ÃÛ¶¹ÊÓƵ

April this year marked the 15th anniversary since we started managing index portfolios tracking minimum volatility indexes for our clients. The timing of this milestone coincides with a period when we observe material outperformance of minimum volatility indexes vs. market: for example, YTD the global minimum volatility index, as measured by MSCI, has outpaced the market by more than 10% (Exhibit 1).

Exhibit 1: MSCI World Min Vol EUR opt vs. MSCI World Index: +10% YTD 2025

Source: ÃÛ¶¹ÊÓƵ Asset Management, MSCI, Rimes.
Note: TR Gross index performance data in EUR as at 24 April 2025. Past performance is not a reliable indicator of future results

Such relative performance patterns are not surprising: the strong defensive properties of low volatility strategies tend to be particularly well pronounced in times of heightened market volatility, as is the case at present. While low volatility strategies tend to lag market cap weighted indexes in ‘bull market’ times, they tend to provide good downside protection in weak markets, and, over the long term, they would typically provide similar returns as the market on average but with materially lower volatility (Exhibit 2).

Exhibit 2: Minimum volatility indexes have provided good downside protection historically

MSCI World Min Vol annualised relative return vs. MSCI World: 1989-2024

Note: TR Gross index return data in USD. 'Up years' ('Down years') defined as years when MSCI World delivered positive (negative) annual return in USD terms. Data for minimum volatility index contains live and back-tested data sourced from MSCI. Past performance is not a reliable indicator of future results

Investors interested in capturing the low volatility factor via index exposures could opt for either a mean-variance optimisation-based index, where both stock volatilities and correlations are considered in the construction methodology, or a heuristic index, where stocks are selected and weighted by risk/volatility metrics.

The low volatility anomaly – low volatility stocks outperforming high volatility stocks – has been widely researched, documented and persistent in practice across various geographical regions and time frames. It is known as an ‘anomaly’ because it contradicts conventional finance wisdom of a trade-off between risk and return, i.e., in order to earn higher returns investors need to assume higher levels of risk. Amid the Global Financial Crisis, MSCI launched the Minimum Volatility index series in 2008, which was followed by launches of many other low volatility index series constructed by other index providers.

From construction methodology perspective, these can be viewed in two groups: low volatility indexes constructed via a mean-variance optimisation, and low volatility indexes constructed via heuristic metrics. A mean-variance optimisation approach (e.g., MSCI Minimum Volatility) is based on a covariance matrix (inputs for the matrix include stock volatilities and stock return correlations) and a factor risk model. A heuristic metrics approach (e.g., MSCI Risk Weighted) is based on selecting stocks that have historically been less volatile and weighting them by their inverse volatilities, measured by metrics such as standard deviation or beta. In contrast to optimisation techniques, heuristic weighting techniques do not take into account stock correlations.

Despite different construction methodologies, the goal of all low volatility indexes is similar: to offer lower volatility compared to market cap indexes without materially compromising returns over the long term. Ultimately, the choice of a low volatility benchmark for an index portfolio is based on investor preferences and priorities. Some of the points investors might want to consider when selection a low volatility benchmark include: whether they would prefer a more complex or less complex index construction methodology, whether they would prefer a benchmark with higher Sharpe Ratio or higher Information Ratio, or whether they would prefer a benchmark with higher or lower tracking error vs. the market cap index.

In general, optimisation based low volatility index strategies have tended to perform better than heuristic based during market downturns historically, i.e., they have displayed better defensive properties in such environments, as they take into account both stock volatilities and their correlations. Another handy feature of optimisation based low volatility indexes is the possibility to select the currency of optimisation. When we launched our first minimum volatility index strategy 15 years ago, we partnered with MSCI and pioneered and implemented the idea of optimising in the same currency as the investor base currency (until then optimisations were carried out in USD only).

Optimisation currency is a key determinant of the constituents and volatility of this type of index. For example, construction of MSCI Minimum Volatility indexes begins with selection of a Parent Index and a base currency to perform the risk minimising optimisation. Optimisation for different base currencies tends to produce different index holdings. Exchange rate fluctuations add to volatility, hence optimising to the same currency as the investor base currency produces the lowest predicted risk. As the index objective is to reduce volatility measured in the base currency, considering currency risk as part of the optimisation process would tend to result in the most effective index for the investor, as shown in Exhibit 3.

In the matrix we compare the risk, measured in USD, EUR and GBP, of three MSCI World Min Vol indexes: optimised in USD, EUR and GBP. Lowest volatility is displayed by the index optimised in the same currency as measurement currency of the volatility. For example, the volatility, measured in USD, of the Min Vol index optimised in USD (11.23%) is lower compared to the volatilities of the EUR optimised (11.25%) and GBP optimised (11.26%) indexes. Similarly, when the volatility is measured in EUR, the lowest volatility is displayed by the EUR optimised index, and when measured in GBP – the lowest volatility is displayed by the GBP optimised index. These results are consistent with the objective of MSCI Min Vol indexes to reduce volatility measured in the currency used for the optimisation.

Exhibit 3: The importance of optimisation currency in minimum volatility indexes

Annualised standard deviation measured in USD, EUR and GBP: November 2001 – March 2025

Source: ÃÛ¶¹ÊÓƵ Asset Management, MSCI, Rimes.
Note: TR Gross index data in from 30 November 2001 (earliest date when data is available for all examined indexes) to 31 March 2025. Annualised monthly standard deviation of index returns is used as a measure of risk. Data for MSCI Minimum Volatility indexes contains live and back-tested data, sourced from MSCI. Past performance is not a reliable indicator of future results.

Given the strong run of low volatility indexes so far this year vs. market cap, investors might be wondering if they have missed the entry point, or if there might be further upside potential. Current valuation levels of minimum volatility strategies remain compelling relative to market cap and to historic averages (Exhibit 4).

Exhibit 4: Minimum volatility indexes current valuation is compelling

Relative P/BV: MSCI ACWI Min Vol vs. MSCI ACWI

Note: Data for relative price-to-book (P/BV) ratio, historic average, plus and minus 1 standard deviation to 31 March 2025. Data for MSCI factor indexes contains live and back-tested data sourced from MSCI. Past performance is not a reliable indicator of future results.

Timing markets is hard. Instead of trying to predict when to enter/exit a particular index strategy, investors might want to consider holding a combination of pro-cyclical (e.g., value factor) and defensive (e.g., low volatility, quality factor) equity factor index strategies over the long term (Exhibit 5), to supplement and diversify their market cap index equity exposure. If equity markets continue to be as volatile as they have been so far this year, allocation to Min Vol indexes could be a sensible strategy for index equities.

Exhibit 5: Factor indexes performance tends to be cyclical

Factor indices vs. market: 1 year monthly rolling relative return

Note: Data from 31 January 2001 to 31 December 2023. TR gross index performance data in USD.
Data for alternative indexes contains live and back-tested data sourced from index providers (see 'Index data' page). Past performance is not a reliable indicator of future results.