Tatu, please tell us a little about what your job involves and the key responsibilities you have.
I lead our multi-asset trading desk, which consists of five traders including myself. While I remain actively involved in trading whenever possible, I also oversee regulatory compliance, risk management, process development, and the seamless integration of our trading operations within the broader investment organization.
What types of clients does OP Asset Management provide services for?
Our largest client segment by far is Finnish retail investors, primarily through our mutual funds. In addition, we manage institutional mandates and wealth management portfolios. We also oversee mandates for OP Financial Group’s insurance and life assurance companies and manage the liquidity reserve of the bank.
What range of instruments are you generally working with?
We operate across a broad range of financial instruments, including equities, developed market credit, covered bonds, government bonds, emerging market hard currency bonds, CDS, swaptions, IRS, FX spot/forwards/swaps, and listed derivatives.
How would you describe the key objectives and guiding principles of your trading desk and the dealing activities it undertakes?
Our mission is straightforward: to deliver best-in-class execution outcomes across all asset classes for our clients. We also aim to generate alpha by advising portfolio managers on execution strategies, liquidity sourcing, market structure, and instrument selection. In addition, we play a central role in managing counterparty relationships and implementing allocation changes across fund-of-funds, institutional, and wealth management portfolios.
Working with multiple assets can be a complex undertaking. How do you address the various technology and workflow challenges involved?
The multi-asset setup requires both skilled traders and a resilient infrastructure. We currently use three core trading platforms by asset class, which also act as backups—an often overlooked but critical aspect of business continuity. While we’ve explored multi-asset EMS platforms, the cost-benefit equation hasn’t yet justified implementation. Our priority has been workflow automation: ensuring that traders receive orders through the correct systems and formats. This requires strong integration between the portfolio management system and our trading platforms. In FX, we still see room to improve automation. When evaluating new workflow tools, our focus is on automation capabilities and connectivity to TCA providers, OMS, and counterparties.
Your team recently started to use FX algos. What was the motivation behind that decision?
Having used equity algos for years, expanding into FX algos was a natural progression. The underlying execution logic is quite similar, and we saw a clear opportunity to improve execution quality. FX algos helped to fill a gap in our toolbox, particularly for order types where traditional RFQ or WMR FIX trading protocols aren’t ideal. Our aim was to assess whether algos could deliver better results for certain trade profiles.
What are your main objectives when undertaking algorithmic FX trading and what types of orders are usually a good fit for them?
In notional terms, our largest FX activity is related to hedging for our insurance mandates, primarily in swaps. By ticket volume, however, the bulk consists of spot and short-dated forward trades linked to mutual fund flows. WMR fixing orders still account for a significant share of our passive fund flows and are generally not suitable for algo execution.
This leaves us focusing on larger hedge adjustments, sizeable spot trades, and scenarios where FX execution can be aligned with other asset flows—such as equity trades executed over the day. Since our portfolio managers rarely take active intraday FX views, we can afford to be patient, accept market risk, and focus on minimizing market impact through passive execution. We believe this approach yields better long-term outcomes, even when accounting for the increased arrival price variance that passive strategies may cause.
How do you source your FX algos, and what factors influence your choices?
As we’re still in the early stages of FX algo adoption, it made sense to start by leveraging our existing relationships with key FX counterparties. We conducted thorough due diligence to ensure their offerings matched our needs—both in functionality and in terms of connectivity with our trading platforms and TCA provider.
There’s always a trade-off between having broad access to algo providers and maintaining clean, comparable data. To ensure robust TCA and meaningful analysis, we’ve limited the number of algo providers we use. Our current mix includes global bulge-bracket banks and Nordic niche players to reflect our regional footprint.
Although most providers offer similar algo strategies on paper, we’ve observed real differences in execution behavior—particularly in aggressiveness and liquidity sourcing. Without a sufficiently large and normalized data set, these nuances are easy to miss, which underscores the importance of strong TCA.
Are you happy to let an algo do its work without much oversight, or do you prefer more real-time visibility during execution?
We’re fortunate to work with providers that offer strong real-time monitoring tools which we consider essential. That said, we rarely intervene once an algo trade is underway. Our order sizes typically don’t warrant active management, and more importantly, hands-off execution allows us to build a clean and unbiased dataset for analysis.
Consistency is key when comparing performance across providers, even with sophisticated TCA tools. We’re mindful of the well-known cognitive bias where people draw conclusions from isolated outcomes, so we deliberately avoid overreacting to individual trade results.
You’ve been working on a new multi-asset TCA. How is that progressing, and how is it helping your dealing activities, including algorithmic FX trading?
We’ve been working with our multi-asset TCA provider for about two years. The initial build took roughly a year, and the system is now operating smoothly. While there’s still room to improve data quality and expand instrument coverage, centralizing the data has been a major advantage. It reduces the burden on our technology and data teams and enhances transparency, something our compliance and risk functions particularly value. That said, relying on a single provider comes with trade-offs, as no vendor excels across every asset class.
Crucially, the ability to benchmark performance against peer groups and cost estimates adds meaningful insight. Even aggregate-level EURUSD slippage figures have limited value on their own—it’s the context that makes them actionable.
Do you expect to make more use of algorithmic trading in the future, and what will influence that decision?
Absolutely, provided we continue to receive a sufficient volume of suitable orders. FX algos have become a core part of our execution toolkit and open up new opportunities. One area where we’d like to see further development is in ultra-passive strategies. As mentioned earlier, our preference is to act more like a market maker—providing, rather than taking, liquidity. Even the most passive algos available today tend to execute too quickly when mid-market liquidity appears, and often suffer from some adverse selection when markets are moving. We’d welcome tools that allow us to post interest and truly behave passively. Algo providers: we hope you’re listening.
What advice would you give to other firms exploring algorithmic FX trading?
Start with a solid TCA framework so you understand your baseline execution costs across different protocols. Only then can you meaningfully evaluate whether algos improve outcomes and confirm that through live testing. Talk to your counterparties, peers, TCA providers, and study the algo offerings carefully. In our experience, passive algos outperform traditional risk transfer methods for larger orders and offer better alignment with asset flows, for example, using VWAP-style strategies to match FX with equity execution over the day. That said, if your orders are very small or dominated by fixing trades, algos may offer limited added value.
