Why are interest rates so low (I)

31 marzo 2015 michele boldrin

I have been thinking at the question in the title for a while, unable to find a convicing answer. Today I discovered that Ben Bernanke believes has an answer, which serves, if I do not read it wrong, as the opening post of his new blog. Needless to say it does not convince me, if anything because his was one of the many hypotheses I also considered and then disregarded. But BB is BB, hence, lacking a better theory and remembering Bob's wise dictum - it takes a model to beat a model - let me use BB's model to try putting order in my own thoughts.

No disagreement on the basic facts: nominal - to a lesser extent also real - interest rates on various forms of debt are at an historical low around the world, the "developed" one first and foremost but, sensibly, also elsewhere. BB recall this in his blog post and there is no reason I add, at least for now, further facts to sustain his claim. Interest rates are low "worldwide".

We also agree on the fact that, while the Fed may move short term nominal and real interest rates with a substantial degree of freedom - if the Fed moves the overnight rate from 3% to 10% when inflation is at 2% then, at least for a few days or weeks, the overnight real interest rate will be close to 8% while before it was at 1%, no doubts about it - it cannot really affect long term real rates if not marginally and certainly not for a long period of time. Hence, when the 10-year real rate goes near zero and stays there for many months or even years, the Fed's open market operations, QE or not QE, are only a minor factor. 

Enter then the Wicksellian Rate (WR), which other would call the "general equilibrium rate of return", a mythical single price of loans at which all factors of production (K and L, first and foremost) are fully employed. BB recalls here a well established "theory" (rather, common sense) according to which the WR is expected to be high in rapidly growing economies (high demand for funds relatively to supply) and low in stagnant or slow growing ones (high supply of funds facing few investment opportunities).

At this point the stage is set for the main argument: the Fed is supposed to pursue - or follow or try to catch up with - the WR through its policy instruments. When it feels that in the "real economy" the WR is low, it lowers rates, and when it feels it is high, it does the opposite. Here two sets of questions come to mind.

1) Given that the WR is supposed to establish itself, somewhat spontaneously, as the equilibrium rate - "perhaps after some period of adjustment" as BB says - why are the Fed's actions relevant at all? BB does not explain this an I am puzzled. If the WR is some natural rate to which the economy will converge, why bothering? Does the Fed, by acting on the very short-term part of the yield curve, speed that process of convergence up? Do we have a theory for that or at least evidence that that is so? No idea. Given that the economic system, according to BB's view, would in any case converge always to the WR, why not leaving the Fed Funds rate (or overnight rate, roughly the same) at some fixed nominal value (I support zero, but that's another story and it is not relevant now) and let the economic system adjust either through a slightly higher/lower inflation rate and/or by moving the nominal rates along the yield curve? What does support the idea that the Fed is supposed to play "catching up" with the WR? I suppose some model out there says that because prices do not move fast enough (i.e. inflation does not adjust quickly enough) the Fed's moving of nominal rates should speed up convergence. But - assuming the evidence supports the idea that the prices of goods and services do not move fast enough, which I do not buy - do we have any evidence that nominal interest rates in the financial markets would also move too slowly without the Fed's intervention? I am not aware of any such evidence ...


2) As the graphs above show, the Effective FF rate moves a lot, and that is the Fed's main tool to play "catching up" with the WR. Now, what leads us to believe that the Fed knows better than the rest of us where the WR is going? Ok, those numbers are nominal and the real FF rate has moved less, still quite substantially, over time. Was the WR really all that volatile in the US economy during the last 60 years? The second chart, below, shows the real rates (the rates adjusted for inflation using the CPE price index). They also moved a lot and in the same way: should we infer that the US economy was a slow growing economy in the 1950s and 1960s and a fast growing one in the 1970s and 1980s? And, has it really been growing slower than ever during the last 4 or 5 years? Somehow I doubt it ...


