Weird Is Normal

This post was originally published on Pieria in December 2013. Since then, the idea that the long-term real equilibrium interest rate must be equal to or lower than the long-term sustainable growth rate has become much more mainstream. I am just amazed that anyone ever thought it could be otherwise. A long-term real interest rate persistently above the sustainable growth rate cannot possibly be an "equilibrium" rate. As I show in this piece, it can only be maintained through rising inequality. It is by definition ponzi and therefore unsustainable. Periodic financial crashes are inevitable in any system in which growth does not cover the interest on debt. 

Three years ago, Nick Rowe produced this post describing a “weird world” – a world in which the equilibrium interest rate is at or below the long-term growth rate of the economy, rather than above it as we are used to. In such a world, bubbles are inevitably created as investors search for positive yield. This is also the world described by Larry Summers in his "secular stagnation" thesis.

But I don’t think this world is weird. I think it is, or should be, normal. We have been living in a weird world of unstable Ponzi schemes that eventually crash and reset. Let me explain.

I’ve argued that investors have no right to expect real returns on safe assets (insured deposits, government debt and, increasingly, land/property) that are persistently above the real growth rate of the economy. I hadn’t done the maths to prove it, but intuitively I felt that an economy in which returns on risk-free savings were persistently above the growth rate could not possibly be in a stable equilibrium over the longer-run. There would be continual transfer of wealth from younger to older and from income-dependent to the asset-rich. The younger and poorer would gradually become either more impoverished or more indebted, while the older and richer became wealthier. Eventually the system would become unsustainable and there would be a crash in which asset holders lost considerable amounts of their wealth.

Andy Harless, it seems, agrees with me:
Most economists think that the natural real interest rate is normally positive. I have my doubts, but never mind, because I'm ditching the whole concept. Once we start correcting for expected normal growth rather than expected inflation, we are clearly not dealing with a natural rate concept that can be presumed to be normally positive.  If we are talking about a risk-free interest rate, then the need for physical capital returns to compensate for risk would make it very hard to achieve an equilibrium with the interest rate as high as the growth rate, let alone higher.
So the equilibrium rate of interest should be at or below the growth rate of the economy. This means that when growth is very low – as it has been for the last few years - the equilibrium rate of interest would probably be negative. Mark Carney said exactly this about the UK economy in a 2013 speech:
As its policy stance reflects, the Monetary Policy Committee (MPC) judges that the equilibrium real interest rate has been, and continues to be, negative.
If the equilibrium rate of interest is negative, then real returns on risk-free assets should also be negative. If they are not, the result will be depressed output, unemployment and deflationary pressures. If the political power of holders of risk-free assets results in an interest rate policy that is too generous to them, therefore, it is bad for the economy as a whole. The implication of Carney’s remark is that UK savers holding risk-free assets (insured deposits and government debt) must continue to suffer capital erosion due to negative real interest rates for the foreseeable future. Looking at it another way, holders of risk-free assets must be taxed on their holdings. We can regard this tax as a charge for the government’s guarantee of those savings. Or, if you like, it represents the value of an unseen asset called “safety”.

To my knowledge we have never regarded “safety” as a positive asset. But the concept of “risk” as a negative asset is at the heart of risk pricing. The price of an asset is discounted by its perceived risk. The most valuable assets are those that are completely safe – there is no risk of loss. The least valuable are those where there is a high probability of loss. Investors expect to be compensated for the possibility of loss, so riskier assets give higher returns. In bond markets this is expressed as the relationship between price and yield: the higher the price, the lower the yield.

But suppose we priced the absence of risk – i.e. safety - as a positive asset? This is in effect what Harless is suggesting:
But the concept of risk preferences models a subjective good – call it “security”
Instead of taking complete absence of risk as the natural position and adding a risk premium, we take average risk as the natural position and discount from it. This acknowledges that some risk is a normal part of life and complete safety is to a degree both abnormal and costly. By using the growth rate of the economy as the natural position, Harless further suggests that that risk is necessary for growth and safety carries an economic cost. The returns on genuinely safe assets should therefore be lower than the long-run growth rate of the economy, the difference being the negative return on an extremely valuable asset called “safety”.

