Wednesday, May 30, 2007

Spreads and China

I'm not going to spend any time talking about the Fed minutes from today. I don't think there was any information contained therein. I think if you are looking for any kind of Fed move at all, you have to wait until at least the baseball playoffs get started.

Instead I'd like to discuss flows from China and the US Bond market. The CSI 300 Index fell 6.8 percent today, as the Chinese government increased a tax on stock transactions three-fold in an attempt to cool the economy. The US stock market looked like it was going to open much lower, based on early morning futures trading, but the S&PP 500 actually set a new record by the end of the day. Meanwhile, corporate bonds were soft all day, although I heard auto bonds improved by the end.

I don't really want to talk about the impact foreign flows are having on interest rates themselves. Foreigners now own 52% of US Treasuries, according to this story from Bloomberg news. This same article quotes a famous Fed study from 2005 which estimated rates would be 150bps higher if it weren't for foreign involvement. Let me just say that number is an exaggeration of reality. Its non-sensical that foreign holdings would ever go to zero. The real question is what would happen if foreign holdings went from 50% to say, 35%? That's more within the realm of possibility.

The bigger question, in my mind, is in regards to MBS and high-grade corporate bonds. In these sectors, Asia really has been the marginal buyer for the last 2-3 years, and without their involvement, its likely spreads would be considerably wider. What if Asia stopped buying spread product?

First, let's talk about what would have to happen. The primary reasons why Asia is a large buyer of US assets is...

1) We run a large goods trade deficit with Asia.
2) China has pegged its currency to the US dollar.
3) High quality US bonds are viewed as a good investment for the dollars held by various Asian interests.

So in order for Asian buyers of bonds to dissipate, at least one (probably two) of the above must change. So let's discuss the prospects for any of the above. For purposes of this discussion, I'm focusing on the probability of a meaningful, near-term change. Long-term, gradual shifts would have limited impact on spreads near-term.

1) In order for the trade deficit with Asia to reverse in any meaningful way, US consumption probably has to decline. The reason for the trade deficit in the first place is that Asia is a low-cost producer of many goods. While it would be possible that relative inflation causes wages in China and elsewhere to rise such that they are no longer low cost producers, I can't see this happening fast enough to have the kind of impact we're talking about here.

2) China does seem interested in loosening its dollar peg. But China's incentive to have a relatively weak currency would be in tact no matter what happens with the peg. If their goal remains a weak currency, but they attack that goal with a different strategy, I question how large an impact on Chinese demand for US dollar assets this would cause.

3) Here is where it gets more interesting. Asian investors were once strictly US Treasury buyers. Now they have diversified into MBS and high-grade corporate bonds. They could continue to diversify into other assets at the expense of high-grade US bonds. Recently, the Chinese government announced they would be investing $3 billion with private equity firm Blackstone. It is also possible that Asian investors increase their use of euro-denominated assets at the expense of US assets, although the dollars they get from goods trade still has to flow back into the US, so exchanging dollar assets for euro assets doesn't automatically mean there is less demand for dollar assets.

So as I'm reading the news and reading between the lines, its the diversity issue that worries me the most. Now we have to gauge how quickly China, Japan, etc. might move into new asset classes. I'm of the mind that this will take many years before Asian investors have moved a meaningful amount of money away from the bond market. In the end, this will have an impact on spreads. After all, it took years for Asian investors to exchange Treasuries for other bonds too, but its obvious there has been an impact on spreads.

So what to do? As I've written in the past, its hard for PM's to just not own spread product for an extended period of time. As I've also written, a large percentage of PM's overweight credit as a means of adding alpha.

I think credit is more vulnerable to a shift in the wall of liquidity than MBS. First of all, credits generally have more duration. So each 1bps widening in credit has a larger impact that the same widening in MBS. In addition, MBS is constantly paying down principal, allowing the investor to reinvest. Corporate bonds eventually mature, but at a much slower rate. So should spreads slowly widen in both MBS and credits, which is my hypothesis, the MBS investor gets to reinvest at wider spreads. This mitigates the impact of the widening. MBS pays more in yield, if you believe the current prepayment assumptions. A generic Fannie Mae 30yr 6% MBS has a yield spread of 118bps currently, according to Bloomberg. A generic A-rated corporate bond has a yield spread in the 80-100 range, depending on the name.

Finally, and having nothing to do with Asian liquidity, MBS don't get taken private by TPG. That's kind of nice isn't it?


Anonymous said...

Fascinating analysis.

I think the only missing element is the likelihood that the American consumer will get very sick, very fast.

What probability should we assign to that event?

Personally, I think any assignment below 50% is unwise, and 50% is a frightening number.

Accrued Interest said...

Sure... but it would have to get pretty bad to have a big impact on the the level of Asian liquidity. Say that consumer spending declines by 2% overall. That'd be a pretty big move, but enough to erase the liquidity?

Anonymous said...

On a related hypothetical, suppose that the Chinese equity market crashes (not unlikely IMO). Normally one sees a flight to Treasuries during foreign financial crises, but in this case, is it possible that China actually sells Treasuries to raise cash to help prop up the domestic market, for fear of angry mobs of local speculators attacking ChiCom headquarters? Or will the crash, if and when it happens, have little effect at all here?