Quote from Trend Following:
Those documents are but one example of "people" as the real issue, not debates about "changing markets." People make decisions. Some make consistently good decisions, some bad. Think about this: Acceleration Capital will at some point in the future be a data point for some noted Ivy League academic. Faith's firm will be held out as one of the "unlucky" ones who did not make it. Acceleration Capital will be used (along with other misguided miscreants) as a basis to argue that trend following doesn't really work once you control for survivorship bias.
As far as "changing markets"... Good systems respond to changing market conditions. If one is constantly tweaking a system to supposedly meet changing market conditions...that is not much of a system.
Also, I can't help but chuckle at this busy body like tendency of yours to supposedly put the thread back on track... especially after rants about discretion and Amaranth.
OK. Now that you mention about systems...
Which of the following would be considered a "tweak"?
1. Optimization and implementation of a parameter. For this sake, let's say it's like referenced bar for a Breakout like x in calculating HighestHigh(x).
2. Taking the mean bars of the last 30 swings and applying it on the HighestHigh(x).
3. Crabel like filtering. ORBs filtering signals based Macro chart conditions.
4. Tweaking the models using Greeks... "calibrating" based on the underlying.
5. Windowed Parameter.
6. Out-sample averaging.
... I ask this because the term tweak can mean alot of things. All of the above can be considered a tweak. So before I start some discussion, it would help if you can tell me what your basis / definition of a tweak is.
How about we start there?