Psychology in the Bacteria Market [1115.1998]

(continued from the last rant..)

"The stock market is not your mother." Meaning: don't expect it to take care of you.

I spent most of last night creating a ledger object to record transactions, and started work on a generic "Broker" object to handle buying and selling. (ftp://manifestation.com/monty)

That is the easy part. The hard part is beating the market. I went to the bookstore in search of information about Day Trading. There was some. It wasn't comforting. The quote above was one author's argument against automated trading systems.

But I think I've got a new approach. For one thing, I'm not an analyst. I don't know how to value a stock. I'm interested in analysis for my own manual investing, but I know there's no way I'll be writing AI strong enough to analyze the market and make good stock decisions based on that analysis.

While conventional investing deals with bulls and bears in all their complexity, I'm dealing with bacteria. Simple, stupid, and most importantly, small.

I visualize the bacteria market as highly chaotic: where cheap, volatile stocks fluxuate by pennies throughout the day. What makes the bacteria market different from small cap and day trading (I hope) is that we invest only small amounts, seek only a few dollars profit on an investment, and work with many stocks simultaneously.

Now, to automate this, we might take a lifelike approach. Our little traderbot only has so much life energy - so much cash. Say it has $30 cash and $0 equity. At my brokerage (currently webstreet), each trade is about $15. So the bot begins with enough life energy for one buy and one sell, and no room for mistakes!

Well, what's a bot to do? If it wants me to keep running it, it had better make me some money. Sitting still is suicide. (As an alternate motivation along the tiny organism theme, if it wants to grow big enough to split into two copies of itself, it has to make money!)

So, first it scopes out the environment. The Stockpoint Investing Tools look like a fun, free source for initial research. Our bot might decide to look for some highly volatile stocks under $1.

Now, it becomes a predator. It sits and waits for one of the target stocks to dip low enough in price that it becomes a good buy. Say that stock XYZ drops to $0.50. Our bot latches on, buying $15 worth.

After a $15 buy, the bot is most vulnerable, because it had to pay the additional $15 commission. The bot now has $0 cash, and $15 equity (30 shares of XYZ). Selling is suicide, because bailing on the stock costs another $15 commission, leaving it with $0 cash, $0 equity. Unable to move, the bot switches to parasite mode. It must wait until the stock rises high enough to allow the bot's escape.

How high? To make all of this worth it, let's say the bot wants to make $5. That means it has to sell off its equity for $50 ($15 to cover the commission, $30 to return to its old level of energy, plus the $5 goal). The stock has to jump to $50/30, or $1.67.

Interesting! When I first calculated that, I had the target price at $2.17 because I mistakenly counted the commission twice. The "buy" commission is covered in the "$30 to return to its old level of energy"). In any case, even this tiny leap is too much of a risk, so I'm going to pretend such returns (the kind the bot filtered for to begin with) are impossible, and stick with my original ending:

Ouch. Rest in peace, little bot. Maybe jumps like that happen all the time, but I doubt it. Let's have a moment of silence for our friend, and in the next rant, we can do some tweaking.

(continued in the next rant..)