A smart investor cannot go purely quantitative or qualitative
One axiom of financial planning is that one size doesn't fit all. People have different risk:reward profiles and even the same individual will have different R:R profiles at different stages of life. This is true for investment gurus as well. But the gurus are more self-aware, focussed and consistent.
Over time, all successful active investors develop their own preferences. Many have shared their logic and philosophy. Analysts have also "reverse-engineered" famous portfolios to figure out what works. Often somebody works out an algorithm to reproduce the guru's selection methods.
For example, Graham placed great emphasis on book-value and low price-book value ratios. Buffett places emphasis on stable earnings growth. Another investor might seek turnarounds. A fourth picks high dividend yields.
Easy access to data and an abundance of computing power makes it easy to write programs to select stocks by given criteria such as these. Such programs can be applied mechanically to the entire population of listed stocks to spit out the names that pass the filter.
This sort of querying is common and maybe even indispensable, to survive information overload. Some 700-plus highly liquid stocks are listed on Indian exchanges and it would be interminable (and boring) to sort manually through all for specific characteristics.
Many software packages offer such querying tools. There are even online websites that offer such querying tools. Some are free. Some are paid. Among the better-known globally are Morningstar, Yahoo! Finance and CNBC – all offer multiple pre-set filters.
Most querying systems are based on simple boolean algebra and these can be mastered very quickly. One can create hundreds of such screens. They can be based on balance-sheet analysis, on P&L analysis, on technical criteria, or a combination of all three.
Can such screens actually throw up portfolios that would meet the approval of the gurus? This is a simple mechanistic method, which is easily replicable. However, most successful investment methods aren't rocket-science and it is clear some (very few) investors do consistently earn higher returns while apparently sticking to very simple methods.
So being simple and easily replicable doesn't necessarily mean being ineffective. However, a screen is only as good as the understanding of the person who created it. If that person doesn't have a clear understanding of the rationale behind selection methods, the screen may throw up garbage.
Assuming that the basic program is solid, a screen is still only a first filter. Very few gurus invest purely on the basis of numbers without an assessment of management quality and some understanding of any given business. In the absence of knowledge about these details, it's difficult to judge vital things like sustainability of growth rates, ability to ride out recessions, etc. These are soft factors that cannot be programmed easily.
Even if they are blunt instruments, stock screens can be useful in several ways. First of all, they give the user a rough idea of how different gurus think and thus, an indication of what their concrete picks would be in a given market. By back-testing across different market cycles, one can also get an idea of which methods work best in which phase of the market.
Again, back-testing such portfolios will also give a rough idea of the possible historic returns and risks involved in investing in a given way. For example, a focus on small high-growth stocks may lead to high returns with extreme volatility while a focus on dividend plays will offer more stable but also lower returns. A turnaround-chasing strategy may yield great returns for some years coupled to massive losses sometimes.
Knowing the historical performance helps an investor figure out his comfort zone and fine-tune his thought processes. He might realise which guru's style he is most comfortable with. If he's really good, he will adapt and adopt features to suit his own needs.
There's a balance to be struck here. Purely quantitative analysis doesn't take note of soft factors. Purely qualitative analysis is very laborious when dealing with mounds of data and risks missing obvious databased signals.
In the end the active investor has to develop his own philosophy and he has to search for deeper insights that go beyond the data-processing that any mechanical stockscreen is likely to provide. A well constructed screen is not a bad way to start looking for stock picks. It is a good way to get to know yourself and your preferences and prejudices.
