Widow-Maker Part 2: Monetary Policy

Though central banks have a lot of control over the monetary system, they suffer from what can be thought of as 'Policy Target Whack-a-Mole'.

This means that, while central banks are largely omnipotent when it comes to controlling any one variable - and they really can control one economic variable at a time completely if they like - when they use this power over that one variable to set a target rate, they have very little control over other related variables or secondary targets - which can work against their stated policy intentions. This tends to give observers the impression that central banks have little control, leading to the wrong conclusion regarding sovereign debt markets.

A case in point is one that I've talked about before - QE. The common understanding is that QE is supposed to lower both longer term yields and raise inflation to help the employment situation and recovery. One source of confusion is to think of the unemployment rate and recovery as the policy target - this is a goal of QE, but it is twice removed from the specific target, a sort of second derivative if you like.

Even yields and inflation were not the true policy targets, though they were discussed at length as intentions that would result from QE, there was no specific target for yields to fall to, and the inflation target was fairly vague. A central bank could control either one of these variables if it were truly determined, but because the Fed did not state "we will buy unlimited bonds until the yield on the 10 year is X.XX" neither of these secondary goals was really achieved.

Part of the reason is that higher inflation and lower bond yields are contradictory goals for market participants, and the Fed did not limit the functioning of a free speculative market. I discussed at length why QE actually raised yields on long term government securities while it was running. From a trader's perspective, if the central bank says it is going to try and increase inflation and support risk asset prices, a sensible trader will not fight them and sit in risk-free fixed interest, missing out on the equities/high yield rally and seeing inflation lower your real return - so portfolio allocations moved away from bonds under each QE program and yields rose despite the secondary target of lower yields. The reason Fed buying did not offset this is because the Fed is entirely a price-taker at auction, and must accept what the private sector offers - it does not have a mechanical pathway to drive yields lower under the current structure of QE.

The true policy target of QE is to purchase a quantity of assets - 85 billion per month in the latest program. The Fed achieved this target perfectly, purchasing a quantity of assets it deemed appropriate. Few question that the Fed can purchase any quantity of assets they wish, because the Fed is the monopoly issuer of reserves and can create them simply by expanding its balance sheet. QE achieved its target, but perhaps a quantity of assets was the wrong target to specify.

So what if we think through some other possibilities for QE. What if, instead of targeting a fixed quantity of assets, the Fed states an intention to purchase unlimited assets until an inflation target of 3% is achieved. There should be little doubt that the Fed can buy as many government securities as it likes, and perhaps it would be forced to purchase a very large quantity or even all of the privately held bonds. In such a circumstance, suddenly it doesn't seem like the target will be so hard to achieve. If nothing else worked, legislation can be changed for the central bank to purchase private sector securities such as the case with the BoJ or Bank of Israel - it's not unheard of. Simply put, the Fed could achieve the 3% target if it really wanted, but at what cost?

Central bank whack-a-mole is the problem. If the policy target is now an exact rate of inflation, the central bank no longer has control of the quantity of assets it needs to purchase to achieve this. If it requires purchasing too large a portion of outstanding government securities, we can start to imagine some scenarios where the economy is seriously destabilised. If a moderate QE program has pushed market participants further out on the risk spectrum, potentially causing unnatural asset valuations, a situation where the Fed takes out the majority of risk-free assets could leave the usual investors in those assets - such as pension or mutual funds - with no choice but to move in to highly speculative areas which suffer from volatility and decreasing asset quality.

Not only this, but the exchange rate under an unlimited QE policy could move very sharply, such as the Yen under Abenomics. Bond traders and investors too would demand higher yields from the government to hold Treasuries, because they firstly will want to be compensated for higher inflation, but also because the central bank would be supporting a speculative orgy in more risky assets, which bondholders would miss out on.

Many would assume that the central bank purchasing such large amounts would lower yields, however this is not a given, because the Fed purchases on the secondary market in separate auctions to auctions that the US treasury issues it's securities under. To entice bidders in to the primary funding market, yields would likely rise as dealers/investors don't usually invest charitably on purpose. Not only this, but unless the Fed sets a yield target, it is a price taker - dependent upon what yield the private sector offers at competitive auction. The Fed is only a price maker if it abandons its inflation target and targets yields instead. Banks or primary dealers can be forced to make a market for treasuries, however state controlled banking is unlikely to be an ideal choice in a capitalist economy unless things got desperate, and that is not without its own problems.

