Recent Posts

Sunday, August 21, 2016

Central Bank Direct Lending Does Not Add Much To Policy Space

A new front has opened up in the "Helicopter Money" debate: the belief that direct lending by the central bank somehow constitutes "helicopter money," and that it opens up new space for policy action. In my view, it just creates another way for central bank-focused mainstream economists to discredit monetary policy further, and for the central bank to lose money.

S W-L on the BoE TFS

Professor Simon Wren-Lewis discusses the new Term Funding Scheme (TFS) of the Bank of England (BoE) in "Negative rates, helicopter money and the Bank of England."

He recognises the complaint that negative interest rates punish savers (which is why some non-mainstream economists view the effect of low interest rates on the economy as being ambiguous). He argues:
We normally think about monetary policy as changing the [emphasis S W-L] interest rate. If rates are cut, that benefits borrowers but is bad for savers. But suppose the central bank gave money [emphasis mine - BR] to private banks, on condition that this money was passed on in the form of lower rates to borrowers. If it did this, but did not change the [emphasis S W-L] interest rate, that would be helping borrowers but not hitting savers. The Bank introduced such a scheme yesterday, called the Term Funding Scheme (TFS). What is more, this subsidy for borrowers is financed by creating money. Eric Lonergan argues that the ECB is doing something similar, and if there is any insight in this post I owe it to him (but if there isn’t it is my fault not his!).
(I will discuss the Eric Lonergan's arguments below.)

I will immediately note that I am not an expert on the TFS, but based on the descriptive comments of the BoE, the name of the programme describes exactly what it is: a scheme in which the BoE lends money (reserves) to banks for a period of time to counterparty banks.

Although we often say that banks "give" loans to their clients, this is not the usual usage of the word "give," as in a gift.  The usual conception of "helicopter money" is that the central bank hands out gifts of money -- there is no obligation to pay the money back. A loan has to be paid back, so it does not match up with the usual usage of "helicopter money".

In Simon Wren-Lewis' conception, the borrowing banks are going to promise that they will narrow their lending spreads in return for accepting the low interest term funding. In other words, the entire premise is based upon the good faith of modern bankers -- good luck with that.

Realistically speaking, it would be suicidal for a bank to narrow its lending/borrowing spread across its entire balance sheet in return for a bit of cheap term funding. The only reasonable outcome would be that certain borrowers would get better lending terms from the bank, up to the size of the programme. Realistically speaking, this is just going to be a subsidy for a handful of borrowers (assuming that the banks do not just pocket the gains from a slightly lower funding cost on some borrowing).

This is a form of a directed subsidies towards some borrowers. Such a policy was standard practice when "industrial planning" was in vogue; whether it will survive challenges from the idiotic "free trade" treaties that nations have signed is an open question.

Furthermore, such policies already exist -- and are completely ignored in mainstream models. Developed countries generally allow corporate borrowers to deduct interest expense on taxes (and in the United States, household mortgage interest), while most owners of bonds hold them in tax-advantaged vehicles (pension funds, etc.). The economic cost of borrowing is at the after-tax rate of interest, while savers receive the pretax rate. This effect is completely ignored in mainstream models.

This is another wonderful illustration of the myopic nature of mainstream economic analysis. A tiny programme run by the central bank that will affect a handful of borrowers allegedly creates new policy space, while a similar policy run by Her Majesty's Treasury that encompasses the entire corporate sector is completely ignored.

Not a New Policy, Just a Dangerous One

When I first wrote this article, I believed that the TFS was a  form of unsecured lending by the central bank. (Oops -- see comment below). Unsecured lending adds a third option for central bank operations, on top of:
  1. outright purchases of assets (typically fixed income); and
  2. lending against assets (repurchase agreements or "repos", or discount window operations).
(Accountants might argue that repos are not loans, and they may or may not be correct according to GAAP. From an economic point of view, they are collateralised loans.)

(UPDATE: This article had been written when the programme was announced, but publication was delayed as I was driving across Northern Ontario last week. I could not find many details on the TFS, and I assumed that it was unsecured lending. As Nick Edmonds pointed out in comments, this is not the case -- TFS loans are indeed secured in a conventional fashion. As a result, this programme is not particularly novel, and only of interest to British banks.)

However, this third method is really just a variant of lending against collateral -- in this case, the value of the collateral is zero. Does this add to policy space? Not really. Any bank that lacks unencumbered assets that are eligible for rediscounting or repo operations is a bank that is already hurtling towards oblivion.

It is also bad policy. Although central banks are indeed banks, they live a sheltered life. They do not have teams in place to work out defaults. (I argued in my article on "Overdraft Economies" that it would be a good idea if central banks did have such credit expertise, but even then, it should take the form of secured lending.) Lacking lending expertise, they need to stay at the front of the line of creditors, by only lending against top quality collateral. Let the private sector fight over the scraps in a liquidation event.

One might argue that a central bank cannot go bankrupt, so who cares about credit losses? This ignores the reality that central banks are somewhat unaccountable, and if they lend billions to a chupacabra ranch operator who then hightails it with the cash, that represents a real transfer of wealth that could have gone to citizens with slightly more legitimate needs. Elected politicians are the ones with the mandate to make such distributional decisions.

