As most know by now, over the past month or so, pressure on the currencies of EM deficit countries has intensified again. Goldman's EM research group, however, remains negative on EM FX, bonds, and even stocks suggesting using any strength, like this week's exuberance to add protection or cover any remaining longs. Central banks in most of these countries have become more active in attempting to stem pressure in the last two weeks. But with a Fed decision on ‘tapering’ looming, investors have also become more cautious and are now focused on the parallels with prior crisis periods. In what follows, Goldman provides some concise answers to the questions on the EM landscape that we encounter most often, confirming their longer-held bearish bias.
Goldman Sachs: Quick Answers To Tough EM Questions
Q. Why have EM assets performed so poorly over the past several months?
A more challenging environment as global rates rise and China downshifts
Much of the recent poor performance of EM assets is linked to two dynamics that are likely to be with us for some time:
(1) A post-housing-bust normalisation in the US. As the US shifts to above-trend growth and the Federal Reserve gradually reduces its unprecedented monetary stimulus, this places upward pressure on EM rates and downward pressure on currencies. For those with financing needs – and above-target inflation – managing this kind of adjustment is likely to remain particularly challenging.
(2) A downshift in Chinese growth as the authorities focus on credit excesses, the anti-corruption drive and ensuring the sustainability of medium-term growth. This puts downward pressure on China-linked assets, especially commodities and EM equities.
Through most of this year, one or other of these forces – and sometimes both – has been in play. The most recent leg of EM asset underperformance, focused on FX weakness in Indonesia, India and Turkey, was clearly related to the first and to the expectation of Fed tapering later this month.
Arguably, the big picture story is even simpler. The US housing slowdown, DM demand collapse and subsequent deleveraging – as well as the unprecedented monetary easing to deal with it – forced a series of adjustments on many EM economies (Exhibit 1). They boosted domestic demand to offset the external shortfall, through monetary, fiscal and sometimes direct credit stimulus. Currencies appreciated as capital flowed in – a pressure that was partly mitigated by reserve accumulation and partly by sharper cuts to interest rates than domestic conditions alone would have justified. This mix was generally a success in cushioning the EM universe from the DM slowdown. But it often came with stronger FX, significantly lower interest rates, booming credit and non-traded inflation pressure, and a marked deterioration in current accounts. Part of China’s growth downshift is linked to structural forces, but the credit problems that have been in focus recently also stem from the aftermath of the late-2008 shock.
Now that the US housing market has turned the corner, a positive US impulse is bringing about the reverse adjustment and higher rates (Exhibit 2). But the reversal is likely to be harder, as capital inflows retrace more quickly and also because the starting point is more challenging. EM current accounts have deteriorated to their weakest levels in years, credit has boomed, inflation has remained above-target, currencies are too strong and interest rates are too low. Moreover, as we have highlighted elsewhere, EM economies and assets have benefited from a number of tailwinds over the past decade (many of which were dependent on the impact of China’s high growth) that are likely to fade or even turn into headwinds. These include a progressive deterioration in the terms of trade for commodity exporters and an erosion of low inflation credibility. In that sense, these are ‘not your older brother’s emerging markets’ (see Global Economics Weekly 13/23).
Q. Given the size of the recent moves, is the pressure on EM assets mostly behind us?
Unfortunately, no. The normalisation of global rates has only just begun.
The twin headwinds from the post-housing-bust normalisation in the US and China growth downshift are likely to be around for some time to come (Exhibit 3). While we are a bit more optimistic about China’s cyclical picture in the near term, we have outlined the medium-term risks from their long credit boom (see The China credit conundrum, July 25, 2013). Furthermore, we doubt those risks will diminish markedly in the next few years. We also think that the process of US growth and rate normalisation is likely to be a dominant dynamic in the global economy for much of the next two or three years, with US 10-year yields steadily heading back towards a more normal level at a pace that may still be faster than the forwards discount. The flipside is that the rebalancing of demand, and the FX and rate adjustments this brings with it, still have a long way to go.
Even after the recent poor performance, most EM assets still do not look particularly ‘cheap’, although EM equities are closer to that point. This implies that our strategic view of the EM landscape is still quite cautious – as it has been since the start of this year – and we think we are in the middle of a rebuilding of risk premium in many places.
