Recent Posts

Wednesday, May 11, 2016

Implications Of Negative Interest Rates

Chart: 5-Year JGB Yield
I crashed a wine and cheese event that was associated with the CFA Institute conference in Montréal, and ended up in a discussion with an asset allocator about the implications of negative interest rates. This is a topic that I have not written too much about, but this is because I see the implications as being very limited. The only interesting implication is that it makes the position of fiscal conservatives who worry about debt-to-GDP ratios even more foolish.

We can look at the implications from four different angles:
  1. investing in bonds;
  2. central bank policy;
  3. psychological analysis of macroeconomists;
  4. fiscal policy.
This article covers these areas in turn.

Investors - Shoulda, Woulda, Coulda

From the perspective of an asset allocator investor, negative bond yields represent the "pain trade" (the market movement which causes the most grief for investors). They spent the past cycle trying to find "low cost" ways of getting short the bond market, and the market went and rallied on them further. Meanwhile, the duration of their liabilities is lengthening, and so their asset-liability matching position got worse.

They are stuck in what is known as the "shoulda, woulda, coulda" analysis mode -- "we shoulda started matching our liabilities in [insert year before 2014]." As most experienced investors know, this mode of analysis is painful, and completely non-constructive.

Going forward, government bond yields are so far below return targets that the entire concept of liability matching is called into question. The only reason to own bonds is that the alternatives are worse. A German bund -- regardless of the quoted yield -- looks a lot more attractive than a deposit in the euro area banking system from the perspective of the return of capital, when we consider the pathological nature of euro area policymaking.

In other words, it does not matter whether the 10-year yield is 1% or -1%, if you believe that the alternative is a 50% haircut on risk assets.

There is the issue of "equity duration" -- are equities at risk if discount rates rise? I am fairly skeptical about the concept of equity duration, as I do not think the risk-free discount rate has mattered much for equity valuation since the mid-1990s. If you look at infinite horizon dividend discount models, there are three terms that matter -- the risk-free discount rate, the nominal growth rate of dividends, and the equity risk premium. In the current environment, the latter two terms dwarf whatever the risk-free rate is doing. In the United States, the growth rate of nominal GDP is 4%, and so it is reasonable to assume a growth rate for dividends somewhat in that ballpark. The uncertainty around that the growth rate is much larger than the movements in the risk-free discount rate.

Central Bank Policy

The ability to push the policy rate to a negative level has made it possible for other central banks to maintain a negative yield spread versus the U.S. policy rate. The idea being that this will help suppress the value of their currencies, and support their trade balances.

If and when the Fed hikes rates further, this justification disappears, and policy rates will probably revert back to zero.

Economist Psychology

I took an undergraduate elective course in psychology, and one of the few things that stuck in my mind was some of the results from behavioural psychology.

In one set of experiments, chickens were held in cages, and given rewards when they pressed a button. The chickens would learn to follow patterns, or even just learn to press the button repeatedly if the rewards were given on random presses. However, if the rewards stopped appearing, the chickens eventually figured this out, and they stopped pressing the button.

Mainstream economists who are big believers in negative interest rates are playing the role of the chickens in a real-world experiment. How many failures of interest rate policy to revive growth will it take before they begin to revise their world view? Or will they keep pressing the "lower interest rates" button forever?

This will be the basis of very entertaining doctoral theses in 2100.

Fiscal Sustainability

Negative interest rates do a very good job of blowing the whole "high debt-to-GDP ratios will cause default" theory out of the water.

Since these theories are incoherent, it is hard to summarise them. The idea is that if the debt-to-GDP ratio is "too high," the government has to default (or something like that).

If the nominal interest rate is negative, the greater the amount of debt, the greater the "revenue" negative yields create. No matter what the level of the primary deficit is, the debt outstanding will always stabilise at some level. (The primary deficit is the fiscal deficit less the effect of interest payments.)

About the only way to get an "out-of-control" debt ratio spiral is to assume that nominal GDP is shrinking. (This is what so-called Neo-Ricardian theory would suggest.) However, such an outcome is incompatible with the government running steady nominal primary deficits. The implication is that the private sector would entirely disappear, which is not an outcome that even remotely resembles real-world behaviour.

