By Gary Stone, Chief Strategist, Bloomberg Tradebook.
The French phrase déjà vu literally means, “already seen.” Have you ever gone to the cinema and felt that you have seen the story before in a different film? Well, Dances With Wolves (1990), FernGully (1992), Last Samurai (2003), District 9 (2009) and Avatar (2009) are different films, however their plots all resemble one another.
Sequels are another example. Story or plot “resemblance” is actually a common occurrence in the film industry — perhaps because producers believe that if the story was successful before, it will be successful again. Over the past 19 years, the equities markets have evolved and transformed. Now, a similar evolution is under way in the Foreign Exchange market. Will the FX story resemble the equities experience? Of course, the market structure and the details will be different but, like films, this story feels eerily similar.
The evolution of the equity market started in 1996. Over the ensuing 17 years, trading technology became more sophisticated. Now, other asset classes and markets across the globe appear to be following a similar evolutionary trajectory (Figure 1).
Equities started as a manual market where orders were communicated to voice brokers. Market makers in the OTC Nasdaq marketplace and members of the NYSE used electronic bulletin boards to display their IOIs or “axes.” Sell-side brokers developed and distributed in-house electronic platforms to communicate to their buy-side customers to capture their flow.
The FIX Protocol, emerged — standardising how platforms could “communicate” with one another. This enabled vendors to develop broker-neutral platforms and solve fragmented buy-side workflows. Electronic order routing then began as a workflow and straight through processing solution.
As transparency improved, so did trust and empowerment. The buy-side sought greater control over orders. Different pricing models (maker/taker) and innovative order-type algorithms (e.g. Iceberg, Pegging and Discretion) to help control execution were developed. Competition in the marketplace emerged, thus fragmenting liquidity. Soon “smart order routing” aggregated the fractured liquidity and the early adopters on the buy-side started to deal directly in the marketplace.
The empowered marketplace that we now know in the equity markets gained significant momentum when algorithms that controlled all aspects of the execution were released. The buy-side and sell-side, rather than controlling each slice of a large order, were able to use algorithms to try to achieve a desired result. Early adopters significantly reduced implementation costs with these algorithms. The Association for Investment Management Research (AMIR/The CFA Institute) and mutual fund boards began to discuss electronic trading and algorithmic execution as part of “best practices” in buy-side trading rooms.
FX is standing on the shoulders of the equity experience. Many of the execution innovations developed in the equity market are being “tropicalised” and are starting to be adopted by FX market participants. During 2012, FX electronic trading grew both in usage and sophistication. According to a 2012 Streambase Systems survey of more than 240 institutional FX traders primarily located in EMEA and the United States, 69% of participants said that they used multibank platforms for execution, up 17% from 2011. However, the marketplace is now starting to make the leap beyond simple order routing. This is why the FX “film” is starting to feel eerily similar to the equity evolution.
Both the Streambase survey and the Bloomberg Tradebook FX experience suggest that that buy-side and sell-side participants are demanding more algorithmic execution capabilities to enable them to control how they want their orders to be transacted in the market.
The FX market is a highly fragmented market of different pools of liquidity. Banks have developed their own pools of liquidity and many ECNs have emerged as alternative sources. The Streambase survey noted that “liquidity aggregation algorithms” (smart order routing) were among the most commonly used execution algorithms by buy-side (54%) and sell-side (64%) survey participants. More advanced order handling execution management algorithms — algorithmic trading strategies that manage all aspects of an order to achieve a desired result — are also starting to gain traction.
Algorithmic trading strategy usage topped 48% in 2012, a 14% increase from a year earlier. The survey also suggested that further growth can be expected as 75% of the buy-side participants said that they plan to use or increase their use of algorithms for execution. Algorithms can be tactical or benchmark driven. The survey noted that the buy-side has begun to leverage the more sophisticated algorithms including; (passive) Floating (30%), time-slice (29%), TWAP (22%) and VWAP (29%).
Tradebook’s experience is consistent with the Streambase’s findings. More and more traders are using algorithmic trading strategies with increasing degrees of sophistication to implement their institution’s investment approach. Tradebook’s clients’ total notional executed with algorithms grew by more than 23% from Q1 2012 to the end of Q1 2013. Buy-side traders used tactical algorithms such as IF/Done, One-Cancels-Other, Economic Event and Target orders that control when an order is released into the market. Usage of more sophisticated algorithms such as Reserve Scale Back (efficiently manages the accumulation/distribution of a position/order), Discretion, Passive Pegging, Time Slice and TWAP also grew significantly.
The increased adoption of multi-bank trading portals and execution algorithms also indicates a growing buy-side demand for transparency and control. Whereas in the past, FX was a transaction that occurred as a settlement/back office function, institutions are now empowering their FX traders to take a more active role in execution. We believe that there are two reasons for the migration of FX trading to the front office.
First, reducing slippage. For example, a US-based fund buying a non-US stock must convert dollars to another currency for settlement and clearance. With the challenging investment environment, many funds found that if the stock was purchased in the morning, the currency could move significantly by the time the FX was done in the afternoon. The stock could be executed efficiently but significant transaction costs could emerge from currency slippage. Many on the buy-side became aware of this and, as a result, the FX portion is becoming less of an afterthought. Second, the lawsuits filed by two large pension funds claiming that their custodial banks took advantage of their FX orders were a catalyst for institutions to tighten their FX execution practices.
The progression of trading sophistication in FX resembles the equity plot. If you overlay the FX evolution with the equity experience, possibilities of where FX may end up begin to emerge. It could be argued that the evolution of the equity markets had five distinct stages: fragmentation, algos, the rise of high-frequency trading, dark pools and re-regulation (Figure 2). After liquidity fragmentation, smart order routing aggregated liquidity into a virtual marketplace. Then, algorithms that managed execution results in a complex ecosystem emerged. Algorithms sped up the market, while decimalisation regulation caused spreads to narrow — making manual market making untenable. High-frequency market makers stepped into the void created by the exit of the profit-challenged manual market makers.
Many equity market structure analysts point to Regulation NMS as the catalyst for high-frequency trading and dark pools. In our opinion, this isn’t exactly accurate as the rise of HFTs and dark pools was occurring in the OTC Nasdaq marketplace prior to the regulation. Regulation was needed, however, to eliminate rules that enabled manual market makers to hold up the faster electronic markets in the NYSE-listed markets. Once Regulation NMS harmonised the rules of the road, HFTs started to make markets in NYSE-listed stocks and brokers began to internalise their NYSE-listed flow (prior to Reg NMS, dealers were only allowed to internalise their OTC Nasdaq flows).
In the FX market, surveys show increasing use of algorithmic trading strategies. Spreads are narrowing. Manual market makers are getting squeezed. Bloomberg Tradebook’s experience is that more high-frequency market makers are entering the market. Analysts from research firm Celent estimate that high-frequency trading currently accounts for about 28% of FX trade volume. It is expected to grow.
Over the coming years, using the equity experience as a backdrop (Figure 2), it is somewhat easy to see how the FX market may become more automated with new and different types of liquidity centers vying for market share and how traders will become reliant on algorithmic trading to implement investment strategies.
For many FX market participants who witnessed the equity evolution, it feels like we have seen this film before. The details will be different but the resemblance may be uncanny. As New York Yankee great Yogi Berra said, “It’s déjà vu, all over again.”