Thursday, February 13, 2020

Imperial Circle Part 2: The Long US Asset/Short Global Growth Feedback Loop

This post will not be based on any specific trade ideas but rather on what appears to be one of the more dominant macro themes out there and one I have been spending a lot of time looking at. Most of the things here are very well known, the world is very long US assets. However, I don't think many people are looking at the interconnectedness of it all, especially as it relates to the real economy. 

The most obvious macro theme is and has been being long S&Ps, USD and bonds. There are many reasons for individually each of these have done well. The question I am interested in is what has them so interconnected, self fulfilling and sufficient in terms of durability.

The basic point is, there are two ways in which the not only do bonds/S&Ps/USD do well on their own, but actually create a positive feedback loop for each-other that has proven to be remarkably durable. The consequences have been massive US asset outperformance and lower global GDP mostly via the dollar. 

This is what is unique about this version of the imperial circle, it leads to divergent outcomes between global growth and US asset prices. The US has been rewarded for being the only place able to absorb this massive global savings glut, mostly from Asia. The knock on has been, money floods the US, USD rises and then hurts global growth. Not only do these coexist but they have become self reinforcing in a way. Anyway, this my attempt at connecting the duality of money flooding US risk markets and the dampening effect the strong dollar has on global growth as they are clearly connected.  

Two interconnected factors, capital flows and economic effects:

1) Capital flows

The $8t Asian saver (Asia IIP) has neither yield nor capacity to deal with the size of domestic savings, so they need to be exported. What has capacity and yield, the US' balance sheet in both corp and UST. So what happens is simple, Asian savings flood the US. However, these pensions/lifers in Asia would rather not take currency risk, and post 2018, FX hedged yields in terms of USTs were actually negative relative to domestic JGBs KTBs etc. 

So two things happen as the US is still the only place that can take this capacity. Asian savings move out the risk curve and takes more naked FX risk. And they did both. We know this in places like Taiwan were lifer assets are 150% of domestic GDP, now over 20% of their book is FX unhedged. 

So when FX basis crushed yield pickup in the US, Asian savings transitioned into IG/take more FX risk. 

(data from IMF GFSR 2019)




This chart is a ratio of Lifer assets relative to size of domestic corp bond market according to the IMF. As I've seen Mark Dow post on twitter, there is a massive global asset shortage, it's very pronounced in Asia and it has forced this money into the US. Pretty much for all of surplus Asia, the size of lifer assets is +10x the size of their domestic IG market, with Japan the extreme at over 20x. 



2) Doom loop in terms of lower global output

There seem to be two key parts to the dollars relation to the global economy. The way this ties into asset shortage part is, demand for US assets has been an important factor in USD strength. 

One, we know USD tends to do well in times of weaker global growth because the US is by far the least levered DM economy to China. DM exports still account for +20% of GDP, in the US that number is closer to 12%. One of the reasons that structurally growth is so low is because global trade is not the engine it once was and DM is just as levered to it. This is why the market is crying out for fiscal, there needs to be a composition shift away from NX to C+I in terms of GDP. Anyway, the point is, the dollar tends to outperform when global GDP is weaker because it is less exposed to global trade.

The other angle is, the dollar itself can hurt global trade and output. This is Hyun Song Shin at the BIS and Gita Gopinath when she was at Harvard. Gita takes the global invoicing angle and Hyun looks at it from a financing perspective. The focus of Hyun's recent paper in October is, a broad appreciation of the dollar dampens international trade by weighing on the operation of credit intensive global value chains (GVCs). 

From Hyun, USD inverse relationship with global trade.

shin speech.png

If we combine these two angles of (2), one that the dollar outperforms when global growth is weak due to lower beta, and that the a strong dollar can "impose" global trade weakness via the credit supply channel, a pretty brutal doom loop is formed. The dollar rallies because global growth is weak and then imposes pain on GVCs which makes global growth even weaker and the dollar rallies more etc.

As Mark Carney said at Jackson Hole last year, this is a natural problem with the US share of global GDP shrinking but USDs role is not.

This chasm between US as percentage of global GDP and the role the dollar plays is structurally disinflationary. 

"These dynamics are now increasing the risks of a global liquidity trap. In particular, the IMFS is structurally lowering the global equilibrium interest rate, r*, by: - feeding a global savings glut, as EMEs defensively accumulate reserves of safe US dollar assets against the backdrop of an inadequate and fragmented global financial safety net; - reducing the scale of sustainable cross border flows, and as a result lowering the rate of global potential growth; and - fattening of the left-hand tail and increasing the downside skew of likely economic outcomes." - Mark Carney 

(data from Mark Carney JH speech)



Combining the rush of foreign capital into the US with the strong dollars global effect

Now we add in (1) and the fact that all this excess savings is going to basically to one place, and you have a new sort of imperial circle but one that is slightly different mechanistically than the one Soros diagnosed in the 80s. But, one that is also creating this sort of feedback loop, however this one is connecting the financial economy and the real one in a very divergent manner. 

Soros Imperial Circle: combination of high rates, fiscal and strong dollar all together create this virtuous circle.

Current dollar feedback loop: a strong USD, lower global growth and the global savings glut feed off each other in way that suppress real interest rates and elevates asset prices.

The current regime offers stronger durability

Eventually what broke the imperial circle was excess dollar strength. However, there is something fundamentally different now, the low neutral rate world has been effectively making sure this dynamic of flooding US markets continues. So what happens:

Real money buys US financial assets:

Step 1: credit tightens and stocks go up

Step 2: dollar goes up

Step 3: Even though risk is bid, bonds actually go up in price because they have to price the second derivative, strengthening USD and the negative effect that has on growth.

