Back to all posts

21 February 2018

Introducing our new default Style Research Style Skyline™

With hundreds of factors in circulation, and more creative smart beta offerings and factor-based funds in the marketplace, investors may be forgiven for being a little confused. The Style Research Style Skyline™ has long helped investment professionals cut through the noise by providing easy to understand views of key factor exposures and their impacts on portfolio risk and performance. Now we are excited to announce the latest version of the default Style Skyline, fostering even greater insight for today’s dynamic market.

From its inception, over twenty years ago, the Style Research Style Skyline has offered deeper and broader views of Style Tilts™ than other style analyzes. Style boxes, for example, miss the variety of exposures found in equity portfolios. By over-simplifying the underlying investment processes, misclassifications and other real challenges arise when trying to differentiate between funds. Rudimentary or proprietary factor combinations may lead to false and misleading conclusions.

As innovators in the industry, we first introduced the Style Research Style Skyline™ in the 1990’s, with easy to interpret color-coded style groupings to foster insights about different types of factors. At that time, other analytics providers often showed just value and growth, with these positioned at opposite ends of the same spectrum. They were (and many still are) effectively equating poor value with ‘growth’, forcing securities and funds to be either value or growth but not both. The original Style Skyline went further than just value and growth criteria, incorporating factors such as momentum, size, beta and financial leverage.

The Style Research approach has continued to include distinct yet accepted factors, providing a much truer assessment of portfolios and allowing for more subtle combinations of styles.

Added factors, new style categories, sharper insights

Our clients are seeking more flexibility in how factors are categorized, reflecting changes in market dynamics and investor preferences. To meet their needs, and the demands of the industry as a whole, we added 37 new factors and three style categories to this latest evolution of the Style Research Style Skyline™, along with functionality for clients to reclassify certain factors based on their own preferences. While more factors and discretion over their grouping offer greater opportunity for building tailored Style Skylines, many clients also prefer to use a default as the foundation of their analysis. We have therefore created a new ‘standard’ Style Skyline, incorporating the latest research on relevant factors.

Style Research Style Skyline
Figure 1: The new Style Research Style Skyline™ adds three new style categories – yield, quality and volatility – and additional factors (click for larger image).

A closer look at the Style Skyline changes

The new category and factor choices reflect the best elements of current analysis practice demonstrated by unbiased quantitative research.

Value and Yield

Value is the least changed category. The factor “EBITDA to EV” replaces “EBITDA to Price” to better incorporate a company’s economic value. Dividend yield is reclassified into a separate, new ‘yield’ category. Though still correlated with value factors, dividend yield has exhibited idiosyncratic behavior in recent years. In a low interest rate environment, investors seeking income have bid up the prices of high yield stocks, resulting in their returns being distinct from value stocks from time to time.

Figure 2 below plots the average rolling five-year correlation between dividend yield factor returns and those of the remaining value factors. During the 1990’s and through the dotcom period, the correlations were strongly positive and only began to fall in the mid-2000’s, even dipping into negative territory post-financial crisis. More recently the correlation has picked up a little, but yield is still somewhat decoupled from the main value factors.

correlation between dividend yield factor returns and value factors
Figure 2: Average rolling five-year correlation between dividend yield factor returns and other value factors (click for larger image).

 

This detached behavior warrants the creation of the separate yield group. Increasing interest in share buybacks and the total return to shareholders support the inclusion of the total shareholder yield in this category. Income has long been one of the standard ways to classify funds. With the growth of interest in this area, this new style group will enable a closer examination of funds exposed in this area, independent of their valuations.

Growth and Quality

The growth category has undergone more changes. Historic dividend growth joins earnings and sales growth measures, and all are harmonised onto a five-year look-back period. Forecast 12-month earnings growth completes the set. Two factors, formerly classified as growth, have been moved to the quality group: ROE and net profit margin. They join financial leverage, and earnings and sales growth stability in that category. Quality is a far less coherent group than other styles, with no general agreement between investors as to the most appropriate measures. However, the factors chosen for the standard Style Skyline include popular measures of profitability, stability and balance sheet strength measures of quality.

Figure 3 below shows the relationship between ROE returns with growth (in green) and quality (purple) factor returns. While it is evident that, in general, the correlation between ROE and quality has been higher than with growth, the picture reversed in the last couple of years, and we have moved ROE into the quality group. We concur with common practice to bracket profitability with measures of quality such as low leverage and stable earnings. Clients preferring the original classification can, of course, continue to use the legacy Style Skyline or benefit from the flexibility Style Research solutions provide to create bespoke Skylines by defining their own desired classifications.

ROE returns against growth and quality returns
Figure 3: The relationship of ROE returns against growth and quality factor returns (click for larger image).

 

Momentum

The decision to move the forecast revisions factor to the momentum group is supported by Figure 4 below. Whilst there has consistently been a positive correlation between its factor returns with those of growth factors (green), the link with momentum (black) is much stronger. Earnings revisions are closely related to earnings momentum. The addition of this factor to the total return based momentum factors serves to broaden that category into a more general view of momentum.

forecast revisions against growth and momentum factors
Figure 4: The relationship of returns of forecast revisions against growth and momentum factor returns (click for larger image).

 

Note also the replacement of medium-term momentum with the more commonly used 12-1 momentum. This omits the latest month, to mitigate short-term reversal effects.

Size

Market cap continues to be the default measure of size, in its own group.

Volatility

Given the popularity of low-risk strategies, we have created a new category for volatility to house the market beta factor and measures of volatility over one and three years. These form a coherent group.

Behind the scenes of styles and factors

Taken together, these adjustments to the default Style Research Style Skyline reflect today’s investment landscape. Many other factors can be selected by clients, and other categories, such as ESG or economic, can be included. The new default Style Skyline remains straightforward whilst covering the main themes relevant to today’s investors.

In addition to being in line with current market practice, the reclassifications discussed above are supported quantitatively. We conducted a cluster analysis to determine how closely different factors are related to one another. For example, in the US the analysis reveals that the highest similarity is between momentum factors, and separately amongst the revisions factors. In turn, these group together with each other, not with the growth category. Volatility factors also form a tight group. And, as would be expected, value factors are very coherent. But measures of yield cluster into a related but distinct block of factors.

Stability factors correlate strongly within themselves, but pair up weakly with the profitability ratios. The factors within the growth group are also reasonably well correlated, though it is fair to say that the dividend growth factor is less integrated, especially when compared with forecast earnings growth.

We’ve summarized the latest data for the final groupings in the new standard Style Skyline in Figure 5 below, which shows average intra- and inter-category correlations over the last ten years for sector-adjusted US factor returns.

Figure 5 (click for larger image)

 

Our Style Skyline’s refreshed factors and classifications support what we hear from our clients in terms of what they need for analysis of funds, peer groups and markets. They are supported quantitatively and capture the main factor drivers of equities in consistent and complementary groups. The new default Style Research Style Skyline™ gives clients a comprehensive and easy to understand view of factor exposures for individual portfolios, funds of funds and entire peer groups. Style Research provides the ideal framework for managers seeking to differentiate their investment processes, and investors and selectors to pinpoint portfolios with targeted factor biases.

 

Request a demo