The Global Savings Glut, a Modern Policy Failure

Back in February I put out a piece called the "Imperial Circle part 2, the feedback loop between rising US asset prices and slowing global growth". From this angle, I wanted to look at other imbalances either financially or economically and have found a lot of them come from this original savings glut that Bernanke highlighted back in 2003. However, while a lot of these factors are structural, many of them are a function of a policy choice, which is worth remembering as this crisis will likely trigger policy changes. However, without changes, these forces are immense and will continue to exert themselves over markets and the global economy. The goal of this note is to combine a lot of macro themes under the umbrella of a meta theme, there is too much global savings.

One of the reasons the long US short RoW trade has dominated is, many DM countries have the trade on, either implicitly or explicitly, or both! Going forward, the question is, is the virus enough of a force to change these dynamics. I think it is, but it will take time. The French+German proposal today could be the start of that change.

Sections:

- How did we get here. Globalization ended in 2011 and no one adjusted. Export based policy and high savings rates reinforced each other even as globalization forces weakened starting in 2011. The dollar was both rewarded as the place that could accommodate this excess savings but also reinforced the dynamic by inflicting pain on export based economies.

- Too many savers post GFC. Everyone wanted to save, governments pulled back, households were in balance sheet recession (Koo) and corporates had very few attractive investment options. If everyone is saving, someone must be dissaving in a big way. The US did and was rewarded for it. We are operating in a world where there is a massive excess of capital vs. productive places to put it. Which is why valuations on high quality assets able to absorb this savings is so high.

- This dynamic neutered monetary policy. Via slower global growth and an immense demand for safe assets, the neutral level of interest was crushed. In that world, monetary policy is not really easing but just keeping pace. If policy can't get inside r*, its adjusting, not easing.

How did we get here

Going back pre 2008. China wanted to be the world's manufacturer but it didn't want to take the exchange rate adjustment that came with it. And if China didn't want to accept a higher exchange rate, the countries that were selling to them, surplus Asia and Europe, weren't going to be keen to have one either. So there is a gap between purchasing power and money coming, that imbalance is worked out via higher savings rate and a continued rise in the current account balance.

So as trade grows, and countries fight against exchange rate adjustments, rising incomes don't get spent because they are effectively constrained by an artificially weak exchange rate. The world then had a choice, rebalance surpluses into buying more US goods via a fair exchange rate regime or just send into the US via demand for financial assets. They chose the latter.

This is one of the reasons the status quo has persisted. US manufacturing never got a chance because it's constrained via a strong FX and it is why there was never a meaningful pickup in consumption as a percentage of growth in the surplus world. Relative exchange rate regimes reinforced this dynamic of, surplus world doesn't spend and the US doesn't save.

The negative dollar spillover

The dollar problem from this became more obvious as Chinese demand began to structurally fall as the credit impulse weakened. Much of the global economy had a fairly simple model, export to china and recycle surplus' into USD. Exchange rates stay tame and the money is better in USTs/US IG/ Tech sector etc. than anywhere else.

However, the financial side of the economy began to inflict pain on the real side. Money comes into the US, the dollar goes up, and that slows global trade. But it also set up a more troublesome effective doom loop. The dollar would rise, inflict pain on global trade, and then rise even more because the US is a relatively closed economy and less exposed to global trade. And what has been the only way out of this doom loop as post GFC global trade has been relatively weak, Chinese credit expansions.

It is not a coincidence that the only time we have seen real sustained USD weakness since the GFC is post China stimulus episodes (arrows meant to mark three most significant Chinese credit expansions).

Globalization died in 2011, no one adjusted

One of my favorite charts is from Hyun Song Shin at the BIS, ratio of world goods exports to world GDP. It shows a pretty remarkable fact, globalization was dying before Brexit, the Trump election and the trade war.

Despite this post crisis shift, growth composition for many advanced economies has been incredibly sticky with exports still making up over 30% of GDP in much of Asia and Europe.



