Talking about the relative value of

PMC-Sierra

(PMCS)

and

Coca-Cola

(KO) - Get Report

, or

JDS Uniphase

(JDSU)

and

Ford Motor

(F) - Get Report

is a good way to make sure you don't mindlessly jump from sector to sector with every shift of a very scared and very volatile market.

But relative value doesn't tell us if a specific stock is worth 70 or just 50. Relative value doesn't give us the target prices we need to make intelligent buy and sell decisions.

As much as investors may need them right now, the ups and downs of this market don't make setting a target price easy. And any targets an investor sets don't seem to last for long. In the last few weeks, for example, I first raised my target price for

Vitesse Semiconductor

(VTSS)

to 110 by June, and then on April 25, I lowered it to 84 by October. I've done a similar dance with

Citrix Systems

(CTXS) - Get Report

-- up to 115 by June in a March call, and then down to 88 by October in an April 25 call. And I doubt I'm done with those targets, either.

But I am seeing signs that the same stocks are trying to find a trading range. These issues seem to be hitting the same lows when the market hits its lows and the same highs when the market hits its highs. I know that these trading ranges may not hold -- the odds are still good that we'll see lower lows sometime this summer -- but I think we can use these current trading ranges to build conservative target prices for the end of 2000. In turn, those target prices can give investors some guidelines for buying and selling in this very difficult market.

I'm not alone in having this problem, by the way. Wall Street can't make up its mind about what a specific stock is worth either. On April 11, for example, two good Wall Street firms had target prices of 185 a share and 85 a share, respectively, on

Commerce One

(CMRC)

.

Keeping Up With Daily Changes

I know why my system is resulting in volatility in my target prices. A key part of the data I use for calculating a target price is showing huge changes almost every day right now.

Setting a target price based on a company's fundamentals, which is what I do, is a two-step process. First, you have to project earnings for some point in the future -- a year or two ahead is my preference. Second, you have to make some estimate of what investors will be willing to pay for those earnings in the future. Multiply the two together -- future earnings per share times future P/E ratio -- and voila! a target price.

Projecting future earnings is never easy, but that part doesn't get markedly harder during a volatile market, either. During the recent technology stock crash, for example, earnings per share for technology companies have by and large come within pennies of the forecasts. And since the economy hasn't changed course during the market meltdown, CEOs and CFOs have been able to give remarkably clear guidance to Wall Street analysts about upcoming earnings. In the overwhelming majority of cases, analysts have raised their estimates by a few cents a share after listening to the company conference call. Robertson Stephens, for example, raised its estimate for Citrix Systems' 2000 earnings to 86 cents a share from 84 cents, and raised its estimate for 2001 earnings to $1.06 a share from $1 after the company reported on April 19. Those aren't nicely positive moves, but they aren't huge changes.

All the volatility is concentrated in the other half of the equation -- in the

price-to-earnings (P/E) ratio. Just look at a stock like

Mercury Interactive

(MERQ)

. When I added Mercury Interactive to "Jubak's Picks" on March 10 at a price of 114 a share (ouch!), the stock was trading at 197 times projected earnings per share (EPS) for the next four quarters. On April 25, at 75 a share, the stock traded at 114 times projected EPS for the next four quarters. The effect of the drop in its forward P/E ratio from 197 to 114 dwarfs Mercury Interactive's positive April 13 earnings report and the subsequent increase in projected earnings per share for the next four quarters from 57 cents to 66 cents.

Setting a price target based on a hope that the stock will again trade at 197 times earnings is wishful thinking. (It's not much better than setting a 134 target based on a belief that the stock will trade at that level a year from now because that's the 52-week high.) I think this selloff has been severe enough already that the market isn't going to automatically return to its old psychology. That old 197 forward P/E ratio is just not relevant any more.

Target Prices For the New Economy

So how do you go about setting a target price for Mercury Interactive if the old P/E multiple isn't a useful guide to future valuations?

The stock's recent trading range is a good place to start, I believe. On March 27, the stock fell below its 10-day moving average at 104 and dived almost without stop to 62 by April 4. After thus rather decisively breaking its long-term upward trend, the stock has bounced around with the market. On good days when the market is rallying, Mercury Interactive has moved as high as 85. On bad days for the market as a whole, it's fallen to the high 50s. Most of the time, the stock has moved between 60 and 75 a share.

If the stock continues to do this for a while, it would be in a phase called consolidation. This back-and-forth movement in a range helps build a base for a stock's future advances. Investors gain confidence in the stock -- it doesn't fall off a cliff, taking their investments with it, at every market dip. By testing the top of the range, the stock convinces investors that it can move higher. This confidence-building is often resolved by a move to prices above the top of the range.

This process usually takes months after a stock has taken a severe beating, and the longer the pattern builds, the more confidence I feel in it. In the case of Mercury Interactive, for example, the pattern is still too new and too fragile for me to be confident that the stock is going to consolidate here rather than drop some more.

Still, I think you can use this pattern to help rough out a target price for a stock.

What's a reasonable target price for Mercury Interactive in December 2000? Well, in that month investors will be looking at projected earnings of 90 cents a share. (I get to that number by beginning with the current consensus estimate of 78 cents a share for 2001 and then assume that Mercury Interactive will continue to report 10% earnings surprises each quarter in 2000. Those surprises drive up the base for future estimates and gradually ratchet expectations for growth higher as well.)

If I simply used the pretechnology crash forward P/E estimate of 197 with that earnings projection, I'd come up with a December target price of 177. That would be nice, but as I've already noted, I think building an investment strategy on a quick return to the excesses of this March is asking for trouble.

