Maybe a Pine Tree

The Start of a Wonderful Thing

Rating: 0 not yet rated(0 votes)    Vote:

You are logged in as | Logout

Welcome.
I thought I would design a virtual portfolio that takes advantage of the virtual portfolios that I have created before. If I were to manage a fund of funds, and choose stocks from a variety of funds and assign sector allocations accordingly, I thought it would be an interesting exercise at doing exactly that. I also realized that the process might not win the competition, but it would be an excellent process of discovery. So, I made an algorithmn that viewed the virtual funds I had at Marketocracy and it did a few things with those funds:

1. Assign a score to each fund based on longevity/skill, consistency, and return/risk. With the tools available on their website, I figured that I would have enough information available to decide which (and how much info) from each of my funds I should use.

That process gave me a list of each fund, a score and some idea of how much some of the funds should be used.

2. Decide on sector allocations. Which sectors should I over/under weight?

By analyzing sector allocations and combining scores, I was able to piece together a quant method of which sectors I would under/over weight. At this moment, Financials, Telecom and Materials would be overweighted; whereas, Information Technology, Staples and Discretionary would be underweighted.

3. Decide which stocks will be chosen from the portfolios.

Some may have thought that 'sector portfolios' would get the first choice, but I designed the procedure in such a manner that stocks wouldn't necessarily come from it's 'sector portfolio'. Sometimes a stock was chosen from such sector portfolios, but not always.

4. Put the stocks in the portfolio.

You might ask, what's the sell discipline? In step 3, it decided which stocks would make it in. If a stock was excluded, it wasn't immediately kicked out. That would depend on how far down that list it went.

Some may wonder, how did I choose the stocks in the first place?

Well, it's basically quant work taking academic research, using some valuation, followed by a bit of catalytic analysis.
In general I like: low PE, low price/book, insider buying, low short interest, low accruals and positive earnings momentum.

I'll borrow this line that I've written before:

"Quant screen-valuation approach utilized. Stocks screened with parameters chosen from academic research -- relating to valuation, earnings quality, momentum, smart money, and external financing. For the valuation I have used 'eVal' by Richard Sloan and Russell Lundholm. Another great valuation product out in the market is www.afgview.com. Want some superb screening software -- use Factset. Look at www.earningstorpedo.com for some potential poor (and good) stocks. Look at www.alphaseeker.com for some academic research. If you gain money, good for you... if you lose money, your fault."

I hope during this process to enlighten you with some of my stock ideas as well, but do your own due diligence.

I'm doing this for fun and hopefully I'll learn a few things along the way.

Comments: 0