Hi Q community
I have been giving some thought to the whole concept of regime change, change of market dynamics or you might call it "why does what used to work not work anymore?" The point of this post is to open a discussion of this phenomenon and possibly share research (either your own or others) and ask questions and openly think about how to profit from it, if possible.
There are different kind of regimes: E.g. the regime of bull market or bear market, or economic expansion or recession. The latter two are widely correlated with certain moves in stocks (up and down), but economists usually see the correlation post factum, which for a practitioner or algorithm creator is useless at best, and dangerous at worst as it can lead to overfitting.
Market pundits often speak of the market being in a "momentum" regime or a "mean reverting" regime, but also regarding fundamental data can the market be in a regime that favors value stocks (however you may define that) or growth stocks and a whole host of other different regimes. Just as with the economists knowing that stocks go down in a recession (ironic bravo) the market pundits may know that a certain market have favored certain factors for a while, which leads me to a few questions:
Do you think that:
1) It's possible to predict or at least diagnose what kind of market regime we'll be in or are currently in?
2) If possible, how do you measure it, i.e. what kind of tools would you use?
3) How do you avoid overfitting, in the case let's say you can predict a recession is coming, but who knows if it will be like the last one(e.g. especially finance stocks going to the dumps)?
For some more information see this which I have nothing to with and only found a little while ago from a quick google search.
Let me know what you think on this so important and difficult subject.
