Weighted and Unweighted Returns: What is the difference between the weighted and unweighted returns in the performance table, and why are they so different this year? The accounts I manage vary in size by a considerable amount (I choose not to disclose the specific values). The weighted calculations give more influence to the larger accounts. For example, a $500,000 account would have ten times the effect on the weighted return than would a $50,000 one, and the actual spread of my managed accounts is much greater than ten to one. In the unweighted calculations, an ordinary or simple average is computed, so each account has the same effect on the overall average return regardless of its size.
Prior to this year, the two averages were fairly close, but that has not been the case so far in 1997. The main reason is that the larger accounts, most of which have been with me for less than a year, tend to be more aggressively managed, which is the account holder's choice. More aggressive management means that the account is more likely to hold only one Select fund at a time, which usually produces a higher return, but with higher risk. The current trade in Energy Service has done much better than the funds that have also been held in diversified accounts, which has increased the gap. Moreover, some of the larger accounts use margin (see below), which has also tended to increase returns, but with additional risk.
The natural question is which computation best represents the actual experience of my managed account clients. The answer is that it depends on the particular client and no single account is likely to have returns that exactly match either of the averages. Those with aggressively managed accounts that concentrate in a single fund and possibly use margin should have returns closer to the weighted averages. Those with accounts holding two or three Select funds at a time are more likely to have returns closer to the unweighted average. I will supply returns of representative accounts upon request.
Effects of Timing and Margin: What are the effects so far of the new timing and margin alternatives? Earlier this year, I began to offer alternatives to the basic management program that are designed to alter the reward vs. risk characteristics of my managed accounts. It will probably be a year or two before a definitive assessment is possible, but we can take a look at how these alternatives have worked so far.
The first use of a timing filter to decide when to enter and exit the market began this past March. There was a sell signal in March, which had the effect of delaying the investment of new funds, and a buy signal that I acted upon in early May. The net effect was almost negligible with respect to the return, but the timing filter did prevent larger intermediate drawdowns. The timing model avoided some of the drop in the market, but by the time it issued a buy signal, those in the market had recovered the drawdown. This is what I expect the timing to do: avoid some drawdowns, but probably reduce overall return in the process. The brief real-time experience was better since there was no reduction in return.
The effects of margin use are harder to evaluate. Some accounts were willing to use margin at the beginning of the year, but the model that signals when to do so remained negative until early May. (That model is much more selective than the one for deciding whether to be in or out of the market.) The margin model turned negative in late June and positive again in early September. Accounts using margin have benefited so far by doing so, but the model missed the large run-up in Energy Service in July and August. The negative signal in late June turned out to be right in one respect-the broad market was little changed over the next two months, but wrong in another respect since the Energy Service purchases in July did not buy additional shares using margin.
The overall effects on investment returns of using margin are fairly complex. New shares purchased on margin may have to pay loads of up to 3% depending on the size of the account, which is an immediate negative factor. Also, interest must be paid on the margin debt, which is another detriment to returns. In order to preserve the load paid and the margin borrowing power, the margin debt is not paid off when coming off margin, and the Select money market fund is owned instead. This can also reduce the return since the total interest on the money fund may be less than the interest paid on the margin debt. On the other hand, owning additional shares of a rapidly appreciating mutual fund can easily overcome these negatives. I have not yet done a thorough accounting of the effects of these factors. However, due to the strong markets, accounts using margin have done better so far this year than comparable accounts that do not use margin. Keep in mind that retirement accounts are not eligible for margin and that use of margin increases the risk.