The only way of coming up with a bad outcome is to assume that somehow there will be a rapid inflation. How such an inflation would be sustained in the face of the government taxes being levied as a percentage of nominal income, and if not all government spending is indexed, is very unclear. In any event, a rapid inflation would wipe out the debt-to-GDP ratio, which eliminates the alleged initial problem.

(c) Brian Romanchuk 2016


  1. Brian,

    I am not sure debt-to-GDP ratio is relevant. As even right-wing ideologue like Tyler Cowen knows quite well that what what matters is debt-to-national wealth. Defict spending increases private wealth if there is no leakage. For country like Japan, there is virtually no limit to government debt given its annual large current account sulplus.

    1. Sure, the ratio is not relevant, but that did not stop Reinhart and Rogoff. Obviously, different fiscal conservatives have their own theories, but worries about the debt-to-GDP ratio does appear to be a common denominator. I am just attempting here to respond to the general argument, and not any one person's views in particular.

  2. Very helpful post.

    I was curious about the short section on central bank policy. I would have thought that the exchange rate effects were just one motivation among others for negative policy rates, not the main one. But you may be right, I don't have a strong view here. What I am wondering about is this:

    Given the lack of connection between the policy rate and other yields (longer risk-free rates, risky rates, returns on equity, etc.) why does the policy rate have such a strong effect on exchange rates? We should be open to the possibility that it actually doesn't, but certainly it seems to. y working hypothesis would be that the main or at least the marginal participants in foreign exchange markets are not asset owners in general, who are adjusting the mix of currencies in their portfolios, but foreign-exchnage specialists who don't hold other assets. Does that seem right?

    1. I have a small section on the effect of the policy rate on the currency in my upcoming report (unless it is savaged by the proofreader, should be ready within two weeks). I think the section appeared as a first draft on this site; I discuss how the policy rate differential affected the Canadian dollar. But for a fuller explanation of my view, it will be available shortly at the nominal cost of $3.99 (🙂).

      But the shorter version is as follows. The fundamental driver are portfolio flows. I think the reason policy rates appear to be a driver is a signalling mechanism -- central banks are assumed to have a decent handle on domestic economic conditions, so rate hikes signal that the economy is growing. Things might be different in an inflationary environment like the 1970s.

      As for who is doing the trading, the vast majority of the volume is the result of foreign exchange traders doing hyperactive trading. Since they work with forwards, they create huge notional positions which bloat their importance. That said, they do not have the balance sheet capacity of the equity/bond portfolios, and so all their activity is aimed at getting in ahead of the fundamental flows. The implication is that forex specialists can swing the currency around on a day-to-day basis, but they have to respect the underlying flows. For example, if international investors decide that Canadian equities are a great investment, no matter what forex traders think about CAD, it is eventually going to go up.

      Finally, you also need to take into account corporate activity, like buyouts. During the tech boom of the 1990s, the USD was supported by foreigners being suckered into buying US tech firms.

  3. Fiscal sustainability is closely tied to current account sustainability as the two balances are connected by an identity.

    Fiscal sustainability by itself means less: fiscal balance is more a reflection of the current account balance on an average, since the private sector balance is a small positive on an average.

    So it's more a question of debt sustainability to foreigners which one cannot claim to be sustainable at all times.

    About your point on inflation, it isn't true that inflation isn't a problem in debt sustainability. The following situation is a bit academic but still illustrative of issues:

    Suppose there's a large fiscal expansion and prices rise. Since households have a wealth target, they will try to attempt to compensate the real wealth loss due to high inflation by saving more. But fiscal balance depends on households' saving behaviour.

    It seems contradictory to the intuition of large expenditure in a period of rapid inflation but can be checked with a stock flow consistent model. In Godley/Lavoie's book, there's an example of how debt/gdp can blow due to large inflation.

