China and the Renminbi, an EM FX Macro perspective

Photo by Eric Prouzet

21 April 2023

This week I have been focusing on China and the Renminbi, it’s a big topic, that will crowd out some of the usual analysis. After a brief summary of the past week, I will talk about the following:

• Broad Thoughts on China

• RMB currency regime

• China's balance of payments dynamics

• Further flow analysis

• China Macro

• Longer term trends

• Valuation metrics

• Market monitors (the usual set with a China focus)

The past week has been light on narrative-shifting new information and the FX market has washed around accordingly. I maintained positions short USD against MXN and BRL and watched them bounce without feeling compelled to adjust course.

There were macro data releases of note. On Tuesday we had a slew of positive data out of China with a solid round of GDP data, showing particular strength on the services side. There was a substantial upside surprise in retail sales and a modest downside surprise in fixed asset investment. This speaks to the narrative of an upswing based on pent-up consumer demand rather than driven by government spending. The credit data last week suggests this is also part of the story here, only less dominant than in previous cycles.

The glacial turn in US initial claims continued Thursday with a 245k print. The week ended with a pickup in flash readings of the US S&P Global PMIs, with manufacturing rising above 50 for the first time since last October and services also making gains.

Central Banks

In China, we saw the PBoC maintain policy rates at 2.75%. The soft low inflation picture might merit easing on its own, but I would imagine the PBoC would like to monitor the current post-zero-covid upswing before acting. The Bank of Indonesia maintained policy rates at 5.75%, which fits with the regional trend.

Four of the countries I cover have central bank meetings next week and all are likely to be in wait-and-see mode.  These are Colombia, Hungary, Russia and Turkey. In the case of Turkey, looming elections will delay any action on building imbalances.

Broad introductory thoughts on China (for what they’re worth)

(Those who follow China regularly can skip the first couple of sections)

To me, the People’s Republic of China is a fascinating new experiment based on what the West understands to be a flawed system of central control. If I were an alien living on another planet I’d be keen to see how things turn out. As a member of this planet, I’m concerned about the seemingly inevitable conflict with the West, and the consequences for living standards on both sides. Not to mention the risks surrounding the future of the island that shall not be named. In a world that requires a sharp acceleration in global cooperation to tackle climate change, the growing tensions between the US and China will likely have fatal consequences for many in countries with little influence on either.

Under Xi, the experiment with central control enhanced by 21^st-century processing power has a lot going for it. They have been adapting on many fronts to deal with past fragilities. On the currency, enormous surpluses combined with enhanced capital controls facilitate an iron grip. Both mean that internationalisation efforts are inconceivable despite the flurry of recent opinion pieces to the contrary. Growth has been steered higher without a financial crisis for significantly longer than any of my many bearish past hedge fund colleagues believed possible. The levers of growth relied upon for over a decade have been real estate and infrastructure. Both these levers continued to reliably deliver a boost to growth for longer than most expected. It’s possible that the real estate lever is now materially damaged, although perhaps it’s too soon to say. The huge pile of foreign assets amassed has enabled overseas lending across the emerging world. This has already put China front and centre of a number of debt restructurings, with more likely. I can’t do this topic justice in one note, so forgive any areas I skirt over. In that spirit, I’m going to try to stick mostly to my FX focus as usual. With that, I understand that with a managed currency like the RMB, politics is a key part of the picture. I have many thoughts about how things might go wrong in China, but there are far superior minds to describe the risks. I heartily recommend anything written by George Magnus, who I used to follow when he was an economist at UBS. In particular, in his excellent book, Red Flags, George Magnus tackles all of the key issues.

