A Quant Approach to TAA: Winning by Not Losing

I published a paper last year to focus on the topic in the title of this blog post - "A Quant Approach to TAA". It took inspiration from the Policy Portfolios of the Harvard and Yale endowments to come up with a simple strategic asset allocation - 20% each in US Stocks, Foreign Stocks, US Government Bonds, REITs, and Commodities.

This portfolio can be easily implemented with low-cost ETFs or mutual funds, and rebalanced every so often. It then applied a simple tactical overlay to reduce risk and drawdowns - i.e. winning by not losing. This technique would have had you out of US Stocks the end of 2007, out of Foreign Stocks the end of January, and out of REITs way back in June of last year. And of course you would still be happily in 20% bonds, 20% commodities, and 60% cash.

I have received a handful of emails asking to post the monthly return series online, and below is a PDF you will have to click on to enlarge and view. As you can see, the returns in 2006 (when I first circulated the paper) to 2008 out-of-sample are quite representative of the previous 30+ years. 36 profitable years and counting - including this year (so far), a pretty difficult year for those who are long stocks. (All figures below are gross of any management fees, commissions, taxes, or bid/ask spreads. The original paper also understated the cash returns a bit and is corrected here.)

[click to enlarge]

Comments

[SmartETF:]

I have no idea how a 20% mix allocation per asset class can be constured as Quantitative but the results are compelling. I will read the link to the whitepaper.

2008 Mar 11 11:45 AM

[granger:]

Great info as always. Thank you. I will be downloading the paper. Thanks again.
(I will probably have to read it a few times, but I kno willbe worth it)

2008 Mar 11 05:59 PM

[Indexor:]

Thank you Mebane, I especially enjoyed your paper and its references. I have used a very similar approach for a number of years, first with no-load mutual funds and now with ETFs, although not equally-weighted. While it has eliminated tremendous amounts of risk from individual client portfolios, while providing market rates of return, most do not appreciate the value. Only during the 2000-2002 Bear market did clients appreciate being in cash instead of stocks.

2008 Mar 11 06:13 PM

[etf_newbie:]

Mebane - I found this article quite useful. Its sensible and simple for a novice investor like me looking for simple strategy to adopt.

I was considering looking at this in more detail but had a question -

In your article you mention the 10-month SMA. Most chart sites like Yahoo or stockcharts.com list the 200-day SMA, but i gather that is not exactly the same as a 10-month SMA. Right?

2008 Mar 16 08:16 PM

[etf_newbie:]

If you know of any site that lists the 10-month SMA, Pls let me know.

Thanks for sharing this article.

2008 Mar 16 08:17 PM

[etf_newbie:]

Yahoo finance for e.g. provides the 20-day and 200-day SMA. But not monthly averages. I suppose the results could vary using the this. I wonder by how much.

2008 Mar 16 08:33 PM

[tpoto:]

Nice results. Now that the S&P 500 has inverse -2x and +2x ETFs,
it would be interesting to see how this strategy would have done employing either of those two based the 200 days signal.

Or a more conservative approach might be to go to 50% cash
and 50% -1x inverse ETF when price is less than MA for that
amount that was allocated to S&P (vs. going to all cash)

What ETFs would you suggest for the other asset classes?
(e.g., DBC for commodities?)

2008 Mar 19 04:50 PM