Passive versus Aggressive Exploring the underlying decision making behind FX algo strategies

Nicholas Pratt asks how buy-side firms are using FX algos to execute their passive versus aggressive trading strategies

For the most part of this millennium, buy-side firms have been looking to take more control of their trading processes, from direct market access to the use of execution algorithms, firstly in equities but now also for their FX trades.

Buy-side traders have also developed more sophisticated trading strategies and techniques due to greater awareness of market impact, opportunity costs and their respective impact on best execution. This has become especially important during the market volatility we saw as a result of the Covid 19 pandemic. So not only are buy-side traders using algos more, they are using algos to do more than simply slice and dice their trades.

Consequently algo developers have looked to improve their offerings to match buy-side clients’ demands. One area of development has been around the use of algos to perform either passive or aggressive execution strategies. But how well do algos deal with this distinction? Does everyone share the same definition of passive versus aggressive, especially between buy and sell-side participants? And is the decision to trade passively versus aggressively one that should be made by algos at all, or should it be left to traders themselves?

Risk versus reward

“Passive versus aggressive execution is a question of risk versus reward,” says Alexander Barzykin, FX eRisk quantitative analyst for HSBC. “In the context of placing a single child limit order, the reward of a passive placement is spread capture and the risk is volatility which may come into play. Similarly for parent orders, passive execution aims to explore cost saving opportunities while facing higher risk of deviating from the target benchmark.  Aggressive execution style aims to minimise this risk while somewhat sacrificing transaction cost savings,” says Barzykin.

Different execution algorithms use different performance benchmarks, such as arrival price or time weighted average price, and manage risk differently, adds Barzykin. “For example, HSBC’s Liquidity Seeking algorithm is inherently passive while the execution style controls the depth of order placement. HSBC’s Implementation Shortfall algorithm, on the other hand, can actively manage risk by crossing the spread while the execution style controls the level of volatility at which the algo will aggress.”

There are various factors that determine traders’ level of execution risk appetite and which then influences their strategy selection, says Farzana Nanji, EMEA head of eFX sales at HSBC. The factors include market volatility, spreads, market trends and the size of the order.

“If the client’s main objective is to minimise footprint and they have discretion over execution duration, then a passive strategy would be a more likely choice,” says Nanji. “For a client who benchmarks execution against arrival price, an implementation shortfall strategy would be a strategic choice.  For algo execution, the client takes on the execution risk, which is also a factor when selecting the strategy – the more passive the strategy, the higher the execution risk,” says Nanji.

Given the effect of market conditions on execution strategy, the unprecedented rise in market volatility and widening of spreads that we saw in March as a result of the impact of Covid 19, did see a greater reliance on the use of algos, says Barzykin. “Wider spreads and diminished liquidity created particular challenges for larger transactions.  Taking all of these factors into consideration, clients turned to algos to help navigate their way through an increasingly fragmented FX market, opting for smarter execution,” he says.

“A use case that emerged during this period was for clients to start out with a passive strategy, testing liquidity and observing any impact, with the ambition of capturing spread.  However, if volatility significantly increased, then they could resort to a strategy that cleared the risk more aggressively,” says Barzykin and he refers to HSBC’s Liquidity Plus product which provides the ability to switch between strategies either manually or on the basis of a defined market level.

“During this period of time, both volatility and spreads increased,” he says. “What is interesting to note is that, at least for some instruments, long-term volatility increase was relatively less pronounced than that of short-term, or high-frequency volatility. So besides lack of liquidity for larger order sizes, this effect could also explain significant increase in algo execution.”

The new generation of algos involves a more curated form of liquidity, which monitors a number of factors within each of the execution venues for a more optimal execution, says HSBC’s Nanji.  For example, it monitors fill rates by looking at the percentage of orders that have been filled, based on the total number submitted. It also monitors round trip times and market impact.

“The liquidity curation process becomes a vital component of the algo performance, especially given the fragmented and delocalised nature of liquidity within the FX markets,” says Nanji. “HSBC algorithms dynamically evaluate the market microstructure in real time and make optimal decisions around passive and/or aggressive allocation of orders at different venues, to provide an overall better execution experience for the clients.”

