Canadian Money Forum banner

1 - 20 of 29 Posts

·
Registered
Joined
·
246 Posts
Discussion Starter #1
Let me re-iterate the title again. It is rules based investing and not trading.

I just finished reading "What Works on Wall Street". It talks about buying a portfolio based on a rule without second guessing it. WWWS's most of the strategies consist of 50 stocks, but it has applied some of those strategies for 10 and 25 stocks too.

I am seriously considering putting 10K towards 10 stocks based on a rule or 12.5K towards 25. One other important point is unlike trading, you are not in or out. With this, you are in the portfolio all the time, its just the selection is completely rules based and changes annually. So it is irrelevant that market is at all time high.

Has anyone been investing in a portfolio based on a preset rule. Considered it? Thoughts/Comments?
 

·
Registered
Joined
·
246 Posts
Discussion Starter #3
25 stocks sounds like a lot to me... let alone 50! What's the "rules" of the strategy as described in the book?
One of the first rules he mentioned was Dogs of the Dow. I dont have the book now (had borrowed from the library). All the rules were applied on last 70ish years of data. The most successful were counter-intuitive for e.g I remember one which was filter all small cp stocks with PS < 1. Rank these by their 1 yr price chg and buy the top 10. Although it seems against common sense to buy stocks that have gained the most, the data shows otherwise.

The book has more than 25-30 of these strategies, all mechanical.
 

·
Banned
Joined
·
3,080 Posts
Has anyone been investing in a portfolio based on a preset rule. Considered it? Thoughts/Comments?
Sure, lots of people invest this way. Run a preset stock screen once a year, or whatever frequency your strategy calls for. Buy all stocks in the screen in equal weight. Hold for a year. Rinse and repeat. Best done in a tax-sheltered account.

If you want to get full-nerd credentials:
* call yourself a quant
* describe your investment style as quantitative investing

The tricky part is getting access to a good quality stock screener. Free stock screeners suffer from poor data quality. Professional stock screeners like Bloomberg are hugely expensive. You have to find a sweet spot: an affordable stock screener that returns reliable data. Search the forum for previous stock screener discussions.
 

·
Banned
Joined
·
3,080 Posts
AAII web site:

James O'Shaughnessy
“What Works”: Key New Findings on Stock Selection

William Bentley from Kentucky posted this comment to the article:

Early last year, I attempted to put O'Shaun's very good work to practice, adding to it some of the work done by another researcher. I used Stock Investor Pro to generate data, first filtering to meet the authors' market cap filters. I downloaded the data into Excel. Very few of the exact metrics were provided; some of those could calculated in Excel, in many cases data was missing, requiring that I try to collect the raw data from other sources. After sorting, I then computed composite metrics using multiple factors. In summary, you can't screen directly using SIP, and the amount of work required to do on your own is quite substantial. Add to that the inability to effectively back test, and I shelved this concept. Too bad, I think the author is on to something.
I think that describes your typical experience with a non-professional screener. DIY quantitative investing is hard work.
 

·
Registered
Joined
·
658 Posts
I've read WWoWS as well and like what he's "selling". Practically, it sounds like it should work and the author backs it up with a lot of stats and tables. I've not pulled the trigger on any of his models, but did consider it.

Another "rules based investing" is the Magic Formula Investing (MFI). Designed by Joel Greenblatt the book is "The little book that (still) beats the market", it's based on buying cheap stocks and make good use of their capital. I'd suggest reading it too. Simple, short, and pretty easy to understand the concepts he's proposing.

This is my starting point for my USD portion of my portfolio. I use the screener on www.magicformulainvesting.com to get a list of MFI stocks, and buy 6 of them every 3 months (selling the previous 6 purchased 1 year prior).
 

·
Registered
Joined
·
2,108 Posts
Hi, Amitidi

I have a rule & that is do my own thing be true to that which my eyes tell me is true & not to se the world through the eyes of others.

Like almost all traders/investors you are breaking the rule of not doing your own thing.

From experience we all have different knowledge & understanding. I have differnt knowledge & understanding in the market then you as you have differnt knowledge & understanding then I do. We both have differnt personalities Emotions are automated value responses issuing from the subconciuos within a context of an individuals knowledge is for him or against. Every emotion has a kenetic element as an imputus to engage in action related to the particular emotion involved. If your understanding & knowledge is saying buy & the other persons method is saying sell. Your emotions will cause you to protect that which you value most highly.
Emotion will over ride trying to follow someone elses method if your personalities & understanding & knowledge is differnt.