George wrote a piece on last month’s National People’s Congress (NPC) where the Chinese Communist Party (CCP) set out its main economic targets and priorities. He reports that post-zero-Covid there is a focus on getting the economy back on track. In that, they seem less confident than the market in the ability of pent-up demand to do the job for them. They specifically mention the need to ‘boost market confidence’. This desire seems to be contained within the constraints of the current framework of SOE dominance and a firm hand on the private sector’s rudder. If they are concerned about growth, they sound more resolved on the risk management side, unsurprisingly given the fragilities created in the real estate sector that became so apparent to all last year. There seems to have been some turnover among senior players, and a perceived strengthening of Xi through the promotion of loyalists.  The 5% growth target for 2023 seems modest when compared to market expectations. The Budget report allows a modest rise in the fiscal deficit to 3% this year. The IMF forecasts from last week’s World Economic Outlook show their assessment of the fiscal outlook, and I include a regional comparison above.

China currency regime

The current structure dates back to 2015 when the People's Bank of China (PBoC) announced a policy of setting a central parity rate each morning based on the prior day's close. On August 11, 2015, the PBoC facilitated a 3% devaluation which, whilst small by normal FX standards had a significant impact on global markets partly due to expectations of further falls. This was later interpreted as a policy error by many due to the ensuing capital outflows triggered by the move.

CNY can be monitored against the CFETS (China Foreign Exchange Trading System) basket, which uses the official set of weightings. The CFETS website provides the below alternative reference indices and publishes weekly data.

With any managed exchange rate regime, a country is afforded the opportunity to enjoy a competitive advantage if the rate is maintained at an artificially low level. Large current account surpluses and the robustness of capital controls will define the likely effectiveness of efforts to maintain a valuation distortion.

Despite having not been actively involved for a few years, it’s my assumption that the market considers the daily parity rate when compared with that which is implied by the basket as a signal of policy intentions, and there might be some value to trading the momentum of this differential. There are indeed periods of strong momentum in the trading history of the CNY against the basket.

It’s my understanding that offshore investors usually choose to trade the CNH, which generally moves with CNY but with some fluctuations. The PBOC can manage the CNYCNH rate by intervening in the CNH market. Short-term CNH-denominated debt in Hong Kong can be bought/sold by state banks to alter liquidity conditions and impact interest rates, creating a premium/discount on the CNH. These activities and the intentions behind them will impact market participants’ views on the intentions of the authorities and in normal conditions traders would follow the intentions of the mainland policymakers. Nonetheless, CNH rates fluctuate more than CNY rates and that is reflected in the vols. For the remainder of this note, I will mention both pairs (USDCNY and USDCNH) somewhat interchangeably as comments about one can apply to both.

We can look at how USDCNH has traded in the past years by making a comparison to two other ‘baskets’. One is the USD index, which is the USD against a trade-weighted basket. Another is USDSGD. The Singapore Dollar has been managed against its own basket for many years, where policymakers make adjustments to how the currency can move to help meet monetary policy goals. This is an example of how a managed exchange rate becomes an explicit policy tool alongside the price and quantity of money.

Here we can see how the trend appreciation of the CNH against the basket between mid-2020 and early 2022 comprised a period of CNH outperformance during USD weakness followed by a period of USDCNH stability during the early stages of the USD uptrend.

As a final observation for this section, it’s important to note the impact of the Fed’s sharp tightening cycle on the carry implications for investors in CNH. The below chart shows a flip from CNH carrying 3% positive on an annualised basis to 3% negative.

Balance of Payments (BoP) dynamics

There are a number of interesting narratives regarding the BoP, from inconsistencies to stealthy reserve accumulation.

China has maintained an impressive current account surplus for decades. This has typically comprised a large goods surplus and a smaller services deficit. This trade surplus is partially offset by a deficit in investment income. Chinese overseas investments are dominated by reserve assets – i.e. foreign government bonds. The yield on this stock of bonds along with other overseas investments is clearly insufficient to offset the returns on assets the Chinese have allowed overseas investors to purchase over the years.

To take a wider look at the BoP I have isolated some prominent items. The Current Account, Other Investment (from the Financial account), The Financial account ex-Other investment, Reserve Assets and Net Errors and Omissions (which is the residual that balances the BoP). Any corporate accountant would be horrified by the size of this term in many emerging markets.