Diverse execution:

What remains to be seen is whether buy-side traders believe that an algo is the best means by which to make the decision on how passive or aggressive a trade should be. The global asset manager Allianz Global Investors uses a diverse set of execution strategies that cover both passive and aggressive trading, says Andreas Anschperger, director, European head of foreign exchange trading. He does not believe in the binary descriptions of passive versus aggressive execution and instead sees execution strategy as a human judgement on several factors.

“Algo execution will always be a trade-off and a judgement between a key set of parameters – execution urgency, market volatility, risk appetite, client/portfolio restrictions and timing which is down to a trader’s skill,” says Anschperger.

There are factors that influence how traders make those judgements and trade-offs, says Anschperger. These include pre and post-trade analytics and the related transaction cost analysis and respecting various liquidity pools in various currency pairs.

Market conditions also play their part in the trader’s decision making and preference for passive versus aggressive execution. “When envisaging a higher market spread during a time of high market liquidity, a more passive strategy might be advisable, including the opportunity for matching market interest,” says Anschperger.

When it comes to a new generation of algos being engineered to fine-tune passive vs aggressive execution strategies, Anschperger says that it is the widely discussed search for intelligent algos making dynamically use of above mentioned interest and market conditions with the dominant factors such as pools of liquidity, internalisation and dynamic adaptation to changing market volatility.

Trade urgency:

Passive order execution is where a participant adds liquidity to the market while trying to fill an order, with the objective to only trade when someone has an offsetting interest, says  Ronald Lagarde, expert trader FX, APG Asset Management.  In contrast, aggressive order execution is where a participant takes liquidity from the market, buying and selling at market prices and looking for immediate execution.

But when it comes to FX algos, there is a miscomprehension that passive versus aggressive relates  to how fast or slow an algo works, says Lagarde.

While a passive algo can execute more slowly, it doesn’t always have to do so. “Some algo providers understand this but not all. Our main goal is to lower market impact, while minimising opportunity risk. In our experience this can generally be done best when you spread your execution over a period of time, with the use of passive algos,” says Lagarde.

The main influence on the decision to use passive versus aggressive is the urgency of the trade, says Lagarde. “That urgency derives from the order I get from my portfolio managers and the objective of the fund. Next to this there are the market conditions like costs and volatility. If there was high urgency, we would not use passive algos because that creates opportunity cost.

But if you have low urgency, we would use passive strategies unless we have strong reasons not too. We have been using passive strategies first through leaving bids and offers with banks, but also directly on ECNs, although this was manual and cumbersome. It really took off when FX algos became available,” he says.

“The size of the order is also a factor. If you have a large order and you trade with high urgency, you will incur high market impact and it will be self-defeating. It is really about market impact versus opportunity risk. Our main objective is market impact so low urgency trading and passive algos are our priority,” says Lagarde.

Over the last five months, since the outbreak of the Covid 19 pandemic, market factors have become a much greater influence, says Lagarde. “We have seen super high event risk. Prior to Covid 19, we were in a low event risk environment that really benefitted from the use of passive algos. In the first weeks, there was a sharp rise in volatility. Then for a short while, there was a decrease in liquidity.”

This is where a human trader can really excel, says Lagarde. “Spreads widened a lot and it became very expensive for risk-based trading but passive trading also becomes expensive if there are sudden, directional and extreme changes in markets. So what we saw in that period was a big increase in the use of different strategies, with different urgencies, as the market changed.

It was a reminder that you can’t rely on the algos to do all the work for you and to rely on the short-term experience and data that feeds those algos. Trader decision-making became much more important and to trade based on what you see in front of you in terms of cost and volatility,” he states.

After that initial four week period of volatility, conditions have started to normalise, says Lagarde. However, he hopes that the importance of human discretion will be remembered when it comes to the development of the next generation of FX execution algorithms.

“Some algo developers are looking to include the decision between passive versus aggressive execution into their algos. But that is not really what we require. They should focus more on the smart order routing behind the execution strategy and creating more efficient algos, be they passive, aggressive or neutral,” says Lagarde. “The choice between passive versus aggressive remains the choice of the trader because it is based on information that is not included when I send out an algo order.”