Brad Setser: Richard Iley of BNP Paribas – the author, with Mervyn Lewis, of a new book on the US current account deficit --- doesn’t see the world quite the way I do.
I put a lot of emphasis on the importance of central bank financing of the US external deficit; Richard argues that I overstate the role of the official sector. I have also argued that large deficits eventually imply a deterioration in the US net international investment position and a negative "income" balance in the current account. Richard isn’t convinced – and the data has gone his way over the past few years.
But rather than try to summarize our differences, I am just going to turn the blog over to Richard Iley himself. Monday’s current account data will offer me a chance to expand on my own views. There is nothing like a bit of intellectual debate to make what might seem to be a dry data release come alive.
First, let me thank Brad for this opportunity to ‘guest blog’ at RGE monitor; not many would deliberately invite and welcome opposing views in the way Brad is happy to. I only hope I can rise to the occasion.
I wanted to begin with a little deep background ‘colour’ on how my thinking on the current account deficit has evolved. Apologies for the length of this post but I figured I may only have one shot!
Back in 2004, like most macro-economists, I was increasingly alarmed by the trajectory of the U.S. current account. Brad, of course, was in the vanguard of highlighting and amplifying these concerns. The shopping list of concerns over the deficit is a familiar one so I won’t repeat it. Suffice to say that I was so worried that I was prepared to take the ultimate radical step: write a book on the subject! But this project combined with the surprisingly benign out-turns of recent years has slowly produced a conversion on my part. My fears over the current account deficit, while not completely disappearing by any means, have slowly receded somewhat. Certainly, I have become more relaxed than Brad about the implications of, and prospects for, the US deficit.Crucially, working with my co-author, Professor Mervyn Lewis, gave me an Antipodean perspective on the US deficit. ‘Down under’, extreme concern and unsuccessful attempts to target the current account deficit have progressively been replaced with a policy of benign neglect and recognition that, in a world of stepped up capital flows as home bias declines, current account deficits can largely be left to look after themselves. Provided a deficit (or surplus) is largely a private sector phenomenon (which it currently is in the US), it is essentially a matter between ‘consenting adults’ and so not a cause for explicit policy concern. We all have current account positions – some in surplus, some in deficit. If individuals and companies in a given country want to borrow and the rest of the world is happy to lend, why should this be a particular source for concern?
We need to be very clear; this is not to claim that deficits do not somehow matter or certain consequences are not likely to follow from them. Of course, they do. But rather, they need not be a matter of explicit policy concern. To couch it in Australian terms, the ‘larrikin’ will wake up with a hangover while the ‘wowser’ will enjoy a clear head: both tend to get what they deserve. We should only care about their behaviour to the extent, it impacts on us. In other words, there are clear and tangible externalities in the borrowing process. I’m not sure that this case can made on the US deficit, nor is, say, Oliver Blanchard. The emergence of the ‘consenting adults’ view is most associated with two leading Australian economists: Corden and Pitchford. This recent RBA paper gives a fascinating, and extremely readable, overview of the evolution of the debate over the current account in Australia.
Just as the ‘consenting adults view’ offers a useful armature for thinking for believing deficits/imbalances in general terms can be less threatening than first blush may suggest, evidence has steadily accrued over the last couple of years that the US deficit itself is less threatening and worrisome than I believed three years ago.
Above all, developments that seemed automatic back in 2004 simply have not happened. Metrics of sustainability have not deteriorated and have, remarkably, even improved in some cases. Specifically, the US’s net stock of external debt – the so-called net international investment position or NIIP – has increased to just over -$2.5 trillion in 2006 but, at around -19½% of GDP, relative to GDP it is no worse than in 2002! And relative to exports – the ultimate metric of sustainability – it has actually improved from a peak of 205% of export revenues to 165% in 2006. Moreover, a further improvement in 2007 looks on the cards. Nor has the investment income balance deteriorated as widely expected. Three years is a long time to be wrong!
