J.P. Morgan's Big Bet

http://ftalphaville.ft.com/blog/2012/05/11/996131/too-big-to-hedge/
http://blogs.reuters.com/felix-salmon/2012/05/11/counterparties-your-massive-guide-to-jpmorgans-failed-hedge/
http://www.marketplace.org/topics/business/easy-street/jp-morgans-loss-explainer#.T60fG7ppJTE.twitter

Assuming Alphaville is correct that it was a short position in the CDX.NA.IG.9 my questions are

(1) how much of the loss was on the original position and how much of the loss was due to trying to liquidate the position?  If it really was a hedge then why liquidate it?

(2) The Marketplace article has me confused.  If JPM was long the corporate bonds then the hedge is to buy protection but all the articles suggest that they sold protection.   If they were long bonds then JPM's CIO has a lot of explaining to do.

(3) If the trade really was a hedge then was it

        (a) a curve trade like Alphaville contends.  If so then we should observe a huge increase in the notional on the other leg of the trade as well.  Do we?  Everyone seems to focus on CDX.NA.IG.9 but  have seen nothing on the other leg of the trade.   Also how much notional would they have had to have on for a curve trade to lose USD2BB.

        (b) a basket of the index versus the underlyings.  If they lost USD2BB on this trade then how far did the index get out of line before they started to liquidate and how much of it did they actually have on (it must have been a lot)?

        (c) short the protection and short the corporate bonds.   Again how much did they have on?

(3) the chart of total notional outstanding suggests the total outstanding notional on CDX.NA.IG.9 has doubled in the last 6 months.  How much of this was due to JPM?

The one case where Alphaville speculated on the specifics of a trade where I actually knew what the trade was...they were wrong.  I'm not saying they are wrong this time but rather it is sometimes difficult to infer the specifics a trade from looking at market activity.  Eventually the true story will come out.