Dmytro Sheludchenko joined SEK 324 billion ($33 billion) Swedish pension fund AP1’s quantitative division in his mid-20s after graduating in maths and financial engineering, and a brief stint at Sweden’s Central Bank.
Now, aged 29, his position in the fund’s vibrant quant team underscores how leading pension funds are building internal young teams to embrace technology, look beyond short-term implementation to focus on long-term value, and draw expertise from the vast amounts of data in today’s new investment landscape.
“I used to be the youngest in the investment team, but I’ve younger colleagues now,” says Sheludchenko. “My time here has flown as we build up the internal portfolio of quantitative investment strategies in line with the fund’s decision to prioritize in-house expertise within this field.”
AP1 has allocated to factor strategies or risk premia for over a decade, but the portfolio of classic and alternative risk premia (ARP) across currencies, rates and equities is evolving and developing in what Sheludchenko calls a constant and on-going process.
Risk premia strategies in general have a low correlation with traditional asset classes and their liquidity gives a flexibility that makes the portfolio easy to scale up – or down, he says. Longer-term, structural changes are also afoot.
The buffer fund recently separated the portfolio into specific mandates, giving each a more dedicated role within the overall portfolio. Change has also focused on how the allocations are structured. For example, the fund has swapped its original focus on long/short equity factors to long only implementations.
“The portfolio used to be structured around classical ARP strategies in equities and other asset classes with several counterparties,” he explains. “But as time went on, we realised there is a better way for us to implement risk premia. For example, within equities we found that shorting single stocks is too expensive in the long run and ‘eats up’ much of the premia. But we still liked the equity factors so we dropped the short position.”
In another development, the fund has also decided to implement more equity strategies internally for cost reasons as well as to make better use of internal expertise.
During his tenure, the fund has also introduced leverage using derivatives backed by liquid assets to boost factor exposure in multi-asset trend following strategies.
“Our approach to leverage is not that different from what hedge funds do and is applied so that the factors reach an appropriate risk level in relation with to risk targets,” he says. “Our allocation to ARP strategies has resulted in a broader use of derivatives and led to higher requirements in risk management infrastructure to monitor and report risks.”
Most recently, increased investment in infrastructure and data management means the internal team’s capability has grown again. Although some implementation is outsourced to partner banks because it is too complex and demanding to run in-house, AP1 still aims to design the portfolios and see for itself how they work.
“We can’t compete with the capacity of some of the large quant teams. However, we still want to be able to see how [the strategies] will perform, and we do this by building up robust infrastructure,” he says.
In another development, the growth of risk premia strategies has also led to a change in focus in the hedge fund allocation, bringing it in-house and moving away from Commodity Trading Advisor (CTA) structures.
“Our hedge funds now focus on strategies that offer more alpha and approaches that can’t be replicated in systematic factor fashion. The risk premia and hedge fund allocations are very complimentary – although I would say the hedge fund allocation is even more sophisticated than we are.”
Much of Sheludchenko’s role centres on developing expectations for the different allocations prior to actual investment, followed by rigorous comparison of investment outcomes compared to those initial expectations.
“For us the most important part of investment follow-up is to check if the historically observed ex-ante properties and expectations are met,” he says. It is this, rather than returns, that test a strategy’s success, as analysis of last year’s poorly performing trend strategies illustrates.
“The allocation did badly, as the whole CTA industry did, but that was entirely in line with our expectations. It was actually what we wanted,” he said.
Indeed, measuring success not on profits or loss, but on the basis of the portfolio doing what it is expected and predicted to do, is one of the aspects of the job he enjoys most.
“When we look at our performance it’s not really in the context of good or bad. Instead we ask if that performance was in line with our expectations and decisions. Think of it as ordering an expresso and making sure you get an expresso,” he explains.
That process involves shutting out day-to-day market noise like macro news and Central Bank briefings, as well as resisting the compulsion to check strategies’ performance on a daily basis.
“Our focus is on the global picture; if we are getting what we expected to get and if the portfolio performance is contributing to the overall performance in the way we defined.”
It’s a hunt for precision when the range of outcomes and uncertainty are so wide that made quants a “natural fit” with his maths background, he says, noting however that “sometimes things can’t ever be explained – even from a maths perspective.”
The job demands keeping up with the latest developments in academia and current thinking; checking what others in the field are doing and if it could be applied to AP1. Recently this has led Sheludchenko to explore new factors and results from combining different factor definitions, as well as ensuring current definitions are economically sound and correspond to what is expected.
“We still have a long way to go to include more alternative risk premia in more traditional portfolio framework,” he concludes.
Photo of Patrik Nyman and Dmytro Sheludchenko (left).