The price of this “safety asset” may be highly distorted. Investors unreasonably expect to receive returns on safe assets that are at a premium to the growth rate. Remember I said that price is the inverse of yield? That applies to safe assets just as much as risky ones. If the return on a safe asset should be discounted relative to the growth rate, then for a safe asset to return a premium suggests it is either seriously undervalued or not really safe.

In the 1970s, government debt did return a premium to the growth rate, probably because of concerns about the safety of government debt in the post-Bretton Woods turmoil. But yields have been on a downward trend for over 30 years, and government bonds from sovereign currency-issuing countries in good standing (I include Germany in this category because of its dominance in the Eurozone) are now generally trading at a discount to the expected growth rate. This is not weird. This is as it should be.

Not that there is anything wrong with safety. Safety is a good, and we don’t produce enough of it, as Harless observes:
The history of interest rates and growth rates suggests that we have seldom produced enough security, but the deficiency today is clearly worse than usual. 
People need to be able to save safely for their retirement and for other beneficial outlays. Discouraging safe saving by restricting the supply of safe assets is itself costly, because it forces investors to accept risk they neither want nor can afford. We need to enable people to save safely if they wish – not least because if they know they can always retreat into safe assets, they may be more willing to take some risk.

But we also need to charge those who seek safety a realistic price for their disengagement with the risk-taking that drives economic growth. We have always undervalued the “safety asset”. History shows repeated cycles of asset price & credit booms ending in crashes, as we would expect when there are unsustainable returns on safe assets. The Old Testament’s “debt jubilee” can be seen as a way of managing this process – it created an artificial crash every 49 years. But it didn’t solve the problem.

Governments should create sufficient safe assets to meet demand, and price them sensibly. I say governments, though in practice the governments that can produce genuinely safe assets are small in number: the Euro crisis showed all too clearly what happens when government debt that is thought to be safe turns out not to be. But realistically the private sector cannot create genuinely safe assets without some kind of government guarantee. Caballero & Farhi, concerned about “fiscal limits” above which government debt is no longer regarded as safe, suggest private/public partnerships for the creation of safe assets. I’m afraid I think this is silly. If the private sector cannot create safe assets without government guarantee, there is no point in pretending that they can. Let the government create them, and stop imposing arbitrary limits on debt/GDP.

Governments massively increasing production of safe assets (debt) to enable the private sector to save safely does of course raise a concern about the burden of debt service on future generations. But if the returns on safe assets are persistently below the growth rate of the economy, then debt service in the future would be sustainable: there would be no need for higher taxes in future to service government debt. In fact there would be a wealth transfer to future generations, representing the value of the “safety assets” that the current generation bought from future ones. Or, putting it another way, the current generation’s need for safety is met by means of a guarantee from future generations, for which the current generation pays by returns on their savings that are below the growth rate. I hope that makes sense.

Of course, there might be considerable problems convincing savers that it is reasonable for them to pay for safety through negative real returns. And there would probably be even more problems convincing them that risk-free savings should not give a return that they can live on comfortably without drawing on capital. But they have to learn. Safe saving is essentially a Ponzi scheme. If it is to be stable over the longer-run, it cannot give unsustainable returns.

I know this is a very weird way of looking at things. But then weird is the new normal. In fact it has been normal all along. We need to get used to it.

Related reading:

An exact consumption-loan model of interest with or without the social contrivance of money – Samuelson (1958)
A model of the Safe Asset Mechanism: Safety traps and economic policy – Caballero & Farhi
Global safe assets – BIS
Why stagnation might prove to be the new normal -,Larry Summers
The spirit of the season – Mark Carney
When governments become banks
America’s greatest export is its debt – The Week


  1. Having a safe assets interest rate above the growth rate aka rentiers get everything till the system crashes was in fact the normal world from Sumer to the Industrial Revolution. It was a world where parents sold their children into slavery and everything ended either in revolution or debt jubilee. Only tech progress ended that.

    1. It was the world that Keynes wished to end when he talked about "euthanasia of the rentier". Rentiers were squeezed after WWII, but since the1980s they have steadily regained the ascendancy. I don't think tech progress ever did more than temporarily mask their rise.

