[lbo-talk] Once again, food prices

SA s11131978 at gmail.com
Wed Feb 9 10:13:49 PST 2011


On 2/9/2011 12:36 PM, Doug Henwood wrote:


> On Feb 9, 2011, at 12:24 PM, SA wrote:
>
>>> I don't get the idea that there aren't measurable net flows of money in and out of asset classes.
>> I don't see how you could meaningfully measure that, even in principle.
> E.g.: http://uk.reuters.com/article/2011/01/18/uk-commodities-funds-flows-idUKLNE70H01520110118

The article sez:


> Barclays Capital data and estimates suggest that, if prices remain as
> they are, total fund investments in commodities at the end of 2011 are
> likely to be around $420 billion (262 billion pounds). With any price
> appreciation, the total would be higher.
>
> Barcap estimated that 2010 ended at a record high of about $360
> billion invested in commodities by funds -- a year-on-year increase of
> around $90 billion. Of that, about $60 billion was down to flows and
> the rest due to price rises.

Okay, so this is telling us that *commodity funds* received $60bn in net inflows. They used that money to buy commodities contracts, I'm assuming. That made them buyers. But who sold them the commodities contracts? Sellers, of course. So the net flow into commodities contracts was zero. The net flow into commodities *funds* doesn't really represent an inflow into an asset class, because commodities funds shares aren't tradable assets; they just represent me giving a "friend" some money to invest on my behalf. (The anomalous exception here would be ETFs that trade above the value of their underlying assets; that is almost by definition a bubble, I think, although there's a whole literature on that.)

To prove the lengths I'll go to to procrastinate, I just looked at some data on two variables: (1) dollar trading volume on the NYSE; and (2) S&P 500 prices, on a monthly basis from Jan. 2004 to Dec. 2009. The correlation was 0.71, meaning trading volume explains only 50% of the variance in prices. (The correlation between monthly change in volume and change in price was slightly negative.) So we might say that volume explains half and sentiment explains half, but I suspect a lot of the volume is explained by sentiment.

SA



More information about the lbo-talk mailing list