A large balance in ‘other investment’ – within the financial account is not atypical for an emerging economy. In the case of China, FX Interventions performed by state-owned banks can sit in this category. The suggestion is that with more political focus on reserve accumulation, particularly from the US, it might be preferable to disguise this trend. ‘Trade credit and advances’ which are contained within ‘other investment’ are a place where a structural accumulation of FX reserves within non-financial firms can reside. The high level of domestic savings in China also creates investment flows that can end up in this category, some of which will be household FX reserves. The large and growing ‘net errors and omissions’ category also has the capacity to contain within it stealth-reserve-accumulation. These might involve a rise in indirect interventions, off-balance-sheet activities involving FX and informal capital flows (hot money).

I imagine someone has done a much more thorough analysis, but it looks to me like effective FX reserves could be anything up to double what’s officially reported. What this tells us is that the ammunition to defend the exchange rate in the event of a crisis is substantial.  

I stumbled across some work done on discrepancies between the BoP data and the customs data, so I ran a quick comparison myself.

Whilst you would expect some variation, the current difference of over USD200bn is indeed noteworthy. Brad Setzer from the Council on Foreign Relations points out that it’s the current account surplus (not the customs surplus) that goes into assessments by the IMF and U.S. Treasury on currency valuation and exchange rate policy. This discrepancy of around 1% of GDP is enough to move the needle on these assessments, which have political ramifications. The famous proverb that proposes ‘hide your strength, bide your time’, which has been less prominent on the geopolitical front recently, hasn’t totally gone out of fashion.

Further flow analysis (local bank activity)

SAFE publishes data on foreign exchange transactions which in theory could provide some insights into the behaviour of local banks and corporates. I’m straying a little off-piste here and I know there will be those with more experience handling this data, so take the following with a pinch of salt.

Below are spot data which show the balance of what they call ‘settlements and sales’. I understand sales to mean sales of local currency (purchase of USD), and hence the balances they refer to in the tables must be the balance of purchases of local currency.

From the second chart, you can see that the largest aggregate selling of foreign currency is from customers, which will be mainly corporates. This is what you would expect given the structural surplus. This interpretation also makes sense to me given the behaviour of this data during the 2015 devaluation and resulting panic. In 2015 we see an episode with corporates selling local currency and banks forced to provide a temporary counterweight. I thought it worth charting these corporate settlements against the trade balance for context.

This cumulative divergence seems to fit with the plateauing of FX reserves and the rise of ‘other investments’ and ‘net errors and omissions’ in the BoP analysis above.

I ran some regressions to determine if I could find a shorter-term signal for moves in the currency against the basket, I didn’t find anything tangible, although it’s definitely an area for further research.

Recent Comments on Intervention Policy

We’ve seen the impact of past currency interventions, explicitly via the official reserve accumulation and implicitly elsewhere in the BoP. Just this week we’ve had headlines from PBoC Governor Yi Gang about winding down currency intervention. Nikkei Asia His comments during the IMF spring meetings look like a strong nod to conventional orthodox currency policy. This looks like part of a marketing exercise for China’s future role on the world stage. My framework would be that currency policy exists to serve the strategic goals of China. If the priority was to embrace Western orthodoxy, free up the capital account and reduce China’s surpluses then intervention may indeed become a thing of the past. That doesn’t sound like Xi’s China to me.

Discussion of the broader macro picture

Currency policy must run alongside the management of the broader macroeconomy. To discuss this, we must briefly run through the current position. 

China has some of the lowest inflation among major economies giving significant breathing room at the start of an economic upswing. As we can see from the IMF chart on the right, above, mobility monitoring data seems to corroborate the view of pent-up demand being exercised.

Recent PMI data has risen with strong new orders on the manufacturing side. Recent credit data has also been strong (data from (yardeni.com)).

This all amounts to a comfortable position in the short run, with policy stimulus and pent-up consumer demand driving growth in a benign inflation backdrop. Such a strong growth upswing would ordinarily be currency positive for a free-floating currency.

