I've always been fascinated by the idea of a computer-managed investment portfolio. If we could just come up with some rules and investment principles that beat the averages more often than not, we could eliminate human errors and the guesswork element by having an automated algorithm manage our money.

Or so I'd like to think. If it were that simple, everyone would be doing it, right?

Certainly lots of people have tried, with mixed levels of success. Many funds use "quant" -- short for quantitative -- strategies to help them pick stocks. Some of these are little more than stock screeners, set to search for a (human) manager's preferred combination of attributes. Others use super-sophisticated market models and automate all of the investment decisions, with a human "portfolio manager" to fine-tune the model and keep an eye on things.

What most of these strategies have in common is that they apply rules that work most of the time. As you probably noticed, whatever last year was, it wasn't normal. Some quant strategies got clobbered, and those funds now look pretty lousy at first glance.

But even though the market continues to be volatile, things have settled down from last fall's craziness. What are these funds, and are they worth your money?

What quant strategies really are
Quant funds come in lots of different flavors, but for many, the quantitative analysis boils down to two elements: investment selection and portfolio optimization.

The investment selection part is straightforward in concept, though the details of a particular strategy can be complicated. Joel Greenblatt's "Magic Formula" system of value stock selection, as outlined in his (excellent) The Little Book That Beats the Market, is a good example of a simple quant strategy that screens for promising value stocks and trades them in a fixed, disciplined way.

The next step, portfolio optimization, addresses how you mix and match the stocks you like. In a nutshell, if you were to quantify the risk and return of every possible combination of assets and plot them on a graph, you would find an "efficient frontier" of portfolios that maximize return for every level of risk. Optimization fine-tunes the portfolio to put it on that frontier.

So does it work in real life? The short answer is yes, for many funds it works pretty well. But how much of a difference does it really make?

How different is it really?
Let's take a look at a couple of quant funds. Schwab Core Equity (SWANX) is a large-cap "blend" fund that stayed ahead of the S&P 500 in 2008, but has lagged so far this year. Quant Long/Short Fund (USBOX), another large-cap blend fund (but one that can sell stocks short in addition to taking long positions), did well in 2005 and 2006, but has lagged somewhat since. (Blend funds combine value and growth stocks.)

So where are they these days, and how do they compare to a normal actively managed large-cap blend fund? Take a look:

Fund

Top Holdings Include ...

2008 Return

YTD Return

Schwab Core Equity

Wal-Mart (NYSE: WMT), Hewlett-Packard (NYSE: HPQ), ExxonMobil (NYSE: XOM)

(33.2%)

(2.7%)

Quant Long/Short Continued...