Is now time to double down on diversification?

Podcast:Signal over Noise with Ulrike Hoffmann-Burchardi (6 mins)
Video:The three drivers to watch for 2026, with CIO's Jon Gordon (5 mins)
Video: CIO's Jeff Harwood on opportunities in US financials (6 mins)

The past fortnight has been far from quiet. AI-driven disruption concerns have led to sharp stock movements and estimated market capitalization losses in excess of USD 1 trillion, according to Deutsche Bank. Unlike prior episodes of technology and communications services turbulence, the recent repricing has extended to a wider net of sectors where the application of AI is feared to have more disruptive consequences. These span software, legal services, logistics, insurance, and commercial real estate.

Our view on artificial intelligence remains that investors will need to be nimble; responsive to fast shifts and increased blurring across the enabling, intelligence, and application layers of the AI value chain; and highly attuned to what economist Joseph Schumpter called the “creative destruction” that technological innovations bring.

But with more broad-based volatility comes the renewed need for portfolio diversification, in our view, a principle whose efficacy is supported by decades of empirical research.

Diversification helps navigate performance dispersion between and within asset classes. Statistical analysis of asset class performance reveals that no single asset consistently outperforms across all market cycles. For example, the MSCI AC World Index delivered three consecutive years of returns above 20% through 2025, but historical data shows that such streaks are rare. Between 1926 and 2025, a USD 1 investment in US equities (S&P 500) would have grown to approximately USD 21,000, while the same amount in medium-dated US government bonds would have returned USD 110 and US dollar cash (1-3 month Treasury bills) USD 25. However, equity returns were not linear: Following the 1973 oil shock, the S&P 500 required 6.6 years to recover its previous peak. This volatility highlights the importance of holding multiple asset classes to smooth returns and reduce drawdowns.

Diversification depends on how assets move together. The effectiveness of diversification is contingent on the correlation structure between assets. Ideally, investors find assets with negative return correlations—one asset rises in value as another falls. Based on our correlation analysis of 15 years of data, 1-5 year US government bonds have historically provided modestly positive returns for US investors during periods when international equities (Eurozone, Switzerland, and the UK) declined. But diversification is also about finding assets that move less than others in periods of market stress, helping to steady and smooth portfolio returns. We like high-quality government bonds for this reason, as well as their potential to rally in adverse economic scenarios outside our base case. But correlations change through different market regimes, so investors must regularly review their holdings and adjust positioning, in our view.

Diversification can extend to alternative investment philosophies. Investors should not overlook the potential power of diversifying beyond approaches that rely on human qualitative judgment. Recent advances in quantitative finance have introduced systematic strategies that can dynamically adjust asset allocation based on macroeconomic and market signals. These approaches, which rely on statistical models rather than subjective judgment, have demonstrated the ability to reduce behavioral biases and enhance risk-adjusted returns. For example, systematic de-risking during periods of elevated volatility may mitigate losses and preserve capital.

While we believe diversification is the bedrock of successful long-term investing, we believe it is also important not to conflate it with hedging market risks. Diversification aims to reduce specific risks of holding a certain company, asset class, or exposure to a region or sector. It does this by spreading investments across assets, especially those with low or negative return correlations. However, true hedging requires assets or instruments (such as options) whose payoff structures explicitly aim to appreciate when reference assets decline. Among major asset classes, persistent negative correlations are rare, making hedging through diversification alone insufficient. Structured strategies, including capital preservation approaches that combine bonds and derivatives, may offer additional risk mitigation but entail trade-offs such as reduced participation in upward-trending markets and increased complexity.

Diversification is not a tool for exclusively maximizing returns—indeed, diversification can mean holding assets with paper losses at certain phases of the business cycle. But we believe it is an approach for managing risk, smoothing performance, and protecting against unpredictable shocks. Investors should tailor diversification strategies to their unique exposures and objectives, incorporating quantitative and systematic approaches where appropriate. While diversification cannot eliminate all risk, it remains the most robust defense against the uncertainties inherent in today’s complex market environment.