    1. My comments were very brief, and conditional upon negative interest rates. Under that assumption, since the interest cost is negative, I will stand by statement that debt sustainability would not be an issue.

      The question is whether negative interest rates could be sustained if inflation is rising, or the currency is falling. If fiscal policy was coherent with a low inflation objective, I see no reason why not. And fiscal policy is generally coherent in this manner; the only real risks are posed by indexation in government contracts.

      You are supposing a large fiscal expansion, which easily could be done in a fashion that is not coherent with a low inflation objective. (I lump the external value of the currency in with "low inflation objective".) i believe that a floating currency sovereign can achieve most reasonable objectives for fiscal policy and meet the low inflation objective, so long as the real resource constraints are respected. A lot of the "hydraulic Keynesianism" policies ignored those real constraints, and ran into problems.

      Would private sector desires to increase wealth pose a difficulty? Under the assumption that they are forced to allocate a greater portion of their wealth towards foreign currency/private sector assets as a result of the negative interest rate on government liabilities, the flows could work out.

      Finally, although the debt-to-GDP ratio could rise when inflation accelerates, the empirical evidence is that faster nominal GDP growth reduces the ratio. To get this in a SFC model, all you need to do is track the long-term debt separately, and have the initial stock of debt with an interest rate below that of the eventual growth rate of nominal GDP. Compounding the denominator faster than the numerator will reduce the ratio under any reasonable level of the deficit. Given the way that tax brackets are fixed, and most government spending is not indexed at a high frequency, inflation acts to tighten the fiscal deficit in the current environment.

      Fans of rational expectations would say that you could not announce such a policy (as bonds would reprice), but the empirical reality is that the bond market has not perfectly forecast future growth rates.

    2. " To get this in a SFC model, all you need to do is track the long-term debt separately, and have the initial stock of debt with an interest rate below that of the eventual growth rate of nominal GDP."

      Disagree on two levels.

      In the open economy case, if output grows too fast (relative to exports), debt/gdp rises. The interest rate part is a minor thing.

      You can see this in Godley/Lavoie's open economy models.

      About inflation and debt/gdp, my point was that it's not obvious that debt/gdp will not rise as a model in G/L's book shows. That real budget balance has correction terms because of inflation isn't a reason to presuppose that debt/gdp will improve.

      Few economies have negative interest rates on government bonds. That's fine: a lot of them are creditor nations. Second nobody ever thinks that if growth resumes, bond yields of some nations' government bonds will remain negative. So I don't understand the importance of negative interest rates in debt sustainability issues. It's a minor thing.

    3. It's not a major topic of interest, agreed. But it suggests a policy of locking in negative rates, and laughing at debt sustainability analysis.

      Since the negative interest acts as a tax, growth cannot get out of control. Hence, the external situation will also stabilize.

  4. Negative interest rates are like bank fees or government taxes.

    1. Wish people said this more. Most of us already receive negative rates on our transaction balances, and have for many years. It's not some dramatic new development.

  5. Brian,

    SoberLook had a great comment on understanding negative interest rates:

    Too many people don't understand negative rates and what they really mean.

    I even had one pension fund manager tell me they will never buy bonds with negative rates but they might not have much of a choice!

  6. "If the nominal interest rate is negative, the greater the amount of debt, the greater the "revenue" negative yields create."

    Not necessarily. This might apply to new debt but not old debt necessarily.

    Presumably debt already on the books is at positive rates so interest is being paid not received. (You are talking about high debt ratios, so high debt already exists.)

    Negative interest rates suggests deflation in which case the real value of existing debt is increasing.

    Neither of these are good for the debtor.


    1. Even with an existing stock of debt with positive coupons, the amount of interest paid is limited. As new negative coupon debt is added, it will eventually turn the net payment negative.

      The idea that negative rates suggest deflation is the biggest challenge to my argument. It is related to the "neo-Fisherian" argument that real rates are constant, and hence negative rates imply deep deflation. However, it seems unlikely that an economy could have much deflation if the government is running a primary deficit of 3%-5% of GDP per year.

      I probably needed to write a longer article on that topic...


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.