Instead, let's use the forward P/E range traced out by the stock recently. At the high of 75, the stock is trading at a forward P/E ratio of 114. At a low of 60, the stock is trading at a forward P/E ratio of 91. Using those numbers for December 2000 gives me a target price range of 103 on the high side to 82 on the low side.

While both are above the April 25 close at 75 and change, let's think about return and risk for a minute before rushing out to buy.

Compensating For risk

This is a very risky market and Mercury Interactive -- or any stock trading at a forward P/E ratio of 100 or so -- is very risky. So I'd like a nice, fat potential profit to compensate me for taking that risk. A potential 30% by December sounds about right. (I'm setting it so high because there's a chance I won't get it.)

Now let's work backwards from those target prices to a range of purchase prices. To make a 30% return if the stock climbs to 103 by December, I'd have to buy Mercury Interactive at 79 or less. To make a 30% return if the stock climbs to 82, I'd have to buy at 63 or less.

These levels aren't guarantees of anything, of course. I think there's a good chance that this stock will command a forward P/E ratio of better than 114 by December. If I read the year right, confidence in technology stocks should be higher then than it is now. In my opinion, a target of 103 is conservative. But there's also the chance that we'll head lower this summer before we rebound. And no stock can be worth even 91 times projected earnings. But with those important caveats, I've now got a conservative target price and a possible buying range for Mercury Interactive.

Setting More Buy Targets

I can do the same calculation for other stocks that have started to build trading ranges recently.

Network Appliance

(NTAP) - Get Report

, for example, has recently shown signs of settling into a range between a high of 70 and a low of 50. Right now analysts are projecting earnings per share of 56 cents over the next four quarters -- that gives the stock a forward P/E ratio ranging from 136 (when the stock is at 70 a share) to 89 (when the stock is at 50 a share). Looking forward into 2001, analysts are projecting earnings of 82 cents a share. Using those numbers, next December's target price ranges from 108 to 73 a share. Working backward and using a 30% required rate of return, I get a current buying target of 83 to 56 a share.

Here's one more example -- this time for

Broadcom

(BRCM)

, a stock that isn't a Jubak's Picks recommendation at the moment but that I'm looking to add back to the portfolio at some point this summer. I put Broadcom on my wait-to-buy list on Feb. 15 at a split-adjusted 140 a share. Since bouncing off its recent low of 122 1/4 on April 14, it has moved between a high of 160 and a low of 140.

At the high of 160, Broadcom's forward P/E ratio is 186; at the low of 140, the ratio is 163. Using those ratios and projected earnings per share of $1.11 for 2001, I get a target price range for December 2000 of 206 to 181. Working backward from there with a 30% required rate of return, I get a buying target price range of 158 to 140.

When To Make Your Move

What should you do with these numbers? I'd recommend waiting to see these trading ranges consolidate for a while longer, then waiting even longer for clear signs of a breakout from the range to the upside. Let's not kid ourselves -- a reliable trading range isn't built in a week, which is pretty much the time period I'm talking about for Broadcom, for example. These ranges are extremely fragile and there's no guarantee that they'll hold. We're very early in the work of repairing the damage done to this market -- if we've actually entered that period at all. And I still can't build a convincing argument for why technology stocks should go higher in June and July.

But maybe you're convinced that the worst is over. Or you believe that the bounce back will be so fast you won't be able to catch it. If the compulsion to buy has become overwhelming, I'd still suggest that you try to time your purchases to hit the bottom of my buying target ranges. I think those prices offer you the best combination of upside potential from a market recovery and downside protection in case this market slips further.

For example, going back to Mercury Interactive, I think the stock is a so-so buy at 79, the top of my buying target range. Buying at 63 or lower gives me better downside protection and more upside potential. On the upside, I think I'm looking at a worst-case gain of 19 a share by December, and a good shot at gaining 40 a share on a 63 a share investment.

What you decide to do will depend on your risk tolerance, of course, and your opinion of how much further technology stocks will sink and how long it will take them to recover. But whatever your opinion on the market and of the technology sector, I think this method can give you some help in setting targets for a number of stocks.

One Caveat

I wouldn't apply this technique to most of the market at this point. Some stocks are trading in such wide bands that I don't think we can even call them trading ranges.

GlobeSpan

(GSPN)

, another one of my Feb. 15 "wait-to-buy" stocks at a split-adjusted 60 a share, has traded between a high of 90 and a low of about 47 1/2 since April 7. Some stocks show even less of a pattern right now. Just take a look at the charts for

Exodus Communications

(EXDS)

or

Ariba

(ARBA)

, for example. I don't see any reliable signs that these stocks have put in a bottom yet.

I think there's a reason that so few stocks have started to build even a rudimentary trading range so far -- and why almost none of the money-losing Internet stocks are showing any signs of this kind of pattern.

I think you'll find the answer in a fundamental called cost of capital. And I think a look at cost of capital will also explain why P/E ratios -- especially forward ratios -- have been so volatile and are likely to continue to be so. It's not just the manic swings of market psychology. I'll explain why cost of capital is currently so important in my next column.

Jim Jubak is senior markets editor for MSN MoneyCentral. At the time of publication, he owned or controlled shares in the following equities mentioned in this column: Broadcom, Citrix Systems, Commerce One, JDS Uniphase, LSI Logic, Mercury Interactive, Network Appliance and Nortel Networks.Holdings can change at any time. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. He welcomes your feedback at

mctsc@microsoft.com.

More from

MSN MoneyCentral

Jubak's Picks

Markman's SuperModels

Rowland's Watch Portfolio