In Murphy's first book on Intermarket Analysis, he made a strong case for sustainable market relationships. Years later he published his second book on Intermarket Analysis where he introduced the term "decoupling".
If it holds useful anlysis info it could be used as factor in strategy. It appears there are relations and some lag between dif instruments, that may hold potential fututure price info greater than spreads, but most of the relations seem to be in constant change and not much consistency. It seems quite complex how to spot or interpert those, i dont mean only the standard definition of 2 price correlation but other ways to measure or more instruments also. The ammount of possibilities dif instruments could be related is quite great for even computer to process and there are endless possibilites for what to look for. Are there some good books/other sources where may be some useful info?
I think you summed it up as shown above. Any keen insights on relationships are likely to be proprietary and not published. Best approach is probably your own research.
Lately have been testing some loose correlation/arbitrage based ideas. What i have noticed with tests is that timeframes with opportunity in correlation based strategies goes lower with each year as time moves forward and spread is main enemy. How significant spread advantages etc. does competition have? How high is the barrier of entry to get better trade conditions?