These two charts are structurally disinflationary. World GDP has changed, but advanced country growth composition has not. And the problem is, from a policy perspective, the response has been to chase after lost external demand instead of reforms that rebalance the composition of domestic growth more towards consumption.

If we look at European policy making from post Euro crisis on, that is basically what happened. External demand started falling, and the reaction was, we'll try adjusting the currency to rebalance. This is how Europe got to negative interest rates while running primary surpluses. The reaction was to chase demand that wasn't coming back instead of investing domestically. That is basically what nirp is, another way of weakening the currency at the expense of domestic demand (local credit channel). Nirp ends up being a sort of tradeoff between the external and domestic sectors of the economy. The world doubled down on trying to save imported demand instead of figuring out how to grow internally.


Massive savings rates, a policy failure

As the world economy was doubling down on an economic model that was clearly structurally broken, savings rates continue to move higher as investment doesn't seem that compelling in weak NGDP world. The world economy shifted and Asia+Europe didn't get the message, so the imbalance between savings and investment grows even wider. And to add onto this, governments were running fiscal surpluses....



So what do we have now. A balance sheet recession with both the private and public sector trying to save. So savings rates in places like East Asia hit 40% of regional GDP. The gov't wants to run a primary surplus, households and business are either repairing balance sheets or are not seeing attractive investment options because NGDP is low. So where does all this money go..... Financial markets have to absorb it. One problem is, the amount of savings in Asia and Europe was far bigger than the size of their domestic asset markets.

While governments weren't spending, monetary policy was doing QE, removing the few government bonds from the market. And, on average, Asia + Europe lifer insurer assets are over 12x the size of their respective domestic IG bond markets according to the IMF's October 2019 GFSR. So there is no risk free assets and not nearly enough investment grade bonds. So where does the money go if there's no place for it at home, to whomever who can absorb it, which has always been the US.

As Bernanke said in 2007, you want to explain Greenspans "conundrum" here it is. The world saves and funnels it into the US. Term premia never had a chance......




The Japan example

The Japanese financial economy was given a tricky hand. The BoJ owns over 50% of the JGB market. And other than a post Abenomics three arrow blip, bank lending never really went anywhere. So we had this immense transfer of JGB holdings from financial sector to the BoJ. Combine the financial world with massive non bank sector (Lifers etc.) and you get a +3 trillion dollar positive NIIP position. Japanese demand for foreign assets grows every year and it is perfectly logical. Lifer assets are almost 25x the size of their domestic IG bond market and the BoJ owns half the JGBs, what else was there to do.



Another example of this, but with very different characteristics has been Taiwan (5th biggest NIIP in the world). In Taiwan, life insurance asset are over 150% GDP. Creating this imbalance has been a central bank that has repressed the exchange rate to defend the tradeable goods sector and now almost more importantly the non bank financial sector which has built up a very large implicit and explicit FX position. These three Asian economies are sending over 1.4 trillion USD into US credit markets.




The savings glut killed monetary policy

This global savings imbalance has created a big problem for monetary policy.

1) It is a position that is effectively short NGDP. If both private and public sector has a savings impulse, both growth and inflation fall. As the world is highly integrated, those conditions are exported. If r* is falling in Asia and Europe, it will be falling in the US as well.

2) So there are two ways excess savings relate to the neutral level of interest. First, it leads to lower growth as money cannot find productive places to invest or spend. Second, a key part of the calculation for the r* is a the demand for safe assets. So growth is slow and savings leads to a heightened demand for safe assets, interest rates around the world fall.

