Search found 4 matches

by MorganSys
Mon Sep 20, 2004 4:16 pm
Forum: Testing and Simulation
Topic: Robust Optimization
Replies: 26
Views: 20067

What I would need to know:

-How many trades
-Distribution of returns within each bar
-Temporal stability of the distribution within each bar
-Distribution of returns in underlying asset
-Price time series of underlying asset (if only one)
by MorganSys
Tue Sep 14, 2004 10:36 am
Forum: Stocks
Topic: Yahoo finance data
Replies: 6
Views: 7590

there are some data errors, but not many. You can scan for huge % moves and remove the ones that are not real.

Also, I think that the database is subject to survivorship bias.
by MorganSys
Tue Aug 10, 2004 8:34 pm
Forum: Testing and Simulation
Topic: Ideal Timing for Large Lump Sum Investment
Replies: 8
Views: 5697

be careful that there is no selection process towards controlled drawdowns in your optimization routine.
by MorganSys
Wed May 12, 2004 4:19 am
Forum: Money Management
Topic: Ryan Jones' Money Management
Replies: 33
Views: 37752

Criticisms: 1. Equal number of contracts: The proportions of different types of contracts that is appropriate to trade may be found through optimization (not strictly by solving from historic data, because this over allocates to historically good curves, and underallocates to poor curves-but by doin...