Introducing ETP Analysis Service
active management, analysis, exchange-traded fund, exchange-traded product, portfolio

Alpholio™ has recently added the ETP Analysis Service to its platform. The exchange-traded product (ETP) is an exchange-traded fund (ETF), exchange-traded note (ETN), NextShares ETMF®, or other exchange-traded financial instrument.

The main motivation behind the new service is the availability of ETPs that do not track market-cap weighted indices. In particular, this includes “smart beta” (a.k.a. “strategic beta“) strategies that blend active and passive management. Due to the former aspect, smart-beta ETPs resemble traditional actively-managed mutual funds. Consequently, they can be analyzed with Alpholio™’s patented methodology, which constructs a custom reference portfolio of ETFs for each analyzed fund.

This leads to an apparent paradox: an analyzed ETP (which may be an ETF) is to be replicated by a portfolio of ETFs. Why do this at all? Just as with a traditional mutual fund, for several main reasons:

  • To determine whether active management aspect of the ETP adds value on a truly risk-adjusted basis
  • To understand the exposure of the analyzed ETP to various factors. This helps eliminate excessive exposures in the overall investment portfolio.
  • To replicate the ETP’s performance with other ETFs that may have preferable characteristics, such as lower fees, smaller trading premia or spreads, accessibility, etc. Conversely, to simplify a portfolio by substituting multiple ETFs with a single ETP.
  • To discern periods of underperformance and outperformance of the ETP after adjustment for its exposures.

Let’s demonstrate the new ETP Analysis Service in action. First, we will analyze the PowerShares FTSE RAFI US 1000 Portfolio (PRF). This ETP tracks the FTSE RAFI US 1000 Index, which

…is designed to track the performance of the largest US equities, selected based on the following four fundamental measures of firm size: book value, cash flow, sales and dividends. The 1,000 equities with the highest fundamental strength are weighted by their fundamental scores.

To conduct the analysis, we will use the simplest variant of Alpholio™’s methodology, which builds a reference ETF portfolio with both fixed membership and weights. The following chart and related statistics show the cumulative RealAlpha™ for the ETP (to learn more about this and other performance measures, please visit our FAQ):

Cumulative RealAlpha™ for PowerShares FTSE RAFI US 1000 Portfolio (PRF)

Over the five years through July 2016, the ETP added a small amount of value vs. its reference ETF portfolio of comparable volatility. The RealBeta™ of the ETF was the same as that of a broad-based equity market ETF.

The following chart with accompanying statistics presents the fixed composition of the reference ETF portfolio for the analyzed ETP:

Reference Weights for PowerShares FTSE RAFI US 1000 Portfolio (PRF)

The ETP had major equivalent positions in the iShares Russell 1000 Value ETF (IWD), Vanguard Value ETF (VTV), iShares Core S&P Total U.S. Stock Market ETF (ITOT), SPDR® S&P® 500 Value ETF (SPYV), PowerShares BuyBack Achievers Portfolio (PKW), and Guggenheim S&P 500® Pure Value ETF (RPV). Clearly, this ETP had a very strong exposure to the large-cap value factors represented by reference ETFs. (The Other component in the chart collectively depicts additional six ETFs with smaller weights, some of which were effectively zero.)

In the second example, let’s analyze the Guggenheim S&P 500® Equal Weight ETF (RSP). This ETP

Seeks to replicate as closely as possible the performance of the S&P 500 Equal Weight Index, before fees and expenses, on a daily basis.

Here is the chart with related statistics of the cumulative RealAlpha™ for this ETP:

Cumulative RealAlpha™ for Guggenheim S&P 500® Equal Weight ETF (RSP)

Over the five years through July 2016, this ETP also added little value vs. its reference ETF portfolio. Its RealBeta™ was above that of a broad-based stock market ETF.

The final chart and statistics show the static composition of the reference ETF portfolio for the ETP:

Reference Weights for Guggenheim S&P 500® Equal Weight ETF (RSP)

The ETP had major equivalent positions in the First Trust Large Cap Core AlphaDEX® Fund (FEX), iShares Russell Mid-Cap Value ETF (IWS), PowerShares S&P 500 Quality Portfolio (SPHQ), PowerShares S&P 500® High Beta Portfolio (SPHB), iShares Russell Mid-Cap Growth ETF (IWP), and Consumer Discretionary Select Sector SPDR® Fund (XLY). The Other component in the chart collectively represents additional six ETFs with smaller constant weights, one of which was effectively zero.

As could be expected, due to equal-weighting of its positions this large-cap ETP had a significant tilt toward mid-cap stocks, especially of value characteristics. In addition, the ETP had considerable exposure to economic sectors such as consumer discretionary, financials, technology, and industrials.

If you would like to take advantage of the new ETP Analysis Service, please register on our website.


Pin It
Analysis of Harding Loevner Global Equity Fund
analysis, mutual fund

A recent profile in Barron’s features the Harding Loevner Global Equity fund (HLMGX; Advisor Class shares). This $870 million no-load fund has a reasonable 1.18% expense ratio and 21% annual turnover (a five-year average as of the second quarter of 2016). According to the article

The payoff for investors has been 6.6% average annual returns over the past decade, versus 5.0% for the average world stock fund tracked by Morningstar.

The primary benchmark for the fund is the MSCI All Country World Index (ACWI). One of the efficient implementations of this index is the iShares MSCI ACWI ETF (ACWI). According to Alpholio™ calculations, from April 2008 through July 2016 the fund returned more than the ETF in approximately 49% of all rolling 36-month periods, 62% of 24-month periods and 66% of 12-month periods.

The fund’s average cumulative (not annualized) outperformance over a rolling 36-month period was 1.5%. However, the median was minus 0.15%, which indicates that the fund significantly outperformed the ETF in a relatively small number of rolling periods. (A rolling period of 36 months tries to approximate an actual average holding interval. Many investments in a fund do not start precisely at the turn of a calendar year.)

To take the fund’s volatility and exposures into account, let’s employ the simplest variant of Alpholio™’s patented methodology. As other variants, this one also constructs a reference portfolio of ETFs that most closely tracks periodic returns of the analyzed fund. The reference portfolio has a fixed ETF membership and weights. Here is the resulting chart with statistics of the cumulative RealAlpha™ for the Harding Loevner Global Equity (to learn more about this and other performance measures, please visit our FAQ):

Cumulative RealAlpha™ for Harding Loevner Global Equity Fund (HLMGX) over 10 Years

Over the ten years through July 2016, the fund underperfomed its reference ETF portfolio. The volatility of the fund, measured as the standard deviation of monthly returns, was slightly higher than that of the reference portfolio. The fund’s RealBeta™ was close to that of a broad-based domestic equity ETF.

The following chart with related statistic shows the constant composition of the reference ETF portfolio for the fund over the same evaluation period:

Reference Weights for Harding Loevner Global Equity Fund (HLMGX) over 10 Years

The fund had major equivalent positions in the iShares MSCI EAFE Growth ETF (EFG), Guggenheim S&P 500® Top 50 ETF (XLG), SPDR® Morgan Stanley Technology ETF (MTK), iShares U.S. Medical Devices ETF (IHI), and iShares MSCI Singapore ETF (EWS). The iShares TIPS Bond ETF (TIP) represented fixed-income holdings of the fund. The Other component in the above chart collectively embodies additional six ETF with smaller weights (see table).

A similar analysis of the fund over the five years through July 2016 yields even worse results:

Cumulative RealAlpha™ for Harding Loevner Global Equity Fund (HLMGX) over 5 Years

Over this shorter evaluation period, the fund generated a negative annualized discounted RealAlpha™. Its cumulative return was almost 15% lower than that of the reference ETF portfolio, whose composition is depicted in the following chart and accompanying statistics:

Reference Weights for Harding Loevner Global Equity Fund (HLMGX) over 5 Years

This reference portfolio indicates that the fund had a considerable exposure to financials (IXG), U.S. energy (IYE), Japan equity (EWJ), and U.S. technology (MTK and IGV).

In terms of cumulative RealAlpha™, the fund’s performance over the three years through July 2016 was similar to that over ten years (its annualized discounted RealAlpha™ was slightly negative).

In sum, over the most recent three-, five- and ten-year periods, the Harding Loevner Global Equity fund failed to add value when compared to the respective reference ETF portfolios. Despite a sensible turnover, in 2014 and 2015 the fund had long-term capital gain distributions around 2.5-4% of the NAV, which made it less suitable for taxable accounts.

To learn more about the Harding Loevner Global Equity and other mutual funds, please register on our website.


Pin It
Analysis of T. Rowe Price Dividend Growth Fund (Update)
analysis, mutual fund

A recent piece in Barron’s features the T. Rowe Price Dividend Growth Fund (PRDGX). This $5.9 billion no-load fund sports a competitive 0.64% expense ratio and 25% turnover. According to the article, the fund

…has outperformed the Standard & Poor’s 500 index year to date and over one-, three-, 10-, and 15-year periods. So far this year, it has returned 10.21%, versus 8.4% for the S&P 500.

Unlike our previous analysis, which covered a longer-term performance of the fund, this evaluation will generally focus on a recent shorter time period.

The prospectus benchmark for the fund is the S&P 500® Index. One of the long-lived and efficient implementations of this index is the SPDR® S&P 500® ETF (SPY). Alpholio™’s calculations indicate that over the five-year interval through July 2016, the fund returned more than the ETF in only 16% of all rolling 36-month periods, 22% of 24-month periods and 47% of 12-month periods. The median cumulative (not annualized) return difference over a rolling 36-month period was minus 3%.

To adjust for the fund’s volatility, let’s employ the simplest variant of Alpholio™’s patented methodology. In this approach, a reference portfolio of ETFs with fixed both membership and weights is constructed such that its returns most closely track those of the fund. Here is the resulting chart and statistics of the cumulative RealAlpha™ for the T. Rowe Price Dividend Growth (to learn more about this and other performance measures, please visit our FAQ):

Cumulative RealAlpha™ for T. Rowe Price Dividend Growth Fund (PRDGX)

Over the five years through July 2016, the fund subtracted value on a risk-adjusted basis. The fund’s volatility, measured as a standard deviation of monthly returns, was comparable to that of the reference ETF portfolio. The RealBeta™ of the fund was slightly lower than that of a broad-based equity ETF.

The following chart with related statistics shows the constant composition of the reference ETF portfolio over the same analysis period:

Reference Weights for T. Rowe Price Dividend Growth Fund (PRDGX)

The fund had major equivalent positions in the Vanguard High Dividend Yield ETF (VYM), PowerShares Dynamic Large Cap Growth Portfolio (PWB), First Trust Large Cap Growth AlphaDEX® Fund (FTC), PowerShares S&P 500 Quality Portfolio (SPHQ), iShares U.S. Industrials ETF (IYJ), and Vanguard Dividend Appreciation ETF (VIG). The Other component in the chart collectively represents six additional ETFs with smaller weights (15.6% in total).

Over the three-year period through July 2016, the fund still failed to add a meaningful amount of value over its reference ETF portfolio:

Cumulative RealAlpha™ for T. Rowe Price Dividend Growth Fund (PRDGX) over 3 Years

The composition of the reference portfolio over this shorter period was different from the previous one, although the top-six equivalent positions also accounted for more than 80% of holdings.

Over the most recent three and five years, the T. Rowe Price Dividend Growth Fund failed to add a significant amount of value when compared to a static reference ETF portfolio. Similarly to 2014, the fund had a substantial long-term capital gain distributions in 2015, which diminished its suitability for taxable accounts. The relatively low expense ratio and modest turnover work in the fund’s favor.

To learn more about the T. Rowe Price Dividend Growth and other mutual funds, please register on our website.


Pin It
Evaluating Lazy Portfolios
analysis, asset allocation, portfolio

MarketWatch tracks eight lazy portfolios. Each of these simple portfolios consists of three to eleven, low-cost, no-load index mutual funds from Vanguard®. The fund membership and weights in each portfolio remain constant over time. (In theory, this implies that each portfolio would have to be perfectly rebalanced daily. This is not only impractical but also impossible because the fund’s daily NAVs, and hence their new weights, are not known until after the market close.)

Unfortunately, MarketWatch compares lazy portfolios solely on the basis of annualized returns in one-, three-, five- and ten-year periods. Volatility of returns as well as other performance measures are not taken into account. Luckily, Alpholio™ can help – not only does our Basic Portfolio service provide ample statistics, but it also allows for a selectable periodic rebalancing of portfolio positions to their original weights. For the purpose of this analysis, let’s assume a 15-year evaluation period from July 2001 through June 2016, as well as quarterly rebalancing of each portfolio.

