For our May 28, 2004 broadcast, John Reese of Validea.com and money manager Eric Miller of Heartland Advisors talked to Wall $treet Week with FORTUNE's Karen Gibbs about using software to screen for investments. Here are portions of their conversation that didn't make it on the air:

KAREN GIBBS: Well, John, where did the concept of Validea.com come from?

JOHN REESE: The concept of Validea arose several years ago when I sold my first business. I was looking to invest a lot of funds in the stock market. I was starting to methodologically look at what were the best ways of investing, how do you get the best results. I started reading up on what many people consider to be the very best strategies.

I'd actually first started with Peter Lynch, his book One Up on Wall Street. It was very eye opening in terms of being able to follow a methodology and the guidance in terms of picking and selecting stocks. I then turned to the grand book, The Intelligent Investor and found also that was absolutely amazing in terms of telling me what to do. I then proceeded through a number of different styles on this, Martin Zweig for growth investing, David Dreman for contrarian investing, and turned those into computer programs using artificial intelligence techniques, so that I could conveniently push a button, type in a ticker symbol, and it would automatically pull the information, all the data points that were needed for me, and then come up with the evaluation, and furthermore tell me how he actually arrived at that particular decision. I wanted to know point by point for myself how that particular stock scored on the guru's criteria.

GIBBS: Well, let's take a look at the Peter Lynch-based guru portfolio. Tell me, what screens did they use? What is the common metric used?

REESE: Well, first of all, let me explain that these are not common strains. These are actually very complicated logic that's built in. Following the Peter Lynch logic, you can first classify stocks into several different categories, such as fast growth, slow growth, stalwart, and then applies different criteria to each, depending on what they are. It may also apply different criteria depending upon the industry that they're in. It also looks for extra bonus things that can occur, for instance, when the net cash in a share of one share is a third or more of the price of that share.

GIBBS: Does this program allow for unexpected or unpredictable events?

REESE: It does not take into account unpredictable or unexpected events, because by definition they're essentially unpredictable. What it does is it will see the effect of those events on the price and on the fundamentals of the stock, and then make a decision based upon that, whether it's still worthwhile to invest at that price.

GIBBS: You mentioned that you rebalance monthly. That incurs transaction costs. What happens to the buy-and-hold theory?

REESE: Well, let me emphasize that I strongly believe that you don't have to hold stocks for the long term in order to be a buy-and-hold or long-term investor. As long as you adhere to the methodology over a long term, you'll experience the successful results of that.

Now, transaction costs are an interesting question. What we've found is due to technology in the brokerage industry, it's now possible to have accounts for almost an unlimited number of trades for a very small monthly fee. It's almost an insignificant part of trading costs right now.

GIBBS: Eric, you're a classic Graham disciple, but yet in your portfolio I don't see any of the Graham portfolio holdings. Why?

ERIC MILLER: Well, there may be a few of them. One thing that we do, and something that goes back to the founding of our firm and founding of our value fund 20 years ago, is we've decided to specialize in small microcap stock. And that's actually why it's even more important in our view to have a very disciplined approach to that to both take the emotion out, as John mentioned, so you avoid some of the momentum swings that people sometimes suffer through, but also to give us sort of a safety net, because small micro-cap stocks tend to have more volatility than large-cap stocks.

So in order to somewhat minimize that volatility by applying principles of Benjamin Graham, we've found it's been great for long-term investing. Now so many of our names are just not going to pop up on screens, because the median market cap of our portfolio today is probably around $230 or $240 million market cap, very small. So a lot of people when they screen for stocks are not going to screen that low.

Now there are several names in the portfolio that John has that we actually own in different portfolios. I believe he has LaFarge in one of his portfolios. That's about a $3 billion, $3.5 billion cap company, so it's a little bit too big for our value fund, but where we manage individual portfolios and have a little extra room on the market cap, we own that.

But again, what we're trying to do is, we look at low P/E, we look at low price-to-book, we like to see very clean balance sheets. But two important tenets of Graham that we use, the first one is defining intrinsic worth, and especially in microcap, small-cap stocks, what we want to do is determine for every stock we own what a private buyer would pay for this company, and I think that sort of gets to the heart of Benjamin Graham.