  2. Aren't we missing the savings depletion (or accumulation) rate here?

    Interest - Depletion < Growth.

    If savers net spend as a class, the pool of savings shrinks (or grows less quickly than otherwise) meaning less capital and higher price (interest rate) for it, relatively speaking. Conversely if savers net accumulate you need your interest rate below growth by that much (may make zero look a bit high in the right circumstances).

    1. Not really. For the system to be stable, you need r <= g regardless of the rate of capital accumulation. Higher net spending should raise g as well as r. If it doesn't, the spending is unproductive and the system is ponzi.

  3. Hi,

    This is one topic that always sends my head into a spin, interest rates versus inflation rates versus growth rates.

    - I don't know if it is even possible to calculate an "equilibrium" rate of interest. It seems to be more an economists' theory. However, maybe one way of looking at the relationship between interest rates and growth rates would be to consider that most debt is created because the debtor believes that the return on the debt will be greater that the interest rate paid?

    Which implies that the growth rate created through debt should always be larger than the interest rate, so there is a "leveraged" relationship between them???

    - I looked into the biblical debt jubilee, as I was not convinced that the money lending Jews would have ever allowed debts to be simply "written off". Sort of goes against the grain. Indeed, if it was year 48 in the cycle, then they would surely have been requiring an interest rate of over 100% on any new debts issued.

    I found a few interpretations which seem much more realistic, in particular the understanding of Jubilee as the final payoff of a lease.

  4. I’m baffled as to why interest cannot exceed growth. The argument seems to be that interest is a TRANSFER from poor to rich and that no other transfers to transactions can pay for or counter the “interest transfer”. That just ain’t true.

    To illustrate with a very simple Middle Ages type economy which consists just of lords of the manor who rent out houses or small farms to peasant farmers. Assume no growth (as was the case more or less in the Middle Ages). Lords charge peasant farmers 3% of the value of their houses and small farms. Peasants and small farmers earn the necessary money by performing tasks for lords, like maintaining the manor house, supplying the lord with loads of food for sumptuous feasts at which the lord entertains his fellow lords and ladies, etc etc. That system can go on for ever.

    1. Ralph Musgrave:

      "I’m baffled as to why interest cannot exceed growth. The argument seems to be that interest is a TRANSFER from poor to rich and that no other transfers to transactions can pay for or counter the “interest transfer”. That just ain’t true."

      Economically, you are correct. Assuming the sort of economics where, for example, people who lose from international trade can be compensated by taxing some of the gains away from those who gain from international trade, with the taxation and transfer adjustments thereby created being not politically unpopular and not having unexpected affects the economy as a whole. Economists show a lack of enthusiasm for actually designing such systems.

      Much economics, where it intersects with politics, tends to make assumptions that would be considered over optimistic by the standards of managers.

    2. Ralph, you are confusing stocks and flows. Lords of the manor didn't charge a percentage of the "value" of peasants' houses and small farms. They charged a percentage of the annual production of the land. This is indeed sustainable provided that production is constant. But it isn't, of course. Harvests are variable. A tithe of a fixed percentage could be impossible to pay in a bad harvest. Furthermore, since the lords' income depended on the productivity of the land, some lords increased the tithe to maintain their incomes in a bad harvest, thus making peasants' lives even harder. The whole system would be unstable in a democratic society. It was maintained only through violent coercion and strict application of feudal laws. And it broke down irreversibly when the Black Death made labour so scarce that the feudal laws tying peasants to the land could no longer hold.

  5. Great piece and something I've often thought about in the context of e.g. the perception of property as a safe asset due to underwriting of property risk by governments (which feels wrong); and that my retired parents will accept cash interest rates below inflation, rather than take capital risk (which feels right).

    Perhaps it's that interest rates are set, an input if you will, that has led them to become the cornerstone on which risk premia are laid. Growth rates on the other hand are an output and future rates require continual estimation, rendering them a shakier concept from which to apply a scarcity discount.


Post a Comment

Popular posts from this blog

WASPI Campaign's legal action is morally wrong


The foolish Samaritan