Tuesday, October 21, 2008

30-year Swaps: Look's like we've got a bad transmittor

The 30-year swap spread fell as low as 0.5bps today, closed at 3bps. 3bps! Three. One less than the number of legs on a AT-AT. Equal to the number of good Star Wars movies. If you don't believe me that municipal spreads are stupid, at least believe me that a 3bps spread on 30-year swaps is stupid.

To translate... swap spreads is the yield differential between Treasury bonds and the fixed leg of a fixed-floating interest rate swap. Remember that any interest rate swap has to have a bank or other financial institution standing in the middle. With the world scared out of their minds over counter-party risk, how is this spread at all-time tights!?! By comparison, 2-year swaps have a spread of 104bps, and traded as high as 165bps earlier this month.

This strange anomaly is just another example of what happens when leveraged investors are desperate to unload bad trades into a highly illiquid market. Over the last 2 years, there were many "range notes" sold that referenced the slope of 10-year and 30-year swaps. Now that trade isn't looking so hot and people want to hedge.

Lots of people rushing to leave the building, but a small door of liquidity, and a spread of three is the result.

This just reiterates what I said the other day about leverage. The market can't act "normal" when fresh capital is hard to come by. You should still buy bonds based on fundamentals, but bear in mind that it may take a while for fundamental analysis to pay off.


spock said...

I am not a finance pro, but when I look at the fundamentals of bonds, it makes me not want to buy any bonds.

Spreads will not pay my bills. At these low interest rates, bonds still look like a "return-free risk" (Jim Grant's term). So for now I leave bonds to those who care to arbitrage one type of bond against another.

I fear that the conventional wisdom (shift toward bonds as you near retirement) will prove harmful for many people.

Stan Jonas said...

Doesn't look as though you ve ever traded a swap versus a Treasury..

Fixed legs? Here's the sad truth.. the Treasury when "bought under a Repo" which is the only way to do it in the US.. is in many ways riskier in the current enviornment than the Swap..

You counterparty on the Swap goes belly up as say Lehman.. you are at risk in the in the difference in yield from when you did the trade to default day.. petty cash..

But if the counterparty with whom you did a Repo goes away.. you don't get YOUR BOND back.. you lose $100 dollars.. i.e the Princal amount in addition to everything else..

With the recent fail's etc.. Doing Repo on a Treasury with counterparties who may defaults is the riskiest.. of all trades.. Worst yet.. if you lose your Bond.. given a default of a counterparty.. yields are likely to go much lower.. leaving you with even more replacemment risk..

By the way I think the real issue is that the LIBOR side of the swap just doesn't mean anything any more.. with where LIBOR is fixed being anywhere within 50 basis points of the real thing..

David Merkel said...

AI -- Supposedly there is strong demand from pension plans trying to immunize long duration exposure... but 3 bps is ridiculous. They may as well buy long Treasury zeroes.

Nicolas said...

bad example, you don't exchange the nominal here, it's just interest flows.

that's not the case with a bond where you actually give the 100 USD to the company.

Accrued Interest said...

Stan and Nicholas... you are missing my point. If the spread "doesn't matter" because counter-party risk is minimal, then why is the 2-year swap over 100bps? When in fact, over the long-term, 2-year and 30-year swap spreads have been similar.

Or why is the 30-year swap at all-time tights while the 2-year was at all-time wides?

The point is that something isn't right here and the market isn't able to correct itself quickly.

dude6336 said...

Swap spreads in GBP and EUR are also at ridiculous levels for the 30-year. Part of it is probably the coming deluge of 30-year bonds which governments will issue to pay for bail outs. The glut of bonds will drive up bond yields. Swap yields won't follow because nobody wants to mess around with swaps and the swap spread will be back to normal.

It's still an arbitrage opportunity, but nobody wants to take risks to arb things these days because liquidity is hard to come by.

marek said...

Well, 3bp spread for 30Y is certainly not related to credit risk. Otherwise, it should be of the same order as spreads for 2Y and 10y.

I also think that this dispairy is mainly due to hedging or better say demand.

I'll try to explain why I think so:
In general there are 3 combinations of these notes: 10y2y, 30y10y and 30y10y.

Let say: if you have the same position in all combinations of the notes then
you would hedge 10Y2Y by buying 2y selling 10y, 30y10y : buy 10y and sell 30y, 30y2y: buy 2y sell 30y.
I know it is like there is missing other side of the trade and it is probably missing.
So, I would say everybody is interested in 2y, 10y rates and almost nobody in 30y.

If I assume that there are mostly 10y2y notes (in terms of the notional) then just from the spreads one can deduct that there were much more 30y10y notes rather than 30y2y.
And this is also what we observe in reality.

PNL4LYFE said...

I have heard about these range notes, some of which have a digital payout when 10y and 30y rates inverted. Three cheers for infinite gamma!

Yesterday, I heard about another exotic product that is causing large demand to receive long term rates. These are bonds designed for Japanese investors. The coupon on the bond rises when the Yen falls vs the dollar and vice versa. They are typically 30y bonds but are callable every year. As the yen has rallied, the duration on these bonds has increased dramatically so the dealers that sold them have had to receive. The estimate I read was 3bn 10y equivalents per 1% move in JPY/USD.

marek said...

PNL4LYFE it seems you are right ...
Asian countries annouced to create fund pool to protect currencies, equity have fly to "quality" and USD/JPY goes to docks. Interest rates are going to fall to support economy in recession so payoff goes to 0.

It would mean further widening of spreads. So, have fun.

plexist said...

I don't know if you're still following this, but the 30-year swap spread is now at -27. People are claiming this because of hedging of some weird Japanese derivatives (http://www.reuters.com/article/bondsNews/idUSN1752416920081117). Does anybody buy that?

Accrued Interest said...

The "exotic" notes are the same range notes I mention in the article. That is the same excuse every one is giving for the negative spread.

plexist said...

The spread is now at -50! Is it really impossible to hedge with treasuries? I would imagine that even a little basis risk hedging with treasuries instead of swaps would be worth it instead of paying that spread.

I guess maybe the comment by Stan Jonas is correct (i.e., it's impossible to do repo these days because of counterparty risk). I found his comment confusing, though. Wouldn't you buy the treasury and then repo it to get back the cash? Then if the counterparty defaults, you don't get your bond back but you keep the cash.

Accrued Interest said...

I think the problem is that the non-inversion notes don't reference Treasuries, they reference swaps. That's the problem.

Nicolas said...


you have to be careful those days about what an interest rate really is.

If you look today in the market, you have as many interest rate curve as you have instruments.

They all differ by one or two little thing that actually make them really different.

One thing is for instance funding which has huge impact on short term debt instruments.

That makes the deposit and the future market completely different.

Another one can be inflation, which has an antagonist effect :
you can see that as much as you want to hold the cash in the short run, inflation will eventually eat it out in the long run, which might explain your spread going from positive to negative.

can you give the precise instruments that you are spreading (bloomberg code?)

I am not a fixed income expert, being more on the equity derivative side, just throwing ideas around.