Will US Rates Soar if China Dumps Treasuries?

August 31st, 2015 5:31 am | by John Jansen |

This is a very interesting article from the Financial Times which makes a theoretical case that sales of US Treasuries would not cause rates to soar. The author posits that the original result of purchases by China was a stronger dollar which then led to lower inflation and lower real growth and that is the real reason for the initial decline in rates. The author points out that the bulk of purchases by China were in the belly of the curve and that a percentage of their holdings the long end is rather insignificant. He notes that the curve did not steepen dramatically as China accumulated this belly paper.

Via the FT:

China, you may have noticed, has switched rather abruptly from being a massive buyer of foreign currencies to a major seller. Some people — including some relatively influential policymakers — are worried that this switch from suck to blow, as it were, could cause Treasury yields to spike. That fear may be animating some of those who think the Fed should adjust its schedule of rate hikes, or even engage in additional large-scale asset purchases.

We’re sceptical.

Back when the People’s Bank of China was buying Treasury bonds, it was really just selling yuan for dollars it could hold as foreign currency reserves, some of which were then invested in USTs. It’s possible those reserves were accumulated as insurance against a repeat of the Asian Financial Crisis, but it’s more likely the government just wanted to subsidise Chinese companies by suppressing the exchange rate. (A weaker currency raises the cost of foreign products and makes exports more competitive in global markets.)

The PBoC could just as easily have bought other dollar-denominated assets to accomplish this objective, and it did. The important thing was that the PBoC was selling yuan for dollars to offset the unusual combination of a big current account surplus and big net inflows of private capital. Had the PBoC refrained from buying dollar-denominated assets, the value of the yuan would have had to increase to balance the flows of money going in and out. Instead, it pursued policies that propped up the value of the dollar.

All else equal, an overvalued currency reduces nominal income by hitting both trade and foreign earnings. The central bank can offset at least some of the impact on aggregate growth by lowering interest rates, which makes investment and consumption relatively more attractive. The net effect would be a change in the composition of growth away from tradeable goods and services in favour of construction, healthcare, government, and education. That’s exactly what happened in the US in the 2000s.

(An overvalued currency can also lead to a deterioration of the fiscal balance, which explains both why the federal budget went from a big surplus to a big deficit and why economists found that the parts of America most sensitive to trade competition with China saw a big decline in employment and a big uptick in government transfer payments.)

It’s therefore entirely reasonable to think Chinese and other foreign purchases of Treasury bonds lowered domestic US interest rates — but that’s because those purchases lowered the outlook for American inflation and real growth, not because there was something special about buying bonds compared to other dollar assets. If the whole process went into reverse it would be a good thing for the US.

Besides, we have no reason to think China’s total dollar holdings are actually shrinking. China hasn’t decided to undo its old policy of suppressing the exchange rate to support domestic producers — the trade surplus is near its all-time high and the yuan is falling. Rather, the PBoC is simply trying to preserve its existing exchange rate regime by offsetting massive outflows of private capital, most of which are probably headed straight for the US. (The same can be said for any other case of reserve liquidation.)

Suppose the private capital flight and official reserve liquidation perfectly balance out and there’s no net change in total Chinese holdings of dollar assets. UST yields would still spike if both of the following statements were true:

  • Private Chinese investors and the Chinese government have radically different asset allocation preferences
  • Foreign buying and selling of US government bonds affects US borrowing costs in ways other than the straightforward currency channel

We’re willing to believe the first proposition. Anecdotally, Chinese subjects (rightly) want to invest in foreign real estate that gives them residency rights in places with respect for human rights and the rule of law. Why would they want to hold lots of low-yielding fixed-income claims in a currency that doesn’t match their own liabilities?

Our disagreement with the fear-mongers comes down to the second point. US Treasury debt is the most liquid market in the world. Arbitrageurs can offset much, if not all, of the buying and selling by price-insensitive reserve managers. The burden of proof should be on those who think otherwise.

Imagine there were some significant impact of foreign bond-buying on US rates separate from the consequences of having an overvalued currency. If so, the effect should show up as a distortion in the shape of the yield curve, since the portfolios of reserve managers don’t perfectly line up with the Treasury’s issuance.

People worried about a yield spike due to Chinese reserve liquidation must necessarily believe that Chinese reserve accumulation was at least partly equivalent to the Fed’s large-scale asset purchases. Those supposedly distort the yield curve by pushing down longer-term interest rates relative to what you would expect from adding up expected future short rates, although the way this is calculated is subject to debate and the net effect isn’t exactly obvious. Remember this?

Anyhow, it just so happens some people thought foreign bond-buying messed with the US yield curve during the 2004-6 rate hike cycle. Recall, more than ten years ago now, Greenspan’s confusion over the “conundrum” that short rates and long rates weren’t moving in the same direction:

Long-term interest rates have trended lower in recent months even as the Federal Reserve has raised the level of the target federal funds rate by 150 basis points. This development contrasts with most experience, which suggests that, other things being equal, increasing short-term interest rates are normally accompanied by a rise in longer-term yields…Heavy purchases of longer-term Treasury securities by foreign central banks have often been cited as a factor boosting bond prices and pulling down longer-term yields.

Put aside, for this post, how weird it was that Greenspan claimed to be confused by something he had seen so many times during his tenure as Fed chairman. (Don’t worry, we’ll get to it in a subsequent post.)

Focus instead on the hypothesis that foreign official purchases of US bonds made the yield curve flatter than it otherwise would have been, specifically by pushing down longer-term interest rates — a claim distinct from the idea that reserve accumulation pushed up the value of the dollar and worsened the outlook for nominal income growth. If this claim were correct, we could imagine reserve liquidation having the opposite effect.

The problem with the theory is that the Chinese and others concentrated their purchases at the front end of the curve. According to the US Treasury, about three-quarters of the foreign official holdings of USTs had a maturity of 5 years or less, as of June 30, 2006. By contrast, only about 3 per cent was invested in USTs maturing in 10 years or more. As of June 30, 2014, the latest available data, the distribution is almost identical:

When Greenspan wondered about the “conundrum” back in 2005, the concentration at the short end of the curve was slightly higher, at around 78 per cent.

You might counter that what really matters is the difference between the maturity distribution of USTs owned by foreign governments and the maturity distribution of all USTs outstanding, which looks like this:

It turns out that reserve managers are downright allergic to long-dated Treasury bonds, and disproportionately invest at the front end. That was true in 2005-6 and it’s even truer now. (All the data on US debt maturity is in the downloadable spreadsheet here.)

As we noted above, foreign governments, in the aggregate, only keep around 3 per cent of their US Treasury holdings in bonds that take 10 years or more to mature, even though these instruments have consistently constituted about 13 per cent of the total over the past decade. Foreign reserve managers are also underweight the long end even if we focus on the narrower category of bonds in the 10-20 year sector: 1.4 per cent of their portfolio in 2014 vs 3 per cent of the total outstanding.

This means that if foreign purchases of Treasury bonds had any effect on US interest rates separate from the currency channel, it would have shown up in the form of a steeper yield curve caused by relatively lower rates at the short end and the belly compared to the back end of the curve — the exact opposite of what everyone was complaining about!

To put this in today’s context, if US interest rates will be affected by foreign reserve managers liquidating their holdings in some way separate from the currency impact, we should expect rates to rise at the short end and stay flat at the long end. Yet the worrywarts focus on the prospect of a steepening curve driven by rising rates further out. Go figure.

Be Sociable, Share!

Post a Comment