Toward a New Model of Diversification

There are a handful of world class investors whom I will go out of my way to see, hear or read. One of these heavyweights is of course Oscar Rogers. I have not seen anything from him in a while - this link is from late October and I believe is his latest public commentary. In that clip, he gives an impassioned plea for them to fix it.

I think Oscar might be on the verge of throwing in the towel, despite this article from the NY Times saying that diversification does still work and that investors should not give up on it.

The idea is simply that in a short term, panicked flight to safety, yes everything goes down, but the proper mix did go down less. The article never mentioned any sort of defensive strategy like the things I write about. It is no shock that a mainstream media article would omit such an idea, but as mentioned in this week's video hopefully you did something along these lines in the current bear market or have learned to for the next bear market.

I still believe in diversification. The notion of how to construct a diversified portfolio is evolving, however, by necessity and thanks to new investment products (not all new products are bad, even if a lot of them are flawed). A few of my recent articles for TheStreet.com and a couple of posts on my blog have tried to seek out different ways to structure portfolios to have much less risk and reward - which of course means the need for a much higher savings rate.

The inspiration is that this decade will cause people to want to give up on the stock market. Well, anyone wanting to give up still needs to save and get some sort of return on their savings even if a "normal" 70/30 allocation is not how they will do it.

One idea I'm trying to pull together for an article is along the lines of picking six or seven asset classes (including cash), picking one or two broad based funds for each one and from more of a bottom-up tactical strategy, entirely sell out of a fund (asset class) that goes below its 200 DMA, or any other defensive concept, and go back in when it goes above. While you are out you are out - not chasing into something else. I imagine this concept would have gone into cash asset class by asset class over the last however many months and now be 100% cash, or close to it.

The asset classes might be:

• Domestic Equities
• Foreign Equities
• Commodities
• Domestic Bonds
• Foreign Bonds
• Inflation Protected
• Absolute Return
• Anything else you can think of

The reason not to use a target date fund or asset allocation fund is that most of the time diversification will work and breaking out the asset classes gives a better chance to capture the zigs and zags as they occur.

For domestic equities, for example, this strategy could use the Russell 3000 Index Fund (IWV). It went below its 200 DMA last December. So the idea is to have sold when that happened and then just wait until it goes back above. It still has not gone back above, so that part of the allocation would be in cash. Most of the different segments would have gone to cash a while ago.

This is obviously not that original on several levels, but for someone on the verge of giving up there is not that much work. Pick one or two broad based, index-like funds for each segment, set up something to alert you when each holding breaches its 200 DMA and check your email. The funds used would ideally just be a proxy, not an attempt to add value versus a proxy. IWV is the US market, something like iShares Aggregate Bond (AGG) is the bond market and so on.

One issue, and there are many, is that IWV is about 40% below its 200 DMA, which is a lot of potential gain to give up. While that is a drawback, the 200 DMA is headed lower and in another couple of months or so the slope down for it will steepen dramatically and so a crossover could occur long before a 40% rally - or not.

As a reminder this concept would be aimed at people that want to give up but know they can't, and I assume that if you take the time to read a stock market blog "people that want to give up" does not include you. So while there are flaws galore, it is better than nothing.

Comments

[sprintracer:]

Roger,
Check out Mebane Faber's published paper on his site worldbeta.blogspot.com... He suggests using a portfolio very similar to you're suggestion but based on a 10 month sma and monthly prices. Basically the same as the 200 day sma. His paper includes 5 asset classes with backtesting to 1972 and he updates the current returns on his site for some out-of-sampling data.

2008 Dec 08 08:48 AM

[squark62:]

hi roger,

what you are describing is how i feel i've been forced to learn to trade so as not to lose money. the market has confounded the best over the last 20 mons or so and that's about how long i've been seriously learning to invest/trade. infact, i've since learned that i've worked out a simple system to be market or delta neutral. by the time oct/nov '08 came around i managed to make gains essentially the way the system has developed to this point. so learning to stay basically even has done wonders for my confidence going foward.

mj

2008 Dec 16 11:12 PM

[donzelion:]

"I still believe in diversification." I believe in arithmetic - "buy low, sell high." Way I see it, diversification is just one of many strategies to achieve the real goal - a good strategy in most cases, but no strategy works in every environment, particularly not when a single factor (e.g., leverage inflating asset classes) crosses beyond every sector.

Hence, "anything else you can think of" ought to focus on what nobody will ever link to financial markets. My candidate? Human capital. Not "anything you can think of," but the skill and capacity to "think" - skills, training, credentials, and social networks. Human capital isn't readily converted to pay the rent, but it's easily the most non-correlated asset class out there.

Jan 10 03:12 AM