Buy cheap and sell dear." That's the age-old formula for stock market success. Very simple -- if only we could tell when "cheap" and "dear" had arrived. The trouble is, stock market prices are always relative, and only hindsight determines the turn. This year, hindsight shows that the market in general was dear at the end of June, especially among high-tech fast-growers, who hit all-time highs and since then have fallen into ignominious decline.

A few randomly selected examples: Paradyne traded at 51 3/4 before retreating to 36 within a few weeks; Prime Computer shot up to 49 1/4, then wilted to under 25; Wang B went from 45 5/8 to 29 3/4; Cullinane Database Systems hit 68 3/4, then slumped to 47 1/2; Evans & Sutherland brushed 39 and dropped to 25 1/2; Dysan went from 31 3/4 to 22 1/4; Safecard Services plunged from 24 3/4 to 13 3/4; Aeroflex Labs from 29 1/2, to 13 5/8.

These and scores of others were sharp and frightening declines -- some, like Aeroflex, shed over 50%. Even this magazine's fledgling INC. Index wilted (p. 134) after a glowing stretch during which INC.'s fastest-growing companies outperformed all other stock market averages by extraordinary multiples. But are such prices now "cheap"? In terms of historic P/E ratios, the answer is decidedly no, particularly inasmuch as what bargains may seem to abound are among INC.-like fast-growth issues traded primarily over-the-counter. The Media General Financial Weekly, a newspaper that devotes itself to chopping market data into bits and stuffing them into statistical sausages, indicates that P/Es have risen dramatically in O-T-C companies, but only moderately on the New York Stock Exchange, and hardly at all on the American Stock Exchange.

For example, the top 5% of O-T-C P/E multiples averaged 57.8 in early August versus 40.9 the previous year. This rise pushed the frontiers of P/Es 41% higher. The same figures for the NYSE were 27.5 versus 22.6 for an increase of 22%; on the Amex they were 48.1 versus 43.8 for an increase of only 10%. (By way of comparison, it is interesting to look at the 26 companies that have gone public over the last three years through Hambrecht & Quist, one of the country's leading underwriters and a venture capital firm that specializes in quality high-tech growth issues. Recently, these companies were selling at 21.1 times H&Q's estimated 1981 earnings.) And the Amex was timid in another unexpected way: The exchange, whose listings are considered moderately speculative, showed a decline in P/E ratios of 50 or over from 2.6% of listed companies in August 1980, to only 1.8% this year. The number on the O-T-C, on the other hand, grew from 2.3% to 3.1%.

It's apparent that on this basis, the O-T-C market was, indeed, over-extended. Neither professionals nor amateurs are yet comfortable with P/Es that flirt with the 40 to 50 range, even though earnings growth rates are unprecedentedly high. For example, of the sharply dipping stocks above, Paradyne's compounded five-year earnings growth rate was 80%, Prime's was 95%, Wang's was 60%, Cullinane's was 56%, Evans & Sutherland's was 62%, Dysan's was 34%, Safecard Services' was 86%, and Aeroflex's was 43%. No slouches, these outfits -- a fact we shall return to.

The problem was that the market had entered an intermediate-term bear phase. Some signs among the many were:

1. Poor advance-decline figures. The cumulative difference between stocks that advanced on the NYSE over those that declined established new lows for well over a year.

2. The high-low differential -- the difference between yearly new highs and yearly new lows -- failed dismally in June to exceed its March peak, which, in turn, had failed to exceed an earlier one. This trend unmistakably indicated interior market deterioration even as the DJI was in similar over-1,000 territory each time. Then it tailed off badly into negative ground in July and August when there were suddenly more new lows than highs.

3. Margin debt -- the amount of money borrowed by investors from brokerage houses to help pay for their stock purchases -- reached all-time high levels. This is not necessarily a negative in itself, but any marked increase in margin debt is cause for concern, in that when a decline seizes Wall Street, margined stock is called by the broker as its customers' stocks become worth less as collateral. Thus more selling pressure is added to the general malaise.

4. Institutions had reduced liquidity -- a negative factor described in this department last month. That institutions are traditionally out of phase (they seem to commit themselves fully at market tops) is somewhat a self-fulfilling prophecy, since if institutions use up all their buying power in accumulating stocks, a major bullish force is removed from the market. But low institutional liquidity is a time-honored bearish factor, nonetheless.

5. The new issues sector of the market was badly overheated. Just about anything with a "tech," an "onics," or an "etics" tacked onto the end of its name fetched a premium right out of the gate. The classic case was Genentech, whose initial public offering range was indicated at 25-35. The stock traded as high as 88 the first day, after coming out at 35.

6. Volume decreased on the downside.Low-volume declines indicate lack of buying interest. They may peter out and buyers may rush back in quickly, but more often than not low-volume declines, if they continue as July-August did, are prefatory to increased selling volume as more and more stock owners get tired of waiting it out. Bear markets usually begin with low volume, but end in selling frenzies.

7. Public short selling remained at low ebb, suggesting that amateur investors weren't worried -- yet.Bear markets end as the public begins to sell short in earnest, wrongly convinced that prices are going lower still.

But bear market or not, quality issues of growth companies must be rebought. There isn't any reason to ask when to buy, because you never know when demand is going to strike from professionals who either missed out on the first few rounds or who sold at higher prices simply to take profits, but still like the company's fundamentals. A case in point (as of this writing) is Paradyne, which, despite further incursions in other stocks, recovered smartly from its low at 36, and within a couple of weeks was back to the mid 40s. Professionals aren't waiting for chart bottoms; they're making chart bottoms -- fast.

It is important not to feel that these stocks have had it, just because they've been trimmed. Many of them have more than tripled since April 1980, when the high-tech, fast-growth push began to accelerate. Paradyne's low within the past 60 weeks was a mere 18 1/2, Prime's was 22 1/4, Wang's 24 1/8, Cullinane's 22 1/2, Evans & Sutherland's 18, Dysan's 14 3/4, SafeCard Services' was 4 5/8, and Aeroflex's was 3 7/8. It's easy to forget that these stocks, and hundreds like them, have come a long way in barely over a year. Their declines may have seemed precipitous, but, again, "cheap" is relative.

What's more important is to understand that at every bear market bottom in the 1970s, technology and fast-growth companies have led the recovery. Most technology companies did not return to their previous lows in 1974 or 1978 or 1980. And it's a safe bet to assume they're not going to do it in 1981 either -- not by a long shot. The emphasis, however, should be on quality issues -- good management with proven products and bright, innovative ideas. A key to unearthing many such issues is to see which ones, in fact, did not decline at all in the recent general setback. If they could hold against a selling tide, they're good candidates to spurt once the buying begins again.