foreign exchange trader from a Wall Street bank to improve its knowledge of the asset class. (The company declined to identify the person because the FX project is still at an early stage.) Flow already takes FX prices from other market makers to hedge its ETF positions. Now it wants to make prices as well, a step that would see it become a market maker for currencies.
Dijkstra discloses that, after FX, Flow plans to expand into bonds. The plan is the same as for currencies: Hire a specialist from a major bank to get started, then turn things over to the existing trading desks.
Announcing its first-quarter results in May, Flow reported that it hadn’t lost money on a single trading day in the preceding 34 months. It attributes the stellar run to its use of deterministic modeling, which produces definite outcomes, not probable ones. Most of its rivals calculate prices by means of statistical, or stochastic, modeling, producing hedges that probably (but not definitely) protect them from any downside.
Deterministic modeling has its downsides: It’s less profitable, the math is harder, and if any mistakes are made with the inputs to the model, the outputs will also be wrong. That means Flow has to trade more than its rivals to compensate for its smaller margins. When you make only 0.028 percent on a trade—Flow’s average for the first quarter—you need to make a lot of them. “It’s about shaving off the one cent,” Dijkstra says. Rietberg adds: “We are transferring risk. We trade against A and B. We make a very, very small margin on these products.”
Can it really be that straightforward? “On the surface, it sounds simple,” says Kevin McPartland, managing director, head of market structure at Stamford, Conn.-based Greenwich Associates. “But it must be much more complicated under the surface. It has to be. It almost feels like a philosophical question to me: Can you ever be perfectly hedged? There’s never no risk. There always has to be some risk of market dislocation.”
Every ETF trader has to hedge positions using the underlying securities or futures contracts. That exposes Flow to an unavoidable source of risk: A problem with the structure of financial markets could prevent it from buying or selling at a key moment. Kempen’s De Rijk says Flow has a good track record of containing risk even though it can’t be eliminated. “They never seem to have any losses due to errors in hedging or pricing,” he says. “Their models and estimates are very good, because otherwise we would see them making some mistakes. I think the errors are very small, but there will always be some profit-and-loss risk.”
Aside from risk, De Rijk identifies a second difference between Flow and its rivals: Flow buys and sells the ETFs that no one else will touch. It specializes in funds where at least some of the underlying assets aren’t trading due to time-zone differences. The IShares MSCI BRIC ETF, for example, tracks an index of stocks from Brazil, Russia, India, and China. Chinese stock markets are never open at the same time as Brazil’s BM&F Bovespa, but Flow can still trade the fund. Flow will adopt a similar strategy when it moves into fixed income and FX by concentrating on hard-to-trade bonds and currencies outside the 10 most traedwinded.
Difficult-to-price ETFs are also the most profitable to trade. Flow kept its lead in European ETFs even after Optiver and IMC BV—the third Amsterdam-based speed trader—began making markets in the funds. The giants of U.S. ETF trading, such as Citadel Securities LLC, mostly concentrate on their enormous home market. U.S. ETFs are sitting on $2.7 trillion—at least four times more than their European counterparts.
The third reason for Flow’s dominance of European ETF trading came to light when Flow listed its shares on the Euronext Amsterdam stock exchange two years ago. At the time, the firm had to make certain disclosures, including financial documents revealing that since 2008 it had built a network of major trading counterparties. These “direct counterparties,” as Flow calls them, use the firm to trade large blocks of ETFs over the counter. It takes time to build such a network of pension funds and asset managers. Flow has some 550 of these counterparties. Optiver, which started copying Flow in 2015, has almost 200.
Big Dutch financial players tend to congregate around World Trade Center Amsterdam, a large office complex south of downtown. Optiver, IMC, and Kempen are there; ABN Amro Group NV, the Dutch bank that clears trades for Flow and other high-frequency traders, is nearby. Flow, by contrast, shares a building with several newspapers at a reclaimed dockyard 6 kilometers or so to the northeast. Why isn’t Flow clustered with the others? “Because we’re not part of the Establishment,” Dijkstra says.
At Flow, traders who want to skip the Heineken can hit the fully kitted-out gym off the trading floor. On one wall there’s a photo of Dijkstra and Rietberg with Edwin van der Sar, the imposing Dutch goalkeeper who won soccer’s UEFA Champions League twice, once with local club Ajax and once with England’s Manchester United. The “Ice Rabbit,” as he was nicknamed for his unflappable temperament and quick reflexes, is the most famous of a group of people from the world beyond finance who’ve come to the docklands to coach the firm’s board on how to win.
After all, Flow Traders isn’t a bank. To avoid thinking like one, sometimes it needs a little outside help.
Hadfield covers market structure in Europe. With assistance from Tanvir Sandhu and Joost Akkermans.