Investment Strategies / Mechanical Investing
No. of Recommendations: 10
Wouldn't you know it. VL download prices are garbage, again.
Elan
No. of Recommendations: 0
What checks do you do to determin when their data is bad each week?
No. of Recommendations: 4
Since you mentioned it, I did some spot checks. I only just noticed that the files with names May 16 (released May 12) and May 23 (released May 19) and June 6 (released June 2) are also garbage. In addition to the one that you mention, the one named July 4 released June 29.
Can you confirm you're seeing the same?
All prices seem to be bad in all of those weeks, much like the errors last year but unlike the errors earlier this year that had the total return fields wrong.
The total return fields are wrong in (at least) the update files with these names:
Jan 10
Jan 17
Jan 24
Jan 31
Feb 2
Feb 14
Feb 21
Feb 28
Mar 7
Mar 14
The quick check for that TR problem is that either
* the Total Return 13-Week and 26-Week fields are both blank.
* the Total Return 13-Week and 26-Week and Year-to-Date fields are identical
Jim
No. of Recommendations: 1
Since you mentioned it, I did some spot checks. I only just noticed that the files with names May 16 (released May 12) and May 23 (released May 19) and June 6 (released June 2) are also garbage. In addition to the one that you mention, the one named July 4 released June 29.
Can you confirm you're seeing the same?
I could, but it's a hassle and I'm lazy.
I always do a fairly cursory check when I download the file. I look at the Current Price of two or three stocks with my own spreadsheet in which I have the Friday closing prices of stocks that I own. I also glance at the Total Return fields to make sure they are populated and seem reasonable.
Elan
No. of Recommendations: 2
I could, but it's a hassle and I'm lazy.
I get it : )
My spot check went like this:
I have all my files for a given year in a given directory exported as tab delimited .txt files.
I use Unix command "cut -f4,13,20,184-189 *.txt | grep MSFT"
4=ticker, 13=price, 20=P/E, 184-190 are total returns 1 week through 1 year with the little used YTD figure in the middle.
That gives me the stock price, P/E, and a bunch of total return fields for Apple, week by week.
Any file that gives blank total returns will jump out. Second most common is that multiple lookbacks on a given row have identical numbers, which is essentially impossible.
Microsoft is handy because when the prices are wrong, they are wrong by a mile, being two digits instead of three, and the P/E ratio falls to single digits.
There may be other fields which are garbage from time to time, of course. My list of bad weeks up thread shows the ones that (by the process above) are known bad...the rest might be good, or might have other types of errors that I haven't discovered.
I'm not actually using the data except for statistical analysis these days, having shut down my MI portfolio in March. I keep it up more just by habit, and because I paid for multiple years of subscription in advance. I'm a data pack rat.
Jim
No. of Recommendations: 16
I'm not actually using the data except for statistical analysis these days, having shut down my MI portfolio in March. I keep it up more just by habit, and because I paid for multiple years of subscription in advance. I'm a data pack rat.
I hear you. I'm not in exactly the same situation, living in the U.S. with my financial life anchored here for better or worse. I've been very gradually divesting from individual stocks because of the decidedly dark future I see for this country - for the duration of this presidency and possibly much longer. I've also seen no advantage for mechanical stock picking, compared to index investing, for a few years now. At my age, the thrill of trying to beat the market by 2-3% per year, and failing at it, is no longer there. So if and when the market is ever attractively priced again, I expect to redeploy my cash into index investments. Like you, I keep doing the weekly rankings altruistically by force of habit, and because I've had a monthly automatically renewing VL subscription at half price for 25 years now. And like you, I'm a data pack rat and I can't let go. :-)
Elan
No. of Recommendations: 8
the thrill of trying to beat the market by 2-3% per year, and failing at it, is no longer there.
Again, I hear you.
The more times I read that, the funnier it is : )
Certainly the biggest insight I had many years ago, when I used a lot of different investment strategies, is: stop doing the ones that clearly don't work.
Since that's too difficult a goal, I revised it: at least don't allocate a lot of money to them!
I've made money with MI, but it hasn't been my best success for a long time. My annualized rate of return on daily dollars of capital at risk in Berkshire and Berkshire derivatives is something like 26%/year for just under 25 years. Partly from buying low and selling high (I'm slightly net short at the moment, kind of a short strangle), and partly from the leverage embedded in calls.
Jim
No. of Recommendations: 0
Elan wrote: Wouldn't you know it. VL download prices are garbage, again.
What I do is compare the prices of two of the largest stocks in their data to their full history. In this case, IBM and AAPL. The prices of the most recent file are from 4/25/2018. So yes, they are garbage for our purposes.
No. of Recommendations: 0
I've also seen no advantage for mechanical stock picking, compared to index investing, for a few years now.
I ran into someone who uses O'Shaughnessey's strategies and has returned 20% CAGR for 25 years.
I started doing the same, and have done well so far, but it is much too soon to see if they will work.
No. of Recommendations: 0
In this case, IBM and AAPL.
IBM? Not that it matters much in this context, but it's ranked 32 by mcap in the VL database. I like to check a couple of stocks that have done large splits, like 10 for 1, in the last couple of years, because the price discrepancy pops right out.
Elan
No. of Recommendations: 0
I ran into someone who uses O'Shaughnessey's strategies and has returned 20% CAGR for 25 years.
I started doing the same, and have done well so far, but it is much too soon to see if they will work.
Too soon? What Works on Wall Street was published in 1997. Everything since then is post-discovery.
Elan
No. of Recommendations: 0
I've also seen no advantage for mechanical stock picking, compared to index investing, for a few years now. At my age, the thrill of trying to beat the market by 2-3% per year, and failing at it, is no longer there.
Have you also stopped your 6/3 option portfolio? If yes, how were results until you stopped? If no, how were result the last 3 years?
Let me say, you should be in the Guinness Book of world records for being able to follow this strategy pretty much mechanically for what, almost 25 years?
No. of Recommendations: 0
Mark19,
Would be interested to know which strategies in particular you decided to give a go recently.