The other thing we want is a safety net, another very classic Graham tenet. And that safety net might be tangible book value, it might be a lot of cash on the balance sheets. And that's again I think particularly important for small-cap investors, because small-cap stocks don't have the ability to run to the bank or issue commercial paper if things go south on them. So you really don't want to play around typically, in our view, with high-leverage type of companies in the small-cap arena. And you want to buy time on earnings from an earnings standpoint. So we find the safety net, a lot of cash on the balance sheet, you know, less-leveraged balance sheets, very important from our screening process.

GIBBS: Well, certainly small-cap stocks are under researched, but is it a definition that they are then undervalued as most people are looking at the large-cap names and therefore have taken out all of the discount and have fairly priced them now, Eric?

MILLER: I think it's clear if you went back to 2000, sort of the end of the Internet tech mania, small-cap stocks were frankly never cheaper. And in fact, if you'd looked at our portfolio, I think in the year 2000, we wound up, or maybe it was 2001, we had 12 companies in our portfolio go private. And in the 1990s, in aggregate we didn't have 12 companies go private. So they were incredibly undervalued then, and it was just a matter of you knew you were going buy names, you knew they were cheap, you knew you were going to make money. It was a question of sort of timing.

Now small-caps have just finished their fifth year of out-performance, so clearly the valuation gap has narrowed. I think right now small-caps and large-cap stocks are generally in somewhat of an equilibrium. But the beauty to us of small-cap investing is there's 5,000, 6,000, 7,000 of them out there, and so while in aggregate perhaps the small-cap market is fairly valued in our view, you can still find, we own 250 names in the value fund, and so in our view we can still find 250 that offer great upside opportunity.

GIBBS: Do you have any of those great opportunities that would go against the grain, or in this case, against the Graham strategy of investing?

MILLER: Yeah, sometimes, again, we use Graham's, and actually we have our own what's called 10-point grid. It's somewhat of a similar screening process I think of what John has where, again, we actually give a yes or no on low price to book, whether it's low price to cash flow, whether it's a low P/E. But we're not, and generally we do that because that does take some of the emotion out and hopefully sort of limits some of the mistakes we're going to make, but we don't use that as a hard and fast rule. And there are going to be some stocks, our number one criteria is catalyst, and that even comes before the financial criteria.

GIBBS: Okay, now you know a lot of companies come out and they restate all these earnings, all those little quantitative numbers that you look at. How is it going to back testing, your program?

REESE: Basically the numbers are originally stated what they originally used, so it takes into account the fact that some percentage over time actually do change and do restate their particular earnings. It is possible that those will turn out to be bad investments. One thing I can't say is that these systems pick 100 percent good stocks. I would say that at best they're picking 60 to 70 percent of the stocks go up and go up well, certainly much better than they go down.

GIBBS: And, Eric, one other question about a stock in your portfolio, Mesa Airlines. Now you heard what Warren Buffett had to say about airlines. Mesa is in a very hostile, competitive environment. Why is that stock included in your portfolio?

MILLER: Well, again, Buffett is right and anybody is right, you don't buy and hold airlines for 10 and 20 years. You're not going to make money in it. But you can still use them as trading vehicles. And I pay probably a little bit less attention necessarily to valuation tools for airlines as opposed to sentiment. And again, if you actually look at any chart of an airline stock, you go back to 2001, the stocks were gradually going down as the economy was slipping into recession, and obviously then September 11th hit and the stocks got absolutely crushed. So sentiment in the airline industry on September 12th and through a few weeks there in September 2001 was horrible. But obviously that was a great time you want to buy against sentiment, and if you looked at almost every airline, it more than made a recovery shortly off of that low. But then again, you can't get carried away with a business. Most of them have backtracked fairly significantly. But so in going against sentiment, Mesa is certainly a good example.

The sentiment on Mesa is awful, and why it is is because their biggest customer is US Air, and everybody is saying US Air is going to go out of business, and these guys are going to suffer. Well, they would have suffered far more several years ago when they had a weak balance sheet. But again, this goes back to our balance sheet discipline. Mesa has $5.40 of cash on the books, almost covering all of their debt. So even in the worst case scenario of US Air running into some serious problems, Mesa is going to be around.

We like the regional airlines much more than the big national carriers. The regionals tend to have better balance sheets, be a little more nimble, definitely have lower cost structures, which is extremely important. So I'm not saying we're going to hold, I'm not saying we're going to hold Mesa for five years, but I think we can make some money on the trade right now.