Indexology^® Blog

The fixed income market has historically been relatively illiquid in India, as well as globally. The Indian bond market is smaller than other Asian markets like China and Korea, but it is more liquid than they are, yet it is still largely inaccessible to retail investors. The nature of trading that is almost entirely over the counter, leading to price opacity, and the large issuance size of the bonds combine to make it out of reach for the average Indian investor. Bond exchange-traded funds (ETFs) may be able to solve these issues, which may be part of the reason bond ETFs have soared in popularity in developed markets recently.

The widespread perception is that bond ETFs help bring liquidity to the market. First, bond ETFs allow institutional and retail investors to partake in a larger pool of fixed income securities than they normally would have easy access to. Multiple bonds can be bought in smaller chunks defined by the ETF price and size, catering to all appetites, offering a solution to the large issuance size problem. In the case of Indian government securities, ETFs provide the smooth rollover benefit, where the most recently issued bond replaces the earlier issue with no costs associated for the investor, proving to be cost-effective for the average retail investor to manage. In addition, corporate bond market ETFs can be designed to capture desired duration and yield, which can make targeted exposure far easier to achieve.

Bond ETFs can offer low execution costs and allow price discovery. Buyers and sellers can offset each other, removing the need for the frequent buying and selling of underlying securities. As bond ETFs increase in popularity, the buying and selling of ETFs can far exceed that of the underlying securities, contributing to an overall increase in market liquidity. Price discovery happens as the price of the ETF should be a reflection of the index underlying it, which in turn is determined by the weighted sum of the underlying securities. In addition, the act of trading on stock exchanges, unlike for underlying securities, brings transparency to an opaque market. While investors can buy and sell ETFs as a single block, they don’t have to trade in the underlying securities. Authorized participants are permitted to trade in the underlying securities of the ETFs, with the ETF sponsor fulfilling the important role of keeping the net asset value of the ETF in line with the value of the underlying securities. Even during times of market stress, a bond ETF would be at least as liquid as the underlying securities. Finally, ETFs mandate the publication of the underlying securities to be made public daily, which makes the investment even more transparent than a bond fund.

Recently, S&P Dow Jones Indices published an analysis on the use of exchange-traded funds (ETFs) by U.S. insurance companies in their general accounts. This blog post provides additional details on the investments in Bond ETFs by insurance companies.

The original analysis noted that of the USD 27.2 billion insurance companies invested in ETFs, companies invested USD 7.9 billion in Bond ETFs. Insurance companies in every state—except Alaska, Delaware, Maine, and North Dakota—invested in Bond ETFs. However, companies in five states—Massachusetts, Indiana, Illinois, California, and Texas—accounted for 66% of the Bond ETF investments (see Exhibit 1).

As the paper noted, of the assets invested in Bond ETFs, companies chose to use the Systematic Valuation (SV) designation for 37%, or USD 2.9 billion, of the ETFs. SV is a “bond” like accounting treatment that has the potential to reduce volatility in statutory financials. Companies in 17 states accounted for all of the investments designated as SV. The top five states —Indiana, Massachusetts, Alabama, Connecticut, and Illinois— accounted for 86% of the SV investments (see Exhibit 2).

The overall U.S. use of the SV designation was 37%; however, of the Bond ETF investments in these 17 states, 49% had the SV designation. Exhibit 2 shows the distribution of Bond ETF investments in each state that used the SV designation.

Insurance companies decisively selected whether or not to use the SV designation. If they chose to use SV, they tended to use it for all their Bond ETFs. The companies that used SV designated on average  99.1% of their Bond ETF holdings as SV. Indeed, most of the companies designated 100% their Bond ETF holdings as SV.

The companies that used the SV designation also invested more heavily in ETFs.

Quality is a factor that is frequently disputed and debated. Academics and practitioners often argue whether quality is a factor at all in the traditional risk factor framework. Often times, the debate stems from the fact that there is no one consistent, overarching definition or metric to measure quality.

