It's time again for me to do all sorts of cool things and I was wondering if any of you would be interested in the following project.

If you are interested in seeing this project done, please recommend this post so that I can see if this type of endeavour would be worth the time for me to compile.

The project:

To take my Charity Port v1.1 (see below) and to generate a 1986-2000 spreadsheet showing the daily fluctuations of every stock, every screen and the whole 21 position blend for 14+ years. This would allow you to take a cross section look week by week or month by month at a fairly decent multi-screen blend to see how different screen complement each other in different time slices.

My slightly revised charity portfolio is the following:

Blend:
20% annual (Spark 1-5)
20% annual (Key100 1-4)
20% semiannual (PEG 1-4)
20% monthly (RS26/RS13 overlap 1-25:4)
20% monthly (PEGRSW/PEG13 overlap 1-10:4 pad both)

The changes from the last Charity Portfolio are:

1) using Key 100 instead of Key EPS because of slightly better returns (thanks Ganymede). This is probably statistically irrelevant but I'm going with it anyway because the Spark screen picks the supercaps and the RS screen picks the biggest momentum stocks. In my opinion, Key 100 is a better choice for this particular blend because it is more likely to shift back and forth along the Spark-RS spectrum to pick the more favourable stocks better than Key EPS due to the fact that it can pick up smaller large cap stocks and is not limited to the same super cap universe as is Spark and Key EPS (which picks only from the top 30 stocks by market cap). In short, it has more power to 'vote' by more heavily weighing Spark-type stocks or RS-type stocks depending on whether in a particular year, MI is highlighting the super caps or the smaller large caps.

2) using the Spark 5 instead of Spark 4 because of the 1997 DELL effect (1997 Spark 4: -20%, 1997 Spark 5: +25%). Dell was the #5 Spark stock in 1997.

I feel that the Spark 5 ('86-98 CAGR/GSD: 37 / 30) is more representative of the Spark screen than the Spark 4 ('86-98 CAGR/GSD: 36 / 39). These differences are almost entirely due to the fact that one good stock (#5, DELL) in the Spark 5 carried the 1997 return.

You can call it curve fitting or looking through rose coloured glasses, but given that the last sort is a Market Cap sort and that the 3 stock version returns only 34%, Spark is a screen that is better served with more stocks than less.

Anyhow, My idea is to generate a massive 25-30 column by 3400 row spreadsheet that shows daily changes in this portfolio. It will take me a couple of weeks to put it together.

Recommend this post if you think it is a good idea.

Cheers,

Peter