Let’s start with the most complex Aronson Family Taxable Portfolio that consists of 11 funds. The following chart shows the cumulative return and related statistics for this lazy portfolio:

Aronson Family Taxable Portfolio - Cumulative Return - 15 Years

The fixed-income portion of the portfolio comprises inflation-protected securities (15%), long-term Treasury bonds (10%) and high-yield corporate bonds (5%). The portfolio’s holdings also include domestic (40%) and foreign (30%) equities.

The alpha and beta of the portfolio were measured against the broad-based U.S. stock market ETF, and not just a large-cap index, such as the S&P 500®. Because high-yield bonds generally have a substantial correlation to equities, it could be expected that the portfolio’s beta would be approximately between 1 – (0.15 + 0.10 + 0.05) = 0.7 and 1 – (0.15 + 0.10) = 0.75, which it was at 0.73.

The key measures of risk-adjusted performance are the Sharpe and Sortino ratios. Unlike the former, the latter penalizes portfolios with a large downside deviation.

Finally, the maximum drawdown measure is the maximum percentage loss of the portfolio value from a peak to a subsequent trough. Given the chosen evaluation period, this typically means a decline in each lazy portfolio’s value from October 2007 to March 2009.

The following chart shows rolling volatility (measured as a standard deviation of two years of monthly returns) and accompanying statistics for the portfolio:

Aronson Family Taxable Portfolio - Rolling Volatility - 15 Years

As could be expected, volatility of the portfolio significantly increased during the financial crisis. In general, a good lazy portfolio should maximize returns, minimize volatility, and reduce the magnitude of volatility changes over time.

Similar charts and statistics can easily be generated for all lazy portfolios. They are not published in this post to limit its size. Instead, here is a summary table of statistics:

Portfolio Annualized
Return
Alpha Beta Sharpe
Ratio
Sortino
Ratio
Maximum
Drawdown
Aronson Family
Taxable
6.96% 0.22% 0.73 0.53 0.76 42.08%
Fundadvice
Ultimate
Buy & Hold
6.49% 0.22% 0.62 0.55 0.81 36.58%
Dr. Bernstein’s
Smart Money
6.12% 0.18% 0.65 0.51 0.75 38.00%
Coffeehouse 6.89% 0.25% 0.63 0.59 0.86 36.16%
Yale U’s
Unconventional
7.89% 0.31% 0.69 0.61 0.88 42.94%
Dr. Bernstein’s
No Brainer
6.12% 0.11% 0.83 0.43 0.63 44.48%
Margaritaville 5.86% 0.14% 0.71 0.45 0.65 41.29%
Second
Grader’s
Starter
5.92% 0.05% 0.94 0.39 0.56 49.08%

The Yale U’s Unconventional portfolio had the highest risk-adjusted returns, as measured by the above Sharpe and Sortino ratios. This was likely due to the relatively large positions in REITs and long-term government bonds, both of which benefited from falling interest rates. Please also note that at times, correlation between returns of the REIT and total stock market mutual funds was quite high (which reduced portfolio diversification), as illustrated by the following chart:

Rolling 36-Month Return Correlation between VGSIX and VTSMX

The Coffeehouse portfolio had similar characteristics. Compared to others, this portfolio also exhibited the smallest maximum drawdown.

The Fundadvice Ultimate Buy & Hold portfolio had the third best return-to-risk profile, as well as the second lowest maximum drawdown. While bond funds in this portfolio had short and intermediate maturities, its total fixed-income component was significant, as was the case with the previous two portfolios.

