March 9, 1998

Q. Last summer a friend of mine bought Vanguard Specialized Health Care fund at 62, which was the high at the time; today the shares are at 78. The one sector that money is pouring into is the drug sector. Mergers among the drug stocks will continue at an accelerated pace. Not only do some of the drug stocks own medical equipment companies, but, more important, they own or have marketing agreements with biotech companies. We both agree that drugs have been and will be the sector to be in until we are able to cure man's illnesses.

P.F., Braintree

A. Yes, we do agree that the drug sector has terrific long-term prospects, and I frequently recommend these funds to add spark to a long-term portfolio.

Where we differ is when it comes to the short term. I have been advising people taking positions in health care funds to use dollar-cost averaging. Why?

Well, look at some of the numbers in the latest summary from Morningstar Mutual Funds. The average health care fund portfolio showed a price-to-earnings ratio of 35.9. The price-to-book value ratio was 7.9. Both of those numbers are the highest for any sector or class of mutual fund.

The average yield for health care funds was a meager 0.7 percent. And health care funds had a year-to-date average gain of 6.11 percent, which strikes me as a pretty quick start for a year in which 8 percent profit improvements are expected.

Moreover, while long-term results have been terrific -- the 10-year record of 20.04 percent average annual gains trails only the financial sector, which posted a 21.91 percent average -- the sector is far from invulnerable to weak markets.

You need look back no further than 1992 to 1994, when consecutive results for the sector were a 4.70 percent loss, a 3.69 percent gain, and a 4.26 percent gain, trailing the Standard & Poor's 500 Index in the first two years, and only slightly beating it in the third.

In short, the higher they fly, the further they can fall. I still like health care funds for the long term, but suspect that investors might need a strong stomach to take the shorter-term ride.

Q. My wife and I, age 51 and 56 respectively, would like to retire in seven years, and we have three questions relating to future planning. How far into the future is it reasonable to try to make projections for investment returns and inflation? What estimate of future inflation seems reasonable to factor in at this point? (We would guess 4.5 percent.) What long-term estimate of conservative portfolio growth might be reasonable for a portfolio invested 50 percent in stock mutual funds and 50 percent in bonds? (We would guess 7.5 percent.) D.S., Weston

A. Projecting investment returns can be done with increasing confidence as the time-frame increases, rather than the opposite.

For example, if a 30-year-old is considering retirement savings, you would reasonably expect a diversified stock portfolio to return in excess of 10 percent, and it might even match the 11 percent-plus long-term record of US stocks.

You can continue to apply this thinking for projections that go 20 or even 15 years out. But for shorter time frames, things get murkier.

Although a 10-year investment program is considered long-term, it's quite possible for a 10-year stock program to go pretty sour. Remember the '70s? The Dow Jones industrial average entered 1973 at 1020 and didn't return to that level for good until 1982; at the end of 1983 it stood at 1046 -- a gain of only 2.5 percent over the course of a decade.

(To be fair, these figures don't include dividends, but as I recall, dividends proved small consolation to the investors of those years.)

Such choppiness, of course, is why most retirees put so much money in fixed-income holdings: They know that stocks will do better over the long haul, but they simply can't afford to gamble on how long that haul will be.

As for inflation, I have been using 4 percent for the last couple of years. That should be an adequate estimate.

As for the results of a portfolio balanced between stocks and bonds, for a long-term projection I would go a little higher than you, estimating 10 percent from the stock side and 6 percent from the bonds; thus, an 8 percent average.

But again, the waters get murkier when you consider shorter time frames. This is why I suggest that retirees divide their portfolios into two distinct segments: a fixed-income element to produce an income stream, with all or most of the yield withdrawn regularly, and a stock portfolio, with all distributions reinvested.

Using such an allocation, the bond side likely will deliver the required cash, even if the markets sour. On the stock side, the money is designed to protect against inflation in later years or retirement.

Note: In my column of Feb. 9, I discussed the pluses and minuses of the situation of E.K. of Nashua, who was debating conversion of a traditional IRA to a Roth IRA.

Under such a conversion, the IRA owner must pay taxes on the amount transferred to the Roth IRA, although if the move is made during 1998 the tax bill may be spread out over four years.

Once the money is transferred to a Roth IRA, the owner can then make withdrawals in retirement with no tax liability.

In contemplating the transfer, E.R. noted that if the move were made, some of the money to pay the taxes would have to be taken from the IRA funds.

James E. Molloy, of New England Securities in Waltham, notes that I neglected to point out that the portion of the funds that was not transferred to the Roth IRA would be considered a premature withdrawal, and thus would be subject to the 10 percent federal tax on a premature IRA distribution.

Although E.R. is not a Massachusetts resident, Molloy added that residents of this state should also note ``another snare in the Roth IRA. The Roth is not recognized by the Commonwealth of Massachusetts as a qualified plan. This means any capital gains would be subject to a 12 percent state tax.''

Presumably this is something that can be corrected by legislation, and therefore is a good thing for people to talk to their legislators about.