Improving diversification with smart beta

Factor investing, also known as smart beta, has become increasingly popular as investors realise they can harvest factor-driven excess returns and diversification over the market capitalisation-weighted benchmark through a simple, transparent, and rules-based approach. 

Based on Research Affiliates’ research, the six factors of value, low beta, profitability, low investment, momentum, and size produce a substantial diversification benefit across multiple return drivers. Some factors have conveniently been placed together, such as profitability and investment, which are labelled as indicators of quality companies. The diversifying aspect of combining factors with different risk and return characteristics and low correlations helps investors “weather the storm” during adverse market conditions.

But a question arises when deciding on a multi-factor portfolio: which factors should be included? We find the right balance is achieved by making a decision based on the trade-off between the effective harvesting of the factor premium and low-cost implementation. 

Related News:


Research Affiliates has analysed the six factor-based smart beta strategies by constructing simple long-only investable portfolios. We start with the large-cap universe of US stocks, except for the small-size strategy. For each we select the best stocks based on the corresponding characteristics. For example, we construct the value portfolio by choosing the top 30% by book-to-market ratio. Within the portfolios, we weight the selected stocks by capitalisation, except for low beta which we weight by beta ranking.

The portfolios are rebalanced annually each July with the exception of momentum and low beta, which are rebalanced quarterly. 

We found that, on average, the six factor-based smart beta strategies deliver enhanced returns, with an average annualised excess return of 1.86% over the study period, July 1973 to December 2018, as indicated in Table 1.

Table 1: Performance of long-only factor-based smart beta strategies, United States Jul 1973-Dec 2018

Source: Research Affiliates, LLC, using data from CRSP/Compustat

The correlations of the six factors’ excess returns are mostly low or negative. We did find, however, a more positive correlation between the investment factor with value and with low beta.

Historical data tells us that the momentum and size factors appear to offer a substantial diversification benefit for the multi-factor strategy. The momentum factor has a negative correlation with three of the other factors—value, low beta, and investment—and a slight positive correlation with profitability and size. The size factor is relatively lowly correlated with the other factors, and especially seems to be a particularly strong diversifier of the profitability factor. It seems that small companies are spending time growing, rather than being profitable. 


It is very important to look at factor investing from a practitioner’s, rather than an academic’s, perspective. The difference is that while research can find a relationship between a financial characteristic of companies and future returns, investors are only interested in those relationships that can deliver an excess return in real-world portfolios.

Implementation requires thoughtful analysis given the notorious reputation of the high transaction costs associated with many factors, with the greatest offenders being momentum and the small-size factor. The explicit costs of implementation, such as brokerage, are generally well managed. We focused on the implicit component of implementation cost, which can be measured by the market impact of the trade, or in other words, the movement in a security’s price due to trading.

Understanding the importance of lowering implicit implementation costs, we have analysed them closely. We incorporate means of reducing the market impact of a strategy directly into its design. 

The market impact of a portfolio rebalancing can be attributed to several factors. The first, and most familiar, is the strategy’s turnover. The more you trade, the more it will cost. But we also need to consider a strategy’s portfolio volume, liquidity, turnover concentration, and tilt.

Portfolio volume is the aggregate of median daily trading volume of all stocks you hold in your portfolio. Lower portfolio volume, that is, when the portfolio holds more illiquid stocks, increases implementation costs.

Liquidity refers to how quickly and easily a security can be bought and sold. Stocks of larger, better-known firms with greater market capitalisation are typically more liquid than stocks of smaller, lesser-known firms with lower market capitalisation.

Turnover concentration measures how spread out your trades are across the securities held in your portfolio. When turnover (trading) is focused on only a few securities, and in relatively large size, in your portfolio, the trades are typically more expensive to execute than if trades are more equally spread out, and generally of smaller size, across all securities in the portfolio.  

Lastly, tilt measures the illiquidity of a portfolio as compared to the most liquid possible portfolio. When your portfolio holds more illiquid stocks, which are more costly to trade, the portfolio will have higher tilt. 

Intuitively, we can understand that high portfolio volume, low tilt, low turnover, and low-turnover-
concentration strategies are associated with a low market-impact cost, while low portfolio volume, high tilt, high turnover, and high-turnover-concentration strategies are associated with a high market-impact cost.

The average turnover across the six strategies we analysed from 1973 to 2018 is 61.6% and the average estimated trading cost of the strategies is 127 basis points (bps), assuming US$10 billion ($13.6 million) in assets under management (AUM), as shown in Table 2.