Our strongest conviction is that, on average, EM currencies – particularly in deficit countries and those with weaker institutions – will weaken further against the USD. But that in turn is likely to come with higher real (and nominal rates) in many places, as the declines of recent years reverse. We think that in this environment, EM sovereign (and, perhaps more obviously, corporate) credit spreads probably still need to widen more. In EM equities, which have underperformed since late 2010, there is likely to be lower downside than in other EM assets. Relative to the other asset classes, China’s growth dynamics may ultimately play the larger role, rather than the US rate adjustment (Exhibit 4). But given the pressure on EM rates, currency and credit, we still think the risk-reward in equities over the next year or two is higher in the US and DM in general than in EM.
Q. How far could these moves go?
Different approaches give different answers but all suggest the moves in FX, rates and credit can extend further.
Working out where various EM assets ‘belong’ or could move is a key challenge and something we plan to focus on more in this publication in the coming weeks. Different approaches give different answers. But they generally support the idea that there is still plenty of room in many places.
One simple way is to look at how much the shifts in EM currencies and rates have reversed either since the US housing market slowdown started in late 2005/early 2006 or since QE began in November 2008. Exhibit 5 shows the shifts in real trade-weighted currencies across EM since then. While there is a lot of variation, rates in many places are well below and currencies well above those levels.
This is a useful benchmark, but there is little to suggest that EM currencies were at the ‘right’ levels in 2005-06. We can focus instead on the kinds of currency adjustments that may be needed to accompany the narrowing of current accounts in some of the places where they have deteriorated most. Such exercises are quite sensitive to assumptions, but they generally show that quite large (20%-30%) TWI depreciations could ultimately be needed in many of those economies. By the same token, EM policy rates are on average around 200bp below where a US-style Taylor rule would set them given their output gap and inflation dynamics.
Benchmarking sovereign credit is harder. EM sovereigns have already decoupled substantially from the tight relationship that they held with US corporate credit through most of the post-2008 period. But even after the latest pressure, sovereigns are generally well below the levels seen in the periods of Euro area stress in 2011 and 2012. We think we are likely to see those levels again in the next year or two.
Q. Will it be another 1997-99-style EM crisis?
No, but different problems could surface instead, particularly if monetary tightening hurts domestic banks and corporates.
The recent sharp pressure on the IDR and INR in particular has brought back memories of the Asian and broader EM crises of 1997-99. In those cases, initial currency weakness led to a spiral of further currency declines, widespread corporate and banking stress, and deep recession.
The key question at this point in time is whether the currency depreciation that has already begun – and is needed – creates a feedback loop that turns out to be destabilising. In the Asian crisis and its aftermath the answer was a clear yes. The main reason was that companies and banks had taken advantage of currency pegs and low offshore interest rates to borrow short-term foreign currency abroad in large quantities. This meant not only that rollover risk was high. But currency weakness, once it began, destroyed corporate and bank balance sheets and crushed domestic spending, increasing the pressure for capital flight. With very high leverage rates, interest rate defence of the currency – when it ultimately came – also created serious private-sector pressure, which turned the defence into a double-edged sword.
This time, that dynamic is much harder to see (Exhibit 6). Foreign debt levels – including short-term debt – are much lower than they were then, current accounts in general are smaller and reserves are much larger. Only Ukraine runs a traditional peg, and even Turkey and Indonesia (which have lower reserve/debt ratios than many others) are well below the Asia crisis levels. Funding conditions may become more difficult for current account deficit countries, but the rollover risk is probably manageable. So the real question is whether there is an alternative story in which feedback loops set in, or whether currency weakness can be allowed to happen (perhaps alongside some modest restraint of domestic demand). There is more to be said about this, and we plan to expand on it in the coming weeks.
The most obvious channel through which this can occur is high inflation related to currency weakness. The risk is more acute in places where non-traded-sector inflation is already high, and where accepting currency weakness may erode inflation credibility further. The direct inflation impact of currency weakness should – in theory – be temporary, so for those with substantial output gaps and a below-target starting point, the pressure here should be absorbed. But in places such as India, Brazil, Indonesia and Turkey, inflation is already running above targets and above politically acceptable levels. In these countries, we are likely to see continued efforts at slowing the pace of depreciation with front-end tightening.
