Why investors should diversify sources of returns
(MoneyWatch) Historically, investors have tended to focus on earning the risk premiums offered by traditional asset classes. With stocks, it has been the equity risk premium; with bonds, the term and default (credit) premiums.
But over the past 20 years, an abundance of literature has shown that there are various other premiums that may be captured in financial markets.
For example:
© 2012 CBS Interactive Inc.. All Rights Reserved. But over the past 20 years, an abundance of literature has shown that there are various other premiums that may be captured in financial markets.
For example:
- Size -- small-cap stocks have provided higher returns than large-cap stocks
- Value -- value stocks have provided higher returns than growth stocks
- Momentum -- stocks that have had higher recent returns tend to continue to outperform over the near-term, before reverting to the mean in the long-term
In addition, some evidence suggests that there is a low-volatility premium. In the literature, the existence of these premiums is widely acknowledged, with the debate no longer focusing on their existence, but only on why they exist at all. The author of the paper, "Strategic Allocation to Premiums in the Equity Market," using long-only portfolios (the kind most investors use), examined the historical evidence and concluded that there are significant benefits to diversifying across the various sources of return, even if the future premiums are much smaller than they have been historically. The reasons are that there are high tracking errors between the various premiums, and the correlations of the various premiums are either low or even negative (as is the case with the value and momentum premiums).
For example, based on very conservative forward-looking premiums (well below historical levels), a portfolio optimization aimed at maximizing the Sharpe ratio produced high allocations to the various premiums, as did a simple equal-weighted 1/N (number of premium) portfolio. These results also confirmed that significant diversification benefits can be obtained by combining different premiums into one portfolio, as the tracking error of the multi-premium portfolios are significantly lower than the levels we observed before for the various premium portfolios in isolation.
The author concluded: "Investors can substantially improve the risk/return characteristics of their strategic asset allocation by considering not only the classic equity premium, but also other premiums present in the equity market. Moreover, even when one expects premiums to be smaller in the future than in the past, optimal allocations remain sizable."
Popular on MoneyWatch
- Reverse cell phone lookup service is free and simple
- Why geniuses don't have jobs
- Microsoft slashes Surface prices to lure buyers
- Look who doesn't deserve financial aid at NYU
- Have you mastered the art of listening?
- Chrysler expected to make Jeep recall refusal official
- Top 10 professional life coaching myths
- 6 things that feel productive, but aren't
Can you have too much of a good thing?
When you tilt toward SCV are you looking for the smallest and deepest value funds/ETFs you can find? For example, something like Guggenheim S&P Smallcap 600 Pure Value (RZV) or Bridgeway Omni Small-Cap Value (BOSVX). These are much smaller and have lower price to book to say a more traditional pick such as Vanguard Small Cap Value. Have studies looked at the relative merits of going further and further out on the small and value spectrum?
Yes there is good research here. The value premium is not linear, it is larger in the smallest corner of the market. Of course that means that those stocks are also higher risk, meaning risk doesn't increase linearly either.
Best wishes
Larry
Larry
Your portfolio is low beta, but do you recommend low beta/volatility asset class selections instead of, or in addition to, small and value. There are some new low beta ETFs in some of the asset classes you mentioned. For example emerging markets low volatility vs. say EMV or EMS.
Your portfolio is low beta, but do you recommend low beta/volatility asset class selections instead of, or in addition to, small and value. There are some new low beta ETFs in some of the asset classes you mentioned. For example emerging markets low volatility vs. say EMV or EMS.
Best wishes
Larry
Yes tracking error can be big problem.So the question is are you willing to accept not doing as well as everyone else when things are going well in return for not looking anywhere near as bad when things are going poorly? If you ask that question almost all will say YES. Then what you must do is get them to not confuse strategy (which is right or wrong before the fact) and outcomes. No one complains when we buy life insurance and don't die even though they know they transferred large premiums to the insurer.
Best wishes
Larry
We are very careful with "new" ideas, spending lots of time on researching the ideas and also the implementation costs--strategies that appear to work on paper often don't work in real world with costs. In the case of low volatility this may be true. I have written a post on that which should appear shortly. But the short answer is no.
On Momentum, we have spent a lot of time on the issue. And while there appears to be some benefit from adding a separate long only momentum allocation, the benefit is very small if you already have a very well diversified portfolio. Remember that because you are "long only" most of the return is equity risk, so don't see much dampening of the volatility of the portfolio, and it is not tax efficient strategy (though as bad as many think because it throws off lots of ST losses). Having said that we believe incorporating momentum screens into existing funds makes a significant difference, and thus a big value added. DFA and Bridgeway funds do take momentum into account.
The factors are unique, or independent. In the case of small and beta though it is about 0.4. Very low to basically zero for value to beta and value to small. Momentum is negatively correlated with value, so that makes it particularly attractive for value tilted portfolios. Hence we like that DFA and Bridgeway use momentum screens.
Using TSM is the most efficient way for use as a core portfolio for investors that don't have access to DFA's core funds. But I don't personally use TSM or the core's as my portfolio is the extreme tilt, only US SV, ISV and EMV, allowing me to hold the least amount of beta risk, and thereby cutting tail risk, while accepting more tracking error risk (a big psychological problem for many).
As to how much weight, that would depend on issues like how convinced we are the premiums will persist (behavioral or risk based), how volatile they are, and how the correlations work. In other words, we would look at how the assets mix with the rest of the portfolio and then tailor the decision to each individual client's unique situation.
Best wishes
Larry
I personally use the "Larry" approach with my own money as well. You are correct with the psychological problem. When I explain what this portfolio would have returned in the late 90's vs. the S&P 500 (as an example) most people shutter. I can't imagine how they would act when everyone else around them is getting "rich" and they are not.
Have you started incorporating low volatility yet?
Also I'm curious how much weight you give to each factor going forward. I know you can optimize looking backwards, but looking forward do you equal weight or something else?
Fwiw Everyone's different, but I've been generally using 40%TSM 40%SCV 20%Momentum (in tax deferred accounts).
Using TSM as the the anchor hedges that size, value and momentum can be negative for extended periods. As more factors become discovered and products made to efficiently capture them can TSM be eliminated since the factors are generally uncorrelated?
Of course you'll have to add more FI (TIPS) to balance total risk.
Thanks