I know I post a lot of pictures from The Big Picture but I can’t help it! They are always stunning:
Our GTAA ETF is set to launch this Friday or the following Tuesday (should find out in the next day or two)!!
In the meantime I thought I would answer a question as it relates to the construction of portfolios that I get all the time.
Q: ”Does it make sense to simply use your 5 asset class model as published in your 2007 paper, or portfolios with more holdings as suggested in your book (10 and 20) and your upcoming ETF (50 – 100)?”
I think the two biggest decisions an investor can make are 1) how much of their wealth to put at risk and in what asset classes, and 2) whether to actively or passively manage those risks. I think these two decisions get you most of the way there, but I also think there are some dials one can tweak to improve the risk adjusted returns further.
As long as the underlying assets are not perfectly correlated then it makes sense to break out the allocation into further granularity. An example I always give is the commodity space. While we used the generic GSCI (which is about 70% energy) in the paper, it makes sense to me to split out the commodity allocation along groups such as energy, base metals, precious metals, and agriculture. Corn and wheat may be flying through the roof while copper or oil make be going sideways or declining. (This is just an example of course, as we could get into a much longer conversation about how best to tackle commodities whether long/short, momentum based, or with contago/backwardation considerations but that is not the point of this post.)
The point of the post is to demonstrate how breaking out an asset class into more granularity works empirically. In the below example we are going to create an artificial MSCI EAFE index. Here is a fact sheet on the index. We will take the top 5 countries that are in the index by market cap (Australia has since jumped into the top 5 in the past year but I went with an old fact sheet but the results should be identical):
Japan 21%
UK 21%
France 10%
Germany 8%
Switzerland 8%
As you can see in the below chart, the countries had varying performance since 1973 with France being the best performer at 11.72% and Japan being the worst at 6.8%. The red line is a simple buy and hold of the 5 countries, equal weighted and rebalanced monthly, and from the perspective of a US investor.
Then, if you apply the timing model on this “index” like in the white paper, you get a nice reduction in volatility and drawdown similar to the published paper (using the 10-month simple moving average).
However, what if instead of timing the whole index (our synthetic EAFE), you timed the individual components separately? Below is an equity curve as well as the table that details the additional bump in both return as well as reduction in risk you see here.
Now, you can envision how extrapolating this granularity to other asset classes and styles would be even more beneficial and could be one of the improvements that are possible on top of the published model. It is unrealistic to try and trade 50-100 ETFs in certain portfolios due to the complexity and trading costs (per share vs. per ticket) and this is one of the many reasons we are launching the GTAA ETF this week (or next).
If you read the paper we linked to a few weeks ago you know that institutions are just as bad at timing as you are.
Now there is an article in the WSJ that goes to show pensions are shifting their allocations away from equities into bonds that yield next to nothing. The difficult part of this equation is that most of these funds still expect a 6-8% return on their portfolio. Not sure where that is going to come from out of the magic alpha ether.
Will their timing be fortuitous? History is not on their side.
Economist link to public pension funds here.
I think this is a brilliant website.
Some of my favorite compilations so far and books that I ordered today (dammit):
Risk Management – Graciela Chichilnisky
Catastrophe: Risk and Response
Frozen Desire: Meaning of Money
Investment - Marc Faber
The Economics Of Inflation – A Study Of Currency Depreciation In Post War German
Collected Works of Jules Verne
Crises - Gary Gorton
History Of Crises Under The National Banking System (1910)
Crashes – Charles Morris
Saving the World – Nicholas Kristof
Science – Lewis Wolpert
Other:
“The decline has been painful for the Japanese, with companies and individuals like Masato having lost the equivalent of trillions of dollars in the stock market, which is now just a quarter of its value in 1989, and in real estate, where the average price of a home is the same as it was in 1983. And the future looks even bleaker, as Japan faces the world’s largest government debt — around 200 percent of gross domestic product — a shrinking population and rising rates of poverty and suicide.”
AAPL, MO, and MCD all hitting new all-time highs…
—-
I just attended another great Value Investing Congress in NYC, and here is a roundup of some links:
From Abnormal Returns:
Key takeaways from the Value Investing Congress. (ValuePlays and MarketFolly)
The short case for St. Joe (JOE) from David Einhorn. (Money Game, Dealbreaker, Market Folly)
An extended profile of activist hedge fund manager Bill Ackman. (Reuters)
and from the article Facebook for Finance:
“In practice, though, sharing has long been one of the most important ways that fund managers discover new investment ideas. Private “idea” dinners and gatherings like the twice-annual Value Investing Congress, launched by Whitney Tilson and John Schwartz, have long been a staple of the business. In a 1988 academic paper, written back when social networking meant working the cocktail party circuit rather than friending somebody on Facebook, Yale University economist Robert Shiller and then–Harvard economist John Pound found that roughly 53 percent of the institutional investors they surveyed attributed their initial interest in a stock to another investment professional. When the inquiry was limited to a small sample of stocks that had experienced rapid price increases, 75 percent of the investors traced the origins of their ideas to fellow fund managers.”
—-
Seven Deadly Innocent Frauds of Economic Policy
—-
The originator of the “razor blade” model of selling didn’t follow it. (HT: AF)
—-
I’ve used a simple betting strategy to win our office NFL pool most weeks. It simply fades the consensus picks. Online books and betting websites have been publishing this data since the early ’00s, and now it looks like there is some empirical evidence that backs up this supposition from the website Sports Insight. Now, these %ages will not help you in Vegas (you need to win roughly 55% of the time to overcome the vig) but they may help give you an edge in your office pool. Note that since your competitors are likely following the consensus, any win will likely distance you from the rest of the field as well creating an outlier that should help to separate points from the pack.
The website lets you download the data (for a $) so you can run your own quant analysis. Report back with any interesting findings!
Nice interview with TCU and Investure folks on CNBC.
—-
French Fama on moving averages: “An ancient tale with no empirical support.”
This really surprised me. I have read dozens of academic papers in support of trendfollowing (including my own) as well as a ton of unpublished literature. It is a strange quote from Fama, who finds that momentum is the most predictive of his four factors and is on paper saying that he is open to analysis “if it is in the data”.
MarketSci has been doing a ton of great work here in the past few weeks.
—-
A great post by Holt over on AllAboutAlpha regarding hedge fund transparency. Hedge fund paranoia over disclosure is unfounded.
The Impact of Mandatory Hedge Fund Disclosure
Abstract:
In this paper, we examine the use of hedge funds’ 13(f) filings by market participants. While many argue disclosure could harm investment funds, we find hedge funds largely benefit from disclosure while providing little private information to the marketplace. We detect abnormal trading volume around disclosure dates and also find significant, positive abnormal returns immediately after disclosure, suggesting the presence of copy-cat traders. We also find some hedge fund companies have significant volume changes on their positions prior to their disclosures. A long-short portfolio of these companies’ expanded-contracted positions purchased prior to the disclosure date earns positive, significant abnormal returns through the disclosure period. Finally, we find no evidence disclosed holdings offer long-term investors access to profitable information.
A lot of goofy quotes in here but I like this one:
“No strategy is so good that it can’t have a bad year or more,” Asness says. “You’ve got to guess at worst cases: No model will tell you that. My rule of thumb is double the worst that you have ever seen.”
Paper from last year that I am just seeing:
Absence of Value – Stewart et al
I will be in NYC next week – drop me a line if you want to meetup!