Valuation

On a REER basis, CNY has been broadly stable since 2015 and currently sits approximately 5%, 3% and 2% below its 2, 5 and 10-year averages. Comparison with historical average REER levels are usually a useful long-term indicator, as we frequently see a mean reversion of the REER. In the case of China the Western consensus is that the currency was historically kept significantly undervalued, and hence judging the extent of the normalisation is the real task.

Different models will derive a range of valuations for the RMB. With the large current account surplus a ‘macro balance’ approach would likely tell us the currency is still undervalued. I’d be interested to see what a behavioural (BEER) model would infer.

The 10-year relative yield correlation to the FX has been strong recently, and whilst these relationships are unstable and prone to regime shifts, we do appear to currently be in a regime where the rate differential has been a good guide.

Longer term trends

In a note I put out earlier in the year, I discussed China’s demographic challenges. These are well-understood and documented so I’ll be brief.

From the perspective of the RMB, the literature tells us that these demographic trends matter. The consensus opinion is that rising savings rates that result from an ageing population as we see in China increase the demand for domestic services such as healthcare. This puts upwards pressure on relative prices between non-tradables and tradables, which should cause an appreciation of the real exchange rate through the Balassa-Samuelson effect. China’s problem is not uncommon at this juncture. Where it differs from some other countries with similar demographic challenges is that most of the comparable countries are wealthier and at later stages of development.

From the below forecasts by the IMF of GDP per capita, you can see the impressive development of China. However, there must be a concern that, with its demographic dividend firmly in the rearview mirror, and has already benefitted from urbanisation, strong growth is going to be hard to maintain. This is especially so when you look at how China’s debt has been rising in an absolute sense and relative to its neighbours. Government debt (left – below) only represents a slice of the overall pie.

Estimates of total credit to the non-financial sector are much higher, with suggestions that this exceeded 300% of GDP in the wake of covid. The right-hand chart, from an ECB report, gives a breakdown (the latest data here is from 2020).

Market

To conclude this attempt to run through the FX backdrop and outlook for China let’s take a quick look at some market metrics. I’ll combine this with the usual presentation of my weekly market monitors, but this time highlighting China.

Correlation

CNY is a participant in the recent collapse in the broader USD correlations with USD/ASIA. Here I show EURUSD, but a similar pattern is true for the majors. As we observed last week the RISK/CARRY/OIL dynamic is the dominant market theme in the short term. Broader CNY correlations are in the table below. These show the relationship since 2022 has been weaker in 2023, which is heavily affected by the recent breakdown.

Carry

As we can see from the latest carry and vol data, the vol-adjusted negative carry associated with USDCNY is substantial and becoming more pronounced as a result of the recent decline in vols. The strong performance of carry as an FX theme this year has presumably attracted many to use CNY as a funder in any RV basket. It’s a funder in the one I monitor in these notes.

Vol and Skew

Despite being a negatively yielding pair, USDCNY vols trade higher with spot, like much of EMFX, delivering a structural skew higher (to USD calls).

As we can see above, the recent decline in vol. has come in the absence of a move lower in spot. The skew has been fairly stable of late and looks unremarkable when compared with its neighbours.

When looked at together, the highly negative carry and the upside skew makes structures like call spreads attractive to USD bulls. If one is confident of the current regime or sufficiently well hedged elsewhere then some form of ratio call spread would look attractive, or an RKO if you’re into that kind of thing.

The current setup is that the local macro picture looks positive for CNY, but as is the case so often in China, the way to play the growth upswing is with local rates or other less-managed assets that benefit from the dynamic. As discussed in earlier notes, these have often been commodity currencies in the past.

CNH market model

Finally, as I have done before I have taken non-China market prices and produced a component model to give an impression of short-term value. This implies no recent divergence from what the model would expect, it mainly comprises G10 and EM FX rates.

Signing off

If you’ve made it all the way through this unusually long note, then thank you for your attention. Please share the note with anyone you feel might enjoy it. As always, any feedback and engagement would be very welcome.

In the meantime, if you’re trading, be disciplined and be lucky,

Stephen