An interesting observation from recent sessions is how concentrated the market currently is around a small number of themes rather than broad participation across all FX pairs.


For example: strong AUD participation, weak CHF and JPY flows, selective commodity momentum, while broader market breadth still remains mixed underneath. What becomes difficult in those environments is not finding trades. It is understanding which trades are effectively the same exposure expressed through different instruments.


Correlation clustering starts becoming more important than raw signal count. Another interesting aspect is how often highly active markets still produce structurally poor conditions underneath the surface. Sometimes the cleanest analytical outcome is simply identifying environments where participation quality is low. Curious how others approach this problem: Do you focus more on individual setups, or on understanding the broader structure connecting them together?  (photo is combined data from wfdquant, which I found recently useful)