the use of two data points to support the counter argument was deliberate to show how silly a statement such as there is almost no correlation is
there is clearly correlation (otherwise supply and demand as concepts are redundant), however it is also clearly not a perfectly inverse correlation
the move from 1mt in stockpiles to well <100kt is maybe an indicator of willingness of participating parties to embrace JIT manufacturing principles as much as it is about any conspiracy about there being 800kt in a shadow inventory system, however that also lends itself to the system being finely balanced, so news such as over the last few days of glencore trimming it's annual production could have strong influence on price direction
directionally, what volume do you think is held outside of LME/SHFE? are you saying that has risen proportionally with the drop in official inventory, or that it has always been held outside the view of the official numbers, what is the net situation (if the shadow was constant, there would still be a net lower inventory situation)
