A follow up regarding my aforementioned "leash"

A friend of mine today wrote me an e-mail regarding my recent post on the importance of keeping your trades on a leash. He asked a pretty good question, which sort of made me feel stupid for not mentioning it in the post:

"So, how DO you keep your trades on a leash?"

Excellent question. Here's what I do, and by no means should you take it as gospel. It's a weird little system that I've worked out over several years of trial and error. (By the way, if anyone out there has any suggestions for improving this method, I'd love to hear them.) Here goes:

• Before entering any position I determine how much money I'm going to allocate. Because of the relatively low margin, compared to a futures contract's true value, I always try to overshoot the margin requirement by a substantial amount. My target is to allocate at least enough money so that I'm operating at an effective leverage of 75% or less. The volatility of the market, in my opinion, makes being as leveraged as possible horrendously risky.
• Immediately upon entering a new position, I determine my maximum loss. For me, my personal rule, is 20%. That is to say, I refuse to lose more than 20% on any given trade. I do some math and determine at what price I'll have lost 20% of my allocated funds, and set that as a stop with my broker (a buy stop if I'm short, a sell stop if I'm long). As a personal rule, I will NEVER move this stop away from my position (that is to say a 20% loss is truly my personal maximum).
• I also determine what my exit point is; the price I want to exit this contract assuming it moves in my favor. I then set this as a limit order with my broker. (NB: I should say here that one of the best pieces of investment advice I ever received has caused me to be far more modest with my exit points than many traders would likely expect.*)
• Throughout the day, I alter my stop order if the price has moved in my favor.
• Unless I have new research to strongly suggest otherwise, I do not alter my limit price whatsoever.

Here's an example with a fictional commodity. Suppose I want to go long on March Widgets, which have a contract size of 5000 boxes at a price of $1.00 / box. My broker requires an initial margin of $1000 to enter into this contract. I decide that I will be allocating $2000 to the trade, for an effective leverage of 60% (I'm only buying one contract). To lose 20% of my $2000 investment, the total value of the contract would have to decrease by $400, which means a decrease in price of 8¢ ( [$1.00 * 5000] - [$.92 * 5000] = $400 ) Thus, I would set a sell stop at $.92 / box. From my research, I expect widgets to go up as high as $1.10 / box. I then set a sell limit at $1.10. Throughout the day I am constantly checking the current price of Widgets. If the price ever goes up, I move my sell stop up accordingly (whatever amount would result in a 20% loss from wherever the price currently is). This continues until I either get stopped out, or until Widgets hit my limit price, whichever comes first.

Some of you may have noticed that what I'm describing regarding moving the stop order is technically a "trailing stop" and many brokerages allow you to do this automatically (the stop order moves to some percent above or below whatever the price is currently). Personally, though my brokerage allows it, I do not use this sort of automation. The reason is that, as I've discussed before, the market may make a weird hiccup out of nowhere and then go back to normal trading. Such a hiccup might, if your trailing stop was automated, move your stop up or down, then immediately stop you out at a loss before going back to whatever the price was before. However, since I don't move my limit, if that hiccup happens to trigger my limit order, I'll happily take the money, thank you very much.

*That advice was: "Don't worry about trying to be exactly right. If you're basically right most of the time, that's much better than being exactly right occasionally."