Two other key factors have also worked to increase my degree of comfort over the deficit. The $ has obviously now already fallen a long way, particularly against the G-10 currencies. The fact that it has fallen so far without its drop becoming a rout to my mind has greatly reduced the chances of a sudden plunge or ‘Wile Coyote’ moment. For the latter to happen, exchange rate expectations would, I suppose, become extrapolative rather regressive. This can always happen but, given that ‘adjustment’ in the US deficit is now clearly underway, surely this becomes less, not more, likely. From its peak of -6.8% of GDP back in late 2005, the deficit has already narrowed by around c.1¼% points with this year’s $ depreciation and deteriorating US growth prospects promising more to come next year. And without the impediment of near record oil prices, the adjustment process would be even more advanced.
All told, the US deficit, now clearly retreating, looks more sustainable and has been more easily financed that seemed plausible a few years ago. The road ahead is still a long one but the journey home is now well under way. Rather the deficit prompting a painful recession as feared, it seems that US recession (or at least an extremely weak economy for some time to come) will prompt continued adjustment in the deficit.
The imminent release of the Q3 current account data – next Monday – will provide us an important update on these trends. Given we already know the trade deficit data for Q3 ($692.6bn annualized vs. $713.7bn annualized in Q2), expectations for the overall deficit are in a relatively narrow range. I expect a deficit of around of $188bn (market consensus currently $182bn). Relative to GDP, this would leave the deficit at 5.4% of GDP; the smallest since 2004Q3. While a Q3 deficit of this size would represent a relatively modest improvement of almost 0.2% of GDP from Q2’s level, it should nevertheless underline that adjustment is progressing.
Importantly, the petroleum deficit widened by an annualized $20bn or so in Q3 meaning that the ex-oil deficit improved by an additional 0.1% of GDP better. The oil price remains the ‘joker in the pack’ of current account adjustment in the coming months. A sharp fall, say sustained $20 per barrel drop, would dramatically accelerate the current account adjustment already in train. Here’s a useful ready reckoner. Oil demand seems to have stabilized around 3.7 billion barrels per year. A $20 drop would therefore generate a $75bn or roughly 0.5% of GDP improvement in the trade balance. Nice if it happens! Of course, this cannot be relied upon. The conventional wisdom is that the current account deficit needs to fall back to around 3% of GDP to be sustainable over the longer run (assuming a 6% nominal GDP growth rate, this would stabilize the US’s NIIP position at 50% of GDP), meaning another 2¼% of improvement in the current account balance will ultimately be required. The adjustment seen thus far is only about 1/3rd of that required so I an under no illusions that the road ahead is a long one and that the $’s fall so far may not yet be sufficient to ensure long-run sustainability.
The other two components of the current account are, of course, the investment income balance and unilateral transfers. My expectation for a c.$188bn deficit in Q3 is based upon a roughly unchanged $10bn surplus on investment income and a -$25bn deficit on unilateral transfers (roughly the average of the last four quarters). It is the continued resilience of the investment income balance in the teeth of external indebtedness that remains controversial and on which Brad and I first crossed swords just over a year. Importantly, the longer it stubbornly refuses to slide into the red, the less the trade deficit needs to improve to ensure sustainability so the outlook for this component of the current account is of great importance.
The US’s uncanny ability to both continue to generate a positive net investment income flow from a position of substantial external indebtedness and also to escape the full balance sheet consequences of its record current deficits essentially derive from three inter-related factors.
the enormous scaling up of both sides of the US’s external balance sheet with both assets and liabilities soaring relative to GDP as gross capital flows have mushroomed, magnifying the impact of outperformance of US-owned assets;
the very different risk characteristics of these burgeoning pools. The US overseas assets are dominated by ‘real economy’ FDI and equity assets, its liabilities much heavily skewed towards bonds;
the ‘privilege’ flowing from the dollar’s reserve currency status, which allows the US to borrow in its own currency and invest in other, leaving the US implicitly short its own currency.
The US’s strategy has worked exquisitely in recent years. Global growth has been strong (boosting equity and FDI returns), inflation has been low (capping interest rates) and the $ has generally depreciated. In turn, the US’s net investment income balance has taken on a Janus-like quality, with accelerating net returns on FDI more than outpacing steadily increasing net interest payments to the rest of the world.
A year ago I saw little reason to expect this propitious situation to end anytime soon. Nor has it. And conditions look set fair for yet another repeat performance in 2008. US growth (and so presumably US real economy assets) will under-perform growth in the rest of the world while interest rates and possibly the $ are falling. Looks like the bus has been cancelled for another year and all the time, as stressed above, adjustment in the overall deficit continues.
