The 10,000-foot philosophy behind the strategies is to view markets through the prism of proprietary risk premia. A multi-asset long/short approach extracts risk premia from single asset class strategies in equities, bonds and currencies, as well as cross-asset strategies. It mixes systematic macro and quantitative single stock equity.

The portfolio aims to have some sort of balance between procyclical and defensive risk premia, with the intent to neutralize equity beta during tail/extreme market environments. Moreover, the portfolio also uses some more directional strategies, which should average out at neutral over a full cycle. The approach is to some degree systematic, but unusually for an alternative risk premia (ARP) strategy, there is significant discretion over allocations to risk premia and asset classes, based on fundamental analysis and in order to manage risk.

Two differentiators: trend and value

Discretion is only one distinguishing feature since the sorts and mixes of models are also different, and so too is the return pattern. For instance, most ARP strategies, measured by the SG Multi Alternative Risk Premia Index, profited in 2022 and 2023 when Nordea’s strategy was down. Conversely, 2020 was strong for Nordea but a losing year for that ARP index.

Nordea does not comment on competitor products, but we can identify two of the most obvious differentiators. Some ARP strategies apply price-based trend, and traditional equity value measures, which are not part of Nordea’s suite of thirty risk premia. Nordea follows trends in fundamental data in cross asset momentum models, which can trade both within and between asset classes. What remains of Nordea’s value signals in equities is very different from a generic price to book value indicator.

“Our proprietary valuation measures now even include more qualitative indicators and some value-based equity signals were deleted after the 2010 to 2011 debt crisis,” confirms Hansen, who started his quant career in London before returning to Denmark.

Proprietary risk premia

“We like to think that all of the premia are proprietary, though they do draw inspiration from academic research,” says Hansen. The headline number of risk premia seems very large at thirty, though Hansen admits it might be somewhere between twenty and thirty, depending on how they are defined and counted. Some twelve of the premia are also used in Nordea’s flagship long only total return fund, Nordea 1 – Stable Return Fund, which started in 2005, shortly before the absolute return strategy. “Every strategy needs to fit into our investment philosophy of fundamental and quantitative trading, with daily liquidity,” says Hansen. The portfolio return drivers could be seen as a mix of traditional and alternative risk premia, and alpha, which can also be split between traditional beta, exotic beta and alpha.

Defining nuances around beta neutrality through the cycle

The four “Super Strategies” – equities, fixed income, currency and strategic risk balancing – are expected to be collectively beta neutral, balancing risk-on/risk-off and also strike a balance between “recovery” and “recession”, which can be a different axis. “Sometimes, risk-on maps to recovery and risk-off maps to recession, but they are not always correlated. It depends on the time horizons. Risk-on can mean different things and it can be measured in different ways,” says Hansen.

Meanwhile, the directional strategies tactically vary beta, but should not have any structural beta over time. “Over the long term, through a full cycle, the aim is to avoid any structural traditional beta at the overall strategy level. Some individual strategies might have a degree of beta, but it may be a temporary tactical wager, and in any case the intention is to hedge it with other strategies, which could include currency trades, implied volatility overlays or dedicated statistical arbitrage. Over time, we should be beta neutral,” says Hansen.

There is intended to be some asymmetry to the exposures. The normal objective is to have zero downside beta but some upside capture, to create a positive convexity return profile. The exact sensitivities will vary through market cycles and fashions. “In 2023 the portfolio unusually had zero to slightly negative beta on the upside. This has happened before, around 2008-2009, 2011 and 2018, when the market became very one dimensional,” says Hansen.

In April 2024 the portfolio appeared to have net long equity exposure of 65%, approaching the highest ever levels, but this is deceptive. “On a beta adjusted basis, it is near zero. The reason is that more defensive high-quality equities have recently demonstrated a beta near zero, because the equity market is so bifurcated and unbalanced,” explains Hansen. Nordea has very high conviction in its low-risk basket of stocks for two simple reasons: they trade at a discounted valuation to the overall market and have higher expected returns. The equity book has had some of the MAG7 names (Apple, Amazon, Alphabet, Meta, Microsoft, Nvidia, Tesla), while it has been somewhat short of Europe’s Magic 8 (ASML, Air Liquide, LVMH, L’Oreal, LVMH, SAP, Schneider Electric, Siemens).

The directional/alpha bucket may sound partly like global macro, but it also combines relative value and equity factors. “There are elements of global macro, and we also have some momentum strategies. In the relative value space, we trade factors such as growth and defensives,” says Hansen.

Correlation and leverage

Proprietary correlation models support portfolio construction. “The strategy relies on diversification benefits to reduce forecast volatility from about 56% to 15%. When correlations spike up the range of the volatility target leaves some latitude for volatility to expand,” says Hansen. For the Alpha MA 15, forecast volatility ranges between 10% and 15%, and leverage varies between 7x and 10x to stay within this target range.

The 7x, 10x and 15x volatility target strategies are named based on the higher end of the target volatility range. They mainly trade the same strategies with different volatility targets.

Discretionary rebalancing and risk reduction

The strategy is substantially systematic, but limited discretion can be exercised to vary strategy weightings and the overall risk target. For instance, in April 2024, there was highest conviction in equity premia, moderate conviction in strategic risk balancing premia, and lower risk in fixed income, currencies, cross assets momentum and reversal premia. One example of when discretion was exercised was to reduce overall risk ahead of the Russian invasion of Ukraine, which also increased portfolio turnover in 2022.

“We have the tools, and we trade ideas and implement them with some qualitative input,” says Hansen. This risk optimization process is not the same as mean variance optimization. “The process is conceptually somewhat like Black-Litterman, but is also much simpler,” reveals Hansen.

Convictions per bucket range from 1 to 3, but this does not mean a 3 conviction would be sized three times larger than a 1 conviction. There is a different sort of scaling.

Hansen declined to provide performance attribution value added from discretionary rebalancing versus a counterfactual of equal weighted strategies and constant volatility targets. “We cannot cut it like that. There is always a mix of quantitative and qualitative analysis. When the hot water and cold water flow into the bathtub it is all water,” he said.

However, the degree of discretionary input should not be exaggerated. “Strategic risk balancing forecasts over longer periods than a typical TAA mandate would and does not have as much freedom as a completely discretionary macro approach,” says Hansen.

Strategy evolution

In addition to some changes discussed, “Some risk parity-based strategies were added and developed more robustly after 2014,” says Hansen. New strategies are incubated inside MA with a low-risk budget initially, but it is not used for seeding standalone products as such. “Losing years of 2018, 2022 and 2023 did not necessarily lead to any major changes, though we will look to improve in some areas showing weak spots,” says Hansen.

2024 outlook

For several years, Nordea’s big picture macro asset allocation outlook is that a long only traditional 60% equities/40% bonds portfolio may not provide effective diversification, especially in a higher interest rate climate when equity/bond correlations have historically been higher.

The shorter-term tactical outlook has been majoring on the low risk, high quality equity trade, and has some defensive and tail risk strategies positioned for a potential correction.

Mean reversion in returns

From inception in 2006 to March 2024, gross returns have delivered a Sharpe ratio of 0.53, which is higher than world equities or global bonds. After two losing years in 2022 and 2023, this is at the lower end of the target range: a Sharpe between 0.5 and 1 is implied by the return objective of cash plus 7-10% with volatility of 10-15%. Over earlier time periods, the Sharpe has been at the higher end of the range. The strategy has exceeded its higher-end range of cash plus 10% in 52% of rolling five-year periods and has met its lower-end range return target in 89% of rolling five-year periods.