Serious Money: When is the right time to buy technology stocks? From a conventional, analytical point of view, is there ever a right time to buy technology stocks? Or is the tech sector doomed to remain forever the playground of the chartists and traders, rarely paying heed to anything that traditional equity analysts put into their spreadsheets and models?

Last October, Serious Money made one of its better calls and suggested that Google - one of the tech sector's bellwether stocks - was a buy (and could even be matched with a sell on Microsoft).

This was an unashamedly high-risk bet - definitely not the kind of investment for the faint-hearted - based, it has to be said, on little more than a hunch. A relatively cursory look at Google's underlying valuation would have put any conventional equity investor into a cold sweat.

The simple point to be made about Google remains the same now as it was back in the autumn of last year: the share price will go up for as long as the company continues to defy gravity with its astronomically high growth rate.

As soon as growth starts to slip, my guess - not a terribly profound one - is that the stock will collapse. But so far so good.

Since mentioning Google, its share price is up nearly 50 per cent and Microsoft is down nearly 8 per cent (but owners of Bill Gates's company will have benefited from the huge special dividend paid last December).

Revenue growth, in particular, has continued to beat all expectations, with Google reaping the benefit of the migration of advertising to the internet.

Of course, much of that rise in Google's share price has come in the past 10 weeks or so.

From the end of last October to the end of March the stock didn't actually do very much. In fact, the relative lack of volatility over this period is a touch surprising for such a highly rated company.

And Google's fortunes have, to a certain extent at least, been mirrored by the tech sector's ultimate index, the Nasdaq.

For the first four months of 2005, tech stocks, by and large, did not do well, particularly when fears about an extended soft patch for the US economy were at their highest.

It is one of those little market paradoxes that many technology companies are deemed to be in both "growth" and "deep cyclical" style categories.

But since the start of May, most tech stocks have had a very good run, as global growth fears have abated somewhat. News from other industry bellwethers such as Intel and Texas Instruments has generally been quite good, at least in terms of meeting profit expectations and company guidance about near-term growth prospects.

Other key indicators for the sector, such as the price of certain types of memory chips, have also been sending out positive signals.

Some analysts are saying that prices for Dram (dynamic random access memory, a type of memory used in most PCs), for example, bottomed in the first half of May.

The Semiconductor Industry Association has also been making market-friendly noises.

Just how reliable these sorts of indicators are is a moot point. My guess is that the leading indicators used by many technology analysts send more false signals than any of the indicators used by economists to forecast the economy.

But the industry is as keen a devourer of news on Dram prices as it has ever been, reliability issues notwithstanding.

Other technology-related companies seeing good rises in share prices include the mobile phone giant Nokia.

Tech stocks have also been boosted by the current fashion for preferring growth over value. Style shifts come and go and some analysts believe that growth is due a prolonged run.

The argument here rests on two observations.

First, value has been running for five years so growth is due its turn.

Second, so many stocks are now valued the same it is inevitable that higher growth companies will now be favoured. Personally, I'm not sure about either of these arguments but many market participants seem to be convinced.

Perhaps we are also witnessing a genuine increase in demand for many tech products. Third-generation telephones might finally be making a breakthrough.

All those flat-panel PC monitors bought in the last boom at the end of the 1990s are probably due for replacement - mine is certainly looking tired. Nokia and Intel's recent news about joint WiMax development is something worth watching.

But weren't tech companies behind all of the stock market woes of the late 1990s and early part of this decade? Surely all those problems have not been solved? Absolutely.

Based on any conventional understanding of valuation, most technology companies remain expensive. The trend in those valuations, if not share prices, remains down.

The old market cliché has it that nothing moves in a straight line and tech stocks are no exception. The current price upswing is merely a cyclical blip that leaves the longer term downtrend firmly intact. For some companies, the valuation anomaly will be corrected by earnings growth; others need their share prices to decline further.

The moral of all of this is that only traders and private investors who enjoy a straight punt should get involved in tech stocks.

There is lots of money to be made, on both price rises and falls, but we need to be able to make lots of correct guesses about the behaviour of the wider economy in general and the demand for technology products in particular.

As it happens, both key drivers are looking pretty good at present, so the sector's run could continue for a while yet. But tech's ability to turn on a sixpence means that the risk-averse investor should leave this one well alone.

Chris Johns is an investment strategist with Collins Stewart. All opinions are personal.