InvestorsHub Logo

OldAIMGuy

10/07/10 12:19 PM

#1318 RE: neko #1317

Hi Neko, Re: Reversion to mean............

AIM does expect the pendulum to swing in both directions. It also has built into it a "positive feedback loop" that tends to move the bracket's swing slightly upward with each cycle. This can help to compensate for a slight upward slope to the overall market price.

Automatic Investor, I believe, allows you to adjust that feedback loop, making the machine either less or more aggressive relative to one's interpretation of the future. For instance, if you felt the market was going to be range bound for many years, you would want to reduce the feedback to near zero. That would keep the buying and selling parameters nearly identical cycle after cycle during the range bound period. On the other hand, if you felt the market was going to grow at maybe 8% per year on average, then leaving the feedback in place or modifying it slightly might make the program more effective in that environment.

Of course all this depends upon one's ability to see far into the future.

Best regards, Tom

aptus

10/07/10 11:58 PM

#1319 RE: neko #1317

Hi Neko,

PI 3.0 should run fine on XP Pro and XP Home. It won't run on XP Media center or below since it uses the .NET Framework 4.0.

Your hardware should be okay, as long as it has the minimum recommended specs: 1GHz CPU, 512MB RAM, 600MB free disk space (if you don't have .NET 4.0). If you already have .NET Framework 4.0 on your machine then you'll need less than 10MB of free space.

Yes, many systems rely on reversion to the mean, but the implementations differ.

PI's Value Trading Algorithm actually uses a modified version of that concept by looking at groups of highly-correlated equities and trading based on their relative values -- rather than simply relying on a strict reversion to the mean. Therefore stocks may never actually revert to their individual means but you would end up trading around the group's (or a sub-group of that group's) mean.

Value investing and AIM loosely follow the reversion to the mean concept too, but, again, the implementations are different.

I hope that helps. Let me know if you have any other questions.