However, that response is not pain-free either. While it may succeed in stemming depreciation pressures – and the evidence suggests that decisive action is often needed – it could extract a heavy toll on domestic activity, moving market weakness instead to domestic corporate credits and equities in places where credit growth or leverage is high. And if investors sense a reluctance to go down this route for that reason, currency pressure itself may increase. So this could force some unpalatable choices, even without direct FX exposure on balance sheets, although the immediate risks would be to banks and corporates, not to sovereigns.
Q. What could stop the pressure in the coming weeks?
A dovish Fed, a more decisive EM policy response, and better China data would all help in different ways for different assets, although probably only temporarily.
Notwithstanding our structurally cautious view on EM assets, in light of the recent sharp moves a key question is whether to press the short view tactically or if a bout of relief is likely. Much of that comes down to assessing how much the two dynamics that have beleaguered EM lately are likely to be in play, aside from domestic policy responses in EMs themselves.
For the EM rates and FX universe, US rate relief is the most plausible relief valve. Our own views of the upcoming FOMC are on the more dovish side. We expect tapering – as does everyone else – but we also expect it to be accompanied with strengthened forward guidance. And although we are structurally bearish on US Treasuries this year, we are tactically neutral here. It is also possible that the path of US yields around the tapering announcement mirrors the experience of past QE announcements, i.e., a sell-off ahead of the announcement followed by stabilisation or even a modest rally after it. If this comes at or around the same time as the announcement of a new Fed Chairperson who is more dovish than expected, a period of stability in US yields could bail EMs out of the current round of pressure.
On the EM side, a much more decisive and credible tightening at the front end could also put a stop to the FX pressure. We have seen hikes in Indonesia and Brazil in the last two weeks, reserve policy in Turkey, and some fresh measures in India yesterday to boost foreign currency inflow. But if investors continue to sense a reticence to take more decisive action with policy rates, given the costs to the domestic economy, it may be hard to stop currencies from moving further. And the steepness of curves implies a high investor threshold for hawkish surprises. Overall, we still think the medium-term risk is for higher yields in the face of improving US growth and that it makes sense to keep duration short.
Somewhat better signs of data from China are arguably more risky for a blanket ‘EM short’ in the near term. So far, the shift on this front has been small (see Asia in Focus, September 3, 2013), but for the first time in several months the market’s view of China growth is improving. And although EM PMIs have not yet mirrored the improvement in DM PMIs, we think they will, to a degree, especially in countries such as Korea, Taiwan, Poland and China, which are most plugged into the global trade cycle. EM equities would be the most obvious beneficiary of that kind of cyclical shift, more than rates and credit. It is also possible that the same dynamic could see pressure on some China-sensitive currencies (the CLP, MYR and possibly the ZAR) ease temporarily. Despite our overall caution, we do think a period of better EM equity performance is now possible, like the one we saw in the last few months of 2012.
Q. How do risks vary by country?
Brazil, Turkey, South Africa, India, Indonesia face the largest challenges, but Thailand, Malaysia and Chile may also have some difficult adjustments ahead. The CE-3, Korea and Mexico look better positioned.
Brazil, Turkey, South Africa, India, Indonesia and Chile have been at the epicentre of pressure at different points in the summer on account of their current account deficits (Exhibit 7 and 8). This is likely to remain true. Given the adjustments globally that we have described, the biggest vulnerability is likely to come from those that a) need weaker currencies, b) already have inflationary pressures, c) have seen significant domestic demand and credit-related booms and d) have weaker institutional capacity.
Emerging markets with more solid current account dynamics are likely to outperform. Korea, the CE-3 and Mexico have (for varying reasons) managed domestic demand conservatively, with restrained import growth. Some of these are also most likely to benefit from their high exposure to G3 economies, and the continued signs of strength there.
Between the two extremes, there are countries whose fundamentals are better in a static sense, but that have deteriorated significantly as of late and could experience market pressures, particularly on their currencies. We think Thailand and Malaysia fall into this camp. Israel could also join the front-end tighteners if the ILS stops strengthening and policy makers focus on the stronger domestic housing and labour market picture.
Differentiation may not be a theme for all seasons. During broad and rapid EM sell-offs, we are unlikely to see assets rally in some countries. At best, we can look for KRW-like stability. During times of broader stabilisation, the weakest markets should continue trading at weak levels, while the better-placed ones should recover.