Monday, March 30, 2009

A Few Notes on The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets

In their 2009 book, The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets, Mebane Faber and Eric Richardson "profile the top endowments and then examine how an investor can hope to replicate their returns while avoiding bear markets. The focus [is] on practical applications that an investor can implement immediately to take control of their investment portfolio." Mebane Faber "is the portfolio manager at Cambria Investment Management where he manages equity and global tactical asset allocation portfolios" and a co-founder of AlphaClone, an investing research web site. Eric Richardson is Chairman and founder of Cambria Investment Management. The book has a complementary web site that links to source materials. The principal messages of the book are:

Chapters 1-3: The top-performing university endowments (Yale and Harvard) are worthy of emulation by individual investors. These endowments have excelled through active management of portfolios diversified per Modern Portfolio Theory across a wide variety of asset classes (U.S. stocks, foreign stocks, bonds, cash, real assets, private equity and hedge funds), with disciplined rebalancing toward "Policy Portfolio" weightings that tilt toward equity-like assets. Active management means selecting the best assets within class either directly or indirectly by selecting the best asset managers.

Chapter 4-6: Individual investors can on their own easily emulate parts of the endowment portfolios with a simple Policy Portfolio comprised of low-cost exchange traded funds (ETF) designed to mimic the returns of relevant asset classes, rebalanced annually. This easy emulation does not include: (1) private equity and hedge funds; and, (2) active selection of specific assets within class. Disciplined rebalancing toward Policy Portfolio weightings is important to performance.

Chapter 7 -8: Individual investors can easily enhance this simple endowment emulation, as follows:

Exploit the return momentum effect for each ETF asset class proxy in the Policy Portfolio by allocated policy-dictated funds to the ETF (to cash) when the ETF is above (below) its simple 200-day moving average.

Amplify the momentum effect by rotating all funds monthly to the one, two or three ETF asset class proxies with the highest returns over the past three, six and 12 months (average of the three).

Extend the diversification of the Policy Portfolio by further segmenting asset classes as enabled by available ETFs, allow shorting of asset classes (per the momentum rule) and/or exclude bonds.

If borrowing costs are low, apply leverage.

Actively select specific assets to populate asset classes based on the holdings of top fund managers (such as Bershire Hathaway, Greenlight Capital and Blue Ridge Capital) as reported in quarterly 13F reports to the Securities and Exchange Commission.

The following chart, constructed from data in Table 7.10 in the book, compares the annualized returns for the Harvard and Yale endowments and for three enhancements of the simple endowment emulation over the period June 1985 through June 2008. The variability ranges for each are one standard deviation of annual returns. The chart also shows the worst-year return. Results indicate that the emulation models are competitive with the endowments. However, results for the three enhanced emulation models are "gross returns, so management fees, taxes and commissions would eat into returns a bit."

Note that, over large parts of the period of this analysis, simple and cheap asset class proxies (such as ETFs) were not available to investors. Also, trading frictions in "the old days" (pre-decimalization, pre-discount broker, olden days mutual funds) were higher than now. Costs of constructing and maintaining the three enhanced emulation models during these subperiods may have eaten into emulation returns more than "a bit." Small accounts may not have been practically able to achieve the desired level of diversification.

Two other considerations may affect future performance of the enhanced emulation models:

With asset class diversification, rebalancing and timing greatly simplified through proliferation of ETFs, the adaptive marketplace might permanently degrade the models by disrupting old long-run correlations and timing premiums. (The extreme stress test would be the "if everyone does this" case.)

Much of the logic in the book (like Modern Portfolio Theory) assumes the normality, or at least statistical tractability, of financial market returns via mean, standard deviation and Sharpe ratio metrics. Actual return distributions may in fact be too wild to rely on these metrics, with the stress tests considered in the book (such as "Worst Year" during June 1985-June 2008) not "wild" enough to realize a breakdown.

In summary, The Ivy Portfolio offers investors a well-reasoned, well-documented, easily understood and easily implemented approach to long-term, self-directed portfolio management based on disciplined asset class diversification enhanced by momentum. While investors should expect to underperform the modeled level of returns, the approach has considerable support from formal research.

For reviews of a few other books and information web sites (as well as tests of some trading strategies), see Blog Synthesis: Reviews of Books and Web Sites. See especially our blog entries of 10/23/08, 2/27/07 and 2/15/07 for related approaches.

No comments:

Lunch is for wimps

Lunch is for wimps
It's not a question of enough, pal. It's a zero sum game, somebody wins, somebody loses. Money itself isn't lost or made, it's simply transferred from one perception to another.