Picking The Diamonds Out Of The ETF Rough

by: David Trainer

Hav­ing too many choices can be intim­i­dat­ing. And there are def­i­nitely lots of choices when it comes to ETFs. For exam­ple, in the equity mar­ket alone, there are 30-plus tech­nol­ogy sec­tor ETFs, or 35 "large cap value" and 20 finan­cial ETFs. A very healthy selec­tion abounds for every cat­e­gory of ETF.

The prob­lem is that these ETFs are not made the same even though they may be in the same cat­e­gory. There are major dif­fer­ences in method­olo­gies between funds, which results in dras­ti­cally dif­fer­ent hold­ings even within a given sec­tor. See Fig­ure 1.

Fig­ure 1: Hold­ings Count of Tech­nol­ogy Sec­tor ETFs

Click to enlarge

Sources: New Con­structs, LLC

What is an investor to do? How do you pick the ETF that will deliver the best performance?

That ques­tion was posed to the major ETF issuers at Mornigstar’s ETF con­fer­ence in Chicago a cou­ple weeks ago. The issuers admit­ted that the land­scape for ETFs is a bit clut­tered and may be poised for attrition.

I was not impressed by the issuers’ plan for address­ing the con­fu­sion caused by hav­ing too many ETFs. Their plan, not sur­pris­ingly, was to invest more in edu­cat­ing investors about their ETFs. In other words, they plan to invest more in sales and mar­ket­ing with the goal of steer­ing more busi­ness to their ETFs.

Like most Wall Street types, the issuers are focused on how to profit from ETFs, not how to help investors profit from ETFs.

Accord­ingly, there is a dearth of qual­ity research on funds. When I asked the issuers if they had any plans to nar­row the gap in qual­ity of research between funds and stocks, they basi­cally said “no, there is plenty of qual­ity research already out there.”

I dis­agree.

As high­lighted by Bar­rons’ The Dan­ger Within, for research on funds to rival the qual­ity of research on stocks, fund research must be based on rig­or­ous analy­sis of all the hold­ings within a fund.

Given that there are lots of funds that hold lots of dif­fer­ent stocks, any effort to pro­vide broad-based cov­er­age of funds would have to bring high-quality cov­er­age of lots of stocks to the table.

Unfor­tu­nately, qual­ity of research and num­ber of stocks cov­ered tend to be inversely correlated.

There are a few excep­tions, even fewer that are truly independent.

Mean­while, too many investors have been led to believe they can trust the ETF label, i.e. cat­e­gory, and believe that all ETFs within a cer­tain cat­e­gory are the same when they are not.

For exam­ple, ProShares Ultra Semi­con­duc­tors (NYSEARCA:USD), our top pick among tech­nol­ogy ETFs, holds 47 stocks while Pow­er­Shares Dynamic Soft­ware (NYSEARCA:PSJ), our worst-rated tech­nol­ogy ETF, holds only 30. The Van­guard Infor­ma­tion Tech ETF (NYSEARCA:VGT) holds over 400 stocks.

There is no sub­sti­tute for rig­or­ous analy­sis of every fund hold­ing if one cares about invest­ment performance.

How else can one mea­sure the poten­tial risk and reward of mak­ing an invest­ment in a fund? After all, a fund is the sum of its indi­vid­ual stock parts.

A chal­lenge to qual­ity fund research, as pref­aced above, is being able to bring qual­ity research to bear on all the many stocks that all the many funds hold.

To give the reader a sense of how dif­fer­ent funds can be, I share New Con­structs analy­sis of tech sec­tor ETFs.

Fig­ure 1 lists the top 25 tech­nol­ogy ETFs and their hold­ings. The take­away is sim­ple: tech sec­tor ETFs are not all made the same. Dif­fer­ent ETFs have mean­ing­fully dif­fer­ent num­bers of hold­ings and, there­fore, dif­fer­ent allo­ca­tions to hold­ings. Given the dif­fer­ences in hold­ings and allo­ca­tions, these ETFs will likely per­form quite differently.

To iden­tify the best ETFs within a given cat­e­gory, investors need a pre­dic­tive rat­ing for ETFs based on analy­sis of the under­ly­ing qual­ity of earn­ings and val­u­a­tion of the stocks in each ETF. See Fig­ure 2.

Fig­ure 2 shows how the 25 tech sec­tor ETFs stack up ver­sus each other, the over­all sec­tor and the S&P 500 based on their risk/reward rat­ings and the allo­ca­tion of their hold­ings by rating.

Fig­ure 2: Invest­ment Merit Based on Hold­ings and Allocations

Click to enlarge

Sources: New Con­structs, LLC and com­pany filings

Pre­dic­tive rat­ings empower investors and advi­sors to make more informed invest­ment deci­sions about funds.

Bottoms-up analy­sis of a fund based on the qual­ity of its hold­ings provides:

  1. Intel­li­gent dif­fer­en­ti­a­tion between funds
  2. Due dili­gence to pro­tect against own­ing funds with stocks that might blow up

For a sam­ple of our bottoms-up analy­sis on stocks in the tech­nol­ogy sec­tor, some of the best names in the sec­tor right now include:

  • Apple (NASDAQ:AAPL): At over 200%, they have the high­est return on cap­i­tal of all large-cap stocks, tech or non-tech. For every dol­lar they put into their busi­ness, they get $2 back per year. Most peo­ple are happy with 15–20 cents on the dollar.
  • Google (NASDAQ:GOOG): With a 44% annu­al­ized return on cap­i­tal and a mar­ket val­u­a­tion that implies almost no future profit growth, New Con­structs finds the tech giant stock has very low expec­ta­tions and high growth potential.
  • Microsoft (NASDAQ:MSFT): With a 72% return on invested cap­i­tal, the soft­ware maven also boasts a val­u­a­tion that implies a per­ma­nent 30% decline in its prof­its. Even if you expected zero growth, the implied value of Microsoft’s shares are slightly over $37.

We also like Accen­ture (NYSE:ACN), Research In Motion (RIMM) and Ora­cle (NASDAQ:ORCL).

Fund Rat­ing Methodology

Our analy­sis is based on aggre­gat­ing results from our mod­els on each of the com­pa­nies included in every ETF as of Sep­tem­ber 27, 2011. We aggre­gate results for the ETFs in the same way the ETFs are designed. Given the suc­cess of our rat­ing sys­tem for indi­vid­ual stocks, we believe its appli­ca­tion to groups of stocks (i.e. ETFs and funds) helps investors make more informed ETF and mutual fund buy­ing deci­sions.

Disclosure: I am long GOOG, MSFT, AAPL, ACN, RIMM, ORCL.