Further, and somehow repeating myself, what does lead us to believe that the rates the Fed is able to move are those moving faster than the rest? No idea. And again, what does lead us to believe that the Fed always gets it right? Back in 2001 the Fed "knew" the WR had suddenly dropped and brought the short term rates down to 1%. Ok, assume that was so ... Roughly three years later it also "knew" it had started to move up, the WR, and decided to play catch up with it until about 2006, when it had gone pretty high. Ok, let us assume that was also kind of true. But then, less than a year later, it collapsed? What a volatile real rate! And, even more, what a volatile "real economy" that stagnates for a few years, then suddenly takes off, then stop growing again ... do "real fundamentals" really move that quickly and suddenly? I am puzzled ...


Maybe these are minor, even polemical, points. Not really (they are relevant for policy) but my interest here is more with the model BB and the people guiding our monetary policy have in mind. The real issues, I believe, are elsewhere and are better left for the next round. Ben Bernanke, at end of his post, says 

What features of the economic landscape are the ultimate sources of today’s low real rates? I’ll tackle that in later posts.

so I will wait for that and keep thinking about the issue which, now even more than before, seems rather complicated and obscure to me. 

19 commenti (espandi tutti)


Vincenzo Pinto 31/3/2015 - 10:28

I'm gonna stick to English as I'm not sure about the translation of some of the terms.

Apologies for the undergrad question, but I'm missing a piece of the GE equations. I don't understand how you go from what a rate of return is at micro-level (equilibrium price in a capital market, meaning the level at which demand of capital equals supply of capital) to its role at macro level - quoting " a mythical single price of loans at which all factors of production (K and L, first and foremost) are fully employed". Not asking for mathematical proof, just the intuition: does equilibrium between supply and demand of capital automatically implies factors full employment in a GE model?

Second point: has a feedback mechanism between low rates and growth ever be considered in econ? My point, let's call it Japanese syndrome: something happens (let's call it irrational exuberance) which leds to a lot of investment projects to be implemented which only have NPV>0 under a best case scenario. Best case scenario doesn't occurr, and all of these projects become unprofitable, leading to bankruptcy and growing unemployment. CB receives political pressure to lower rates, hence making some of the marginal investments profitable again: capital is not allocated from low productivity sector to high productivity sector (e.g. banks' assets are full of "zombie companies"). There's where the negative feedback kicks in, CB lowers and lowers interest rates thus delaying the reallocation of capital. 


michele boldrin 31/3/2015 - 23:07

On the first point: no idea. In fact, I never found the WR a useful concept at all; its "rediscovery" by Mike Woordford is not something I welcomed. One can always do a whole lot of hand-waving and argue that WR, as measured say by the FFrate or by the rate on T-bond of some short maturity X is a sufficient statistics for the whole collection of different equilibrium rates of return on different assets, as it is the "marginal" rate, the one equilibrating the market for riskless investment at the maturity X. The rest, one can argue, is priced by some arbitrage argument. Maybe. If we knew everything else that enters the arbitrage equations this could be done. But I very much doubt this is the case, hence I very much doubt the WR is a sufficient statistics for anything. Even IF it were well defined, which unfortunately it is not. And this is something worth insisting upon: HOW do you measure the WR in the real world and HOW do you relate it to all other rates of return?

Further, "full employment" means nothing if one does not specify the structure of the underlying markets where inputs are traded. Competitive, monopolistic, oligopolistic, whatnot ...

Which is why I believe we need to model financial markets at a disaggregated level and study monetary policy for what it is: first and foremost a mechanism through which the monopolist of liquidity affects the portfolio position of SOME financial actors (those trading in the FF market). 

Second point: very good point, I believe it did play a role in Japan and also in the US, especially when it comes to banks. It kept alive lots of banks that should have gone under. Not sure it also kept alive lots of otherwise unprofitable businesses. But it is worth writing a paper, and you should start one :)


marcodivice 1/4/2015 - 13:50

I agree with Michele. You should start writing

Never written a paper in my life and now I'm way too dumb to start a PhD :)

I have always argued that the zero lower bound is another fantasy produced by the religious mind of the obedient macroeconomist.

I think it will take some time to find out what BB really thinks.

In his first post he just presents the basic facts (which you agree with) and then shoots down the idea that 'it is all the fault of Fed policy', He points out that the 10 yr rate is not under the control of the Fed, which you also agree since you say "it cannot really affect long term real rates if not marginally and certainly not for a long period of time."

In the second post, which appears today, he shoots down the Summers Secular Stagnation Hypothesis, that there is a lack of profitable investment opportunities nowadays.

Like the teller of the 1001 Night Tales, he hasn't revealed where he is going and is trying to keep us interested in hearing more.

(So perhaps your criticism ia a little premature).

... the staircase to heaven will collapse if its initial steps are weak.

What I am pointing out are those parts, in BB's basic argument, that are either unsupported or incoherent.

Speaking of which, let me re-iterate a point many have made in the past but your comment suggests to be worth repeating.

BB, and other high ranking Fed's officials, got to make up  their mind between:

(i)  Movements in the FF rate cannot affect long term real rates for a significant amount of time, and

(ii) The (achieved) goal of QE was, and is, to lower the long term real rates for a substantial amount of time.

 You cannot have your cake and eaten too.

Summing up, the incorrect inference BB wants us to believe is the following:

1) The Fed moves the FF rate to track the WR;

2) The Fed, and other central banks, have lowered both short and long term rates;

3) Hence the WR rate is low worldwide.

In the post I try to cast doubts 1).  Matter of fact, this comments also casts doubts on 2).

From which I conclude that - even assuming the Wicksellian Rate is a meaningful concept, which I very much doubt - the conlusion reached in 3) does not stand. 

Now, Ben says he sees no support for the secular stagnation argument, and I agree. Let's see where he's going. I have been arguing here and elsewhere that, in fact, we do NOT have a convincing explanation for what's going on. And he may well be, without admitting it, in my same situation: searching for an answer :) 

nobody knows the answer. But in BB we must trust, Wait and see.

Thanks for the article, which is actually of great interest for the whole financial industry apart from being an economic puzzle. Just two doubts I have, sorry if they are trivial:

1) What is the role of forward guidance? The Fed under BB has spent lots of effort to try drive market expectations via explicit committment of keeping the FFR at the zero lower bound for the foreseeable future. This guidance appears to have had some impact on market rates, for instance looking at 10 year Treasury rate.


When in mid 2011 the Fed explicitly said that the FFR would have remained at 0.25% for a long period of time, 10 year rates dropped very sharply, and they increased again substantially only in 2013 after the "taper tantrum", when BB hinted that it might be appropriate to remove some policy stimulus. Now, the Forward Guidance appears important to me given the "equivalence" that should hold between spot rates and future rates. If market expects the FFR to stay at 0.25% for more than 10 years, then all my future rates will be at 0.25% and also the 10 year rate should be at 0.25%. Now, this brings the question: why rates are on a declining trend since January 2014 after having incrased substantially in H2 2013? It seems that market expects the Fed to increase the FFR only moderately and not return to pre-crisis levels. I personally don't agree with this view, but there are a lot of market strategist supporting this view, which could influence the behavior of market participants. By the way, this theory would explain why interest rates are also so low in the Euro Area and in Japan: common expectations are that the respective central banks will not move the policy rates for a very, very long time. 

2) Is it possible that we have some "finite size" effects? The amount of high quality bonds in the world is finite. Financial institutions need high rating, liquid fixed income assets for their daily operations. Furthermore, financial regulations after the crisis push institutions in holding low risk assets (Basel III for Banks, Solvency 2 for Insurances). If the central banks buy government bonds through quantitative easing, an insurance will not go into equity, but will rather buy some other form of fixed income assets, preferably sovereign bonds. With decreased supply but rather stable (if not increasing demand), yields are bound to go down. Don't you think that we may witness this effect? This seems particularly true in the case of Germany: a country with negative net issuance of government bonds, the only large market in the Euro Area still rated AAA and with ECB committed to buy a large share of the market, bonds price cannot do anything else than increase.

Again, thanks for the interesting article and for any answer/comments you Michele (or any other reader) will give to this post. 

anch'io sono intrigato dalla 2)

The amount of high quality bonds in the world is finite. Financial institutions need high rating, liquid fixed income assets for their daily operations. Furthermore, financial regulations after the crisis push institutions in holding low risk assets (Basel III for Banks, Solvency 2 for Insurances). If the central banks buy government bonds through quantitative easing, an insurance will not go into equity, but will rather buy some other form of fixed income assets, preferably sovereign bonds.

viene ben evidenziato il ruolo del QE, attrezzo mai impiegato prima in questa misura; michele boldrin lo trascura, credo perchè più interessato all''incoerenza del ragionare di BB e alla probabile inanità di fondo di politiche monetarie molto reattive o anche pro-attive.

i nuovi criteri regolamentari per gli investitori istituzionali sono poi conseguenza dell'esplosione nucleare dei precedenti e fatalmente perturberanno ancora per parecchio il mercato. come corollario, molto lavoro per i risk-manager.

il fantomatico wr? sembrerebbe allora ben più alto, negando l'ipotesi di stagnazione secolare "estiva" :-), e prospettando prossimi adeguamenti violenti: una improvvisa crescita, magari!, oppure inflazione a manetta. 1x2, esclusa la x, insomma.

per un tentativo di risposta al quesito nei tuoi due primi paragrafi.

Su WR non so che dire, a me sembra una cazzata che Mike ha riesumato, uno zombi utile solo a far confusione ed eccitare giovani macroeconomisti poco propensi a leggersi i classici direttamente.

Come tu dici se guardo al settore privato reale il WR non so cosa sia. So che i returns on equities sono oggi sostanziali (magari c'e' un'altra bubble? Si, forse. Ma chi lo sa per certo?) e che probabilmente ci sono migliaia di profitable and risky investments in the private sector that are not being financed for the reasons said in the long comments. So, no idea if what we observe is the mythical WR and no idea if there is "full employment" of K and L, whatever that means :) 

The questions you ask are difficult, not only the go beyond what one (well, me) can do by means of a comment on a blog but, more generally, they enter uncharted territories. So, just my one cent.

1) Forward guidance. Indeed, IF 

- I am CERTAIN that the nominal short rate will stay at X% for N years in all the countries I can trade,

- AND I am certain about inflation rates for the next N years in all the countries I can trade,

- AND I am certain of how exchange rates among all the countries I can trade will move during the next N years,

- AND, finally, I am also certain of how other securities (with different degrees of riskiness) will behave (along both the risk and return dimensions) in all the countries I can trade during the next N years,

THEN, the yield on the N years bond should stay at X% in the country I am currently operating.

This is BECAUSE of arbitrage (modify accordingly in the N at some power other possible cases):

should the N year bond yield more than X% in the home country (change country as needed in the N at some power other possible cases) I can borrow short term an (unlilmited?) amount at X%  in the home country each year for N years and use the proceedings to hold a long position in the N year bond, therefore making an (unlimited) amount of money for sure.

No doubt about this. Combinatorial complexities aside, this is a clear case in which arbitrage arguments deliver the result by plain arithmetic.

But, the strong assumptions I have made to state the proposition are enough to show how hard the "forward guidance" argument is to  deliver the result. 

Because reality is always somewhere in the middle of all these demanding assumptions there is no doubt that "anything can happen" and that, the stronger the common beliefs are about all those variables I listed, the more likely it is to see forward guidance working.

As you pointed out during the period you are interested in, the relevant rates gyrated around quite a bit but kept a tendency to get lower and stay low.  Somehow forward guidance had some effect, but this is far from stable and certain. Agents are nervous and uncertain, not certain, about the future ... 

And yes, I agree: currently market participants expect rates to stay low in most relevant countries, they also expect inflation not to pick up and ... well, I have no idea what they expect about exchange rates because if I had ANY clear idea about where exchange rates will go in the next few years I would  bet zillions on that and become rich (and NOT tell you!) :)

I hope this clears the matter: arbitrage arguments are powerful but market's expectations are an untamed and untamable beast. So, in any case, you are forced to take bets and if you are lucky and your bets are correct you hit riches. More than that I do not dare to say. I have my own opinions about what the Fed and the BCE will do (no idea about BofJ, have not gone there for a while) but they are just that: educated opinions ...

There is no free lunch, guidance notwithstanding.

 2) Finite size effects and institutional/regulation contraints on big players. This is also raised by dragonfly below and ... I agree with you both!

