Saturday, December 1, 2018

War-gaming the Chinese Nuclear Option with US Treasuries

[ In the quite useful parlance of Scott Alexander, the epistemic status of this post is fairly uncertain, so take with a grain of salt.] 

As the economic tension between the US and China slowly ratchets up, I've found myself thinking recently about the financial nuclear option that China has, and how it might play out. 

What I mean by the nuclear option is China strategically using its large reserves of US Treasury Bonds and Bills to cause maximum economic and financial chaos in the US. This is something one occasionally hears about, but the discussion seems to be split between serious economists who blithely assume it will never happen, and doomsayers that think it will be the end of the world. As you'll see, I'm somewhere in the middle. 

Then again, I'm not a macroeconomist, so take my views as just an educated guess. Perhaps the easiest way to do it is just to assume that I'm playing the Chinese, and that I wanted to cause maximum carnage. What would I do, and how would it play out? I am almost certain that I'm not giving the Chinese any new ideas here, and since I'm John Q. Nobody in any case, I don't feel particularly guilty at writing this publicly. For instance, here's a recent statement from a Chinese diplomat [Update: Cui Tiankai is actually the Chinese Ambassador to the US, so this is pretty close to the official position of the Chinese Government]:
Cui said he did not believe Beijing was seriously considering using its massive US Treasury debt holdings as a trade war weapon, citing concerns that such a move would destabilize financial markets.
Translated back from Diplomat language, this means: We'd like to remind you that we could use our massive US Treasury Debt holdings as a trade war weapon. 

So without further ado, here's how I'd play the Chinese side. 

The first thing to realize here is the poker mindset. You don't want to be thinking of your chips as money. The money is spent when you walk in. Once you sit down, the chips are ammunition, used to defeat other players at the table. What you get at the end is the prize, but if you're thinking on every raise about the rent money, you're toast.

So it is here. If I were China, I would assume that the current holdings of US Treasury debt are like an ICBM. We're no longer trying to maximise the value of the holdings, or even preserve the value of the holdings as a strategic asset of China. We're treating the assets as already worth zero. Because, as we shall see, the policies designed to keep the value of the assets high are almost the exact opposite of the ones aimed at causing carnage.

US Treasury obligations play a very important role in the financial system. In many applications, we need to know a "risk-free" interest rate, and typically the rate on short term (i.e. 30 day ) US Treasury Bills is used as a proxy for this. They're denominated in dollars, which are fiat, so the Treasury can just print as many as it wants. Even "print" is a euphemism - press a button, and the dollars electronically appear. So there's very little reason why the US ever couldn't pay its short term obligations. It might choose not to, either because it went crazy (e.g. during the Clinton government shutdown, or the debt ceiling debate), or because the amount of dollars required to print would cause massive inflation whose cost would be worse than defaulting on the debt, but again, it nearly always could pay. And in practice, it always has paid. Which is why short term US T-Bills are treated as a proxy for the risk-free rate in lots of financial models, such as those used by banks.

But this interest rate is determined by supply and demand in the market for Treasury assets. China has accumulated a ton of them, and could dump them on the market at any time. In bond terms, the yield or interest rate* is inversely related to the price (assuming a zero coupon bond, as is the case for short term obligations like T-Bills, though the logic is similar for coupon bonds). Dump lots of T-Bills on the market, the price drops, and the interest rate rises. When the interest rate rises, every bank who has a short term financing gap that they were planning on covering on the overnight lending market is suddenly in a huge hole. Chaos ensues. 

So I'm China. As a preparatory step, beforehand I'd take all my own financial institutions and ensure that they're not holding any US treasuries privately as collateral on anything. Slowly switch to safe stuff for my own accounts - gold, Euro bonds, whatever. The point is that we're going to screw everyone holding treasuries, and everyone else too. But we definitely want to minimise our own banks' exposure. We want to try to get the financial side of the Chinese economy as close to self-sufficient as possible. 

The basic step is dumping Treasury assets. If we do it right, the first asset dump will be a surprise attack that will spook the market as much as possible. To ensure maximum carnage, I'd begin my sales at maybe 3pm EST. The aim is to do it shortly before market close in the US, when lots of financial institutions have to mark their accounts to market at the end of the day. 

So, first step, I would start with a massive dump of Treasuries on the market. Not the whole amount  - one must always keep troops in reserve. But enough to cause a big spike in short term interest rates, and enough to panic everyone in the market.

So, this happens at say 3pm. The other reason to do it close to the end of day is as follows. We cannot raise interest rates forever. Indeed, we may not even be able to raise them for very long. The reason is that whoever is playing the US Fed has an obvious countermove. As soon as they realise what's going on, they'll step into the Treasury market themselves and start buying treasuries to raise the price and lower the interest rate. Remember, they're doing this with printed dollars, which are in an almost infinite supply, up to the point of causing inflation. And if you're the Fed, the tradeoff between inflation versus short term market chaos is like worrying about becoming addicted to morphine when you've just been shot in the leg. The choice is obvious - they'll buy, in whatever quantity needed to prevent interest rates from going through the roof.

If they can do this, what's the play? Well, first of all, there's the surprise attack. We're gambling that they don't necessarily have a plan set up to immediately deal with this and implement the massive purchases necessary. Maybe they do, in which case the first round effects aren't as dramatic as we'd hoped. But we can help ourselves by giving them a limited amount of time before market close - just enough time for everyone to react and price in the carnage, but ideally not enough time for them to respond properly.

The aim is that every bank who's long in Treasury assets and has borrowed against them (which is a lot of them) was accounting for this collateral at a low interest rate, and a high price. Suddenly, with minutes left on the equity clock, they realise they're on track to be insolvent by the end of the day, as the collateral on their obligations is rapidly dropping. Loans get called in. Prices of their equities fall, making everyone else panic, making the market as a whole crash. 

So, that's the aim. How would I ensure this gets played up?

Having done the initial dump at 3pm, I would make an announcement at 3:15 or 3:30 or so. The Chinese government is planning to liquidate all its remaining US Treasury obligations and US Dollar denominated assets, immediately. Moreover, we will be switching to Euro Bonds and Euros.

What's the point of this?

In normal trading times, this would be crazy. You're just inviting people to front run your trades, selling before you sell so you get a worse price and then buying back off you later at a profit.

But in chaotic times, this is ideal. We're trying to maximise price impact, not trying to maximise value. By announcing our intentions, we tell the whole market - lots more treasury sales are on the way. What will they do? Start dumping their own Treasury assets, pronto, and switching to Euro ones. In this way, we're not just using our own Treasury reserves. Now the second round of selling is coming from every other financial institution with a fast-moving trader and a desire to stay solvent. Of course, this still doesn't count for squat if the Fed puts in an infinitely large buy order, but there's a long term point. We want as many financial institutions as possible to stop holding US Treasury assets, so the Treasury is basically having to hold the whole lot themselves. This is functionally equivalent to just printing money to cover the entire deficit. They'll do it if they have to, but they don't really want to.

And in the mean time, now the Fed is not only trading against us, but trading against lots of other people in the Treasury market as well. Which makes it harder for them to just take steps that would somehow freeze us out of the market. 

Anyway, let's assume we're playing against a highly competent Fed. Unbeknownst to us, they have contingency plans in place to send enormous buy orders if the price drops sufficiently, and so we don't get anywhere much in terms of disrupting the Treasury market. What then? Has the plan failed?

No. The Treasuries are only step one. The real action is in the foreign exchange market. The Fed is going to have to provide me as China with lots of US Dollars to purchase all the Treasury assets off me. As they do, what's my response? Immediately start selling those US Dollars and buying Euros. In large quantities, as fast as possible.

In other words, we're trying to tank the US dollar. And this is something the Fed has a much weaker position on. Why? Because while they can print up an infinite amount of US Dollars to purchase treasuries, they can't print up any Euros at all to buy US Dollars. Sure, they have some reserves, and you can bet your ass they'll use them. But now there's a finite target. We may even have some idea of how much we're having to trade against, yet in any event, it's a finite and achievable task. And the more they support the T-Bill, the more they give us ammunition to attack the currency.

Which gets to the other point - why buy Euros? Why not buy Renminbi?

Because China wants a weak dollar, but it doesn't really want a strong Renminbi. When the dollar weakens, the US gets imported inflation on its many foreign goods. It also makes it easier for US exporters, so it's not all bad (we have to assume that trade between the US and China will be totally frozen, so it's just other countries we're thinking about). So a big drop in the US Dollar will cause significant inflation in the US.

China relies on exporting industries, so it generally wants a weak Renminbi. Instead, the aim is to make the Euro strong instead.

You might wonder, would the Europeans actually want this? What if they started printing Euros to resist it? 

Well, they might. But I'm not so sure. What we're really trying to accomplish as China is a shift in the question of who gets to be the global reserve currency. It's not going to be China. But it may easily be Europe. A large part of the Euro project was trying to set up a global counterweight to US financial hegemony. There are probably a fair number of Euro policy-makers that would be quite pleased to see the US dollar get displaced. The main advantage of being the global reserve currency, when we're talking fiat, is getting to run enormous ongoing deficits without creating inflation. If the US can't sell its Treasury debt to global investors any more, it runs a real risk that printing more money will result in inflation. As the reserve currency, so far this hasn't happened. 

Indeed, this was surprising to many people during the financial crisis - the Fed was printing like crazy and buying up all sorts of things. Why didn't it result in lots of inflation? Well, part of the reason is that the dollars weren't circulating back into the US economy in the same way. Foreigners buy up the debt. If they bought Euro debt instead, European countries would face much lower borrowing costs, the famous exorbitant privilege that the US currently has. Not only that, but being the reserve currency means that other countries want to invoice in your currency, and hold assets in your currency to hedge against this risk, and have banking services with your country. So being the global reserve asset fosters the development of your financial sector. If the US Dollar goes out of fashion, expect New York to become significantly less important relative to Frankfurt and London as the place of global financial markets. 

It's not just me that thinks this, by the way. This is approximately the argument in the recent Gopinath and Stein paper. They' don't say as much on the question of how one might shift between equilibria, but they more or less agree on the effects of being the dominant currency for trade, invoicing and financial development.

In other words, we're potentially peeling off the Europeans from the Americans. We're saying, hey, don't just instinctively support Uncle Sam here. We think it's probably in your interests to play along with us.

Why might this also be important? Because the other thing that the Fed will probably do in this scenario is call up every other Central Bank in the world and demand that they start buying US dollars with their own currency reserves. Assume that lots of serious threats are made here. If we're China, we can't overwhelm everyone. But if we can convince the Europeans to think twice before buying dollars, then maybe Japan thinks twice too, and India, and Australia, and before you know it, suddenly everyone is holding Euros instead. Russia certainly would be pleased to see the change, so you can count them out. Bye bye US financial hegemony.

Ideally, if I were China, I'd implement this plan when the US was in a recession, or near it. Because this means it's a lot more costly for the US to take the other option to support the currency - let the bond sales go on and interest rates rise. Because this will almost certainly tank the US economy if they do. 

That's the basic play. I think. As I said, I have a lot of uncertainty as to how large the effects would really be, even under the Holmes plan. Maybe global investors believe China won't be able to displace the US, and so don't sell many Treasuries or US Dollars. Maybe they do, and USG threatens their governments that they'll get Color Revolutioned if they don't demand their banks fall into line. This is pretty close to an act of war, so you'd be crazy to do this as China without expecting serious repercussions. And if the aim is to displace the US Dollar as the global reserve currency, you've pretty much only got one shot at it. Once you've fired all your missiles, they're gone. If people stick with the US, in the long run China might end up weakened, and you've destroyed the value of a substantial foreign currency reserve in the process. 

Anyway, this is a guess. Maybe I'm totally wrong, either in what they'd do, or what the consequences would be.

But let's assume for the moment that I'm approximately correct. One obvious question is, would they actually do this?

It's like a nuke. Generally speaking, no. Firstly, the main aim will be short term chaos. The effect on interest rates will likely be temporary, and the Fed will find some excuse to declare that all the financial institutions aren't actually insolvent at 5pm, no matter what market prices say. The currency play is harder for them to deal with, but also speculative as to how big the effect will be, and if its actually worth it. Maybe the US decides its okay with a weak dollar after all? Many other places are okay with this policy. Stranger things have happened.

Actually, it's not even like a nuke, as much as being like two guys in a bar fight, and one of them is threatening to burn the whole bar down, but he's standing closer to the door, so he won't get burned as much. Chinese financial institutions will suffer too. Chinese exports will suffer, big time. And it's a large gamble to see who else goes along, somewhat like a leadership challenge in a parliamentary system. The votes are taken by all the other countries. If you win, you can win big, but if you lose, you're off to the back benches or worse.

But thinking about it in this way misses a larger point. Are nukes pointless simply because in equilibrium its unlikely they'll get used? Wrong. As long as the threat to use them is somewhat credible, they can make excellent negotiating chips.

And this isn't just hypothetical. This actually happened. 

In 1956, during the Suez Canal crisis, Britain, along with France, decided to implement a military initiative to take back the Suez Canal, which Egypt's President Nasser had recently seized off them.

They felt that, as sovereign countries looking out for their own interests, they could just go ahead and do it.

Wrong answer. 

You know who was annoyed at not being consulted? Dwight Eisenhower. So what did he do?  He called up Sir Anthony Eden, and among other things, he threatened to dump the enormous US holdings of British War Debt, while simultaneously cutting Britain off from borrowing. The pound would tank in value, and Britain would be in chaos. The chaos didn't have to be permanent, necessarily, though it might be. But it would certainly last long enough to permanently end the political career of Sir Anthony Eden.

And so Eden backed the @#$% down.

If there was one day where it became completely clear to all concerned that the US was now the world's financial and military hegemon, this was it.

Maybe the Chinese will never do it.

But you'd be a bolder man than I to think that they'd never threaten it

All I can say is that I hope some of the smart people at the Fed are thinking about this more than I am.