Long time reader Russell120 sent me a copy of John Mauldin's newsletter entitled CAPM is Crap which is an attempt to turn the ideas of Alpha and Beta upside down.
Among other things was a chart that showed taking on higher beta does not result in higher returns. The article also has some positive things to say about fundamental indexing.
Quite frankly I felt as though I was ready things similar to what I have been writing about since I started this site (not claiming anyone stole my work just a sort of like mindedness). I think I can address some of the points made fairly concisely.
Higher beta does not result in better returns. Well there is one problem with this conclusion. A high beta stock is likely to have its day in the sun but eventually it will start to rain. Pick any internet stock from the bubble. Chances are it gave ten years', or more, worth of appreciation from 1998-1999. The person who sold his net stock on Dec 31, 1999 and never went back had a different experience than someone who held on to internet stocks too long. This is true for every stock market fad. Success with the next fad will not come from holding forever; it will come from holding for a couple of years, give or take.
When I write about beta I use the word volatility, not risk. Increasing risk is really not something people want to do, they may want to increase volatility to capture a general rise in the market but there is never a good time to absorb an 80% hit to one of your stocks.
The primary objective of a portfolio is that there is enough money to meet whatever the goal is. Time horizon and how much you have will determine what the portfolio has to do to meet the goal. A 35 year old with $100,000 is probably off to a good start but that person needs to save a lot and needs a normal exposure to the stock market. A 45 year old with $5 million does not need to save as much nor does he need as much stock market exposure.
Let me define exposure here with an example. Take two $100,000 portfolios; one with a beta of 0.80 and one with a beta of 1.20. The latter has more exposure.
I have many posts up about low octane portfolios. This is relevant here. People that save properly probably do not need to go for heroic returns. Getting market equaling returns, or even returns that lag slightly, with much less volatility is an excellent outcome. I would consider anyone that can get away with taking this approach as being lucky.
It seems that fundamental indexing could be a path to this concept as perhaps an anti-alpha. I believe in it in moderation but I think of it differently than most of the articles on the subject. I have disclosed many times owning WisdomTree Energy (DKA) for most clients. I did not buy it because I wanted a fundamental index. I wanted to sell British Pete (BP) and needed to replace it.
Based on my view at the time (this was last November) I felt that single stock risk would not be rewarded so that meant buying a sector fund. In the energy sector I favor more foreign exposure so that meant I wanted a foreign sector fund. I looked under the hood and decided DKA would be my best bet. Fundamental or not was not a consideration.
Based on market history the likelihood of future corrections, bear markets and even a crash or two are very high. The emotional and financial benefit to reducing the impact of one or more of those future events would be colossal.
This topic can't be completely covered in this sort of a post but you get the idea.
Wikinvest Wire
12 comments:
http://www.investorsinsight
.com/otb.aspx
I think this url will work.
click here for Russell's link
Hi Roger, I’m a regular reader of your site and occasional anonymous poster that has finally decided to register – the interaction with you and the other posters is really interesting. Regarding the topic at hand, there is another way of looking at ‘low octane’ portfolios. Unfortunately for the 35 yr old in your example, in an effort to achieve the returns he needs, he will most likely end up with an overconcentration in equities.
If we can agree that most beta sources carry Sharpe ratios of 0.2 – 0.3, we can also agree that various asset classes compensate us on a roughly equal basis for each unit of risk taken. Because stocks carry more units of risk, we get more compensation. Unfortunately, that means to reach our goal we often have to concentrate heavily toward one source of beta (equities).
However, through the use of leverage, it is possible to make bonds, for example, competitive with equities on an absolute return basis – in other words, we can borrow money to make an investment in bonds a higher risk/higher return proposition than they normally would be (look at a leveraged closed-end muni fund for example). Why do this? By leveraging low beta investments it is possible to construct a much more diversified portfolio that performs on par with an all equity portfolio – similar returns, less combined volatility (or same volatility, higher returns). Unfortunately, efficient sources of leverage are often beyond the ken and comfort level of retail investors (and their advisors). Which is why I hope, someday, to see ProShares or Rydex or someone start rolling out some levered ETF’s in areas outside of the major equity indices in which they are currently available.
Although I wish I could claim it, this ‘beta optimization’ concept was pioneered and has been developed over the years by Ray Dalio, founder of Bridgewater Associates – managers of about $165B of pension and other institutional money. He and some of his staff have written white papers on the concept, you might be able to find something on Bridgewater’s site (www.bwater.com). Love the site Roger, I look forward to every new entry.
Hi gobeavs, I remember Roger had a post about those Bridgewater papers you mentioned. I guess you could implement that strategy by investing 50% of your original fund on the ProShares double-long equity ETF and use the left 50% for bonds. But the problem here the history of ProShares similar funds show huge tracking records to the doulbe equity index.....
Actually, you have it backwards - you'd want the leverage in the bonds, not the equities. The point is that you are trying to engineer your investment in bonds to deliver a return similar to the same size investment in equities. Assuming the two asset classes have a relatively low correlation, this should maintain the expected return while reducing risk.
Your point regarding the tracking error is a good one. No doubt the ProShares solution isn't as elegant as using futures or options might be, but it's a step that is probably more palatable to most people. Whether or not the model still holds up with a ProShares type etf would be interesting to see.
gobeavs, I'm puzzled. To use 50% of your original funds for equity with the ProShare's double long equity funds, you actully don't leverage in the equities. You just have another 50% cash left for the bonds......
gobeavs, here's the details of the strategy about how to implement "Rish Parity" from "World Beta Blog":
http://worldbeta.blogspot.com/2006/11/risk-parity.html
You could try balancing your returns, and maybe your risk by weighting various investment types with differing amounts of leverage. But the point of the article is that doing so using the popular tools of financial investment is not problematic.
If beta is useless (as the article implies), then is not a sharpe ratio useless, and covariance, etc?
If these measurements are not a meaningful description of any future expectation then why use them? Would it not be better to find other models?
Roger - Just an FYI, the piece is actually written by James Montier. John has two letters, one that comes out Friday nights that is his own writting and a second that comes out Monday which he presents someone elses viewpoint/letter (with their permission, of course.)
Big Bill
> If these measurements are not a meaningful description of any future expectation then why use them? Would it not be better to find other models?
That's why Fama and French came up with the three factor model - beta, market cap, and book-value/price.
The Bridgewater/Dalio concept makes sense, but putting it into practice would be difficult for most individual investors. There aren't enough leveraged funds in enough asset classes.
uTomK
The French-Fama model is still reliant on Beta and regular distribution of pricing. Mandelbrot (and others) have pretty well shown that the distribution of market pricing is not described by the standard bell curve. Worse, when stock pricing goes of the curve, it generally does so in a downward direction. This makes Beta derived models dangerous.
Post a Comment