Comments on:

By: Kirzner Fervor

Kirzner Fervor — Mon, 12 May 2008 19:55:00 +0000

I tried a study of using 30 year, 10 year, 5 year and so on maturities instead of just the 10 year, but the 10 year has the best risk/return characteristics. However, corporate bonds should be included. They have strong returns relative to risk and not too high a correlation with treasuries or domestics.

I also looked into Emergings vs. World since 1971. I used TRWLDM and TRWLDXM from GFD. The Emerging portfolio historically outperforms the World x/US portfolio and the timing returns look very strong. One thing I noticed is that the returns increase almost monotonically as you decrease the length of the SMA (does not normally happen with other asset classes), suggesting that you wouldn't be curve-fitting by reducing the length down to 5 or 6 (.696 Sharpe vs. .51 on World @ 6% interest) months.

The ideal weight would be somewhere between 50% to 75% of the portfolios allocation to foreign stocks in emerging (6 month SMA instead of 10) and the remainder in the World index. Sharpe will be around .78 for that portfolio (16% return, 12.8% standard deviation).

By: ashpak001

ashpak001 — Sat, 10 May 2008 18:04:00 +0000

You may consider using ILF instead of EEM -it has very high correlation with EEM but is a lot more volatile.

By: John

John — Sat, 10 May 2008 16:04:00 +0000

Yes; I tested a trend following method that I use and it looks quite promising. But I have the same concern – how can we simulate earlier data?

By: Matt

Matt — Fri, 09 May 2008 23:00:00 +0000

Microcap is CRSP Decile 10 from 1972-1997 and BRSIX from 1998 to present.

Emerging Markets is VEIEX from 1995-2006, MSCI Emerging Markets Index 1988-1994, and from 1972-1987 from IFA’s Emerging Value Index with 0.9% annual fees added back in(reconstructed from data found here: http://www.ifa.com/portfolios/PortReturnCalc/index.aspx?i=EV&s=1/1/1972&e=12/1/1987&type=indices&g=1&infl=False&tax=False&wort=0&perc=False&wortinf=False&aorw=1#calc)

I have a spreadsheet with all of this data back to 1972 that I got from somebody on the Diehards website if you want it.

By: Alex

Alex — Fri, 09 May 2008 18:12:00 +0000

What about emerging market bonds?

By: Damian

Damian — Fri, 09 May 2008 17:34:00 +0000

Not really enough history on those ETFs to get a great feel for how it would perform. Based on those, I get an annual return of 2.90% (risk-adjusted return of 20.94%) with 13.85% exposure. The problem is that only the QQQQ has substantial history. All the other trades – and the system has a 57.58% winning rate, but we need more data.

By: Mebane Faber

Mebane Faber — Fri, 09 May 2008 17:28:00 +0000

Matt – where did you get the data for emerging back to 1972?

I assume the microcap is from French Fama?

By: Matt

Matt — Fri, 09 May 2008 17:27:00 +0000

Well, most of the leveraged ETF's would qualify as high risk but I don't think they would work for what you are trying to do.

My pick would be narrow sector ETF's like Biotech or Nanotech for high risk asset classes.

Also, if you run a simple backtest of your AA portfolio (not timing)(20% S&P 500, 20% EAFE, 20% GSCI, 20% Long US Bonds, 20% REIT) and swap microcap and emerging markets for the two equity slices, your CAGR goes up from 11.48 to 13.99 with Sharpe increasing from 0.62 to 0.73 (data from 1972-2007, rebalanced annually).

By: Mr. Monopoly

Mr. Monopoly — Fri, 09 May 2008 17:21:00 +0000

“Any other high risk ETFs that could be used?”

DBA, GLD?