Parsing the US investment position

A few weeks ago I wrote about the unusually strong US net international position. I missed Brad Delong’s rejoinder:

Where Ashok Rao shies at the jump here is in failing to specify where he thinks the market failure is, and how to correct it. Is the demand by foreigners for safe dollar-denominated assets an improper one? And why today is it only the U.S. government–rather than, say, Apple or Wal-Mart–that can tap this funding source? Or is there a deeper problem in that Apple and Wal-Mart could tap this funding source but really do not want to–that they already have all the capital and funding that they think they can use? These are the questions that people are worrying…

I’m going to shoot at the hip here because the honest answer is that I’m really not sure.

To define the terms, I’m not really sure “market failure” is the right concept to think about here. We are talking about certain events led by single actors with a lot of agency and market power (foreign central banks, state owned exporters, finance ministry policy directives, among others). But we can begin to answer the question: are Treasuries unfairly valued given a concerted foreign interest in maintaining high dollar reserves.

There are multiple equilibria here, and that’s exactly what China doesn’t like. On the one hand, China (and its less reserve-oriented neighbors) would like to be free of the “dollar trap” – and yet any move that would motivate such a shift would open domestic central banks to incredible balance sheet risk. So the current equilibrium is maintained by foreign taxpayers who are unwilling to eat the fiscalized risk of prevailing monetary policies.

The next question becomes why aren’t other liquidity providers offering competition to the US Treasury (which, per my argument, would be “undersupplied” in some sense). The simple answer is US corporations with the credibility to offer safe, liquid debt are already sitting on piles of cash without any meaningful investment opportunities – they themselves are funders, and do not need any funding source. Moreover, for risk and liquidity reasons these bonds are undesirable from a central bank perspective. US Treasuries are traded 8-10 times as frequently as all AAA corporate debt, on a volume basis. The liquidity and risk is reflected in a historically high AAA10Y spread:

The spread spiked in the late-1990s for the reasons previously outlined – Asian markets suddenly rocked by the instability of large external debts and low dollar reserves instituting policies to maintain defensive central bank positions. The mid-2000s decline is a little puzzling – as Asian countries continued to pour into Agencies and Treasuries – “reach for yield” hypotheses and a declining trade position might serve to explain the decline in spread – though this is still an open question.

The important point is that we’ve returned, and never really recovered from, the 1990s environment of relatively high spreads.

So the first question is whether this spread by itself is a “market” failure. From a Chinese perspective, I can’t answer the question – though I imagine the freedom to trade huge quantities without rattling the markets is part of the problem, especially since all AAA credit isn’t really securitized into a hyper-safe and hyper-liquid AAA tranche.

The simpler answer seems to be that people want the US government to be a bank. This is a role it fulfilled with glee in the mid-2000s – not just through its fiscal deficit, but the flurry of agency-backed mortgage origination which, in a foreign investors eye, was a guaranteed dollar deposit (the high liquidity basically means it was always callable). This is a role it didn’t play before 2002 and is one it hasn’t played since 2007. In a sense spreads are returning to the old normal in which the US government is refusing to be a bank. (Note this would also be politically hard to maintain without the guise of Agencies as it effectively requires the issuance of debt without any corresponding increase in spending or decreases in taxes).

In that case, there’s nothing fundamentally improper about the situation today – the big change isn’t an attitude from Asian central banks as much as the withdrawal of US liquidity provision thereof.

This doesn’t answer the question of whether Asian policy is correct, but it does make the question sort of irrelevant – foreign demand (i.e. the alleged global saving glut) isn’t what seems to matter here as much as the decrease in government-guaranteed securities. (Or, more accurately, in their expected stock some n years from now).

Here’s a tangential puzzle. Equity prices are relatively high (implying, given a constant risk preference, an increase in expected earnings growth rate) and yields are relatively low (implying, given a constant capital share of output, a decrease in expected earnings growth rate). So one of the “givens” must be incorrect. Yet if risk premium really fell meaningfully a number of previously unprofitable investments should be in the black, on the margin, encouraging further spending and borrowing by the Apples and Walmarts of the world. That’s not what happened. Maybe, as Tyler Cowen has prompted (if not advocated) to me, it was a fall in labor’s share – but that’s a decades-long process and probably can’t explain short-run dynamics.