Jan

22

The RSP (equal-weighted) S&P index ETF is well-known. Less known is the RYE (equal-weighted energy sector ETF). It has only existed since about 2006.

Equal-weighted ETF's give a larger weighting to smaller-capitalization stocks and, to the extent that individual stocks approach zero, they engage in the Rocky pastime of "scaling down to oblivion". That is, If cap weighted indices "ride the trend," equal-weighted indices sell the winners and add to the losers on each rebalancing.

Might anyone have some insights about whether such a practice is inherently superior or inferior over time? And especially for a (distressed) sector index?

Kora Reddy writes: 

But the academic literature suggest otherwise: "equal-weighting is a contrarian strategy that exploits the "reversal" in stock prices" (see this pic).

Except in Australia, equi-weighted outperformed the cap-weighted in major countries.

Gordon Haave writes:

I wrote about this 6-7 years ago when the first Wisdom Tree stuff came out and they were talking about how equal weighted was superior to cap weighted and showed the back-tested numbers. All they were really saying is that "over time small caps beat large caps" which isn't exactly news.

To call a equal weighted index and "index" is itself misleading. A cap weighted index is "the market" or some approximation thereof. Theoretically every single market player could go passive and be in it. You can't do that with an equal weighted index (or at least not without distorting prices).

As to your idea of how they have to double down on the loses that is somewhat limited by the fact that once the name falls out of the index it is dropped.

Larry Williams comments: 

Along that line Our work shows it is better to invest equal dollar amounts vs equal share amounts

Gibbons Burke adds: 

I know a fund which used to invest 90% of client stake in SPX via SPY.  A couple of years ago they switched to 10% equal dollar investment in each of the nine sector select spdr ETFs, with the intent of rebalancing to equal dollar allocation annually. They found, in testing, the strategy provided an average of 200 bps of boost each year over the cap-weighted all-SPY investment.

anonymous writes: 

Regarding a depressed sector, is there any truth to the adage: "Buy the stock that has gone down the least, and also the one that has gone down the most". The strong stock will come back smartly and the oversold weak stock will come up from being smashed on a higher percentage then the middle of the pack.

So if this is true you could design your own basket of strong stock leaders in the depressed sector mixed with oversold beaten down stocks that pass a screening survival test.

Erich Eppelbaum adds: 

Theoretically speaking, re-balancing a portfolio by using the winnings to buy more of the losers is at the heart of the only portfolio selection methodology that I know of that mathematically guarantees to asymptotically outperform the best stock included in the portfolio (See Thomas Cover's Universal Portfolio seminal 1991 paper): pdf link.

I don't know if in real life the portfolios resulting from this methodology are inferior or superior over time to those created by rebalancing based on allocating more to the winners (such as a market cap weighted portfolio); I would assume that any result would depend heavily on the rebalancing costs and slippage (the liquidity of big vs small stocks matter, especially when trying to push size), and I would assume that the slippage incurred in a market cap weighted portfolio would be less than that incurred in a equal weight portfolio (less small company shares to buy/sell).

In reference to a previous post, another thing to consider is that perhaps there are many effects at play other than the small-cap "more-risk-more-reward" effect. For example, a sell-the-winners-buy-the-losers methodology could be profiting partly by say the volatility harvesting effect described by Claude Shannon.

This brings up another question: The volatility harvesting effect becomes greater as the volatility of the portfolio's underlying stocks increases. In the stock market, volatility usually increases when the market falls. Could this mean that an equal weighted/rebalanced portfolio would outperform a market cap weighted portfolio during bad times? and would the opposite be true during good times? Would be interesting to test… 


Comments

WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '9963' AND comment_approved = '1' ORDER BY comment_date

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search