|
TraderBot: The Next Generation [1115.1998] (continued from the last rant..) Buying 30 shares of a volatile, $0.50 stock proved too much of a risk for our last traderbot, especially with a $15 commission on trades. The stock had to jump to $1.67 for our bot to make a $5 profit. Before computers and discount brokerages, this kind of nano-trading would have been impossible and stupid, not just risky. But I think today we can improve our bot's chances. For starters, let's knock the commission down. Assuming we have $2,000 to open an account, we can switch from Web Street (no opening balance required) to Datek Online. In addition to another source of free online quotes, we knock $5 off each trade. Datek commissions are $9.99 for all trades under 5,000 shares. We'll say $10 to keep things simple. Now, our bot starts with $30, finds the $0.50 stock it likes, and buys $20 worth: 40 shares. To end up $5 ahead, it sells when the stock reaches $1.13 (($10 commission + original $30 + $5 profit)/40 shares) Still risky. Let's see what happens if we apply what Robert G. Allen calls "Leverage" in his book, Creating Wealth.. The idea is simple: borrow money to invest, so that you can increase your returns. Let's say our $30 bot goes up to its hive (social insect mode!) and asks for an extra $30 of life energy. It then has $60 cash and $0 equity. If it buys $50 of the $0.50 stock (100 shares) at $10 commission, it can escape with its $5 as soon as the stock reaches $0.75. (($10 + $60 + $5)/100) I can live with that. Great! We've made $5! But in the process, we've risked $60. If one bot dies, other bots will have to make 12 successful $5 runs to replace it. So the odds of that bot dying better be a lot less than 1/13! I think they are. For starters, a bot can't die unless its host stock dies. The company has to go off the market, or reach 0 and stay there. I think that will be a rarity. In any case, we could only look at stocks that have maintained a high volatility and low price for an extended period of time. Probably the more common occurance will be the stock dropping and not going back up. Thankfully, it can't drop too terribly far. Lets pretend our last example (buying 100 shares at $0.50) fizzles and drops to $0.10. There's no reason we actually have to sell. The bot can sit on that stock for weeks until it spikes to $0.75 again. It can check the price a hundred times a day if it needs to, while other bots are profiting from other stocks. Since this isn't a major investment, I can afford to wait. But because it's so volatile, I probably won't have to wait very long. In the worst case, I'm stuck with 100 shares of a worthless stock. But suppose I decided not to wait. Suppose I decided I could live with losing $5, but nothing more. After all, we bought at $0.50 because the volatile stock swung low. If it keeps dropping, let's tell the bot to bail! Our goal now is $55. Uh-oh. That still requires the stock to go up. In fact, the stock HAS to go up, or we lose $20 right from the start. Aren't thought experiments fun? In any case, we've established that the more money we risk, and the lower the comission on trades, the less the stock has to move in order to make a profit. The less money we risk, the easier it is for other bots to replace a bot trapped in a bad deal, but the more the stock has to move. Some questions for the future:
(continued in the next rant..) |