Rule 101 in investing is develope your own method that gives you an edge that is based on your knowledge, understanding & fits your personality.

The number of independent thinkers is a lot, lot, lot, lot smaller then most realize i.e Galileo & the rest of masses thinking the world was flat. Ever watch a flock of birds rarely does one bird stays when the rest flies away when something spooks the flock.
 

·
Registered
Joined
·
499 Posts
I've done some of my own research for this method before. I find this topic very interesting and could possibly generate superior returns. Like what Goldstone said, it's hard to find a screener with quality, up to date information, which is essential for quantitative investing. Recent information, of course, is always available through a variety of sources, but it can be time consuming to extract quarterly or even annual information from a variety of companies' financial statements. I have used variations of quantitative methods before (hybrid methods, if you will) where I do a preliminary screen based on a few metrics, using up to date information. After, I pick the top X number of companies and use a few qualitative metrics to eliminate companies from the mix (i.e., companies that I feel have a competitive disadvantage in their industry, management issues, or if my selection is too concentrated in a particular industry). I like this method for my investment objectives, but of course it is not fool-proof.

One word of caution when doing quantitative investing is to look at your list of results a bit beyond your desired number of holdings. By that I mean if you want to hold 10 stocks, look at your top 15 or 20 results. A current holding may end up 12th on your screener and could likely end up in your top 10 on the next screener - this saves you transaction fees. The TSX isn't a very big index - you will find a lot of repeat stocks on your screener, unless you include microcaps.
 

·
Registered
Joined
·
4,008 Posts
Let me re-iterate the title again. It is rules based investing and not trading.

I just finished reading "What Works on Wall Street". It talks about buying a portfolio based on a rule without second guessing it. WWWS's most of the strategies consist of 50 stocks, but it has applied some of those strategies for 10 and 25 stocks too.

I am seriously considering putting 10K towards 10 stocks based on a rule or 12.5K towards 25. One other important point is unlike trading, you are not in or out. With this, you are in the portfolio all the time, its just the selection is completely rules based and changes annually. So it is irrelevant that market is at all time high.

Has anyone been investing in a portfolio based on a preset rule. Considered it? Thoughts/Comments?
Matters what the rules are.
I do think $500 positions are too small.

I also question why you quoted "investing".
 

·
Registered
Joined
·
3,247 Posts
The problem with mechanical investing (rules based or whatever name one wants to use) is the guy in charge of it. Namely you. Whatever mechanical strategy you look at you will undoubtedly want to see some backtested data before you use it.

Two problems arise with this. The first is a concept known as data mining. That is the situation where if one sits at a computer long enough and sifts though enough data, one will inevitably find relationships that have worked better and better then what you are currently using. Not only does this cause one to keep tweeking the strategy, but also increases the chances that the strategy you are using was nothing more then a fluke. The 1 in a million flukes are out there. With a fast computer and time on your hands, it won't take you too long to find them. They look a lot like the ones that are not flukes ... with just a little higher backtested rate of return.

The 2nd problem with backtesting is that the strategies where their results have shown good "recent" performance, will be the ones that look like they have the best long term performance, as well (this is no different then mutual funds where recent performance can significantly boost long term numbers). When you review almost any mechanical strategy you will undoubtedly find many periods of time where the strategy did not outperform, just buying an index or some other strategy. Since it outperformed in the long term, many investors tend to ignore this and want to use it in the hope of obtaining those long term results. The problem is, when the bad year or bad years come immediately. If you get underperformance out of your new strategy in the first year and next for example, it will be very unlikely that you will stick with this strategy for even that length of time. 2 years is about 500 stock market trading days where you get to see a simple index kicking the behind of what you are doing. If you breathe and pulse blood like any other human, you will undoubtedly tweek this strategy or abandon it completely, long before it starts to work. Since most of the strategies that will look good to you will have some recent outperformance, the likelihood of it underforming in the near term becomes much higher...leading to its abandonment or tweek as previously discussed.

So, long story short. Mechanical strategies are designed to take the human element out of the investing process, however, it is the human element of selecting and continuing with the mechanical strategies, that will keep investors from ever benefiting from them.

IMHO you should simply buy an index fund and forget about it. You will do much better...as long as you don't keep trying to find a better index funds. We humans never stop messing things up, do we?
 

·
Banned
Joined
·
3,080 Posts
^^^ OptsyEagle made some razor sharp observations.

