When I worked as a stockbroker, I was always interested to see portfolios that had been put together by other advisors. This was most common when working with people who were considering moving their accounts or transferred their accounts to me. In many cases, the prior advisor had given good advice and the holdings were reasonable.
There was one thing that would bother me whenever I saw it, however, and that was matching accounts. My father was my branch manager, and he used to always create matching accounts. Let me explain: If a client’s portfolio was to contain 50% Canadian equity, 20% foreign equity and 30% fixed income, then the RRSP, the open account and the spouse’s RRSP would each be allocated 50/20/30.
With mutual funds, this made some sense. The trades were free to us and to clients, so although placing trades to rebalance each account meant (in the example above) triple the number of trades, it made almost no difference to us. Further, all account statements and all software views showed only one account at a time, never consolidated across all accounts for both spouses. Because we created matching accounts, it was easy to see that the overall portfolio was always balanced by virtue of being the product of matching accounts.
With stocks, it makes less sense, especially when each trade is charged a commission. Returning to the example above, suppose the RRSP holds 50% of the assets, the open account holds 20% and the spouse’s RRSP holds 30%. Each account would hold fewer positions if the RRSP held all the Canadian equities, the open account held all the foreign equities and the spouse’s RRSP held all the fixed income. (This would have to be done with taxation and investment risk tolerance in mind.) Assuming that the Candadian equities and foreign equity portions were each composed of 10 stocks and the fixed income was composed of 5 bonds, each account would have to hold 25 positions in order to match. Consolidating positions per account would mean 10, 10 and 5 positions. When trades are made, such as when adding new funds or rebalancing, this would result in just one third (1/3) of the transactions and one third of the cost.
The drawback is that some accounts may unexpectedly outperform others and that each account may not appear to be diversified. In the end, however, the holding taken together all form the client’s (or couple’s) portfolio, so diversification and performance are really only meaningful at the portfolio level. There doesn’t seem to be any really good reason to hold the same position (stock or bond) in more than one account at a time.
The strange thing is that I haven’t seen anyone do it this way. I haven’t seen a portfolio where each account held unique positions. Am I missing something? Is there a reason this wouldn’t work? Would it not really reduce trading costs?