A.I.M. Users Bulletin Board

[SFSecurity]

You could use trailing stops but you would need to increase the numer of shares to be traded as the price increased or decreased.

What I'm trying via back testing is a GTC order at about 6% above/below the hold range. If I recall correctly AIM will give a buy/sell signal when the move is about 5% above/below the first market order it gives. Is my memory correct?

If so, then using ~6% would avoid both the first and second buy/sell signal, and get you to act on the third signal. This will not avoid a run of signals that might happen in a strong bull or bear run.

How this might work in practice, assuming that we take a guess as to where the approximate market top is going to be, we calculate where the first buy signal would be given as the market tanks, then take the "typical" draw down, say 25% - couldn't a table for this in a quick search - and select a buy point somewhere between the two as the place to buy in but more than 6% below where you guess the first signal will happen. No, it won't be at the bottom, but it should be close enough that you need not sit at your computer hours every day to catch the actual bottom so that you can get most of the next run up.

The big problem is if the market does not tank enough to trigger any buys at all. Given this type of issue, you can see why it took Lichello reams of paper and too many hours to count to come up with AIM.

As to the number of shares to sell, it is easy to get close enough by selecting a number less than you hold. For buying select a number of shares based on your cash reserves. BTW, this is one major advantage of having a single actual cash pool for all your positions. It is unlikely you'll run out of cash this way. Keeping track of each position's cash position separately is still wise, it seems to me.

I hope this is as clear as mud.