Of course, the US’s luck (skill?) cannot hold for ever. Implausibly large increases in leverage will be needed to sustain the outperformance of the investment income balance over the medium term. Moreover leverage can (and frequently does) cut both ways. The different risk and liquidity characteristics of the US’s balance sheet also mean that different payouts can ensue from different, future states of the world. A scenario of, say, global stagflation and a sharp appreciation in the trade-weighted $ is, of course, the nightmare combination for the US. But this is not my base case.
As an aside, the fact the UK manages to sustain a 1.5% of GDP net investment income surplus in the teeth of a NIIP of -20% suggests that the US’s persistent surplus may not be the unique perversion it is often regarded as.
I’m sure Brad accepts much of this and I know he has posted regularly on the importance of the second asymmetry in sustaining the investment income balance. But I feel strongly that the importance of the first and third asymmetries have been significantly appreciated and deserve much fuller comment. The explosion of gross capital flows and the underlying drivers of financial liberalization, globalization and declining home bias that drive it, in particular, has been unfairly ignored. Not only has this process of ‘scaling up’ been crucial in enhanced the US’s ability to lever its implicitly ‘long’ equity & G-10 currency, ‘short’ US dollar and debt position but it has left the financing of the US current account remarkably stress free in recent years.
As stressed in his introduction, the importance of these flows and the role played by central bank finance plays is where Brad and I genuinely part company. Brad tends to see private inflows and outflows roughly canceling each out, leaving the net financing position of the US largely dependent on an unnaturally stepped up level of ‘official’ central bank flows. I regard this distinction as essentially bogus as it requires comparing a gross flow – central bank purchases – with a net balance – the current account position.
Here’s a simple example. Assume the US needs current account financing of $1 but also invests $1 of FDI overseas. Its net financing need is, of course, $1 but its gross requirement is $2. Assume the US sells $2 of bonds to finances itself. A European insurance company purchases $1 and an Asian the $1. Can we say who is financing the current account deficit? I don’t think so.
Moreover, I would argue that when properly anchored in the context of record gross overall capital flows, central bank flows don’t appear particularly high. Rather than bemoan the US’s supposedly brittle reliance on ‘official’ finance, I think we should focusing on the US’s unheralded ability to attract private capital inflows on a scale unprecedented in history for such a large economy. The BEA data over the last year reveal how enormous gross capital flows have become. In Q2, capital inflows into the US were a mind-blowing $2.48 trillion annualized or 18% of GDP! Never before has so much capital flooded into the US. Nor is this a quarterly flash in the pan. Given the inherent volatility of the quarterly capital account data (something to bear in mind when thinking about the Q3 data), let’s look at average inflows in the year to 2007Q2. These averaged a record 16.3% of GDP over the last four quarters.
To put this in context, gross inflows into the US in the 1990s averaged 5.0% of GDP! And what is the composition of these record capital inflows in terms of the split between ‘private’ and ‘official’? In Q2, private capital inflows were a huge $2.201 trillion (16% of GDP) while ‘official’ inflows were a more modest $280bn annualized; almost ten times as large. Brad is convinced that these BEA data are a serious underestimate and will be revised higher. I have some sympathy with this view but even if they are, the fundamental picture would not be changed.
No-one questions that official inflows are undoubtedly high in absolute terms but the strength of these flows surely needs to be measured against the incredibly elevated level of total capital flows currently prevailing. As a proportion of total capital inflows, the latest BEA data suggest that the size of ‘official’ flows is basically ‘normal’. Their long-run (1977-2005) share of total inflows is about 16%. In the year to 2007Q2, they accounted for just 19%. Looks normal to me!
Like inflows, the scale of US capital outflows has exploded. Capital outflows amounted to an annualized $1.877 trillion in 2007Q2. Smoothing the data over the last 12 months shows that the latest quarter is no aberration but part of a secular trend. Over the year to 2007Q2, they averaged $1.418 trillion or 10.5% of GDP. Referring back to the private inflow data detailed above, we see that, in Q2, private capital inflows exceeded outflows by around an annualized $300bn in Q2 and, over the last year, by some $368 billion ($1.786 trillion of private inflows vs. $1.418 trillion of capital outflows). The ‘Setser’ way to cut the capital flow data over the last year is therefore to argue that the $793 current account deficit over the last year has been financed roughly ‘50/50’ by net private capital ($369bn) and ‘official’ flows ($416bn). And of course if central banks flows are under-recorded, then the role is even more important.