3) This has contributed to a smaller monetary policy impetus. One, it has driven policy rates around the world to the lower bound. But two, it has never really given policy a chance. We know that rate of accommodation (excluding LSAPs/forward guidance) of monetary policy is the stance of policy relative to the neutral level of interest, which of course is unknowable in real time. However, if global savings via slower global growth and an immense demand for safe assets is crushing the neutral level of interest, then monetary policy is not really easing but just keeping pace. If policy can't get inside r*, it's not really easing, it's adjusting. Said another way, savings have forced CBs to cut in order to not be tightening.

r* has become more and more a global phenomena. Data is from Jorda and Taylor, "Riders on a Storm" paper from last years Jackson Hole.



EM has had an uneven relationship with this savings backdrop

The spillover of this global savings backdrop and DM central banks at the lower bound is, the carry trade. In EM, the carry trade was executed in two stages. First, EM's issued in FX denom (Eichengreen original sin), that didn't work. The way EMs fixed this is by issuing a lot more in local currency and given how low DM yields are, INDOgbs or SAGBs, became very attractive. The problem now is original sin redux (Carstens&Shin). It is difficult for these markets to handle this sort of inflow and countries like South Africa, Indonesia, Mexico, end up with around 40% of the local government bonds in the hands of non resident portfolio flows.



So while yes, it is an advantage that low DM rates have forced capital into parts of the world that need it to further their development, it has come at the cost of volatility. These swings in capital flows since the GFC have become the new normal. EM has way bigger issues than capital flows, but these massive oscillations may have ended up doing more harm than good. This is of course nothing new for EM but this backdrop has helped foster a new vulnerability.




Overall: These all seem to be separate macroeconomic imbalances. Slowing global trade, high savings, low r*, EM flows. However, the umbrella in which they all seem to fit under is the world outlined above, a world that saves too much. And what is interesting from both a trading and a macroeconomic point of view, a lot of this was just a policy choice.

Getting out of the pandemic, there are two outcomes for the private sector. One, a liquidity crisis turning into a solvency crisis, or they are saved and develop a massive savings impulse after this ends. This is why all these plans for fiscal involve some sort of debt forgiveness or socializing necessary costs. Policy will be pushed, whether it knows it or not, to "free" private sector balance sheets. 

Macro in a State of Paradox

Sections:

- Paradoxical world of global macro, short reflation and short spoos is not the same trade.

- The European barbell, politics is a left tail.

- UK v Europe, long policy flexibility v short policy constraints.

- Yen with a Fed backstop, a "free" claim on US assets?

- Does the Phillips Curve matter outside of income channel, it does.


Paradoxical world of global macro

There seems to be this tug of war in markets. On one side there is a wall of global savings with very few places to put it. On the other side there's a massive economic shock and output will likely contract around 10% this year.

And this is largely how the divide breaks down. There is a massive imbalance of capital v places to put it. Of course, the large cap v small cap divide breaks down beyond that as there will likely be a cannibalization process as the "strong" push out the "weak." With that said, if the argument is stuck in, valuations are expensive or the prospects for tech are limitless, that misses the broader macro trends.

The economy and market can and for a long time exist in a state of paradox. Are valuations absurd relative to economic risks, probably. Are equity valuations high enough to offset an insatiable demand for high quality US names from real money domestic and foreign, probably not.

So the real question going forward is, how does this tug of war resolve itself. It is interesting that recently the market has been very cautious around the 2900 level. Is that the level where the valuation/economic reality take over, maybe, but it is worth remembering the asset shortage is massive and the imbalance of capital relative to places to put it creates a lot of incentives to keep up with the things that have worked.

Overall: The combination of this dynamic, how economically regressive covid is, i.e. the poorer countries are more vulnerable, has set up this really interesting dynamic. My expression for this framework has been long tech short EM, or on twitter what I called the "cleanest chart in macro." The reality is, it plays on so many macro dynamics, from relative balance sheet capacity, distribution of economic outcomes, and capital flows in a global QE world.

Said in another way, I want be long "good" outcomes in US equities and "bad" outcomes in EMFX.

Tech v EM

The other way I have been thinking about this framework is, being short reflation and being long the market is not at odds with each other. 

The European barbell, politics is a left tail

I have written a lot about a European barbell, an idea basically that Europe is at a key point and it basically has two options.

Two options:

1) The good one: some level of joint issuance either via mutualization or through commission bonds that allows for risk sharing among members. This would not only be crucial in terms of politics and solidarity but it would also provide the world a safe asset in a time of high demand for cash alternatives. This would be bullish EUR.

2) The not good outcome is, the northern countries don’t budge on some level of risk sharing and despite an aggressive ECB and a real backstop in the ESM, political backlash leads to fragmentation that the ECB will have a tricky time combatting. The point is, technicals aside, even with unconditionality, eurobonds and ESM are not interchangeable from a political perspective.

A muddle through might not work, the South has leverage this time

Another new dynamic this time is, the Italians have power and what if Italy throws a punch.

The creditor v debtor dynamic has changed, the periphery is no longer the "bad guy", that’s now the Northern countries and they don’t seem to get it. An exogenous shock has rocked the southern countries, one the Commission admitted to being late to, this isn't a situation where another round fiscal responsibility lectures will go over well.

This is a problem the market seems to be discounting. Of course within the distribution, the most likely outcome is some form of European muddle, as we have seen so far with the Commission recovery fund that does propose some grants. However, the difference this time is, not that a muddle wont work, but it may have tangibly negative political repercussions, and relatively soon. The political terms seem a lot more binary.

Mutualization type = support and solidarity, ESM type even with unconditionality = you don't care.

If Conte is going to try and sell the people on another classic pathchwork European backstop, this sets up a worrying development. Either the Italian government decide they do not want to sell this or they do and it is rejected via domestic political turmoil. I.e. if a bandaid is the European response, it could get pulled off pretty quickly.

The question is, how do you bet on unfortunate outcomes in Europe. The currency is tricky because that surplus is such a floor in risk off. Despite asymmetry, BTPs are hard to short given the ECB bazooka with PEPP, that will likely get bigger. Sure Lagarde left a lot to be desired last week. but the ECB is there to tighten spreads and the market knows it. So there are two dynamics, there is a tail in Europe that is still likely underpriced, but given structural dynamics within the EMU (current account and APP) its not clean to bet it. The answer may be short CEE FX, short EMU beta.

CHFPLN daily chart (LHS). GBPCZK monthly with 5y moving average (RHS).



Central European economies face a very precarious future, especially if the EMU is going to muddle through this. What is the CEE model. They get trade and funds from EU, it eats up domestic slack (tight immigration), raised rates relative to EMU (CZK), carry etc leads to recycling. The problem now is, the export car parts to Germany so that they can export cars to China, is not an ideal economic model. Since 2002, Poland/Czech/Hungary, have seen their nominal export levels rise by 2-4.2x. Integration has been a great trade for CEE.

As these globalization and global trade tailwinds have transitioned into headwinds, these currencies have to get fundamentally cheaper. And this all before their respective local political situations.

UK v Europe, long policy flexibility and short policy constraints.

I wrote last time why GBP is an interesting story. In theory, when UK eases, GBP gets killed because of current account, external debt etc. This time around, GBP has stayed relatively bid, even with one of the more aggressive monetary/fiscal mixes. The thesis I presented last time was that the market could be “rewarding” the UK for being less constrained than say Europe, which gives it a sort of 1931 feel in terms of the UK breaking free of policy constraints.

UK stocks v France (EWQ) & Italy (EWI), some interesting weekly moving averages.




Yen with a Fed backstop, a "free" claim on US assets.

There are a few interesting dynamics within JPY right now:

1) The Fed helped out GPIF. Dollar liquidity has allowed to them fund/hedge and Fed's backstop of IG has protected their positions. Fed will keep bills-ois in tact and try and prevent it from spilling into unsecured markets. This has been the pressure point for Yen and the Fed is on top of it. As long as BoJ is getting liquidity to end users, the pension/lifer sector, JPY should avoid another flare up.

2) Without technical dislocations in FX swap market, is Japan's NIIP position just a "risk free" claim on US asset prices. Now, especially if the Fed does YCC, why would GPIF take more naked FX risk unless instructed to do so by the BoJ (stealth intervention).

3) Everyone is at lower bound, Japanese FX policy from MoF/BoJ was not designed for that.

The question for Japan now is, how do they weaken Yen. The original catalyst for Yen weakness into this crisis was, Japanese bank and non bank sectors have a ton of dollar funding needs, which has explained the massive BoJ take up of the Fed's swap line. But, given how interest rates everywhere have converged to zero and deflationary risks from the demand shock have lowered expected inflation, it is very hard for JPY to weaken.

Yen v surplus Asia could be an interesting dynamic, sort of a lag trade at this point. As we have documented, a lot of surplus Asia is similar to Japan in a NIIP sense. They have a ton of excess savings that overwhelms size of domestic market so needs to be exported. However, unlike Japan, none of these countries have made the painful export adjustment (Japan is still in it). Exports accounting for 50% of GDP is not the way of the future. So many of these Asian countries are in for an adjustment, likely a fiscal one. So yes, both Japan and surplus Asia have massive NIIP positions which should support FX valuations, but relative to Japan, many of them still have to make broader economic adjustments.

This chart is super interesting. JPYKRW.




Does the Phillips Curve matter if income has been replaced?

This is a bit wonky but it fits into my framework for that reflation and rising asset prices are not the same thing. The question is, why has deflation been ruled out.

Few potential reasons:

1) Even if the Phillips Curve is alive, the income channel has been replaced via stimulus. Goldman did a great chart documenting that as a first order shock absorber to spending, stimulus has prevented a disaster in incomes.

2) It didn't happen in 2009. The new Keynesian model told to us to look through current slack. Future marginal costs matter more than current economic activity. I.e. there is this form of discounting which leads to stickiness. And that is why the a DSGE based on NKPC would not have forecasted deflation post GFC.

3) The most accepted reason in policy maker circles is, the flatness comes from a lack of variance in inflation given how successful monetary policy has been in anchoring expectations.

4) the price PC slope has flattened because both wages and prices are changed less often. Once variance decreases, stickiness increases and the response to changing labor market dynamics is smaller.

5) Empirical bias. As the unemployment rate got lower and lower, and past any SEP estimate of NAIRU, the conclusion from policy makers was, the price PC still lives, but the slope is just much flatter than we assumed.

What if policy makers are too confident the PC is dead or irrelevant given income replacement.....

Yes, a lack of variance in prices seemingly has become self fulfilling. But there seems to remain risks that we buried the Phillips Curve at the wrong time. Maybe, it's not dead, maybe it's just asymmetric. And yes, income has largely been replaced by stimulus this time around. But, there is still a lot of behavioral differences between someone who is employed v unemployed. Healthcare, risk aversion, expectations. I.e. you can replace income, but the impulse is fundamentally different.

This a rough example where you take average PC slope during two "best" years of recovery v two "worst" years of downturn. Plenty of problems with this approach, but does convey an interesting point that should be pretty intuitive. Of course, in these episodes the fiscal stabilizers were nowhere near as big as they are today.

This is why I am surprised the Fed has been relatively slow in changing its forward guidance that is far too delphic v ZLB risks. The current idea seems to be it is too early and the Fed at this point doesn't need to convince the market of much.

I think the right framework for where the Fed is going in this regard comes from Brainard's February speech. Her idea is that we should have aggressive outcome based forward guidance along with interest rate caps, creating what Brainard calls, "the commitment mechanism." This is really interesting because both of these policies have merit on their own, especially at the ZLB, but together, they kill two birds with one stone. Cap yields, strong odyssean guidance, and together they reinforce each-other. I think this the framework the Fed will adapt, whether at June SEP or Jackson Hole, but sooner than later.