Multi asset: responding to extreme investment scenarios

2025 was a tough year for active managers, and the path of rapid economic change and geopolitical shifts looks set to continue. This era of uncertainty has the potential to create both melt-down and melt-up market scenarios. Amid such a difficult environment for diversification, where markets are driven by sentiment rather than fundamentals, we find there is only one way to prepare: research, research and research. This Q1 issue of Simply put examines the rise of extreme investment possibilities going into 2026.

The CIO’s perspective: managing challenging return scenarios

In 2025, markets continued to be driven by sentiment rather than fundamentals, valuations don't appear to matter anymore, and investment seems more about chasing performance and timing markets (up or down) on changing newsflow, which is not easy, to say the least. But are we really in a new paradigm, or rather being slowly pulled into a state of complacency and herd behaviour? 

In 2025, sovereign bonds faced persistent headwinds while equity performance broadened, with sequential bursts across Europe, then Asia and finally in the US. This pattern has made performance difficult to capture for active managers, while rewarding passive ones. The chances of adding value are higher when fundamentals have an influence, but in the current environment, there's only a 50% chance of getting the timing right. Yet, I believe active strategies – within or across asset classes – remain essential price searching mechanisms and are still the engine of fundamentally healthy markets. 

While our view for 2026 centres on balanced growth for balanced performance, we must think critically about our actual risk scenarios – not just market risks, but investment style risks as well. Downside risks have the bad habit of actually being at their highest level when they are somehow imperceptible. At the same time, significant upside risks exist too: the artificial intelligence (AI) and broader technology revolutions are creating unprecedented situations where future economic and earnings impacts remain difficult to quantify. 

Read more: Balanced growth, balanced returns in 2026

The uncertainty inherent in this new economic system presents a downward risk scenario where massive capital expenditure could prove less productive than anticipated, potentially leading to an AI complex readjustment. Conversely, AI might deliver underestimated positive dividends for the global economy and financial assets. 

As we move into 2026, the odds of melt-up and melt-down scenarios have increased. Our research-driven enhancements position us to navigate this increasingly bifurcated return environment, providing both the protection our strategies are known for and the participation capabilities needed during concentrated bull markets. The time has come to dive deeper into the world of rising extreme scenario probabilities and how to best address them.

Portfolio positioning: a final quarter of normalisation

The interest rate shock in 2022 created a lasting hurdle for many multi-asset solutions. Solutions that rely on risk as an allocation signal took a longer time to return to normalised market exposure. One of the most important reasons has been the level of rates volatility, a lasting symptom of the post-COVID-19 era. Now, the Federal Reserve (Fed) has partially normalised rates, as have most major G10 central banks except Japan. With that, we’re finally seeing a broad normalisation in the level of risk and volatility across each asset class in our portfolios – as highlighted in Figure 1. 

FIG 1. Risk premia volatilities¹

Risk appetite stayed at a high level throughout the fourth quarter, despite the market pullback we saw in November. And the overall valuation picture remained largely unchanged – bonds are cheap and cyclical assets are expensive – with the main difference being the increase in emerging market equity valuations. Finally, the macro environment remains benign, supporting the normalisation process of global markets.

Taking account of these different indicators, our market exposure has increased over the quarter, and it now closely resembles our neutral allocation. In other words, the normalisation of the economic and financial environment has resulted in our solutions normalising both their overall risk-taking through leverage and the 100% rebased composition of the All Roads allocation.

Read more: 2026 investment outlook: striking a new balance

Macro: what macro triggers cause extreme risks?

Going into 2026, we find ourselves navigating a paradoxical market environment that requires us to reconsider what macroeconomic factors could trigger potential melt-up or melt-down situations. Currently, we're experiencing what appears to be an increasingly Goldilocks-like period, with growth, inflation and interest rates aligning favourably for investments, although valuations across the board are high. This ‘expensive’ Goldilocks scenario raises a critical question: what catalysts might transform this delicate balance into either a rapid market rally or a sharp correction? 

Our analysis identifies three potential breaking points: renewed inflation surprises, a revaluation of stretched asset prices (particularly in the technology sector), and deteriorating liquidity conditions. While the current environment doesn't yet present immediate warning signals for any of these scenarios, monitoring these key factors will be essential for successfully steering through markets in 2026.

Figure 2 shows the recent evolution of our macroeconomic indicators, generally pointing to a mild or benign investment environment: growth remains subdued, inflation appears normalised and monetary policy remains in dovish territory, which is important given the growth fragilities our indicators are detecting. 
 
FIG 2. Nowcasters and the risk of extreme moves by region² 

Considering these elements together, melt-down risks currently appear limited but should not be overlooked. Markets may remain in the Goldilocks zone for now, but keep an eye out for the three bears of inflation, valuation and liquidity returning home at any moment.

Read more: Higher real rates globally are boosting the outlook for large cap equities

Special focus: dealing with melt-up risks

One of the major challenges for risk-managed solutions is the rapid upward market swings that often follow a significant market shock. We experienced this in April 2025, prompted by President Trump’s ‘Rose Garden’ speech, and it proved costly for many actively managed solutions. The best course of action turned out to be to do nothing, which was not an option for us, as downside mitigation is one of our explicit client commitments. 

However, the opportunity cost of being underinvested is real, and we do not sit idly in the face of it. We have thus designed a re-risking solution that activates when needed rather than hoping for a market recovery after a 20% equity decline – history shows the latter is a riskier strategy. 

Figure 3 presents the simulated performance impact of our recovery overlay, which we constructed using selected short volatility strategies. Our analysis reveals that implementing this overlay unconditionally (Raw combination or RawCombo) provides limited incremental value to our investment process, as the exposure largely duplicates risk factors already present in our diversified cyclical asset allocation. However, the overlay demonstrates considerable value when implemented conditionally (All Roads modified or All Roads Mod). This conditional approach addresses an often-overlooked risk management consideration: for active allocators, upside risk can be as significant as downside risk when positioning becomes excessively defensive. 

FIG 3. Simulated performance of the rebound overlay impact³

Our conditional re-risking overlay effectively addresses the opportunity cost of defensive positioning during market recoveries without compromising our risk management principles, creating a more resilient solution across full market cycles.

New research: Capital Market Assumptions from carry methodically approximated

As multi-asset investors, being able to form hypotheses on future asset returns over a long period is paramount. Whether constructing benchmark tilts in traditional capital-based portfolios, or estimating long-term tail measures (e.g. conditional Value-at-Risk) to build risk-based portfolios, Capital Market Assumptions (CMA) have a direct impact on portfolio allocation and thus on future performance. Forecasting returns is, however, famously difficult, and to have an initial approximation of long-term returns, it is common to turn to carry measures.    

To compare the different models, we consider a dataset of monthly returns and carry measures from January 1988 to November 2025 on a cross-asset universe. Figure 4 shows how, over a medium time horizon (1-year), carry measures appear to be relatively poor forecasters of future returns that remain dominated by short-term shocks. However, as we move to longer time horizons and price changes are smoothed out by the long-term averaging, we see that the forecast errors shrink and that models including carry measures tend to outperform the naive benchmark model. Additionally, including a mean-reversion component and averaging across rolling and expanding windows delivers better forecasts, while the inclusion of a cross-sectional carry variable does not prove useful. 

FIG 4. Out-of-sample forecast accuracy of the different models at different time horizon⁴  

Simply put, carry measures can help form forecasts for long-term asset returns. However, these measures are not very accurate over a 1-year time horizon as price moves are dominated by higher frequency shocks. Carry proves more apt at modelling returns over a 5-year time horizon.