This is a very important question -- one with which I am struggling greatly.
When I realized I was not reballancing my retirements accounts on a regular basis - actually NEVER reballancing them. I decided to put my retirement savings into Life Cycle accounts that automatically reballanced and adjusted asset class allocations based on my age.
As most of my retirements accounts are with Schwab, the decision was fairly simple -- I put them into SWERX, Schwab's 2040 targetted life cycle account. YTD returns have exceeded it's category average by 1.03% and it's index by 0.45%. At the moment I'm fairly pleased with it. However, the underlying funds are more likely to be actively managed than passive, which means overhead and expenses are diminishing my returns. It would be interesting to compare SWERX with Vanguard's life cycle products - over time Vanguard's lower expense rates (I believe they have mostly index products in their lifecycle funds) should outperform Schwab's. Of course if I added up all of the expenses I'm paying for Schwab's lifecycle funds, 12b-1 fees, commissions the funds pay and other unreported expenses (market impact), moving my investments to DFA might make more sense - even if I were to include the mandatory advisory fees. However, that is another discussion.
The key topic here is this:
Should I manage my investments as two separate portfolios with two separate strategies, each with their own separate asset allocation strategy?
Create an asset allocation strategy incorporating my retirement accounts (using schwab's asset allocations for SWERX) AND my custom allocation strategy for my non-retirement index ETFs and Closed End Funds (where ETFs are not available).
Have two separate portfolio strategies. One for retirement - basically put it on auto pilot - and another for non-retirement accounts.
This is a very serious issues as my retirement accounts outsize my non-retirement holdings by almost 17 or 18 times (most of my non-retirement holdings are still illiquid with variable insurance policies - hopefully they will be liquidated before year's end).
If I go with option 1, I will be basically altering/supplementing the Schwab asset allocation strategy to achieve my own asset allocation model; i.e. I'll be adding various ETFs to bring my overall allocation plan closer to whatever %'s I set for each asset class. Therefore if Schwab seeks to lessen international exposure while I plan to increase it, most of my non-retirement holdings will pour into international asset categories.
Does this mean I would be second guessing Schwab's professional investment analysts - it looks like it. It also means that I may be grounding allocation decisions based on Schwab's periodic reporting of investments - therefore they may have changed their investment mix before I recieve an update. 1 point toward using two separate models.
On the other hand, my aggregate investment performance (what I get at the end of 30 years) will directly reflect all of my investments combined together. 1 point toward combined models.
However, my non-retirement accounts are very small - one may argue - too small to be meaningful or effectively rebalanced. I cannot change Schwab's asset allocation model, but I can reballance within my own non-retirement accounts, therefore, I need to have multiple asset classes within my non-retirement accounts to rebalance. The great disparity in size between my retirement accounts and non-retirement accounts may so deeply minimize certain asset classes presence in my non-retirement accounts (perhaps ones over represented in SWERX) that I will not have enough holdings in those classes to rebalance to or from. 1 point against combined models.
At the moment it makes sense to invest my money in two separate portfolios, however, ultimately I should really consider a move to DFA.
Altruist, a money manager that uses DFA charges 0.6% of assets under management. Most domestic DFA funds run expense ratios of 0.25% - 0.35% with some of the more unusual international asset classes running up to 0.9%, which is still quite low compared to the average mutual fund. Ultimately the average fund fee would be at or under 1% when incorporating the 0.6% asset management fee from the professional advisor - which is comparable to what Schwab is charging me to manage it's life cycle fund (plus any hidden fees 12b1 etc...). If I used DFA and a professional investment advisor I would have a single plan that aggregated all my retirement and non-retirement investments. The transparency of the account holdings would facilitate effective coordination between the two accounts.
One small pause - and this must be answered with more research - is that I'm unsure how DFA addresses the issue of fundamental indexing - they certainly have numerous value oriented indexes - even ones that are not normaly available. If I go with DFA am I abandoning this interesting investment approach? Does it matter? Has it even proven itself?
Now back to the current matter - it looks like I should manage my retirement and non-retirement accounts as separate portfolios each with their own asset allocation strategy. Consequently, I'll post my allocation strategy and holdings for my non-retirement accounts and incorporate it into a monthly review soon.
As I progress through William Bernstein's The Four Pillars of Investing, I'll modify my approach with lessons learned.
Are you in this situation now? Have you had to deal with this issue? If so, what decisions did you make, please share!