The company chooses stocks by measuring a company's patent quality as a predictor of future stock market valuations.

Though it's not guaranteed to help investors create a winning portfolio, according to company president Francis Narin, CHI Research's method offers "another metric that investment professionals can use to make objective, quantitative analysis" of a company's investment potential.

The method, patented in January as "Method and Apparatus for Choosing a Stock Portfolio Based on Patent Indicators," measures the impact of a company's patents and a company's rate of innovation as a way to determine how well its stock will perform in the market. The method analyzes publicly traded United States companies with at least 45 to 50 patents. CHI Research's goal is to uncover companies that may be undervalued in the stock market but have tremendous investment potential because of the strength of their technology.

The company, which has analyzed competitive technology strength of companies for 30 years and maintains a database of all U.S. and European Patent Office patents, plans to license its stock-picking method to financial institutions and investment professionals.

Professionals can also subscribe to the firm's Investor Tech-Line, a monthly ranking of the 20 most undervalued technology stocks out of 350 companies in 22 industries based on patent analysis. Companies whose stocks are underpriced for their technology would rank higher on the list than those that are well-priced in the market.

The analysis looks at the impact of a company's patent activity in the last five years to determine how frequently its patents are cited in other company's patents. This is considered a determinant of how influential a company's patents are.

In addition, a company's technology cycle time is measured by finding the median age of patents cited in all of the company's recent patents. This is way to identify how fast a company is innovating. Narin said this indicator is especially important in evaluating electronics and semiconductor companies, in which technology advances quickly. "Any [semiconductor] company that cites 10-year-old patents in its filings is in trouble," he noted.

But the same wouldn't be true for an automotive company. That's why CHI Research compares companies in the same industry.

Another indicator is science linkage, or the number of scientific papers cited in a company's patent filings, which according to Narin tells an investor whether a company is building upon cutting-edge research.

CHI Research conducts linear regression analysis of the patent indicators and R&D density (the ratio of research and development spending to a company's sales) and compares the date to the market-to-book valuation of a company. (The market-to-book valuation equals price per share divided by book value of the company per share.) The method then finds the "best fit" line for 350 companies in industries ranging from biotechnology to semiconductors, Narin explained.

In a press release, CHI Research claimed that its portfolio of the 20 undervalued technology stocks has outperformed the S&P 500 in eight out of 10 years and the Nasdaq in seven out of 10 years since 1991.

Over the last two years the company has employed the method for its own pension plans and has seen a greater number of the companies it identified as undervalued rise in stock price, Narin told EE Times.

Though Narin maintained his method has been successful, not all investment professionals are convinced.

Dan Scovel, an analyst who covers the semiconductor industry for Needham & Co. in New York, agreed that patent quality is an interesting data point and it's important that companies have meaningful technology. However, he said, "It's what you do with what you have" that counts. Rather than use patent quality as the sole measure of a company's potential, Scovel said it should be one of many methods used to pick stocks.

The CHI Research method won't be available to individual investors directly, but in a few weeks investors will be able to conduct their own patent analysis by purchasing patent profiles of companies they are interested in from www.economy.com, a Web portal for financial analysis, data, forecasts and information on the United States and world economies.

According to Narin, identifying a company that will succeed is not about how many patents a company has, but the impact of those patents, how innovative the company has been and how fast the company churns out new technology.

"Underneath it all, what it means is that quality technology is really important, quality research pays off, and good engineers really count," Narin added.