DBI - XLY: Consumer Discretionary Dashboard For August

2023-08-21 05:30:09 ET

Summary

• Durables/apparel and consumer services are undervalued relative to their historical baseline.
• The auto and components industry has the worst value and quality scores in the sector.
• XLY fast facts.
• 9 stocks cheaper than their peers in August.

This monthly article series shows a dashboard with aggregate subsector metrics in Consumer Discretionary. It is also a top-down analysis of sector ETFs like Consumer Discretionary Select Sector SPDR ETF ( XLY ), Fidelity MSCI Consumer Discretionary Index ETF ( FDIS ) and Vanguard Consumer Discretionary ETF ( VCR ), whose largest holdings are used to calculate these metrics.

Shortcut

The next two paragraphs in italic describe the dashboard methodology. They are necessary for new readers to understand the metrics. If you are used to this series or if you are short of time, you can skip them and go to the charts.

Base Metrics

I calculate the median value of five fundamental ratios for each subsector : Earnings Yield ("EY"), Sales Yield ("SY"), Free Cash Flow Yield ("FY"), Return on Equity ("ROE"), Gross Margin ("GM"). The reference universe includes large companies in the U.S. stock market. The five base metrics are calculated on trailing 12 months. For all of them, higher is better. EY, SY and FY are medians of the inverse of Price/Earnings, Price/Sales and Price/Free Cash Flow. They are better for statistical studies than price-to-something ratios, which are unusable or non available when the "something" is close to zero or negative (for example, companies with negative earnings). I also look at two momentum metrics for each group: the median monthly return (RetM) and the median annual return (RetY).

I prefer medians to averages because a median splits a set in a good half and a bad half. A capital-weighted average is skewed by extreme values and the largest companies. My metrics are designed for stock-picking rather than index investing.

Value and Quality Scores

I calculate historical baselines for all metrics. They are noted respectively EYh, SYh, FYh, ROEh, GMh, and they are calculated as the averages on a look-back period of 11 years. For example, the value of EYh for retailing in the table below is the 11-year average of the median Earnings Yield in retail companies.

The Value Score ("VS") is defined as the average difference in % between the three valuation ratios (EY, SY, FY) and their baselines (EYh, SYh, FYh). The same way, the Quality Score ("QS") is the average difference between the two quality ratios (ROE, GM) and their baselines (ROEh, GMh).

The scores are in percentage points. VS may be interpreted as the percentage of undervaluation or overvaluation relative to the baseline (positive is good, negative is bad). This interpretation must be taken with caution: the baseline is an arbitrary reference, not a supposed fair value. The formula assumes that the three valuation metrics are of equal importance. A floor of -100 is set for VS and QS when the calculation goes below this value. It may happen when metrics in a sub sector are very bad.

Current data

The next table shows the metrics and scores as of last week's closing. Columns stand for all the data named and defined above.

Value and Quality chart

The next chart plots the Value and Quality Scores by subsector (higher is better).

Evolution since last month

The value score has improved in all subsectors due to price action.

Momentum

The next chart plots momentum scores based on median returns.

Interpretation

Durables/apparel and consumer services are undervalued by 19% and 11% relative to 11-year averages, respectively. The latter has a better quality score. Retailing is close to the baseline regarding both valuation and quality. Auto and components are the less attractive subsector now: both scores are below the historical baseline.

Fast facts on XLY

Consumer Discretionary Select Sector SPDR Fund ( XLY ) started investing operations on 12/16/1998 and tracks the Consumer Discretionary Select Sector Index. It has 53 holdings and a total expense ratio of 0.10%, which is the same as VCR and a bit more expensive than FDIS (0.08%). The fund is well-balanced across four major industries: broadline retail (25.7% of asset value), hotels, restaurants & leisure (22.3%), automobiles (20%), and specialty retail (19.9%). Other subsectors represent only 12% of assets.

The next table shows the top 10 holdings with basic ratios and dividend yields. The portfolio is very concentrated: their aggregate weight is 71.7%, with 41.5% in the top two names. Amazon and Tesla represent 24.6% and 16.9% of the fund's asset value, respectively.

Ratios by Portfolio123

XLY has outperformed the S&P 500 ( SPY ) by 1.8 percentage points in annualized return since January 1999. Despite a higher risk measured in drawdown and standard deviation of monthly returns (volatility), XLY has also a better risk-adjusted performance (Sharpe ratio in the next table).

Data calculated with Portfolio123

XLY is ahead of SPY by 11.8 percentage points in 2023 to date.

In summary, XLY is a good fund with cheap management fees for investors seeking capital-weighted exposure in consumer cyclicals. Fidelity's fund has cheaper fees, but the difference is insignificant and XLY has much higher trading volumes. Therefore, it is a better instrument for trading and tactical allocation. Investors who are concerned by the high concentration in Amazon and Tesla may prefer the Invesco S&P 500 Equal Weight Consumer Discretionary ETF ( RSPD ).

Dashboard List

I use the first table to calculate value and quality scores. It may also be used in a stock-picking process to check how companies stand among their peers. For example, the EY column tells us that a retail company with an Earnings Yield above 0.0529 (or price/earnings below 18.90) is in the better half of the industry regarding this metric. A Dashboard List is sent every month to subscribers with the most profitable companies standing in the better half among their peers regarding the three valuation metrics at the same time. The list below was sent to subscribers several weeks ago based on data available at this time.

It is a rotational model with a statistical bias toward excess returns on the long-term, not the result of an analysis of each stock.

For further details see: