anand srivastava · @anandsr21
4th Jan 2017 from TwitLonger
Lets look at the data here.
I am looking at the net flow/ETF Flow and gold price. Not import/export.
I think the bars above 0 means net buying and below 0 means net selling.
I do not think that ETF flow is more important for price, although it is more important than phyz. XAUUSD is most important but that probably cannot be monitored.
Interesting thing about the ETF flow chart is that it is much more positive than negative till Dec12. Then selling dominates till Dec15. Still buying is positive flow is overwhelming compared to negative flow. This I think is due to ETF creation. I think that a lot of it has been created than has been destroyed. Like gold gets mined but not consumed. So overall flow for gold is also positive.
Lets look at different periods.
Jan 05 - Jul05 - price is stable, net flow negative, ETF stable.
Jul05 - Oct 05 - price rising, net flow negative.
Oct05 - May06 - price rising, net flow positive. ETF positive, peak buying Dec05
Above shows net flow is lagging price. Not clear for ETF flow.
May06 - Jul07 - price more or less stable, net flow positive, except Oct 06. ETF positive.
Aug07 - Mar08 - price rising, net flow positive. Note that net flow was negative during Aug07. ETF mostly positive.
Shows mildly that flow is lagging price.
Mar08 - May08 - price down, net flow positive. ETF negative in Apr08, in sync with price drop. Or maybe huge sale XAUUSD, causing sale in ETF.
Jun08 - Jul08 - price up, net flow positive. ETF positive.
Jul08 - Oct08 - Price down, net flow negative. ETF peak on Sept08.
Possible correlation with flow.
Nov08 - Sept11 - Price up, net flow positive, except Jan11, Feb11, Sept11, Oct11. ETF flow is much more mixed.
Flow gets negative after the price starts dropping. Shows flow is lagging price.
Sept11 - Oct12 - Price fluctuates wildly, Flow remains mostly positive.
Oct12 - Dec13 - Price drops, Flow turns negative from Jan 13.
A 3 months delay in flow reaction of price drop.
This is the most delay that can be seen from the data.
The rest of the data is also similar, where flow is following the price drop. But later the variation is too much. But it is always the flow following the price change, not the other way.