Table 2: Implementation cost of long-only factor-based smart beta strategies, United States Jul 1973-Dec 2018

Source: Research Affiliates, LLC, using data from CRSP/Compustat

The cost of implementing a strategy that has high turnover, high tilt, or low portfolio volume can be quite large. The momentum strategy, for example, has annualised one-way turnover of a very high 159.5%, which generates a comparatively high trading cost of 241bps for a US$10 billion portfolio, even though portfolio volume and tilt are at reasonable levels. When we consider the implementation shortfall, using a naively-constructed momentum factor as a standalone investment strategy does not seem to produce a good outcome. 

In contrast, strategies with high portfolio volume, low tilt, or low turnover typically have lower implementation costs. For example, the profitability strategy, which has the lowest trading cost of the six factors (17bps for a US$10 billion portfolio) is characterised by all three of the low-cost traits. 

The value strategy also incurs low trading costs (56bps) by virtue of its relatively low turnover and tilt. Interestingly, the size strategy, which focuses on small-cap stocks, has a below-average trading cost (91bps). Although the size factor tends to trade small, illiquid stocks as indicated by its comparatively low portfolio volume and high tilt, it has low turnover due to its broad coverage—2,352 names as of December 2018. On balance, the higher coverage of the size strategy almost offsets the cost of trading smaller companies.


Although momentum is very expensive to implement as a stand-alone strategy, and size seems to be risky from a volatility and tracking error perspective, these factors can be good additions to a multi-factor strategy because of their negative or low positive correlations with other factors. 

We also analysed four portfolios to assess how including both of these factors impacts the performance characteristics and implementation costs of a multi-factor strategy. We look at four multi-factor portfolios with an equal allocation to each set of nominated factors:
Portfolio 1: Value, low beta, profitability, and investment; 
Portfolio 2: Four factors in Portfolio 1 plus momentum; 
Portfolio 3: Four factors in Portfolio 1 plus size; and
Portfolio 4: Four factors in Portfolio 1 plus momentum and size.

When we add momentum to the equally weighted four-factor portfolio to create Portfolio 2, because of momentum’s negative or low positive correlations with the other factors, tracking error is reduced by 84bps and the information ratio (IR) improves to 0.57 versus 0.46. The trading cost of Portfolio 2 is surprisingly lower than that of Portfolio 1, 32bps versus 33bps, respectively, at the US$10 billion portfolio level. 

Adding momentum to a multi-factor portfolio increases portfolio volume and lowers portfolio tilt, which lowers trading costs. Because momentum is associated with more-liquid stocks, the additional liquidity compensates for the increased turnover. In addition, momentum’s low or negative correlations with other factors leads to trades initiated by momentum’s rebalancing, which cancels out trades initiated by value or other factors.  

Adding the size factor in Portfolio 3 also yields performance and implementation cost benefits.  Adding size to the four-factor portfolio improves the return by 26bps (13.13% versus 12.87%), while lowering the cost by 7bps for a US$10 billion portfolio. Admittedly, the size strategy has the highest volatility of the six factors in our analysis. Thus, by including size, the volatility of the four-factor strategy increases by 0.6%, whereas the low correlation of size with the other factors reduces tracking error by 0.27%, improving the IR. 

In summary, because of diversification, in both performance and trading activity, adding momentum and size strategies to a multi-factor portfolio can have positive results. Importantly, however, the two strategies must be in the same portfolio so the off-setting trades realise the lower implementation costs.


As investors’ interest in multi-factor smart beta investing grows, understanding how to optimally combine factors is critical for desirable investment outcomes. A good understanding of the correlation and implementation cost of each factor needs to be investigated. When properly constructed and blended with other factors, the momentum and size factors, perhaps surprisingly, are helpful components in a multi-factor smart beta strategy. 

The addition of momentum helps lower tracking error and improves the IR because of negative or low positive correlations with other factors. These benefits are achieved without a large increase in implementation cost because offsetting trades across the factor strategies cancel each other out and because adding momentum improves liquidity. 

The size factor is actually rather inexpensive to trade because of its relatively broad coverage and low turnover. Thus, adding the size factor to the combination of other factors can improve the performance, and lower the tracking error, of the multi-factor strategy given the low correlation of size with the other factors, resulting in a higher IR together with a reduction in trading cost.

Research Affiliates strongly advocates the thoughtful design of a multi-factor strategy, which requires a conscious and deliberate decision to find the most advantageous balance between effectively harvesting the factor premium and implementation cost.  

Mike Aked is director of research, Australia at Research Affiliates.

Recommended for you



Add new comment