Stockholm (HedgeNordic) – Systematic strategies represent a consistent, disciplined, and transparent approach to investing, significantly reducing the risk of human error and emotional decision-making. These quant-based strategies can be cost-efficient and scalable, making them particularly beneficial for large allocators managing sizeable investment portfolios. The First Swedish National Pension Fund, Första AP-fonden (AP1), has committed more resources over the years to developing and growing its own quant team to capitalize on the advantages of systematic strategies.
A team of six led by Patrik Nyman, Head of Asset Allocation and Quantitative Strategies and an employee at AP1 since 1992, is responsible for two mandates. The first mandate focuses on global tactical asset allocation, while the second involves managing a portfolio of developed market equities using a factor-based approach. The tactical allocation mandate is divided into two components. The first component, as Nyman explains, focuses on “managing the overall risk balancing in the portfolio and ensuring that the strategic risk profile is upheld.” The second component allows the team to “utilize the overall risk mandate for tactical allocation to take more model-driven exposure.”
In-House Trend-Following
According to Nyman, this second component utilizes a basket of systematic, model-driven strategies with a strong emphasis on trend-following, including directional, non-directional, and systematic macro-type strategies. Having previously allocated to external trend-following managers, AP1 has opted to implement these strategies in-house. “Trend-following is an integral part of the overall risk-taking within the tactical asset allocation,” explains Nyman. “It serves as a pure risk-based overlay that consumes part of the overall risk allocated towards tactical asset allocation.”
“Trend-following is an integral part of the overall risk-taking within the tactical asset allocation.”
Patrik Nyman, Head of Asset Allocation and Quantitative Strategies at AP1
“Trend-following is a relatively low risk-return strategy, we see Sharpe ratios sub 0.5, but the contribution to the total portfolio’s risk level is limited and offers very interesting properties and features that complement the overall portfolio really well,” notes Nyman. As a result, trend-following has become an integral part of the tactical asset allocation mandate. “We are running different types of trend-following implementations today,” he mentions. “We have complementary strategies in different areas, but trend-following remains at the core.”
According to Dmytro Sheludchenko, a Senior Portfolio Manager within Asset Allocation and Systematic Strategies at AP1 with ten years of experience at the pension fund, the decision to bring trend-following strategies in-house was driven by the need for greater flexibility, efficiency, and scalability. “We previously had external hedge funds in the portfolio that were very good, so no criticism towards them, but we realized it was much easier for an investor of our size to manage trend-following strategies in-house,” explains Sheludchenko.
“…Our goal is to capture what we call trend-beta. This beta-like exposure captures the characteristics of trend-following, which we believe is a valuable allocation for the fund, almost alpha-like, in the long run.”
Dmytro Sheludchenko, Senior Portfolio Manager within Asset Allocation and Systematic Strategies at AP1
“We are aware we are not going to be among the top trend-following specialists, but our goal is to capture what we call trend-beta. This beta-like exposure captures the characteristics of trend-following, which we believe is a valuable allocation for the fund, almost alpha-like, in the long run,” he emphasizes. “Having the flexibility and ability to scale it as we wish is highly beneficial to us, which is one of the main reasons behind bringing everything in-house.”
The Multi-Factor Allocation in Developed Market Equities
In addition to participating in AP1’s tactical and strategic allocation process, the quant team at AP1 also oversees a mandate to run a systematic multi-factor strategy within developed markets equities, including a small-cap-focused portfolio added last year. This quantitative approach dates back to 2012 when, before Sheludchenko joined, AP1 started its first internal equity strategies with exposure to ‘risk-adjusted’ low volatility as a factor. “Later on, we partnered with certain investment banks to invest in alternative risk premia strategies, long/short strategies,” recalls Sheludchenko. Although AP1 valued these strategies, the costs were too high for scaling. Consequently, the basket of alternative risk premia strategies has also been integrated into what we do in-house.
“Rather than simply blending factors, our goal is to construct a portfolio of equities with specific properties that we believe will be most advantageous in the long run.”