I am quite convinced that regulators are pushing many financial institutions to hold assets of a certain class, of which there are very few, and disregard lots of other options. This is because, on the one hand, regulation has become more pressing post 2008 AND, on the other hand, many banks still have very weak portfolio structures and relatively little own-capital.  This was the reason for the credit crunch that hit Spain, Italy and other countries in 2011-2014. I even said that while playing "politician" and advocating drastic re-capitalization of italian banks :)

And, indeed, I agree that the various rounds of QE, by FRB and ECB, have strenghtened this tendency. There is such a thing as "crowding out" especially when countries are, correctly, restraining themselves on the deficit side, hence the stock of "good bonds" outstanding is not increasing very rapidly while the "safety and regulation induced" demand from financial institutions is growing rapidly.

Notice, and I close here, that this whole thing has near ZERO impact on the real economy and tells us NOTHING about what the "Wicksellian Rate" is in the (risky) real investments out there, among guys taking chances in new enterpriseses. Those rates of return are NOT measured by the return on the Bund!

Well, I spoke too fast: it DOES have some effects.

Because not only the amount of safe bonds available is limited but also the amount of resources invested in bank equities is limited, the push toward safe bonds (plus QE) crowds out of the market the demand for funds from the risky guys, those that create growth.  There is not enough saving going around to finance ALSO them at reasonable rates. When you look at it from this point of view QE is actually DAMAGING for the private sector and for growth. Here is, schematically, the argument.

1) Growth comes from risky investments in the private sector; because of regulations such investments consume lots of "capital" if "banks" have to finance them and comply with Basel&Co.

2) Banks do not have unlimited capital, in fact have little capital relative to the short term liquidity available to them. And they MUST satisfy regulatory requirements, Basel and so on.

3) The amount of low risk bonds available is finite and everybody wants to have a chunk of them in the portfolio. QE also buys these bonds, hence their price goes up and returns down.

4) Because of 2) and 3) financial institutions buy safe bonds no matter how costly they are and are unable/unwilling to use liquidity to invest in risky projects, BECAUSE there is not enough capital going around banks to cover the risk of such projects.

5) In the resulting equilibrium, there is excess demand for the safe bonds, their return goes negative and lots of short term liquidity is kept hidle BECAUSE banks do not have enough equities to support a portfolio long on risky, private sector investments.

The latter go hungry and QE is part of the problem. Amen. 

Breaking down the big banks so that they are no longer a systemic risk, can fail and we can afford to lower the capital requirements? I.e. returning to the world of the 1960s-1970s (in Italy, without the public ownership)? How big are economies of scale in banking - i.e. how much efficiency we are risking to lose? Perhaps not too much in the new world, where you can buy IT services on the cloud and outsource the back offices to India. But would the public trust the smaller, leaner banks?

Should anybody be interested, pg. 130 of the Liikanen Report "Economies of Scale—What are the benefits (and costs) of large banks?". In summary, not that many benefits, especially when compared to potential issues (TBTF, regulatory capture...)

It's very complex to me to understand totally, but: could a LTRO have minimized, instead of a QE, the last five point or it would be the same ?

Your risk, if you are a Life Insurance Company(for example) it's called "RISCHIO DI REINVESTIMENTO". Ti trovi con ricche plusvalenze(in valore nominale) che sfrutti(vendendo) per impieghi diversi(rimborsi anticipati ad esempio) ma con l'impossibilità di reinvestire (a parità di rating pregresso e rendimento pregresso) in assets che ti garantiscono la possibilità di onorare quanto promesso alla clientela. E devi assumerti rischi maggiori. Se chiedi ad un assicuratore tedesco cosa pensa di quanto sopra ti risponderà che è un problema serio..molto serio. Problema che siamo vicini ad affrontare imho poichè non vedo effetti significativi dal lancio del QE. Un reale rischio sistemico generato dall'antibiotico QE che dovrebbe, nelle intenzioni originarie, debellarlo.

Ho scritto in italiano perchè non conosco l'inglese. Mi scuso.

Question: How much liquidity and short term liabilities are around the world (m1+M2) ?

Third part is on line

Germany did it.

It reminds me two alphaville posts. here and here. 

ok  it reminds me Blame Canada too but I suppose it has a connection with the political debate here in Italy. 

Inizia una nuova discussione

Login o registrati per inviare commenti