There’s something interesting going on in these Wall Street Journal articles (Money & Investing and Deal Journal) today about how corporate bonds now sometimes trade inside of Treasuries. Or somethings interesting; one thing that’s going on is, like, why the day after the election? One possibility is that the message here – which the Journal is helpfully conveying from bond investors to the government – is “see? get your fiscal cliff shit together or soon you’ll be pricing your bonds outside of Google, and you don’t want that do you?”
There may be a side helping of, like, annoy modern-monetary-theory bloggers; from a certain viewpoint this graphic is a hot mess of category mistakes1:
BUT THE GOVERNMENT HAS A PRINTING PRESS oh never mind. Maybe Exxon does too, I don’t know.
But this is my favorite part:
On Nov. 1, J&J bonds yielded 0.218%, below its perfectly matched Treasury at 0.283%, according to MarketAxess Research. Google bonds in August yielded 0.17%, below its implied Treasury equivalent at 0.271%. And yields on other top-quality bonds, including from Microsoft, are closing in on Treasury yields.
I dunno from perfectly matched Treasuries but here’s that Google bond, which matures in mid-May 2014, graphed against a(n off-the-run) 1% Treasury maturing at the same time:
Things to notice include:
- yeah, it traded inside of Treasuries, and
- like, four times.
Actually it’s better than that. That graph makes it look like it bops along psychotically, moving rapidly up and down along its jagged little line. But you can actually get granular TRACE data for August 14, when it printed at 0.17%. Here is that data; focus on the 08/14/12 line where dealers sold 50M of bonds to clients at a spread of -10.1bps to Treasuries:
That’s three trades for a total of $150,000 worth of bonds, which breaks down as (1) Dealer A sells $50k to customer at ~10bps below Treasuries, (2) Dealer A buys $50k in interdealer market at ~0 spread to Treasuries, (3) Dealer B sells those same bonds in interdealer market (at ~0 spread to Treasuries since this is just the other side of trade (2)), i.e. one trade, triple-counted. So, that’s one customer calling up his dealer to buy $50,000 worth of bonds. On a $1bn bond issue. Of a $219 billion market cap company. That basically issued debt for fun. Here is how I overactively imagine that call going:
Customer: I want $50,000 of Google 1.25s.
Dealer: Um, just $50,000?
Dealer: That’s annoying. I’m going to have to go find those from another dealer. I’m going to overcharge you by like 20bps of yield – T-minus-10 instead of T-plus-10 – so I can make this trade worth my while.
Customer: Okay, so … how much am I overpaying you?
Dealer: Eh, 20bps of yield on bonds with ~21 months left, call it 0.35% of par, so on $50,000, you’re paying me an extra … $175.
Customer: One hundred seventy-five dollars? That’s fine, I have special reasons for wanting a tiny quantity of this bond.
Dealer: Also, just FYI, a few months from now, my overcharging you by $175 for this Google bond is going to be evidence of the decline in the full faith and credit of the United States government.
Customer: That will be odd, won’t it?
I’m not sure all of those bonds that have traded inside Treasuries look like that, though I will say the Google one is the first one I looked at. But – probably? I highly enjoyed Peter Tchir’s imaginary story about credit index trading; the gist is worth repeating here:
[After a bunch of hypothetical trades the CDX index] moved over 2 bps and not one single trade had anything to do with “valuation” or “fundamentals”. Thinking that all moves reflect a change in fundamentals or a large scale change in risk positioning is just wrong in a market that is as illiquid as this – and [IG index CDS] is about as liquid as it gets in the credit markets.
You know what is not as liquid as it gets in the credit markets? Odd lots of a novelty Google bond.
To be fair! There is much to ponder here – the usefulness of benchmarks, the shrinking supply of Treasuries and AAA structured credit that makes high-IG corporate bonds more attractive, the coming demand for high-IG collateral for derivative clearing, the fiscal cliff, etc. etc. Maybe even “Google’s balance sheet is better than the Treasury’s,” though I’m not convinced by that; feel free to fight to the death about it in the comments.2 In any case, though, if policy makers do want to listen to bond markets as they make their policies, I’m all for that. Bond markets are smart, you should listen to them. Just don’t listen too hard. Bond markets don’t actually mean everything they say.
The New Haven for Investors [WSJ]
Treasury Dominance Over Corporate Debt Queried [Deal Journal]
1. Also a mess:
For instance, Procter & Gamble in August sold $1 billion of 18-month floating-rate notes at 0.10 percentage point below three-month Libor. That Libor rate is 0.31 points, so the P&G bonds now trading at full value or “par” are yielding 0.21% versus the comparable fixed-rate Treasury due February 2014 that is yielding 0.22%.
It don’t work that way! That’s three-month P&G vs. fifteen-month USTs! One-year Libor is like 87bps.
2. I mean, it obviously is, as a balance sheet, but remember that printing press. Buy gold!