Dmytro Sheludchenko, Senior Portfolio Manager within Asset Allocation and Systematic Strategies at AP1
Having initially invested in external long/short alternative beta strategies, AP1 has gradually moved to running long-only multi-factor strategies internally. “We implement a more traditional multi-factor equity strategy,” Sheludchenko begins to describe the approach. Instead of “just blending in factors and creating a multi-factor exposure,” the quant team at AP1 selects relevant factors, rates them, and relies on them to determine the types of companies to be included in the portfolio. “Rather than simply blending factors, our goal is to construct a portfolio of equities with specific properties that we believe will be most advantageous in the long run,” explains Sheludchenko.
The Use of Machine Learning
The team’s approach involves adjusting factor weights based on the market conditions and risk level at a given time. Additionally, they have implemented machine learning models to help identify the most optimal exposure to different factors. “We explored various machine learning techniques before the recent surge in popularity of artificial intelligence,” says Sheludchenko. “I prefer not to refer to these techniques as artificial intelligence because they are not. It’s machine learning.” Recently, a new colleague with expertise in machine learning joined the team, leading AP1 to launch its first entirely data-driven machine-learning process about eight months ago. However, this strategy is still based on the same type of data that is used in the construction of traditional factor strategies.
“It’s a risk-based allocation that involves tilting and overlaying with the use of machine learning models to ensure we are not stuck with the same factor exposures over time.”
Dmytro Sheludchenko, Senior Portfolio Manager within Asset Allocation and Systematic Strategies at AP1
While some managers may completely overhaul their strategies with the implementation of machine learning techniques, Sheludchenko believes in a gradual introduction of these methods into existing models. “We view machine learning as an extension of all factor models, more like a factor rotational input,” explains Sheludchenko. Instead of blending everything with equal allocations, “it’s a risk-based allocation that involves tilting and overlaying with the use of machine learning models to ensure we are not stuck with the same factor exposures over time,” elaborates the portfolio manager. “If things change and we don’t observe them, machine learning techniques are designed to help us avoid allocating to factors that no longer work.”
The ESG Factor
As responsible investing gains traction among investors, incorporating ESG (Environmental, Social, and Governance) principles into systematic investing strategies has become essential. “ESG is a very important consideration for us at AP1 and all Swedish national pension funds. By law, we are required to be prominent investors in sustainability,” says Sheludchenko. When AP1 first started running factor strategies in-house, the team introduced a simplistic ESG filter that excluded certain companies based on their ESG ratings.
With AP1 having developed its own internal infrastructure, factor models, and strategies, the team also decided to introduce its own ESG factor. “Within our multi-factor portfolio, we have a dedicated factor that reflects a company’s comprehensive ESG profile,” explains Sheludchenko. This ESG project was developed in collaboration with AP1’s sustainability team, with the objective of identifying “specific features that we believe align best with the fund’s overall strategy in the long run.” This ESG factor has become one of the many company aspects considered when constructing the portfolio.
“Within our multi-factor portfolio, we have a dedicated factor that reflects a company’s comprehensive ESG profile. We view ESG characteristics as an integral part of a company’s attractiveness.”
Dmytro Sheludchenko, Senior Portfolio Manager within Asset Allocation and Systematic Strategies at AP1
“We view ESG characteristics as an integral part of a company’s attractiveness,” elaborates Sheludchenko. The team considers the ESG factor along with value, momentum, and other factors, but when all else equal, “the ESG factor will tilt the portfolio to more sustainable companies, at least how we define them,” he explains. “In this way, we transitioned from using a simple filter to a more integrated approach by introducing the ESG factor in 2020.”
The decision to develop their own ESG factor was driven by issues such as limited data, data accuracy, low data frequency, and the lack of a clear definition of ESG. “We are personally very positive about ESG and sustainability, but there are still issues for us as a quantitative investor,” notes Sheludchenko. “Data is a significant issue. You can now get ten years of good data, but it’s still not enough data,” he points out. “Data frequency is still quite low, and the biggest challenge is the lack of clear definition of what constitutes ESG,” continues Sheludchenko. “We define it in our way, which we feel is the best way, but for us, as a quantitative investor, this will be a challenge unless there is a more standardized way to define ESG across the industry.”
