The Inflation Diversification Problem

2026, so far, has unwelcome echoes of 2022: war, an energy crisis and supply shocks. For markets, these factors have brought a familiar theme back into focus: the interaction of the traditional core components of a multi-asset portfolio.

In what follows, we will share some observations on the stock-bond correlation, look at alternatives to bonds for diversification in periods of equity market stress and examine how equity market neutral strategy returns differ by inflationary regime. We will end with some insights into the capital efficiency advantage of multi-manager strategies when compared with traditional funds of funds.

The stock-bond correlation

Figure 1 shows the correlation between stocks and bonds by decade, based on US equities and US Treasuries (UST). The size of the bubble represents the price-to-earnings multiple for equities, while the x-axis shows average inflation in the US. Negative correlations, the norm over this period, have coincided with low inflation. The flip to positive correlations has tended to occur when inflation is sticky and above approximately 2.5% on average.

Figure 1. Relationship between US inflation and the stock-bond correlation

Problems loading this infographic? - Please click here

Source: Stock-bond correlation is between the S&P 500 Total Return and the Bloomberg Barclays US Treasury Aggregate Total Return indices. Valuation calculated as the percentile monthly P/E from January 1960 to January 2026. Core CPI from the Bureau of Labor Statistics. As at end January 2026.

Why does this matter? For 60/40 investors, 2022 was a test case in the failure of multi-asset diversification: the MSCI World index lost 18% in euro terms, while the Global Aggregate Treasuries index lost approximately 13%. Bloomberg’s 60/40 index lost nearly 17%, the worst annual return since the Global Financial Crisis (GFC).¹

We have already acknowledged the similarities between 2026 and 2022 in terms of resurgent headline inflation, war and energy crises. There are also a few crucial differences: losses in bonds were greater then than we think they would likely be in a sustained 2026 inflationary shock. This is because duration — the sensitivity of bond prices to changes in yield — is far higher when the starting point is near-zero rates (as it was in 2022). The price-yield relationship is steepest at low yields: the same basis-point move causes a much larger percentage price decline when starting from 0.5% than from 5.0%. In effect, the post-2022 repricing has itself created a cushion — higher starting yields mean lower duration and more coupon income to absorb future shocks.

While smaller prospective losses offer some comfort, the more important point is that gains in bonds may not reliably offset losses in equities in the way they have in previous cycles.

None of this is to say that bonds have no role in portfolios. The 60/40 portfolio has endured as the bedrock of asset allocation for good reason. US Treasuries, for example, have offered deep liquidity, negligible credit risk and a historical tendency to rally during growth shocks. But this diversification benefit is not unconditional. Figure 2 below shows the average monthly return of US Treasuries in months when the S&P 500 fell by 3% or more. It is immediately clear that bonds have a mixed track record during equity losses, and that there has been some regime dependency, with poor performance during the inflationary periods of the 1970s, 1980s and 2020s, but materially positive during the great moderation period of the 2000s when inflation was low.

Figure 2. Average monthly return of US Treasuries when the S&P 500 fell by 3% or more

Problems loading this infographic? - Please click here

Past performance is not indicative of future results. Source: Man Group, Bloomberg. Covers the period between 1973-2025.

Equity market neutral as a more effective diversifier

We carried out the same exercise using a simple average of four Fama-French factors (Value, Risk, Quality and Momentum) as the shock absorber. The results demonstrate that there can be more effective diversifiers for the main risk in a portfolio than an asset whose attributes depend on an unreliable and regime-dependent correlation.

Figure 3. Average equity style returns in months when the S&P loses 3% or more

Problems loading this infographic? - Please click here

Past performance is not indicative of future results. Source: Man Group, Bloomberg. Covers the period between 1973-2025.

What does inflation mean for equity market neutral returns?

If allocators’ principal concern is that inflation is sticky and changes the basis for interaction between the assets in a multi-asset portfolio, and our suggestion is that equity market neutral strategies make a more durable diversifier in equity stress than bonds, it is worth examining how inflation affects the returns of equity market neutral strategies themselves.

For this exercise, we have taken the HFR Equity Market Neutral index from its inception in 1991 through to today, and looked at real and nominal returns segmented by inflation. As shown in Figure 4, moderate inflation regimes – where US CPI is between 2.5% and 4.0% – appear to have been associated with materially better returns than other environments.

Figure 4. HFR equity market neutral nominal (top) and real (bottom) annualised returns by CPI regime (1991-2026)

Problems loading this infographic? - Please click here

Problems loading this infographic? - Please click here

Past performance is not indicative of future results. Source: Man Group, Hedge Fund Research Inc, Bloomberg. Covers the period between 1991-2026.

Our working hypothesis is that sustained higher inflation regimes raise the cost of capital for firms, forcing management teams and boards to restructure in order to address sub-cost-of-capital returns. Japan, until very recently, has been a test case for the opposite phenomenon. Since its bubble burst in 1989, low inflation and a depressed cost of capital removed any incentive to restructure businesses, effectively contributing to sub-optimal capital allocation decisions and the ‘zombification’ of the corporate landscape. In this context, and so long as inflation does not become runaway, we should continue to think of inflation as a tailwind for the equity market neutral approach, adding a string to its bow as a diversifier.