In this case, the central bank has vast power to achieve its inflation target, but the side-effects may not be great. We can extend this type of thought towards any type of central bank policy. For instance, the central bank can set the overnight interest rate at whatever level it wishes, but to set it too high can cause a recession and hardship, even outright deflation if they act recklessly - too low and they risk bubbles and high inflation.

A central bank can also control its exchange rate - if it disregards an appropriate interest rate for the domestic economy, or issues reserves to buy up foreign assets (such as China). The exception to this would be when there is a political crisis such as hyperinflation due to fraud, civil war or government collapse where there is little demand for the currency even at ludicrous interest rates - such cases are more likely in undeveloped economies however, and I'm only discussing developed nations (that are not stuck in the Euro) in regards to the widow-maker.

Whatever the chosen policy target, if it is to be achieved it normally involves losing control over other variables, but the central bank has almost unlimited powers to achieve its target if it really desires - the true limitation is political will, career risk and unwanted side effects. It sounds as if I have tried to call the Fed useless, but my real intention is to show that they can control any variable they like, as long as they deem it more important than any other consideration - the Fed could have lowered yields or raised inflation, if it was truly determined. In a round-a-bout way, we arrive at yield targeting.

Yield Targeting:

Sometimes secondary goals can align with a main policy target. Sometimes, also, an issue becomes important enough to warrant ignoring secondary issues until it is resolved - both of these apply to a country in the position to be considered for the widow-maker trade. Firstly, a central bank might need to purchase a lot of securities to hold it's yield target - further fuelling inflation - however that high inflation reduces the burden of debt in real terms, so there is an alignment between the two goals. Secondly, preventing default and rising yields can take precedence over keeping inflation low, full employment or any other goal.

The widow-maker trade is based on the idea that a country has at least these problems:

1. High and rising government debt.
2. Record low interest rates despite the high debt - so they can only go up.
3. Low or no growth to grow out of the problem.

When interest rates rise, they will be doomed as interest costs will outgrow tax revenues. If the government tries to raise taxes, they will slow growth. Austerity obviously doesn't work because fiscal multipliers can be greater than 1, so any cut in spending can shrink economic activity by a greater amount and make the debt problem worse. The government can't increase spending dramatically to speed up growth without further worsening its debt case. It's a sticky situation.

So what would make interest rates rise? Inflation is the most likely cause, since developed country debt does not carry true credit risk - the government can always force primary dealers or the banking system to make a market for its debt, so default is always a choice or poor decision making rather than a certainty. This the main reason that there have been no bond vigilantes arriving to punish Japan so far - there just isn't a serious default risk. But what if inflation actually came and interest rates started to rise?

Well, here is where the widow-maker doesn't really work, because higher inflation is exactly what the US and Japan are trying to achieve. Interest rates will almost definitely start to rise if inflation picks up in a measurable and persistent manner. But, so far the only inflation that a program like Abenomics has generated is likely to be transitory. That is, a rapid exchange rate depreciation will increase the cost of imported goods - especially for Japan who must import most of its energy needs - however without rising wages, domestic demand is unlikely to rise substantially and might even fall as more income must be allocated towards imported energy. Consumers might try harder to reduce energy consumption, offsetting price increases. Without rising wages, there is nothing to support rising prices while keeping the same level of consumption. If Japan is to create lasting inflation, the BoJ might need to abandon a fixed quantity of asset target, and move to an explicit inflation target with unlimited scope - as reckless as that sounds - because Japan's shrinking population and workforce keeps downward pressure on nominal growth and inflation, making it a different case to the US.

Still, if/when wages begin to rise, and inflation becomes persistent, then yields will rise on expectations of future inflation reducing the real income from bonds. Partially they will rise because some believe central bank cash rate rises will occur in the near future (to slow inflation), though this is probably premature. Interest expense will begin to rise on newly issued debt and roll-overs. However from the revenue side of the equation, wages rising will move incomes in to higher marginal tax brackets and consumption tax revenue will rise. Not only this, but nominal GDP will grow, making the value of the immense debt less of an issue on a relative basis.