For completeness, here are the statistics for lazy portfolios over a ten-year period through June 2016:

Portfolio Annualized
Return
Alpha Beta Sharpe
Ratio
Sortino
Ratio
Maximum
Drawdown
Aronson Family
Taxable
5.94% 0.01% 0.75 0.46 0.65 42.08%
Fundadvice
Ultimate
Buy & Hold
5.06% 0.00% 0.65 0.43 0.62 36.58%
Dr. Bernstein’s
Smart Money
5.41% 0.01% 0.67 0.46 0.66 38.00%
Coffeehouse 6.33% 0.09% 0.66 0.54 0.78 36.16%
Yale U’s
Unconventional
6.85% 0.10% 0.74 0.52 0.73 42.94%
Dr. Bernstein’s
No Brainer
5.64% -0.06% 0.83 0.41 0.59 44.48%
Margaritaville 4.95% -0.05% 0.73 0.39 0.54 41.29%
Second
Grader’s
Starter
5.76% -0.10% 0.93 0.39 0.56 49.08%

Over this shorter evaluation period, the Coffeehouse portfolio had the best risk-adjusted returns, followed by the Yale U’s Unconventional portfolio, and Dr. Bernstein’s Smart Money portfolio that had a slightly higher Sortino ratio and a smaller maximum drawdown than the Aronson Family Taxable portfolio. This goes to show that the ranking of portfolios heavily depends on the analysis time frame.

We hope that this analysis will give investors additional insights into historical performance of lazy portfolios. Of course, there is no guarantee that this performance will be repeated in the future.


Pin It
Analysis of TIAA-CREF Bond Fund
analysis, mutual fund

Today’s profile in Barron’s features the TIAA-CREF Bond Fund (TIORX; Retail Class shares). This $3.5 billion no-load fund has a 0.62% expense ratio and an elevated 309% turnover. According to the article

[active management] has guided [this] low-cost fund to 4.5% average annual returns over the past three years—better than 85% of intermediate-bond funds tracked by Morningstar and ahead of the 4.2% average annual gains for the Barclays U.S. Aggregate Bond Index.

The current lead manager took the reins of the fund in late August 2011. Therefore, this analysis will span from September 2011 through June 2016.

The primary prospectus benchmark for the fund is the Barclays U.S. Aggregate Bond Index. One of the accessible implementations of this index is the iShares Core U.S. Aggregate Bond ETF (AGG). Alpholio™’s calculations show that in the above time frame the fund returned more than the ETF in approximately 91% of all rolling 36-month periods, 80% of 24-month periods and 68% of 12-month periods. The fund’s median cumulative (not annualized) outperformance over a rolling 36-month period was 2.1%.

A comparison of rolling returns tries to approximate the average holding period of the fund. However, it does not take the fund’s composition and volatility into account. To adjust for the latter, let’s apply the simplest variant of Alpholio™’s patented methodology. In this approach, a reference ETF portfolio with both fixed membership and weights is custom-built to most closely track returns of the analyzed fund. Here is the resulting chart and related statistics of cumulative RealAlpha™ for the TIAA-CREF Bond Fund:

Cumulative RealAlpha™ for TIAA-CREF Bond Fund (TIORX)

Over the entire analysis period, the fund added virtually no RealAlpha™. At 2.88%, the fund’s standard deviation (a measure of volatility of returns) was 0.08% higher than that of the reference portfolio. The fund’s RealBeta™ was close to zero, which implies very little correlation of the fund’s returns to those of the broad-market equity ETF.

The following chart and accompanying statistics illustrate the constant reference ETF portfolio for the fund:

Reference Weights for TIAA-CREF Bond Fund (TIORX)

The fund had major equivalent positions in the Vanguard Mortgage-Backed Securities ETF (VMBS), SPDR® Barclays Intermediate Term Corporate Bond ETF (ITR), iShares Intermediate Credit Bond ETF (CIU), Vanguard Intermediate-Term Corporate Bond ETF (VCIT), Schwab U.S. Aggregate Bond ETF™ (SCHZ), and PIMCO 0-5 Year High Yield Corporate Bond Index ETF (HYS). The Other component in the chart collectively represents additional six ETFs with smaller weights. (Some of the weights in the above table are shown as zero due to rounding.)

Under current management, the TIAA-CREF Bond Fund added practically no value on a truly risk-adjusted basis. An investor could have achieved similar results with a simple reference ETF portfolio comprising just six to nine fixed positions. The article provides a likely explanation

…[the lead manager] is quick to give credit to the 13 managers who run individual “sleeves” of the portfolio.

It appears that all the intense trading of the fund’s holdings (re: turnover in excess of 300%) merely compensated for its substantial management expenses.

To learn more about the TIAA-CREF Bond and other mutual funds, please register on our website.


Pin It
Recent Posts
Recent Comments
Archives
Meta