Marshallian K

I've always been struggling to find a good relationship between money supply and asset prices.

A few days ago I came across this BCA research study (GIS 090306 - Gold), where they compare Marshallian K to asset prices.

One thing to notice is how correct Marc Faber is. That rises in money supply to lead to rises in the gold price (and it appears to be quite a tight relationship). What is also interesting is how money supply does not lead to consumer inflation, merely asset inflation. I need to look more into this dynamic, maybe they are driven by the same underlying inflationary force, but consumer prices have offsetting mitigating effects (like trade, globalization, technological advancements).

However, I see a conflicting discussion of Marshallian K in Niall Ferguson's Chimerica article.

He argues that there has not been a surge in liquidity (maybe a little in the Eurozone, but for the US, Marshallian K has actually been flat from 2003 to 2007). So has there been a surge in liquidity or not? What's actually happening here? Is Niall's data incorrect, or am I not comparing the same series (maybe there are many different Marshallian K's).

I did some more Googling on Marshallian K and surprisingly found very little. I could only find a single paper on ideas.repec.org ! I did however find this little paragraph on the topic by Stephen Roach:

"My favorite gauge of the quantity dimension of liquidity is the so-called “Marshallian K” -- the difference between growth in the money supply and nominal GDP.  In essence, this measures the surplus of money that is not absorbed by the real economy.  Joachim Fels has constructed such a measure for the “G-5-plus” group of industrial countries -- the US, Japan, the 12-country euro area, Canada, and the UK. This measure is based on “narrow money” (i.e., M-1) -- the monetary aggregate that still has the tightest linkage to central bank policy adjustments.  The trend in this version of the global Marshallian K is now ticking below the “zero threshold” for the first time since 2000 -- consistent with earlier turns in the liquidity cycle that have been associated either with recessions (1991 and 2000-01) or abrupt adjustments in financial markets (1994)."