Marc,
I generate ideas and concepts and attempt to build my systems around specific themes. I use all aspects of a trading system - universe, buy/sell rules and ranking and hedge to try to create integrated systems. I’d prefer you don’t tell others how I build systems or classify me having never worked with me.
Not all my systems work over all time frames, but ‘idea and concept’ based approach is the process I follow. A lot of what I call my P123 time is ‘reading time.’
Technical breakouts. Quality-Value. Contrarian. These are all thematic based approaches. I’ve never built a system with you. But, as Quant mentions, I collaborated with Quantonomics on 1 1/2 rankings. We began with research reading and literature review and a very specific focus of what we were trying to do with our system and built the ranking from there.
Having said that, bottom up factor testing is used by quant. departments at JP Morgan, Goldman Sachs and dozens of other firms. These shops often begin with a list of X number (say 100 to 200 or) permutations of ‘common’ factors (like their list of the canned P123 factors). Factors that have some logical basis and have worked well historically. They regularly rerun ‘factor’ reports on their trailing period performance by market cap, sector, liquidity bands, etc. And then send out these 100 plus page reports on trailing period factor results to institutional clients as well as using them to adjust and build systems that try to capture ‘what’s working now.’ This is an integral part of what most quant shops are doing.
Many pension funds are now using ‘factor capture’ portfolio’s in place of higher cost active management. I’ve heard the former head of the Texas pension fund for teachers speak on this. I’ve read the academic materials for b-school classes on this - including books like ‘Simulation and Optimization’ in Finance. Some of them at least. So…it’s a little out-of-bounds to say that factor analysis is not a part of ‘mainstream practice.’
And for the record…The first thing I did on P123 was to attempt to a) clone all the existing AAII screens (I was a member there from 2008 to 2010), b) build ‘basic’ screens around concepts I had read about in books like the ‘Little Book that beats the Market’, ‘The Little Book of Investing’, ‘The Intelligent Investor’, ‘Investments’ by Bode, Kane and Miller, ‘The Pillars of Investing’ by William Bernstein, ‘Global Investing’ by Ibbotson, ‘Unconventional Success’ by David Swenson, ‘Way of the Turtles’ and about a dozen others as well as several hundred or so (mostly) SSRN papers, including a literature review on behavioral finance and papers on best practices in trading system design and all the posts on P123 I could find (including all of your tutorials). I also built out ‘common sense’ systems based on the public ranks and P123 systems and what I’ve seen work in my angel investing and operating a company. Many good ideas failed out of sample (by failed I mean underperformed an equal weighted benchmark in the same cap range).
In fact, the only one of the AII or public systems from 2008 that I’ve tracked that’s continued to work really well for me was Piotroski. My initial P123 versions of Can-Slim, Tiny Titans, basic ‘value mo’ and several others have not faired very well.
And Joseph Piotroski didn’t walk in off the street and come up with his model after 5 or 10 hours of work. First, he earned an undergrad degree in accounting, then an MBA and a PhD and, along the way, he had to read and master all of the academic literature in his field and write countless papers, leading up to a thesis. That’s the academic process that led to him coming up with that ‘one simple idea.’ A decade of learning or so it seems to me. How many ideas did he backtest and discard to get to it? Good ideas take time regardless of process.
As far as this quote from you:
“That sort-of reminds me of the story about proposals in the early 1800s to save money by shutting down the U.S. Patent Office because all the good ideas had already been discovered. It was a bad idea back then (and one that was thankfully never followed.) It’s a bad idea now. And an analogous idea for p123 (don’t bother trying to develop your own ideas; all the ideas you need are already here) doesn’t seem any better.”
Does that really relate to what I said? It’s a silly analogy. I believe medicines will continue to be invented, but I go to a Doctor…I don’t try and learn medicine. If I loved medicine, maybe I would. I believe cars will keep getting better and so will airplane navigation systems…but I buy my cars and plane tickets…I don’t try to build them from scratch or go out and get my pilot’s license. Etc.
We live in a ‘trade based’ economy…where the most limited asset is, frequently, time and focused attention. Is it really wise to think that below average effort will yield above average results in a highly competitive field.
I don’t always put in 20 hours a week. But, I put in 20-40 hours from 2008 to end of 2009. Then 5 to 10 or so hours a week for 3 or so years. Then 30-40 hours a week the second half of last year. And I’m at least of average intelligence and discipline. And had an undergrad business degree with a focus in finance. And had been a COO for 3-4 years.
If MisterChang loves the process and wants to learn this stuff, great. I wish him well and look forward to having him in the community. But, there is a steep learning curve…and he will likely need to invest some significant time to get up to the level of many R2G’s. Whether or not he can beat the top quartile R2G’s over time is very much unknown. The amount and size by which he can beat them is also uncertain.
Best,
Tom
