Yesterday’s move seemed almost magical. Yellen spoke and the markets levitated overnight. Jobless claims were a big disappointment. Revisions hit the prior week’s number and yesterday’s number was much closer to 400k than to the almost mythical 350k the market had become used to expecting. The magic of BLS revisions have ensured that although numbers close to 350k were reported for at least 3 weeks, they have all been re-written as 360k or higher.
The markets briefly fell, but then stories of “gnomes” leaking China GDP started the markets higher again. That 9% GDP print turned out to be illusionary, as the real number came in at 8.1%, and since I see no reason for China to lie about it to make the data worse than it is, the real growth is probably even slower than that.
The weakness in sovereign debt over the past week finally made investors look behind the curtain of Draghi’s LTRO show, and they are underwhelmed. The fact that LTRO does what it is supposed to – ensure banks have access to money – is now a disappointment as too many people had believed that LTRO could do more than it actually could.
Which leads us to Voldemort and Volcker. How much of JPM’s earnings were a direct or indirect result of the activities of the CIO’s office. We may never know, especially the indirect part. I believe the primary trade they have on is long credit via tranches on IG9 vs short HY17 and HY18. The trade makes sense, both as a trade, but for JPM and their business in particular. It explains both the price moves in IG9 and the decompression of HY CDS in an environment that would normally see compression. He is big, but the “prop” trading people are worried about the wrong things. The Volcker rule was meant to limit prop bets on market making desks. Banks do need to take risk. Everyone, the Fed included, wants the banks to lend more, but each and every loan is a prop bet, and there is no profitable way to run a fully hedged lending book. Let the CIO and treasury run the bank.
Correlation desks, including the one at JPM should be looked at closely. If a desk buys a tranche and sells a corresponding amount of “delta” is that not a “prop” bet? As a market maker, they haven’t bought and sold something, they bought one thing, and sold another. Volcker should be looking at those positions and determining how much model risk should be allowed. A correlation bet is a bet like any other (just more complicated). The noise about IG9 is reasonable, just misplaced.
As yesterday’s magic dissipates, it is hard to see anything in the data that justifies the bullishness. I remain wary of the market, and certainly feel better today as being caught too short yesterday was painful. CDS indices are weak across the board. Spitalian 10 year bonds are both trading down today, and 5 year Spain now yields 16 bps more than 5 year Italy. A clear sign that the manipulation has run its course (the size of the Spanish bond market makes it far easier for the ECB and banks to control prices – at least for brief periods of time).
I will be looking to add HY17 or HY18 risk today. Not as a general risk on trade, but because if I am correct and they are part of the “whale” trade, they will be left alone by the big buyer and we can see some of that compression that should have occurred earlier this year. HYG and JNK are both back to “premiums” to NAV that seem unsustainable, particularly given the overall tone of trading so far today.