Precision Allocations: Use a Matrix to Cut Guesswork

In a previous post, we shared with you how building a diversified portfolio can actually result in improved compounded annual growth rate, a discovery made at INVRS.

But how do you actually build that diversified portfolio?

We're going to share a tool, again, developed at INVRS, that can help you plan and execute the creation of your goal achieving portfolio.

A portfolio is like a fractal, in terms of diversification decisions.  You make a high level asset allocation decision, for example, between stocks, bonds and alternatives.  And then within each of these asset classes you make more, in terms of jurisdiction, industry, market cap, duration...it goes on and on.

A matrix can help you organize a lot of decisions including:

• what to buy
• when to buy
• how much to buy

I'm going to use a matrix for bond allocation decisions, but it can be made on any asset class and at different granularities.

This matrix works best when you have two criteria for diversification, all though there is a simple trick to allocate across more.  I'll cover that at the end.

Step One: Determine Allocation Categories and Amounts

In the bond example, assume you want to allocate for credit quality and duration.  Further assume you are interested in short, medium and long term bonds.  With respect to credit quality you are interested in government (GT), investment grade (IG) and high yield bonds (HY).  Remember, these can be anything.  You could make a massive matrix with 30 different years and six different S&P grades.  

The next decision is the allocation with each sub category.  This person wants to hold roughly the same number of short and long term bonds and have double the number of medium term ones.

Within the credit category, they want to hold roughly the same number of government and investment grade bonds and half the amount of high yield.

Step 2: Multiple the Matrix

Hopefully that picture is self-evident.  I think it'll be tedious to try to describe the above in words. 

Step 3: Translate It Into Dollar Amount Purchases

This depends on how large your purchases are and/or how large you expect your portfolio to be.

Let's say that every month you invest $2,000 into your bond portfolio.  Over the course of 10 months, you'd have invested $20,000 in a bond portfolio that was fully representative of your asset allocation objectives.

This is how your purchases might look every month:

• January: spend $2,000 on short-term government bonds
• February: spend $2,000 on short-term investment grade bonds
• March: spend $1,000 on short-term and $1,000 on long-term high yield bonds.
• April and May: spend $2,000 on medium-term government bonds
• June and July: spend $2,000 on medium-term investment grade bonds
• August: spend $2,000 on medium-term high yield bonds
• September: spend $2,000 on long-term government bonds
• October: spend $2,000 on long-term investment grade bonds

And then simply repeat the cadence.  As you can see, this is a very elegant way to plan purchases.

Other points: If you don't want to hold a certain sub-class, you don't have to.  You don't, for example, have to hold long-term high yield bonds.  Just X-out the category.

What if you have more than two criteria that you are diversifying against?

If you have more than two categories, consider which one is the highest.  For example, say you also want to diversify geographically.  That might be a higher level decision than say credit and duration in a bond allocation or market cap and industry in a stock allocation.  You make your high level diversification decision for example, 60% developed nations of which 30% is your home country and 40% emerging economies.  Then, for each high level criteria you apply the matrix.  The secondary level mix does not have to be identical.

Here's a quick example, this time we'll use stocks diversified across sector and market cap.  For developed nations, you may wish to include all sectors and four market cap sizes from mega, large, medium and small.  However in developing nations, you might include all sectors, but only mega caps.