The Alma fund is one of three Winton UCITS offerings, which include the Winton Trend Fund and the Winton Diversified Fund, which is in the process of being renamed Winton Multi-Strategy UCITS. The Winton Trend Fund trades a pure trend following strategy in traditional CTA markets only, while the multi-strategy UCITS fund combines trend following, global systematic macro and quant equity market neutral strategies in a single highly diversified portfolio. 

In common with most trend following CTAs, Winton had a strong first half of 2024 thanks to longs in cocoa, equity indices, and shorts in Japanese Yen and natural gas. However, in contrast to some peers, Winton’s strategy held up well after a tumultuous start to August and profited in 2023, a year when others struggled. Reasons included risk management reducing the drawdown, and non-trend models enhancing returns.

Risk management: a source of differentiation at turning points

The fund has an annualised volatility target of 8 to 12% over the long term and is operated within a multi-faceted risk management framework that co-CIO Simon Judes dubs “hypervigilant”. 

Amid the collapse of Silicon Valley Bank in March 2023, many trend-oriented CTAs had their worst month in years, but Winton lost less than most as one of its stress tests stopped the strategy scaling up its positions. “Long equity and short bond exposures appeared to be diversifying for one another heading into March 2023, which meant other CTAs ramped up positions to hit their volatility target. But this diversification suddenly disappeared, leading these CTAs to large losses, which they locked in when their strategies deleveraged after the event,” says Judes.

This stress test focused on correlation risk and was introduced to Winton’s risk management framework after a CTA crash in February 2007. It also helped to curtail losses during early August 2024, when an unwind in the long equity and short yen trades caused double-digit losses for some trend followers. Winton’s correlation stress test constrained its long equity exposure while a different proprietary metric reduced short Yen exposure. 

“Time and time again, we have seen how prudent risk management and not giving back too much of our P&L at these turning points can be a major differentiator in the long run,” adds Judes.

Non-trend signals: seeking returns in trendless environments

Winton’s pure trend strategy has outperformed the SG Trend Index in risk-adjusted terms over the past three and five years ending in August 2024, but it is the non-trend signals that explain much of the strong outperformance of Diversified Macro over the past three years to August 2024, particularly after 2023. 

Within the non-trend bucket, mean reversion models and multi-asset carry made strong contributions in 2023. Multi-asset carry was introduced by Winton back in 2011 when the established concept in currencies was expanded to fixed income and commodities. The signal is the largest non-trend model because it is seen as highly complementary to trend, while more than a dozen other models are roughly equal weighted. 

Multi-asset carry models work through time and space. Intuitively, investors might see that higher differences between interest rates improve currency and bond carry since 2022. In fact, Winton also generated profits from carry even under zero interest rates in the first half of the 2010s, because forward markets overpriced rate rises that did not materialize, and forwards reconverged with spot. Higher rates have recently been most helpful laterally – by engendering the strongest price trends in currencies for many years. 

Beyond carry, Winton continues to introduce new diversifying signals to try and smooth out the performance of the core trend following strategy, which tends to deliver returns in bursts.

Underpinned by trend following

Indeed, trend following is the fund’s main performance driver. The trend percentage of Diversified Macro was “hard wired” at 75% in late 2020, after several fundamental models overwhelmed the trend style and extinguished years of earlier outperformance in 2020. “The rationale for fixing the trend weight at 75% is to give investors the familiar trend following return profile, but with the added benefit of our research into other macro signals, which can make money in periods when trend following struggles,” says James Gilbert, Winton’s Head of Client Solutions. “We have since seen the merits of this approach, delivering strong industry outperformance while remaining highly correlated with CTA indices,” Gilbert adds.

Looking to the longer term

Winton analysis, first published in 2013, shows slower trend models have seen less Sharpe ratio decay, albeit from a lower base, since the 1970s and performed better over most multi-year periods since the global financial crisis. Winton claims to have been one of the first to slow down its trading from 2007 and observes that the trend following CTA industry has since followed suit for the most part. More recently however, Winton tilted the trading of stock indices to faster models in Diversified Macro to make it more responsive to equity drawdowns, a scenario that investors are seeking to mitigate with their CTA allocation. 

