1. Correlation, expected returns and standard deviation are the key attributes of an asset class; not its name, market cap or how other people have decided to treat it.

2. The percentage of the market represented by an asset class really doesn’t matter. So US large cap is 40% of the global equity market. Who cares? What bearing does that have on your risk and return? Zero. Starting your allocation with market cap and moving from there is no better than throwing darts to determine a starting allocation.

3. Start with your fixed-income vs. equity (or equity-like) decision. That may be a slightly artificial or fuzzy division, but not harmfully so. This breaks your decision down into slightly more constrained chunks.

4. Next make your domestic vs. international decision for both sides of your portfolio. Again, this constrains your future decisions a little bit, making them simpler. Start with a 50/50 assumption and move it only if you can identify a positive reason to do so. (There may be, such as currency risk concerns. But make yourself identify them before you budge. Currency risk has a diversification benefit, so it comes down to your personal situation and goals.) Why start at 50/50? If you have two assets that have identical returns but aren’t perfectly correlated, and that is all that you know, then the mathematical answer is that 50/50 is the best decision. It will reduce the combined standard deviation by the greatest amount. So start there and make yourself find reasons to move.

5. Since you have to start somewhere on actual asset classes, start with the assumption that within each side of your portfolio (fixed income / equity) you will allocate equal portions to whatever asset classes you choose. Same reasons as above. Then depart from that only when you have a positive reason to do so.

6. Get a feel for the total number of asset classes you feel good about using. It’s commonly said that more than 12 provides diminishing returns. I would think that anyone as interested as you would want at least 4, and I accept the standard advice and cap myself at 12. Let’s pick 10 for our example – it’s obviously flexible, but it gives you a place to start. Remember that your portfolio size can limit you here. If you only have $10k, start with 3 or 4, be patient, and complicate it later when you have $25k or more.

For example, let’s say you chose 70% equity and 30% fixed-income. Then you chose 60/40 domestic/international for the equity side (perhaps because you think that doing poorly while the US is doing well would pressure you to abandon a 50/50 allocation) and for simplicity’s sake you are happy with 100% domestic for fixed income. Now set that aside and examine the available asset classes for correlation, expected returns and standard deviation. Of course you are trying to maximize three variables here, so there is a whole set of right answers. There is no “best”. Make sure that your average expected return is enough to meet your goals, that your combined standard deviation is within your risk tolerance, and that correlation is low enough to take a good stab at lowering combined standard deviation in practice. Don’t allow the low correlation focus to get lost in the shuffle, here. That’s the free lunch. Don’t miss the free lunch!

So now you’re down to questions like: “30% of my portfolio will be international equity. I don’t want to exceed 10 total asset classes and I want to diversify each portion of my portfolio, so I have 3 asset classes to choose in this part. What three asset classes should make up my international equity component?” That finally starts to look like an answerable question!

There are still choices to be made, but you can review what you’ve read about each option and pick three that fit your overall goal. Remember, there are multiple right answers. How about 10% Europe Index, 10% Pacific Index, and 10% Small Value Index? If you would like to add an Emerging Markets Index, not to skew risk/return too far but for the low correlation, that might be a good reason to break the equal allocations: reduce Small Value to 5% and allocate 5% to Emerging Markets. Keep your eyes on the reasons. If you *did* want to increase returns, then the right answer might have been to combine Europe/Pacific into a World Index at 10% and to bring Emerging Markets in at a full 10% itself.

Rinse and repeat for the other components of your portfolio. If you feel lost within a component, read more about the asset classes that are available to you until you start seeing factors that make sense to base a decision on.

Let us take the above (Bernstein) : 25% in each of 4 funds. ? You split the US domestic equity equally among LargeCap and SmallCap. Why ? Many well known people say that you should go with the Total Stock Market (TSM) as the core because that is basically the market or else you will get “tracking error”.

But then again TSM has only 10% small cap, so one wonders if they wont get diversification benefit. But then thats the market.

Is Bernstein doing what is known as “small value tilt” where people load up on small-cap/value stocks with the belief(hope) that it will outperform other asset classes over very long periods as it has done in the past ?

is this voodoo ? Basically every advisor comes up with his own way of slicing and dicing the market. People say : “Do it as per your taste. Take TSM and add Small Value to taste”. What is “taste”. It is not like Salt/Pepper. It all seems very random to me.

One answer could be : It doesnt matter. Then why provide so many funds ? Just to confuse people ? I am literally losing sleep over this because I dont like confusion, and dont want to change my AA every year (which is sure to give poor returns). I want something and stock to it, but I just cant make up my mind.