Analysis strategy using market indicators and trading systems I have a number of trading strategies, and trading systems which are subsets of
the strategies (many systems in one strategy).
I also have a number of market indicators created by a portfolio manager which
act like technical trading indicators which indicate his beliefs about certain
market conditions. 
I am trying to conceive of an analysis strategy to show the relationship between
the portfolio manager's indicators and the daily returns of the trading systems. 
For example, the portfolio manager hands to me a group of indicators: Ind1,
Ind2, and Ind3.
He asks for an analysis showing a relationship between Ind1,  Ind2, and Ind3 and
the daily simple returns for a number of systems: Sys1, Sys2, ..., SysN.
Ultimately the goal is to build a model guiding the portfolio manager on which
systems to put into the market based on his indicators (or disprove the
predictive value of any of the indicators).
 A: This question is pretty old, so I'll post only a brief answer for now. If someone would like a longer response, just let me know.
Short version:  You'll probably be best served by backtesting indicators (e.g., seeing historically what trades you would have made, and what you would have made/lost). With that info you can check various statistics such as: 


*

*Annual return 

*Standard deviation of returns (typically daily returns) 

*Sharpe Ratio—Deals with variance in all returns

*Sortino Ratio—Focuses on downside variance (losses)

*Treynor Ratio—Another risk-adjusting reward measure

*Information Ratio—This wikipedia page also has links to some of the other ratios that are interesting, but perhaps less often used

*Drawdowns/max drawdown—How big are the dips as you (hopefully) generate returns on the strategy? How big was the biggest one?

*Jensen's alpha


Those measures will give you an idea of the risk-adjusted performance of each indicator. 
(Also, I had a bunch of other links, but since I'm new, I can only post two links.  Again, later if people want them. Edit: as gung pointed out, I can paste them in as text for now, so...) 
Others links:  


*

*Sharpe:  http://www.stanford.edu/~wfsharpe/art/sr/sr.htm

*Jensen's alpha: http://www.financialwebring.org/gummy-stuff/Jensen-alpha.htm

*Various:  http://www.investopedia.com/articles/stocks/11/5-ways-to-measure-money-managers.asp

*Open-source rabbit hole (and I mean that in a good way):  http://cran.r-project.org/web/packages/PerformanceAnalytics/index.html
That last link takes you to a package for R. That package can calculate all of these ratios automatically, once you feed it the returns. The documentation for the package talks about all the ratios and how they're used. It's a great tool and, even if dense, review.
And for ease, Wikipedia links:


*

*Sortino:  http://en.wikipedia.org/wiki/Sortino_ratio

*Treynor:  http://en.wikipedia.org/wiki/Treynor_ratio

*Information: http://en.wikipedia.org/wiki/Information_ratio

*Jensen's alpha: http://en.wikipedia.org/wiki/Jensen%27s_alpha