Improved opportunity set for trend

Most trend following CTAs argue the QE era was challenging because markets driven by risk on/risk off moves resulted in choppy, range bound price action, where it was hard to find broad based sources of persistent trends. Since 2020 a series of surprises have played into the hands of trend followers: Covid and its repercussions, Russia’s invasion of Ukraine, resurgent inflation, deglobalization and divergence in central bank behaviour.

“All of these shocks are hard for investors to interpret, and trend following does well at times like this because complexity and uncertainty means that information is not absorbed into markets very quickly, which allows large trends to emerge. We saw this across energies, stock indices, crops and metals through 2021; in fixed income, currencies and energies during 2022; and in agricultural commodities and the dollar in 2023,” says Judes. Economic fundamentals do not explicitly feed into trend models, but they capture big economic moves anyway.

If the economy does revert to a 1970s pattern, including persistent inflation and higher rates, CTAs may be well positioned to perform well, based on long-term back tests. Long commodities and short fixed income were helpful for parts of the 1970s, which also included phases when trend strategies would have profited from shorting energy and owning fixed income, as they did in 2014 and 2020.

Trend following will sometimes ride popular speculative bubbles as well as parachuting down panics, but its positioning can be either consensus or contrarian. As examples of contrarian positioning, Co-CIO Carsten Schmitz points out that, “Trend followers profited from the oil price crash in 2014 and from European interest rates going negative, which were not consensus views. We also profited from the collapse in nickel prices in 2023, despite the popular green metals’ narrative”.

Future strategy research and development 

Winton’s two co-CIOs are physicists who assumed their roles in 2020, with founder Sir David Harding having since become Executive Chair. Schmitz, who mainly leads multi-strategy initiatives such as alpha capture, quantitative credit and mid-frequency trading, was previously Head of Research, and Judes, who is developing the long running CTA and equity strategies, previously led Winton’s CTA strategies. 

Approximately half of the firm’s 190 staff work in roles that are focused on developing and managing strategies or supporting their implementation. This longstanding commitment to research explains why Winton and Harding claim to have a long history of identifying early trend following innovations that are today seen as standard features of the strategy. The scaling of positions inversely to volatility, portfolio volatility targeting, continuous models and non-linear response functions can all be traced back to Winton or the early years of Harding’s previous firm, Adam, Harding and Lueck (now Man AHL), which has spawned multiple CTA managers, including Aspect Capital, ISAM and Altis Partners.

Having hitherto recruited non-finance quants from academia and other industries to avoid those inculcated in the Efficient Markets Hypothesis paradigm, a deeper quant investment talent pool means that Winton can now hire professionals with track records and new strategies. 

Winton makes little use of academic finance research and predominantly builds proprietary models with the aim of finding sources of returns that are not yet widely traded. Slower, longer-term models tend to be based on observational research and extremely long back tests, where a clear economic rationale might increase confidence in the signal. Shorter-term models, on the other hand, can employ a more experimental approach as higher Sharpe ratios mean that it is more immediately obvious whether the signal is working as designed. 

Closing the gap

The offshore Diversified Macro fund currently trades more than 120 alternative markets in a dedicated portfolio that UCITS funds cannot allocate to. However, Winton and Alma Capital are working to reduce the differences between the strategy’s offshore and UCITS implementations. 

Winton Alma Diversified Macro Fund already trades some markets dubbed “alternative” by some peers through the UCITS-eligible note, including iron ore, European gas, carbon emissions and smaller agricultural markets. The fund’s trading universe was also expanded further in early September to include higher-capacity, more-liquid alternative markets, such as emerging market interest rate swaps and CDS indices.

Winton generally avoids alternative markets that it expects to duplicate traditional CTA exposures, such as certain developed market interest rate swaps that have an equivalent future or equity sectors.