What is the best implementation?

There are three key ways to add equity market neutral to a portfolio:

1. Single manager;
2. Fund of funds; and
3. Cash-efficient multi-manager

At Man Group, we believe wholeheartedly in diversification within diversifiers, so we naturally favour (2) and (3) over (1). But we also believe that optimal returns require the efficient use of cash and fee budgets; to this end, traditional fund-of-funds stuctures fall short because they layer fees and introduce a cash drag in the structure.

The choice between single manager versus diversified structures comes down to two considerations: first, the scope to forecast returns and the probability of loss in a given year; second, available capacity. On the latter point, given that alpha is naturally scarce, good single manager strategies typically exhaust capacity, making it hard for allocators to use them at scale. On the former point, our own experience of forecasting alpha suggests it is a fruitless process, precisely because the attributes that drive returns are idiosyncratic in nature. Additionally, every strategy will incur drawdowns – if an asset manager suggests otherwise, they are being disingenuous. If your whole diversifying exposure rests on a single alpha source, the probability that you will incur that drawdown is higher than if there are multiple legs to the alpha.

Figure 5 below summarises hypothetical returns for two strategies: a single manager strategy with a high information ratio,² and a strategy comprising multiple managers with reasonable returns but limited correlation.

Figure 5: Hypothetical equity market neutral scenarios: single manager versus multiple managers

Hypothetical return distribution

Problems loading this infographic? - Please click here

Problems loading this infographic? - Please click here

The expected return for each scenario is information ratio × volatility. We set each manager's volatility at 5% and vary the IR to produce the expected returns shown. For the single-manager, we assume an IR of 1.5, giving an expected return of 7.5% per year. For the multi-manager scenarios, we assume an IR of 0.75 per manager, giving expected returns of 3.75% per year.

The lower chart breaks the simulated return distributions into percentile buckets showing, for each scenario, which annualised return corresponds to each point in the distribution. All are annualised one-year returns.

Illustrative statistics are derived from mathematical principles based on these assumptions. This data does not take into account any underlying investments, strategies or market factors and should not be relied upon. Hypothetical returns based on mathematic principles are not representative of any past performance or indicative of future results. Source: Man Group.

Four key observations:

1. A collection of lower return, uncorrelated managers delivers a higher portfolio information ratio than a single higher-expected-return manager
2. Intuitively, the probability of loss is higher for a single manager
3. Returns are more skewed to extremes for the single manager – large gains and bigger losses
4. Diversification helps to improve IR and minimise skewness for the multi-manager construct, but volatility is crushed. And as everyone knows, you can’t eat Sharpe³

If a multi-manager construct can help smooth returns by adding diversification, it becomes an interesting prospect. However, the typical fund of hedge funds implementation has embraced exactly the tendency to crush volatility, because these structures are not typically cash efficient - 100% of capital is invested directly into underlying funds, often with some cash held back on the sidelines, and with a layering of fees on top.

In contrast, multi-strategy funds that allocate directly to underlying strategies, and make use of cross-margining and prime broker arrangements can achieve multiple turns of leverage. The effect of this is shown in Figure 6 below. We have assumed that the fund-of-funds selector has an advantage in being able to allocate to star managers at twice the information ratio of the portfolio managers in a multi-manager fund. This is probably overgenerous. But even with this advantage, the compression of volatility in such a cash-inefficient structure and the layering of fees means that a cash-efficient multi-manager may deliver superior absolute performance.

Figure 6. Fund of funds versus cash-efficient multi-manager

Problems loading this infographic? - Please click here

See Figure 5 for details of calculations. Hypothetical returns based on mathematic principles are not representative of any past performance or indicative of future results. Source: Man Group.

Conclusion: rethinking diversification

The possibility of an elevated inflation regime suggests that multi-asset investors need to rethink how they achieve diversification in their portfolios, given that bonds do not tend to be useful diversifiers when inflation is sticky and above trend.

We have shown that a simple equity market neutral approach has been much more effective over time, and that equity market neutral strategies have in fact tended to enjoy more attractive returns in moderate inflationary environments than in lower inflation.

To us, implementation is best achieved in multi-manager constructs, within which we believe that cash efficiency must be maximised to overcome the dampening effect of diversification on volatility.

AI was used to support data analysis and processing in the production of this article.

1. Source: Bloomberg.
2. Measures a fund manager's excess returns relative to a benchmark, per unit of tracking error. A higher IR indicates more consistent outperformance, reflecting greater skill in generating alpha relative to the risk taken versus the benchmark. For an absolute return strategy where there is no benchmark and all risk is active, the information ratio is total return divided by annualised volatility.
3. Measures an investment’s risk-adjusted return.