Quantitative Equity Investing

About the Site

A little history

Back in early 2007, I found myself with a bit of free time while I was involved with an unrelated entrepreneurial venture. I started to read some books and articles about behavioral finance, which in turn led me to some of the classic academic literature on the excessively (by efficient market standards) positive returns to value and momentum investing strategies. Using Microsoft Excel, some self-taught Visual Basic and a serendipitous discovery of a free source of CRSP/Compustat data (alas, no longer available), I decided to see if I could replicate some of these studies. Having found success, I set out to further research these phenomena, which led me to the work of Cliff Asness (of AQR Capital) on combining value and momentum strategies, research with which I pursued with much excitement.

Lo and behold (to me, at least), value and momentum strategies worked pretty well when combined. Naive as I was, I thought I was on to something known only by few, and started to think about how I could raise some money. Then something happened. Quant funds, employing exactly these kind of strategies, started to "blow up." This enlightened me to two things. On a positive note, it was my first confirmation (remember, little was written about such quant strategies outside of academia, prior to the summer of '07) that my strategies actually did indeed "work." And by "work" I mean to say that my results were highly consistent with what I was reading, especially the now famous (in quant circles, at least) August 2007 research note by Matthew Rothman of Lehman Brothers on the "Turbulent Times in Quant Land."

However, on a negative and somewhat more significant note, the events of that summer opened my (and the rest of the world's) eyes to the fact that these strategies were just a tad bit crowded and that I wasn't so smart or so special after all in "discovering" the merits of value and momentum quantitative strategies. So for the next few years, as the Great Recession took hold, I didn't do much with the research, though it was never too far from my mind.

Recently I decided to revisit these quant strategies. I was curious as to how they have held up during these exciting economic times. I know that assets under management are certainly down from peak levels but also that many firms are still pursuing such quant strategies. I set out to update my own research. But since I needed a new data source, I essentially had to start from scratch. I thought it would be a fun project to automate the process and have the results updated daily. Then I figured, I might as well just put this up on a public website, for all to behold the fruits of my labor. Voila.

So what's the point?

Chances are, anyone who's actually running money in these strategies doesn't need this website, as I would think (or hope, at least), that they have the ability to monitor their performance and the performance of various quant factors on a "real-time" basis. But since there are so few websites out there about quant finance (as compared with the thousands of sites about traditional "stock-picking"), I thought it might be useful for someone out there curious about these types of quant strategies. (Of course, this site is meant for informational purposes only and is not meant as investment advice. I make no promises as to the accuracy of the results. I suppose I had to put a disclaimer in here somewhere).

I also thought that just maybe history will rhyme. That is to say, perhaps unusual performance by quant funds (since they essentially play the role of liquidity providers), might indicate other unusual goings-on with the market, just as Quant '07 was one of the first indications that all was not well with the world's financial markets and economies. If nothing else, this website gave me the opportunity to have some fun and to teach myself some PHP and a bit of JavaScript. And finally, just maybe someone will see this and want to give me a billion dollars to invest (or just slightly more likely, hire me as the industry's only PhD-less quant analyst).

Like Einstein said, as simple as possible but no simpler (or don't laugh at me)

Anyone who does have experience with these strategies should immediately notice that methodologies used on this site are much simpler than those employed by real funds. In fact, they're about as simple as this kind of quant strategy gets: take a factor, rank each stock by that factor and then create portfolios from these rankings. For analyzing the results of a value and a momentum factor together, we simply average the ranking of the two factors. We're not using a true factor model or even a Z-score and there's no style tilting. The long/short portfolios are simply dollar neutral, not beta neutral or industry neutral. All of these, and lots more, are good ideas (obviously), many of which I've previously researched, and some of which I may implement here in the future.

Moreover, the performance results shown here are clearly not going to exactly match the "real-world" trading results based on the same strategies - for a number of reasons. First, the results shown here exclude trading costs (likely significant, since we effectively rebalance daily) and calculate all returns based on each stocks's closing price. On the flip side, however, the quality of the freely available data used on this site is certainly likely to be inferior to the data sources used by hedge funds employing quant strategies. Moreover, no attempt has been made to clean or verify the accuracy of the data, other than excluding some outliers.

I do think, however, that there are advantages to such a simple methodology. Not the least of which is in the ease (relatively speaking, to keep the Einstein theme going) of implementing a completely automated website, with free data and without the statistical tools of a specialized software package or even the simple analytical tools of Excel. But also, and more importantly (for you certainly, if not for me), I believe that (and this is a hunch from a quant hobbyist, not an insider - so take my view with an extra large grain of salt) that while funds have (marginally) different factor models, a simple, long/short market neutral, value plus momentum strategy probably results in a reasonable benchmark of the aggregate performance of "more sophisticated" funds.

What next?

I have a lot of ideas for adding additional functionality to this site (I also have a lot of ideas about the quant strategies themselves - but I'll keep those to myself just in case I'm ever in the position to put them into practice). Here a few things that I am hoping to add to the site when I have some more free time:

I'd love to hear from you...

Lastly, thanks for reading this. If you happen to have any comments, suggestions or questions about the site, methodologies or data, I'd love to hear from you. I can be reached at andrew [at] instituteforfinance.com.