Confession time:

I tried an experiment last year. Real dollars. I sold my entire VTI position in Dec 2012. I replaced it with an 18 stock screened portfolio. I used a free screen published by a great investor I trust. The screen has a great track record. It's a purely quantitative strategy. No human judgement is involved. The guy uses Bloomberg to run the screen; data quality is not a concern.

The portfolio performed beautifully out of the gate. 3 months later, at the end of Feb 2013, my portfolio was 9% ahead of VTI. 9% outperformance in 3 months. Brilliant!!

The month of March wasn't as kind. The portfolio became extremely volatile. 3% up days followed by 3% down days. By the end of March, the portfolio gave back 6% of outperformance. It was only 3% ahead of VTI. I chickened out and sold. Locked in the 3% bonus and went back to the safety of VTI.

And that's too bad. As of this Friday, the screened portfolio is up 36% since inception. VTI is up 26%.

OptsyEagle got it exactly right: "Mechanical strategies are designed to take the human element out of the investing process, however, it is the human element of selecting and continuing with the mechanical strategies, that will keep investors from ever benefiting from them."
 

·
Registered
Joined
·
246 Posts
Discussion Starter #14
Matters what the rules are.
I do think $500 positions are too small.

I also question why you quoted "investing".
Because of 2 things -
1) Time horizon should be at least quarter. Ideally annual so that I dont have to give daily attention.
2) I will be always IN the stocks, no matter how high or low the market is.

Its easy for the discussion to turn into a trading method which this is not.
 

·
Registered
Joined
·
246 Posts
Discussion Starter #15
From what I read from the replies above, there are multiple things to consider (roadblocks if I may call that)


1) Choosing a good screener
2) Backtesting the screener
These 2 need a lot of time and some money for DIY investors.


Trying to play devil's advocate here. How about choosing one of WWoWS? or some other book where the analysis and backtesting ha been done. If you like one of them, then good . If not, then spend time and money.



3) Discipline - One of the cliches of stock market. "GoldStone" confession was an interesting read...
Its easy to say that I would forget the portfolio for rest of the year, but I know a lot of us (including me) will track it everyday and might consider selling it if it is 40-50% up.


4) # of stocks
5-10 is too less for these kind of portfolios and with more than 20 and even $1K per stock, your overall portfolio becomes $20K. Can you put $20K towards this?

5) Also to consider is a hybrid approach like "Canadian" pointed out.

I will read the "The Little Book that..." and do some more thinking. But thanks a lot for all your replies till now.
 

·
Registered
Joined
·
211 Posts
20k towards 10 stocks is a LOT more reasonable than your other two scenarios of 10k in 10 or 12.5k in 25.

Consider the trade fees you will incur. Without a certain minimum in your brokerage account, you may not qualify for the lowest trading fee.

Even at $10/trade, you spend $100 and $250 to set up the 10 and 25 stock portfolios in your initial two examples. That represents 1% and 2% of each portfolio's value, and you then need to earn slightly more than that just to get back to square one. At least with 20k in 10 stocks, you're now down to 0.5% cost. If you are revisiting the portfolio each year, then you may be incurring two more fees each time you sell and buy something else.

If your trade fees are higher than $10, then these numbers are even less favourable.

Especially when the portfolio is small, you may want to focus on keeping expenses as low as possible until there is more $$$ to invest. The larger your portfolio, the less of a drag trading fees will be on your returns.

With 10-20k, I would be considering index mutual funds, like TD's e-series of funds. Otherwise, I would look at the less complex Canadian Couch Potato portfolios and invest in no more than half a dozen index ETFs.
 

·
Registered
Joined
·
658 Posts
I chickened out and sold. Locked in the 3% bonus and went back to the safety of VTI.
This is the bigger problem with mechanical strategies. The ability to stick with it, even when it under-performs for long periods. Is there something wrong with the strategy (IE chicken out and sell it) or is it a small slump? I think with most mechanical strategies, your time horizon has to be years, not days or weeks or even months.

Who cares if after 3 or even 6 months you're under water. If the strategy is strong, then stick with it.
 

·
Registered
Joined
·
131 Posts
This is not worth it. It's just maximizing past returns using more and more complex formulas. It has no predictive power for the future. It's the equivalent of Technical Analysis but using some economic data as input, instead of only price data. Useless. Some strategies will beat the market (some by a lot), some will lose, on average will just match the market on the long run.
Good for brokers and exchanges though, as it generates fees.
 
1 - 20 of 29 Posts
Top