Systematic Value Investing
Being Systematic in a Quantitative Environment--
If there's one thing which I've learned from this past summer, it's the importance of being systematic. Large projects make it less and less possible to just bull through a project with code line by line on the fly. Suddenly you need pseudo-code and flowcharts to map your way to the end before even touching real code. By taking that step away from your programming language, pseudo-code and flowcharts leave you with that much more ability to focus on the bigger picture-- with the intention of zooming back in after you've mapped your path to the finish line out. To write out all that pseudo code it helps to use scalable and robust functions; functions which can handle minor, or even somewhat major, variations in the underlying dataset without bugging out. If you didn't, God knows how long your code would end up being. Also, God knows how much more difficult it becomes to debug! Finally, God knows how tedious things would get as you vary the datasets you are analyzing! Finally, large datasets make it important to have efficient functions so that your functions run in a timely manner. Flowcharts, pseudo-code and good functions support exploratory data analysis and not the other way around.
Application to Value Investing:
I'm of the opinion that these sort of ideas are not at all foreign to value investing, as long as we're able to differentiate between the following two dogmas-- "be systematic" and "be systematic when it's applicable to be that way." Just because I have a hammer doesn't mean that everything is a nail. Hell, I'll be technical when it's applicable to be so. I wish I believed in some 'one size fits all' truth which I could subscribe to. But I don't. I'm more of the belief that if there is some all-encompassing truth does exist, it's more of a tapestry, stitching together a vast number of distinct skills, skillsets, and rules, the breadth of which is so large that we can only hope to capture a sliver. That makes sense to me.
With that in mind I'd like to turn to how value investors and most corporations gather information. This is a realm which seems to me to be less systematically thought about than it probably should be. Let's say your boss goes up to you and says "Dan, I've found an interesting stock and the ticker symbol is XYZ. By the end of the day I want a complete write-up on this stock's business, competition, industry and prospects for the future" (I'm sure we've all had a variant of a situation like this).
It's things like this which I think merit a little more systematicity than they're given, on average. How can I be more systematic in my information gathering? Well, for one, it might be of value to do a thorough search of as many data providers as you can, sampling them all in an attempt to wean out new quality data sources, and then recording those sources, perhaps in some sort of information architecture on a local intranet where your respected co-workers can offer their insights as well with categories and comments. One could also classify all information along multiple lines, essentially creating multiple information architectures allowing you to flexibly cater to the task you may have at hand. Finally some sort of a search component is a must, so that you have yet another way to swipe through your information database, this time with a user-specified keyword.
This sort of ideation, taking a step back from the actual grunt process which we call 'information gathering', might not be pseudo-code in the technical sense, but it sure starts me along a path to create a system from which I can more easily gather and process information on an ongoing basis. The above description might not be a scalable function, but it seems to be a pretty scalable system because it supports and encourages iterative growth as members of the intranet add their comments and input over time. Finally, the above description might not be a robust function, but it seems to be a robust system because of its ability to classify information along multiple lines with a search capability, allowing the user to gather information in multiple ways over the wide range of goals those users may have.
Of course one doesn't need to do all of these things, but hopefully you can see how being systematic can be a good complement to however else you happen to gather information.
Being systematic can help with many other aspects of our investment activities-- it might be of value to think about this from your own perspective.
Comments are welcome!
Categories: theoretical_intersections
If there's one thing which I've learned from this past summer, it's the importance of being systematic. Large projects make it less and less possible to just bull through a project with code line by line on the fly. Suddenly you need pseudo-code and flowcharts to map your way to the end before even touching real code. By taking that step away from your programming language, pseudo-code and flowcharts leave you with that much more ability to focus on the bigger picture-- with the intention of zooming back in after you've mapped your path to the finish line out. To write out all that pseudo code it helps to use scalable and robust functions; functions which can handle minor, or even somewhat major, variations in the underlying dataset without bugging out. If you didn't, God knows how long your code would end up being. Also, God knows how much more difficult it becomes to debug! Finally, God knows how tedious things would get as you vary the datasets you are analyzing! Finally, large datasets make it important to have efficient functions so that your functions run in a timely manner. Flowcharts, pseudo-code and good functions support exploratory data analysis and not the other way around.
Application to Value Investing:
I'm of the opinion that these sort of ideas are not at all foreign to value investing, as long as we're able to differentiate between the following two dogmas-- "be systematic" and "be systematic when it's applicable to be that way." Just because I have a hammer doesn't mean that everything is a nail. Hell, I'll be technical when it's applicable to be so. I wish I believed in some 'one size fits all' truth which I could subscribe to. But I don't. I'm more of the belief that if there is some all-encompassing truth does exist, it's more of a tapestry, stitching together a vast number of distinct skills, skillsets, and rules, the breadth of which is so large that we can only hope to capture a sliver. That makes sense to me.
With that in mind I'd like to turn to how value investors and most corporations gather information. This is a realm which seems to me to be less systematically thought about than it probably should be. Let's say your boss goes up to you and says "Dan, I've found an interesting stock and the ticker symbol is XYZ. By the end of the day I want a complete write-up on this stock's business, competition, industry and prospects for the future" (I'm sure we've all had a variant of a situation like this).
It's things like this which I think merit a little more systematicity than they're given, on average. How can I be more systematic in my information gathering? Well, for one, it might be of value to do a thorough search of as many data providers as you can, sampling them all in an attempt to wean out new quality data sources, and then recording those sources, perhaps in some sort of information architecture on a local intranet where your respected co-workers can offer their insights as well with categories and comments. One could also classify all information along multiple lines, essentially creating multiple information architectures allowing you to flexibly cater to the task you may have at hand. Finally some sort of a search component is a must, so that you have yet another way to swipe through your information database, this time with a user-specified keyword.
This sort of ideation, taking a step back from the actual grunt process which we call 'information gathering', might not be pseudo-code in the technical sense, but it sure starts me along a path to create a system from which I can more easily gather and process information on an ongoing basis. The above description might not be a scalable function, but it seems to be a pretty scalable system because it supports and encourages iterative growth as members of the intranet add their comments and input over time. Finally, the above description might not be a robust function, but it seems to be a robust system because of its ability to classify information along multiple lines with a search capability, allowing the user to gather information in multiple ways over the wide range of goals those users may have.
Of course one doesn't need to do all of these things, but hopefully you can see how being systematic can be a good complement to however else you happen to gather information.
Being systematic can help with many other aspects of our investment activities-- it might be of value to think about this from your own perspective.
Comments are welcome!
Categories: theoretical_intersections
posted by Dan McCarthy at
7:36 PM
0 Comments:
Post a Comment
<< Home