For the most part, exchange traded notes (ETNs) are better off ignored...but this one is a rare exception to that general principle, writes Samuel Lee of Morningstar ETFInvestor.
I’ve basically sworn off exchange traded notes, but when a Nobel-caliber economist creates one, you pay attention.
I’ve made no secret of how much I admire Yale professor Robert Shiller. His research on stock-market volatility and intrinsic value persuaded me that markets weren’t as efficient as I had thought them to be. I’m also a fan of cyclically adjusted price-to-earnings, or CAPE, a market valuation measure he and professor John Campbell popularized.
For those of you unfamiliar with this measure, it’s simply current price divided by the average of trailing ten-year earnings, adjusted for inflation. The original idea actually came from the father of value investing, Benjamin Graham, so sometimes it’s called Graham-Dodd P/E. Other names include Shiller P/E and P/E 10.
The rationale for averaging earnings is to smooth out the fluctuations introduced by the business cycle to get a clearer idea of the market’s “true” earnings power. The measure has been shown to be a good predictor of long-term stock-market returns and is the basis of a lot of academic and practitioner stock-market forecasting models.
A Defense of CAPE
Barclays ETN+ Shiller CAPE ETN (CAPE), unsurprisingly, uses CAPE the market-valuation measure. If you don’t believe CAPE is a good valuation tool, then CAPE the ETN is suspect. Many CAPE critics com-plain it provides an overly glum view of market valuations, because it captures two big earnings busts, the biggest and most recent from the financial panic of 2008.
Well, yes, but it also captures two huge earnings booms. Moreover, the busts were short-lived, relative to history, in large part thanks to aggressive Federal Reserve moneyprinting and fiscal stimulus. Bizarrely, the solution CAPE’s critics usually offer is to focus on 12-month trailing or forward operating earnings.
This is a horrible solution. As a general rule, you extend the measurement period the more volatile a time series, in order to better cancel out the noise. The 12-month P/E, whether trailing or forecast, is procyclical, making equities look cheap during booms and expensive during recessions. I reserve special ire for operating earnings.
It’s one of those things that sounds good in theory, but doesn’t work in practice. Operating earnings, simply defined, is profits adjusted for one-off events in order to better represent the true earnings power of a company.
If a company sells a unit, its reported profits will overstate earnings power. If a company pays a big one-time fine, its profits will understate earnings power. However, companies like to categorize bad surprises as singular events and good ones as recurring, leading to consistent overstatement of profits.
Wall Street analysts play along because their employers do a lot of business with the firms they rate. The prospect of being fired for issuing too many negative opinions has a way of concentrating analysts’ minds on the bright side.
CAPE’s real weakness is that it’s slow to account for permanent systematic shifts to earnings. These could include permanently lower corporate tax rates or permanently greater shares of foreign earnings.
Adjusting CAPE for these factors, however, doesn’t change valuations by much, because corporate taxes have been low for a while and foreign sales have been only slowly ramping up.
According to the US Bureau of Economic Analysis’ National Income and Product Accounts, foreign profits accounted for 18% of all corporate profits in 2002 and grew to 24% in 2011. The aggregate corporate tax rate fell to 20% from 25% over the same period. Together, they might call for a modest adjustment to CAPE.
How CAPE the ETN Works
Even if you’re not convinced CAPE (the measure) is not a good gauge of stock market valuations, you can still learn a lot by looking at how CAPE the ETN is constructed. Shiller had a hand in formulating the strategy, and it shows. It’s built on sound principles.
The ETN tracks the Shiller Barclays CAPE US Core Sector Index, which every month rotates its exposures to four of the nine S&P 500 Select Sector indexes, the same ones the Select Sector SPDR ETFs track.
The “Relative CAPE Indicator” indicates which sectors to pick. It’s based on cyclically adjusted price-to-earnings, or CAPE, which divides real price by average ten-year trailing real earnings.
Because sectors have different earnings characteristics, naively picking the lowest-CAPE sectors would structurally bias the fund to low-growth sectors. In order to get around that, a sector’s current CAPE is divided by the trailing 20-year CAPE average to produce the Relative CAPE Indicator. The five sectors most undervalued according to the indicator are chosen. Then, the sector with the lowest 12-month price return is excluded. The remaining four are equally weighted.
The momentum screen is a nice touch. The cheapest sectors are often the most beaten-down, and because momentum is a powerful force in the markets, automatically picking up the cheapest stuff regardless of momentum often exposes you to extra volatility.
While I think the strategy is sensible, keep in mind for most of its back-tested history, which only extends back to late 2002, it kept mostly to five sectors: energy, industrials, consumer staples, health care, and financials. The momentum screen kicked out financials right before the worst of the financial crisis, boosting back-tested returns.
I’m almost somewhat uncomfortable with the short ten-year back-tested history. While I think Shiller has a lot of credibility as a careful economist, I’d feel more at ease if the strategy were shown to work over multiple decades and in foreign countries.
The ETN levies a 0.45% annual investor charge. Keep in mind that ETNs can pretty much calculate expenses any way they want and bury additional fees in the dense prospectus text. Fortunately, I did not find any idiosyncratic fees or calculations, so you can expect the ETN to lag its target index by 0.45% annually.