If these were the only actions that occurred, it is possible that it could go either way. If the bond market got the jitters, interest expense might grow more rapidly than new tax revenues - causing problems. Of course, no central bank will sit idly by while this occurs, and this is where policy goals align to prevent the widow-maker trade from working - even under rising inflation and yields.

Central Bank Intervention:

I outlined above that central banks have supreme control over the monetary system if the need to act is urgent enough. If there is a risk that by doing nothing, default risk will rise for the US Treasury or Ministry of Finance (Japan), then the central bank will disregard goals such as low stable inflation in order to preserve the functioning of the system. This can involve yield neutral bond purchases (such as the current iteration of QE in both countries) or it can involve yield targeted purchases.

In the first case, QE does not have a specific target for yields, so yields can actually rise under central bank bond purchases. This is an action that will only be taken when deflation is a threat (such as now), as yields are unlikely to rise too far. Under such actions, the central bank can end up owning a large portion of outstanding debt - this is already the case with both the Fed and BoJ, who own a large portion of their own country's debt. The profits of these central banks are remitted to their respective Treasuries, which is to say that when the treasury pays out coupons and principal on bonds held by the Fed, after some minor expenses, most of that money comes right back to the US Treasury - a sort of circular finance. The Federal reserve now holds over 10% of US debt, and the Bank of Japan 20% of its own debt - this debt can be considered mostly irrelevant from a solvency standpoint, since Treasury will effectively be paying itself.

Should there be persistent inflation, the Fed could move from a quantity target towards a yield target. The mechanics of a yield target would be very simple in the situation where a developed country like the US has amassed a huge level of debt: The Fed would offer to purchase any bonds below a certain price level so that a targeted yield was achieved. No one would have any reason to sell for a lower price than what the Fed offered to purchase for, and if they did an arbitrage profit would exist and soon disappear. This yield target would likely be below the rate of inflation, such that inflation might begin to rise - possibly even to double digits - as there would be no or minimal cash rate rises to slow it. If the private sector is discouraged from holding bonds due to fear of loss of purchasing power via inflation, the Fed would simply end up owning most of the bonds on offer - but yields would still not rise above the target.

Offsetting the inflation however, the government is able to implement higher tax rates to focus on debt repayment, and this removes some of the inflationary pressure while simultaneously lowering debt levels. Too often, arguments that say that these governments will default or the bond markets collapse tend to forget that fiscal policy will help the situation as inflation rises.

Because yields can be forced to remain below the rate of inflation, each month under this form of financial repression reduces the real value of the debt burden. During World War II and up to nearly 1950 the Federal Reserve used yield targeting to keep a cap on treasury yields of 2.5%:

After the removal of the yield cap, there was no traumatic and sudden rise in yields for a few reasons, one being that people had just witnessed the Fed hold the yield target without faltering during a turbulent period in which debt was skyrocketing. There was very likely an understanding that the Fed had a lot of control over yields, and this stopped a lot of speculative flight from those who might have otherwise been concerned about capital loss on bonds during rising rates. Furthermore people were also scarred from the crash of 1929 and many were scared out of the stock market for life, preferring to save in bonds - a similar aversion occurred in Japan after it's bubble. This was a period in which the US debt to GDP ratio rose to 120% - higher than today.

In the end, it wasn't necessary for the Fed to hold this yield cap to 2.5% after 1950. Rising nominal GDP through both growth and inflation reduced the severity of the debt, and there was little reason to expect rates to rise catastrophically, so they didn't:

The takeaway from this is that, though the Fed did not need to maintain this policy to hold yields down, it could have if it wanted to - were it faced with drastically rising yields. By doing so it could allow its debt burden to reduce over a number of years in real terms. For anybody wanting to sell or short bonds, there should be a consideration that one of the major risks to the trade is that the central bank simply implements a floor under bond prices when it gets concerned, leading to substantial losses for those traders. This is probably the nail in the coffin: If the trade starts to become too successful, the central bank will implement regime change to control yields, ruining the trade. As well as this, inflation which is anticipated to cause yields to rise, is actually how the situation is resolved from the government's perspective, so the logic isn't totally water tight.

In the end it probably isn't worth taking the risk - I can only assume most bond traders come to the same conclusion and don't bother. The other point is that we have seen this scenario before - the bond market did not collapse after the US became heavily indebted, it had no reason to when inflation offered an easier way out - even with inflation rampant, yields did not rise and there was no collapse in bond demand.