And step 3 shows how this cycle continues in an immensely durable way. The dollar and the bond market effectively keep each-other in check. And of course, given the low neutral interest rate world and asymmetric monetary policy response function, the Fed can backstop that bond market reaction. Which is what we saw in 2019. The Fed had to cut to stand still in an FX sense given what effect the tariffs were having and the Chinese using the exchange rate as a shock absorber. 

So the dollar goes up, which triggers a stronger and flatter bond market, which keeps the dollar in check and excites asset markets. That is the durability element. 

Conclusion:

The US is in the process of a monster sucking in of capital from the rest of the world. This is very easy to see in the IMF BoP data which basically shows a 1:1 match of big positive NIIP positions equaling out in the US' negative one.Basically the world saves a lot and then invests in US because it has the combination of best growth and maybe more importantly, the most capacity.

Global NIIP v US (ex Taiwan, which is fifth biggest in the world)

(data from IMF)



Then we combine the current dynamic of these two forces: the US sucking in global savings and the dollar reflecting/contributing to lower outcomes in terms of global output. It seems like the one of the biggest macro dualities effecting global markets and is it very durable as the feedback loop is very well reinforced. Savings come to the US, the side effect of that is USD constrains global growth and that reinforces the short GDP trade which is long duration and USD. 

The question now is, what ends this regime, rate differentials haven't mattered yet



In theory, the dollar should have been a compelling short for much the past few months as it set up in such an asymmetric fashion. Either growth around the world was going to pickup or the Fed was going to cut more. Interest rate differentials have topped in the developed world and the dollar will have to price that. Neither have really happened, and DM rate differentials continue to narrow. 

In a low r* world, with most of the developed world at the ELB, in risk-off periods the Fed is the only ones with ammunition. This fear was very prevalent last year, so much so that it forced the ECB to get out ahead of this potential narrative by cutting and more importantly trying to push out the perception of the lower bound in order to keep the EUR capped into Fed cuts. This eventually died out after the September GC meeting. 

Looking at things now however, their fear may have have been misplaced to an extent as narrowing interest rate differentials post facto are having almost zero effect on exchange rate valuations. The question is why. 

The basic idea is, we may have transitioned from measuring the degree of rate differentials to once we establish them, it's about leveraging them out the risk curve. FX has become much more sensitive to capacity in terms of savings absorption than rate differentials in pure isolation, especially in DM. Japanese real money is not going to repatriate because the differential between JGB yields and USTs is narrowing, the problem the GPIF has is not only yield, although mismatch is huge problem for real money, but the other problem is there are not enough assets for them to invest in at home. As the chart above shows, in Japan lifer assets are more than 20x the size of the domestic IG bond market......

The broad dollar recently has been more correlated to relative risk asset performance than yield differentials.



With that in mind, what are the two things that can change this dynamic:

1) US idiosyncratic risk that causes US equities to underperform global peers. 

2) Global growth pickup led by China as the US less levered to Chinese growth than everyone else. 

We can see (2) in the correlation between rising USD and lower global r* via Laubach Williams estimates. 

(data from NY Fed)



The dollar is a both a result of a disinflationary world but also is a culprit in it.

This is a bit confusing but I think it is one of the most important points. I think Taiwan is a very good example of this. Basically, within the current account there is war between the financial side and the goods side. The goods side does poorly in this low NGDP world as agg demand is low. However, the financial side is effectively short these outcome via long USD exposure and a massive duration position in US fixed income. 

What would happen if 2017 reappeared and a Chinese led demand spurred a reflationary episode that pushes up global GDP and global trade and inflation. One would think in that world, USDTWD would be a sell as Taiwan being a a key supply chain would be a large beneficiary of this pickup in global trade volumes. However, and even in the context of a good risk environment, that could pose a problem to the Taiwanese economy. The lifers have a huge effective long USD exposure via over 20% of their foreign asset position FX unhedged. Brad Setser has done amazing work on this and he shows that if USDTWD were to drop 10%, the lifers would incur over 12.3b USD of losses and they would need to rollover the hedged part of their book even as they were taking losses on the unhedged part, that reduces their capital.

Taiwan is of course an extreme example of this, especially in terms of unhedged risk as the CBC has used intervention to protect the lifers. However, it is just an extreme example of a prevalent trend, Asian savings is short their own real economy. Surplus economies in Asia still get +50% of GDP from exports.  

In its current form, the second derivative of this trade is in a sense hurting their domestic economies via the dollar. 

Short NGDP = long US assets. This chart is US NIIP v SPX/RoW equities ex US. 

niip v spoos.png

To conclude, to me there are few takeaways from this:

1) The world has a short global growth position on via the US on the financial side, while on the output side is still highly levered to it via high export dependency. 

2) There is a massive global asset shortage that is forcing the most money into the place with the most capacity. It is also rewarding it via the currency. 

3) This phenomena is creating a self fulfilling loop in which US equities, the dollar and the bond market reinforce each-others strength. 

Money comes into the US to buy financial assets => hedging costs become prohibitive so more is unhedged => USD stays strong as their immense demand for it from the massive global savings pool => because of USD strength and the effect that has on the global economy, the bond market stays bid and flat => low and flat curves force money further out the risk curve and this circle starts again.