Gyroscopic Investing

I really learned quite a bit about the flaws of conventional market cap weighted funds like VTI from Episode 18 of Meb Faber's new podcast. Rob Arnott was the guest, and they had a long, deep discussion. I am now really thinking about just swapping out my stock ETFs with fundamental weighted ETFs.

I am at Schwab, they have an Arnott affiliated small family of six ETFs. FNDX FNDB FNDF FNDA FNDC FNDE are the symbols I believe.

http://www.morningstar.com/advisor/v/82 ... flawed.htm

His backtesting shows a monkey-seleted fund beats capweighting. Why?

ochotona wrote:I am at Schwab, they have an Arnott affiliated small family of six ETFs. FNDX FNDB FNDF FNDA FNDC FNDE are the symbols I believe.

The devil is in the details. Most of the "smart beta" funds won't outperform market cap weighting in the long-term because of the higher management fees as well as how they poorly weight the securities and the excessive number of securities bought. It needs to be low cost, weighted by factor and concentrated enough to beat the legacy baggage of taxes, trading commissions and all the other seen and unseen fees in market cap weighted funds.

And keep in mind, factors go in and out of season and may disappear alltogether. You may be better off with a broad equal weight fund like RSP than trying to time or allocate to factors. And at this point only growth and value are still fundamental drivers; momentum is a derivative of growth; value is a derivative of quality; dividends is a derivative of value; low volatility/low beta is a derivative of Consumer Non-Cyclicals. Everything else is just data mining bullshit or Wall Street wouldn't be pushing it.

boglerdude wrote:http://www.morningstar.com/advisor/v/82 ... flawed.htm

His backtesting shows a monkey-seleted fund beats capweighting. Why?

Probably tracking error due to lack of career and institutional constraints. Buy RSP if you want the same effect.

Why do equal weighted funds outperform capweighted

The thumbnail sketch answer is that as the major stocks like Apple, Facebook, Nike, Google (FANG) get more and more valuable, and take up more and more of the market-cap index, they also all get closer and closer to being fully valued, and their forward-looking returns diminish. The market-cap index is distorted, and looks like a few companies, and not like the US economy as a whole.

boglerdude wrote:Why do equal weighted funds outperform capweighted

Size and value exposure. What is "best" depends on all inclusive fees. Most "smart beta" funds are several multiples more expensive than market-cap weighting so the small edge will get wiped out. I favor broad funds and leaving factor concentration to the VP where you can actually do a proper job.

I agree smart beta fees will wipe out any alleged extra returns. That said, the large-cap Schwab alternative weighted ETF FNDX is 0.32% expense ratio, which is not horrible.

ochotona wrote:I agree smart beta fees will wipe out any alleged extra returns. That said, the large-cap Schwab alternative weighted ETF FNDX is 0.32% expense ratio, which is not horrible.

Thats lower than average. How has the performance been vs S&P500?

MachineGhost wrote:

ochotona wrote:I agree smart beta fees will wipe out any alleged extra returns. That said, the large-cap Schwab alternative weighted ETF FNDX is 0.32% expense ratio, which is not horrible.

Thats lower than average. How has the performance been vs S&P500?

WORSE. Since January 2014, 8.42% CAGR for VOO, vs. 6.84% for FNDX. But that's not a long period of time at all. Do you find other ETFs that beat the S&P?

ochotona wrote:

MachineGhost wrote:

ochotona wrote:I agree smart beta fees will wipe out any alleged extra returns. That said, the large-cap Schwab alternative weighted ETF FNDX is 0.32% expense ratio, which is not horrible.

Thats lower than average. How has the performance been vs S&P500?

WORSE. Since January 2014, 8.42% CAGR for VOO, vs. 6.84% for FNDX. But that's not a long period of time at all. Do you find other ETFs that beat the S&P?

RSP is one. I'm sure there are others but I don't have much interest in ETF's without any downside risk management so I don't really pay that much attention to "buy and hold".

RSP Guggenheim Equal Weight sounds good to me. I think I can get it commission-free.

FYI, here is the document which governs the Research Affiliates / FTSE Russell Funamental Indices:

http://www.ftse.com/products/downloads/ ... es.pdf?969

Page 15, is where it all starts:

"4.4 The review of the FTSE RAFI Index Series and the calculation of the fundamental weighting of a company are carried out using the following factors as found in a company’s annual returns:

A. Sales = company sales averaged over the prior five years
B. Cash Flow = company cash flow averaged over the prior five years
C. Book Value = company book value at the review date
D. Dividends = total dividend distributions averaged over the last five years"

ochotona wrote:A. Sales = company sales averaged over the prior five years
B. Cash Flow = company cash flow averaged over the prior five years
C. Book Value = company book value at the review date
D. Dividends = total dividend distributions averaged over the last five years"

So its just growth and value tilting. You have to decide if this is worth the extra fees over the other alternatives. Generally, factor composites work better than a single factor. So that's how I would view this. Do they use the best growth or value factors? Not from the looks of it.

http://m.wealthmanagement.com/etfs/when ... akes-sense

One of the better comparisons I've seen. It is fairly easy to get most of the best of both with simple Mo between the two or if you are hardcore agnostic, just buy 50% of each and you get the average. Simple averaging is a great tool if you can't decide between two options. You won't get the best but you won't get the worst either.

Kbg wrote:http://m.wealthmanagement.com/etfs/when ... akes-sense

One of the better comparisons I've seen. It is fairly easy to get most of the best of both with simple Mo between the two or if you are hardcore agnostic, just buy 50% of each and you get the average. Simple averaging is a great tool if you can't decide between two options. You won't get the best but you won't get the worst either.

That was the impetus for the Equal Weight Market Cap approach I came up with last year. But what bothered me about that was the risk was not equalized among them because small and micro caps are definitely way more riskier than mega caps. Stupid me didn't realize until just this very minute (that I can recall) that the way to go is to risk paritize the market caps. Doh!

They didn't prove it but maybe it is the sector overweighting that matters, not the equal weighting.