For example, some market participants see low accruals (accounting related) or low earnings variability as signs of earnings quality. At the same time, there are market participants who see profitability as a sign of a high-quality company, and they use measures such as gross profit margin, return on equity (ROE), or return on assets (ROA) as quality proxies.

Solvency measures such as the debt-to-assets ratio or debt-to-equity ratio assess the “safeness” and quality of a company.  In recent years, good corporate governance, as part of the greater trend in ESG investing, is viewed as an indicator of quality as well.

As another sign of the academic community recognizing that differences in quality factors lead to differences in stock returns, Nobel Laureate Eugene Fama and fellow researcher Ken French added two new quality factors—profitability and investment—to their asset pricing model in 2015.[1] Therefore, higher-quality companies, regardless of the definition of quality, on average earn higher risk-adjusted returns than lower-quality companies in both cross-sectional and time series analyses.

To demonstrate this point, we used the S&P 500^® Quality Index, which is a composite measure of ROE, accruals, and leverage, as an example. We can see that profitability, as measured by ROE, was rewarded over a long-term investment horizon (see Exhibit 1).[2] Quartile 1—which comprised the most profitable securities or those with the highest ROE—earned higher returns than the other quartiles.

Earnings quality is another desirable characteristic, as shown by long-term premium earned by companies with low earnings accrual ratios over those with higher earnings accrual ratios (see Exhibit 2).[3] Low earnings accruals indicate that earnings are representative of the company’s true earnings power, and that they are more likely to persist in the future.

The financial prudence of a company or its use of leverage is another indicator of quality. Unlike other factors, leverage usage does not necessarily have a linear relationship with returns (see Exhibit 3). While highly leveraged companies can potentially result in financial distress, a low leverage ratio can indicate that a firm may be relying too much on equity financing to finance future business opportunities.

Lastly, we formed quartile portfolios ranked by overall quality score.[4] The results show that higher-quality companies outperformed lower-quality companies (see Exhibit 4).

As with all factors, quality goes through performance cycles. Even though higher-quality companies outperformed lower-quality companies over the long-term investment horizon, there may have been periods when lower-quality companies performed better. In upcoming blogs, we will discuss quality rotation in conjunction with market environments when quality premium is positive.

[1]   Profitability is represented by robust minus weak (RMW), which is the average return on the two robust operating profitability portfolios minus the average return on the two weak operating profitability portfolios, as shown by the following equation.

[2]   In our analysis, we ranked all the securities in the S&P 500 universe on a monthly basis by ROE and then divided them into quartiles

[3]   We define accruals as the change in the company’s net operating assets (NOA) over the last year, divided by its average NOA over the last two years, as shown by the following equation.

[4]   We ranked constituents of the S&P 500, the underlying universe, on a monthly basis using the winsorized average z-score and calculated the back-tested returns of the ranked portfolios. For more information on the quality factor calculation, see https://spindices.com/documents/methodologies/methodology-sp-quality-indices.pdf.

I recently completed several meetings with financial advisors in Kansas and Tennessee. Traveling to meet with advisors in their offices or at events is something that I enjoy doing and informs my love of the work that our team does in advisor education. However, I am troubled by a reoccurring conversation that came up. In some of these conversations, advisors spoke of positioning for “participate but protect.” Having experienced a long bull market and higher-than-average gains in 2017, the positioning is “keep it between the guard rails” to describe how they are investing now.

But are these same advisors really positioned to protect? What changes has a 10-year “age of Taurus” brought to their portfolios? That bull market has slyly and slowly encouraged some to allow equity allocations to creep back up to pre-global financial crisis (GFC) levels of 70% equity weighting or higher. While they should not feel comfortable with that, they shouldn’t feel alone. Some of the most well-known U.S. asset managers offering target date solutions are in that same place. Our most recent S&P Target Date Scorecard indicated that a majority of “To” target date funds were at 70% or greater equity allocation for the 2035 vintage, and the same was true for “Through” for the 2030 vintage. To get to a more conservative 60/40 allocation, one would need to be well within 10 years of retirement (see Exhibit 1).

As a simple and uncluttered way to consider the risk/return of the more conservative 60/40 allocation, I commend Jason Giordano for his recently published Elevating the Aristocrats: Pairing the S&P 500^® Dividend Aristocrats^® with the S&P 500/MarketAxess Investment Grade Corporate Bond Index. His analysis of combining the S&P 500/MarketAxess Investment Grade Corporate Bond Index and S&P 500 Dividend Aristocrats is timely for an asset allocation discussion. In Exhibit 2, he shows the historical hypothetical[1] benefit in risk/return comparison to holding only the quality equity exposure of the S&P 500 Dividend Aristocrats during three market dips that have occurred in the U.S. equity market since the GFC.

A conclusion one might draw from this simplified portfolio example of data and analysis? Rather than lightening up on fixed income exposure to anticipate Federal Reserve actions, the advisor with an eye to protecting the portfolio may want to be more mindful of a prudent balance in their asset allocation.

[1]   The S&P 500/MarketAxess Investment Grade Corporate Bond Index was launched on Jan. 9, 2017, so the data in his paper are back-tested data.

In prior blogs,[i] we discussed the return contribution of mega-cap securities in 2017, as well as the impact of style classifications that may give small-cap active managers more autonomy to invest in significantly different risk exposures. In this blog, we look at active factor risks taken by active managers across three market-cap ranges against the appropriate S&P DJI style benchmarks.

Using the aggregate holdings of the managers[ii] and the Northfield US Fundamental Equity Risk Model, we observed the managers’ active exposures to systematic risk and whether those factor bets were rewarded.

Exhibit 1 shows the median fund’s[iii] active exposure for both book/price and log of market cap, which serve as the proxies for value and size factors, respectively. Funds have noticeably shifted their exposures to those two factors over the past 18 years.

For example, across all three market-cap categories, the median fund started the 2000s with a negative active exposure to size. In other words, the median active fund was invested in companies that, in general, were smaller than that of the respective benchmark. However, over the years, the median active fund’s exposure to size has increased across all market-cap ranges, as shown by the increasing bubbles. By the end of 2017, active exposure to size for mid- and small-cap managers was roughly in line with that of the respective benchmark.

Undoubtedly, the longest-running equity bull market we have been experiencing since the 2008 global financial crisis influenced this gradual shift to neutral weight in the market-cap factor that we observed in actively managed funds. As market-capitalization-weighted benchmarks increased their index values, and with market beta responsible for 312% of average benchmark return (see Exhibit 2), active managers could not afford to have a sizable underweight to the market factor or a significant overweight to the size factor.

Similarly, active exposure to the value factor has also been converging to that of the benchmark. Both mid- and large-cap funds started the evaluation period with high active exposure to book/price. This equated to the median fund investing in companies that were more “value-like” or “cheaper” than their benchmark. By the end of 2017, however, they had a marginal positive active exposure to the value factor. 

It is worth noting that the value factor performed rather poorly over the period from Feb. 28, 2009, until Dec. 29, 2017. Based on the Northfield US Fundamental Equity Risk Model, book/price returned -12.84% over this period. Among the five Fama-French factors, the value factor—as represented by high minus low portfolios formed by book/price ranking—returned -12.83% over the same period, compared with 37.95% delivered by the profitability factor.

It remains to be seen whether the size factor or the value factor will continue their performance cycle. One thing we can be certain of, based on the factor exposures of actively managed funds, is that active managers have displayed dynamic exposures to size and value factors, gradually shifting from active underweight to a more neutral position over time. That dynamic shift was in line with the performance of those factors.

[i]   The Impact of Size on Active Management Performance in 2017: Part 1 and The Impact of Style Classification on Active Management Performance in 2017: Part 2.

[ii]   Fund holdings were sourced from FactSet’s Ownership database on a monthly basis for all available funds within the CRSP dataset. The funds that met the style criteria were then pulled out for this analysis.

[iii] On a monthly basis, the benchmarks’ factor exposures were subtracted from each fund’s factor exposures to arrive at the active exposure. The funds were then averaged across each factor and year to create an average yearly active factor exposure for each fund. The median within each market capitalization, year, and factor was then presented in Exhibit 1.