But this is simply one of several ways to compare the gross flow and net balance data and so make generalizations about the financing of the current account. Alternatively, it would be wholly correct to argue that that central banks’ ‘official’ purchases have financed around 30% of record US investments abroad over the last year while surging private capital inflows financed the remaining 70% and, of course, all of the current account deficit. The implications of presenting the data this way are quite startling. For example, if central banks stopped buying $’s overnight, the US would need to cut its current record spending on overseas assets by around 30% (from 10.5% of GDP to around 7% of GDP and it should be remembered that US capital outflows as recently as the 1990s only averaged 3.5% of GDP) to finance the current account deficit, assuming an unchanged rate of private inflows.
The obsession with ‘official’ inflows into the US seemingly arises from two controversial conclusions. First, that central bank purchases are somehow special, if not outright ‘abnormal’. Flowing on this is the usually tacit but sometimes explicit assumption that central bank purchases may prove more ephemeral or footloose than more inherently normal private capital flows. Both assumptions are highly dubious. A counterweight to the first argument is that central banks may be serving as financial intermediaries, compensating, in the case of China for example, fro immature and callow financial systems.
‘Special’ or not, it also highly questionable whether central bank finance will be more footloose than private capital flows. More likely, with considerations of profit and return less important than for private investors, central banks are unlikely to affect violent shifts in their portfolios. Not only would such an act crystallize significant capital losses on central bank’s existing holding, it would also increase upward pressure on their currencies; the avoidance of which was a key motivation in acquiring US financial assets in the first place. Rather it is more likely that private investors, whose prime motivation in buying US assets is more likely to be purer considerations of expected return, will prove the more footloose. Central bank selling of US assets may exacerbate any possible US financing crisis but it is unlikely to trigger it.
Viewed in the proper context, current account ‘concerns’ should properly crystallize around the US’s reliance on record private, rather than ‘official’, inflows in financing its deficit. The spotlight therefore now turns inevitably to the sustainability of the current record pace of private capital inflows. If the US can continue to receive private inflows anything like on the scale it has seen over the last year then financing strains should continue to be almost entirely absent.
Here the composition of the record inflows over the last year matters. One cause for concern is that short-term bank lending flows, clearly the most footloose element of capital flows, have picked sharply over the last year. Indeed, at 6.9% of GDP in the year to 2007Q2, they have begun to outpace FDI and portfolio investment inflows for the first time in over twenty years. But the apparent reliance of these flows may be less threatening than it appears. Banks of course make money by using their short-term low yielding liabilities (i.e. deposits) to acquire longer-term, higher yielding assets: in the jargon, ‘maturity transformation’.
With deposits flooding in the year to 2007Q2, US institutions have understandably put this money to work in higher yielding foreign investments. The explosion of US investment overseas over the last year is therefore probably the other side of the coin to the record bank inflows that the US has received. If one dries up so may the other, leaving the US’s financing position essentially unchanged. I suspect that this is what the Q3 capital account data on Monday will essentially show; Brad will probably see it differently. The moral of the story must be that the asset and liability sides of the US national (or any) balance sheet cannot be viewed in isolation.
I, as usual, have said more than enough. But two final points in conclusion.
First, the lesson of recent years has been that foreign demand for US assets – both private and ‘official’ – appears more structural and hence more sustainable than anyone thought. The question ‘what is the price of money?’ is a very difficult one for economists to answer. But the marginal productivity of holding $’s appears to higher than most economists believed.
Second, capital flows can reflect either or both of two separate and distinct phenomena – reallocations of portfolio stock and allocations of newly created wealth in established patterns. The former occur in response to changes in current and expected yields and in perceived risks. They are essentially one-shot effects. A key question is whether the $ has now fallen enough to prompt big stock adjustments and if so in which direction. $ positive or $ negative? I know Brad will have strong views.