Thursday 14 February 2013

Shares I have bought

How I started 

Since I started to take the stock market seriously I have always been a stock picker. I have attempted to find shares that, for one reason or another, are undervalued by the market. They seem poised for reappraisal. I chose this approach because it offers the chance to significantly outperform the market without taking undue risks.

I find my shares by back testing. I pick a theory of what criteria should predict which shares will outperform and then test to see if it works.

When I started there were few resources to help me do this and it took a lot of painstaking and boring work to find a reliable method. The theory behind my stock picking filters came from an analysis of 100 years of stock market data published in a book called What Works on Wall Street by J.P. O'Shaughnessy. I back tested my ideas using data provided on monthly CDs published by an organisation called REFs designed by Jim Slater. It still exists and works the same way as it did when I used it 13 years ago. From this I created my own filter which saw me through my first 12 years trading on the UK market. Since that time buying and selling shares has been my sole source of income. I started with what these days is a rather paltry pot of cash but despite that I have lived well.

What I do now

I now have an alternative. In the US a company called Vector Vest offers a fabulous data base which allows the back testing of all manner of ideas. It makes the development of theoretical filters difficult because of its quirky presentation of fundamental data. But it more than makes up for this because back testing is very easy.

In my last post I explained that I had given myself a very well aimed kick in the pants about the beginning of December. I then jumped on board the Vector Vest train. Its site provides a range of almost 300 selection filters for picking shares. They are available for the UK, the USA and a number of other countries. I patiently worked my way through them all to discover which were the best predictors of future share price movement. I started by making a short lit of those which have worked well recently and then tested the members of that short list for strong performance month in and month out.

I did this work for the US and the UK and found the best performing filters for each market. Over the next two months I bought shares based on these criteria and did well, especially in the US where my overall return is up to about 20%. Since my results in the UK struggled to achieve a quarter of that performance I have now switched almost all of my holdings to the US.

I am using a range of selection criteria and they have changed as I have refined my methods of back testing.  Currently I look for four main groups of shares:

  • shares that are undervalued when measured by their forecast earnings and earnings growth. They have had consistent earnings growth in the past, and have seen good share price appreciation which suggests the market is beginning to understanding that the price needs to be reassessed. In addition the share will have shown a significant increase in earnings in the past quarter
  • shares that are increasing in price rapidly, have a rapid earnings growth rate and they moved up in price on the day before purchase on higher than average volume
  • shares in the S&P 500 which have had their prices beaten down to very low levels and might now bounce back
  • Shares that have recently improved the overall quality of their fundamental performance but are low in absolute price
Some of the returns I have achieved have been spectacular. I show below the shares I have bought since January 2. The first group Ihave sold to realize either profits or losses. The return has averaged 1.3% per week. I have mostly ditched under-performing shares so returns are not as good as shares I continue to hold. 

Here is a list (In order to find out more about the shares I have in my portfolio go to Yahoo Finance or Google Finance and search for the ticker symbol I have given):


date of purchase Share % growth per week Date of sale
22-Jan GTN 17.7% 11-Feb
02-Jan BBY 8.6% 25-Jan
02-Jan BBY 5.6% 11-Feb
02-Jan PBI 5.3% 07-Feb
18-Jan SNFCA 3.7% 29-Jan
02-Jan CHK 3.2% 28-Jan
02-Jan AMD 1.3% 24-Jan
03-Jan SPMD 1.0% 31-Jan
02-Jan KSS 0.9% 28-Jan
03-Jan DEXO 0.8% 31-Jan
18-Jan PGTI 0.0% 11-Feb
11-Jan SCMR -0.2% 01-Feb
22-Jan GTIM -0.5% 11-Feb
18-Jan HW -1.4% 29-Jan
16-Jan SKUL -2.0% 31-Jan
10-Jan IQNT -2.3% 31-Jan
18-Jan MHO -3.1% 31-Jan
10-Jan MLNX -3.8% 29-Jan
22-Jan SSYS -10.4% 29-Jan
Average 1.3%
The shares which I am still holding have returned an average of 2.3%

date of purchase Share % growth per week
01-Feb CPSS 15.2%
11-Feb CIMT 14.1%
05-Feb STRZA 11.1%
03-Jan WFR 6.2%
02-Jan BSX 5.7%
04-Feb CPSS 4.3%
02-Jan MU 4.1%
11-Feb VHS 4.1%
23-Jan CMLS 4.0%
04-Feb MEI 3.8%
22-Jan PBI 3.7%
30-Jan GRPN 3.6%
31-Jan FBP 3.3%
30-Jan FBP 2.7%
04-Feb EOX 2.6%
22-Jan CHK 2.6%
22-Jan KSS 2.2%
22-Jan JCP 2.1%
11-Jan HBAN 2.0%
22-Jan BBY 1.5%
25-Jan AAPL 1.5%
05-Feb MXWL 1.4%
16-Jan ALLT 1.3%
30-Jan NMR 1.0%
01-Feb TA 0.9%
23-Jan VIRC 0.5%
23-Jan GBNK 0.4%
01-Feb SWFT 0.3%
03-Jan SOMX 0.2%
02-Jan FTR 0.2%
30-Jan MNST 0.0%
01-Feb DHI -0.1%
22-Jan NEM -0.2%
28-Jan LSI -0.4%
12-Feb TVL -0.9%
25-Jan BTU -0.9%
04-Feb HCA -1.3%
25-Jan COH -1.7%
31-Jan SPNS -1.7%
05-Feb SBGI -1.7%
04-Feb OMX -2.0%
Average 2.3%


Just think what this means. 1.3% per week carried on through the year would achieve over 60% return. 2.3% would achieve well over 100%. 

It is most unlikely that good times will continue for the whole year. But assuming that I make the best of the good times I should still do well. 

In the first 6 weeks of the year I have made over 18% (That includes contributions from some shares bought last year.)








No comments: