World and gold: diversification without guaranteed protection

Fees: a simple structure with contained costs

Recurring charges of 0.18% per year earn A because the equity ETF and gold ETC provide their exposures at low ongoing cost. The fee grade rewards implementation, without judging whether gold is attractively priced. The recent performance comparison in the visual includes the returns of both assets; their difference from World cannot be attributed to fee savings alone.

Brokerage charges, bid-offer spreads and personal tax are outside this reading. Frequent small rebalancing trades can erode the benefit of low fund charges. A simple holdings structure makes a practical monitoring rule easier to implement.

Long-term assumptions: how gold changes compounding

The expected compound-growth gap is -0.78 percentage points per year, small enough to earn A. Most capital still receives the model’s equity growth assumption, while gold’s weak expected correlation helps limit fluctuations. Smoother returns can improve compounding, so the combined result depends on correlations and volatility as well as the expected return of each holding.

Gold distributes no income. Its contribution depends on its resale price, while World provides exposure to company earnings and valuations. These different mechanisms support the case for combining them, but gold does not mechanically replace equity growth. [2][1]

The allocation remains below the model’s expected frontier. Its high grade means it is close to the growth-optimal reference in expected compound return, rather than optimal at its own risk level. That reference can take more risk and use other assets. The result depends on the house assumptions and gives no signal about when to buy gold.

Historical efficiency: an advantage over the measured period

From 22 October 2016 to 22 October 2026, the replay returned 12.91% annually in euros, against 10.79% for World adjusted to the same measured risk. The advantage of +2.12 percentage points per year earns B. The mix delivered more return than its measured volatility alone would explain relative to that reference; the capture does not assign the entire advantage to gold.

Asset-class series fill 0% of the history, with 0% excluded. The engine’s selected histories do not recreate an investor’s contributions, transactions or taxes. This result neither attributes the entire advantage to one holding nor establishes that it will recur.

The historical B and forward A both support this particular mix on their own measures, but use different evidence. The first records a realised period; the second depends on assumed future returns and relationships. Neither establishes that gold will keep contributing in the same way or that its current weight is right for every investor.

Crises: gold does not guarantee a floor on losses

World still occupies 80 % of the portfolio and supplies almost all the loss in the US-equity shock. Gold makes a small positive contribution in this scenario, but cannot cancel that equity exposure. The net decline of -28.6% earns C: improved resilience compared with equities alone, while leaving a substantial crash loss.

The shock propagates through asset-class sensitivities based on long-term assumptions. It is neither a replay of a historical crash, a maximum loss nor a protection contract. A low average correlation cannot ensure that gold rises at the precise moment cash is needed.

SGLN consists of limited-recourse obligations secured by underlying property according to its documentation. Owning the listed instrument differs from holding metal directly. Its structure and lack of income create risks separate from fluctuations in the gold price. [2]

Diversification: different drivers with changing relationships

The expected matrix assigns a correlation of -0.08 between global equities and gold as asset classes. This weak relationship explains the diversification potential. The separate observed holdings matrix shows a positive relationship over the latest available year: fund prices over a particular window are different evidence from long-term asset-class assumptions.

The B is supported by growth that survives the loss of the main equity driver. Gold supplies a separate expected source above cash, while its weak modelled relationship with equities allows that contribution to survive the stress on the dominant source. World still provides most expected growth, which limits the grade. The 2 growth sources label alone says less than the size and independence of their contributions.

IWDA’s optimised physical replication describes how the fund holds assets; it does not make each fund weight identical to its index weight. The archived evidence provides no detailed country or sector concentration measurement to reproduce here. The engine uses a category classification for the gold sleeve where economic composition is missing. [1]

Points to consider

The B overall grade reflects low fees, a small expected-growth gap and favourable historical efficiency. The C for crises tells a less reassuring part of the story: most capital remains in equities. Gold’s contribution to diversification leaves a separate question: how much of an equity-led loss could the investor absorb, and what would explicit protection cost?

1. Gold’s intended role

   Is gold here for long-term diversification, for potential rebalancing, or for both? Those roles imply different tradeoffs around its weight. Price-driven changes in that weight and the dealing cost of restoring it are relevant even without a change in the intended role.

2. Capacity to absorb a loss

   How would a decline of -28.6% interact with planned spending? The scenario illustrates substantial remaining equity risk, without bounding the possible loss. A need to sell either holding during a decline would change the experience of this allocation.

3. Gold’s behaviour alongside equities

   How much confidence does the evidence support in gold’s diversifying role? The recent observed correlation is positive, while the long-term class assumption is weak. Neither establishes dependable crash protection. The ETC’s structure and risks also differ from holding metal directly. [2]

Other portfolios

• World and emerging markets
• One World ETF: concentrated in the US, led by big tech
• World, S&P and Nasdaq: three ETFs, shared drivers