Eric Lonergan on TLTRO's

Eric Lonergan writes:
The ECB understands this [the ineffectiveness of lowering the policy rate - BR]. That is why it turned its focus to the other interest rate – the rate at which it lends to banks under its latest TLTRO programme. Disguised behind horrible acronyms, TLTROs are far more radical than QE, OMT, or SMP programmes. TLTROs are also so obviously legal that the German media barely bothers to mention them.
Here’s why TLTROs are hyper-radical – and will work. What are Targeted Long-term Refinancing Operations? They are loans which the ECB makes to banks at a duration and interest rate of its choosing, for specified purposes. With every TLTRO the ECB chooses the interest rate, the duration of the loan, and potentially, the credit risk. Why is this potentially more important than all other monetary tools? Because the ECB is never out of ammunition with a TLTRO. Just cut the interest rate further, extend the duration, and extend the scope of the lending. For example, the ECB could make 20-year TLTROs at -5% interest rates, available for durable goods purchases, in an amount up to 10% of GDP. Would it work? Just watch the share prices of automakers!
There's a few problems with this argument.
  • It assumes that there are creditworthy borrowers. Even if a loan can be rediscounted at the central bank, the lending back (or securitisation) eats credit losses first. Furthermore, we would need to find borrowers who are willing to invest (or at least borrow to consume). Given the amount of private financial capital that has been desperate to find investment opportunities, there is not a whole lot of pent up demand for borrowing. As a result, announcing a negative discount rate on a particular class of loans will not necessarily create much new demand for credit.
  • This is not exactly helpful for savers. A certain class of loans will drop to an interest rate of -5%, well below the benchmark policy rate. That eliminates a certain amount of assets that could have been bought as bonds or securitisations by savers in the bond market. If the central bank extends this policy, the aggregate yield on assets available to savers will converge towards the "special" financing rate.
  • The effectiveness of the policy is predicated upon the central bank being able to allocate credit in a fashion that is superior to the private sector. Given that the brain trust at the ECB drove the euro area periphery into a depression -- a feat that many thought was impossible in the modern era -- this seems like wishful thinking of the highest order.
  • Narrowly focused sectoral subsidies is an inducement to create bubbles. As the Greenspan/Bernanke Fed figured out the hard way, cleaning up after a burst bubble is hardly a cost-free exercise.
Admittedly, given the disastrous design of the euro area, any form of disguised fiscal policy is better than the policy stance that is in place. This offers us no insight into the operation of less deranged monetary systems in place in the other developed economies.

(UPDATE: Eric Lonergan questioned the idea that there could be credit risk with a 20-year loan at -5% interest. It is possible that he intended to write a 20-year loan, with a -5% annual coupon (with the coupon based on the initial principal balance), with the negative coupon reducing the principal balance. That is, there would be no repayment by the "borrower" at any time (since the principal balance is $0 in 20 years). However, this would not be a -5% interest; it is effectively -100% interest -- the "loan" generate no repayment cash flows. Such a structure has a NPV of $0 at any discount rate (above -100%), and would have to be immediately written down to $0 under any accounting convention. Since the loss is immediate and automatic, one could debate whether this should be considered a "credit loss." Nevertheless, it would represent an income loss to the central bank, and would raise the exact same political problems as a "normal" credit loss.)

(c) Brian Romanchuk 2016


  1. Have you managed to work out why mainstream economists completely miss the auto-stabilisers as well as the tax break on interest?

    ISTM that auto-stabilisers do everything that interest rate jiggling does and more besides. They are better targeted, more flexible, respond faster. Everything borrowing money is not.

    But because they are paid by the Treasury, rather than the Central Bank apparently they don't count.

    1. That's pretty much it. Pretty well the only government funding for macro research is given by central banks, and they want to fund research that says central bank policy is the only game in town.

      Free marketeers also want fiscal policy and the welfare state to disappear. Why allegedly progressive economists go along with this is one of those mysteries I cannot understand.

  2. Brian,

    Some comments on the TFS:

    - Although the aim is expressed as being to ensure that cuts in base rate are reflected in cuts in lending rates, the level of funds made available is actually conditional on the extent of new lending, not on changes made to rates. Granted this is also not easy to truly pin down, but it does go more directly to the intended end result, and it also recognises that banks often expand lending, not by cutting rates but by relaxing credit criteria.

    - This is not unsecured lending. Advances under the TFS must be secured with the same sort of collateral normally required for monetary operations.

    - The fact that the scheme is "funded" by the creation of reserves is a bit of a red herring. It could have easily been structured as a loan of bills, as was the earlier Funding for Lending scheme. As far as I can see, the only advantage in switching to reserves (and possibly the reason for doing so?) is because of the recent change to exclude reserves from Leverage Ratio calculations.

    1. Thanks for the colour. With that added information, this does not sound like much other than a technical tweak to central bank operations, which is going to have no measurable effect on the real economy. Obviously, the central bank needs to adjust to a changing regulatory regime, but this is not a new paradigm for stabilisation policy.

  3. Brian - My point is not that CBs can make loans which have zero credit risk - although to all intents and purposes they can. Rather, the point is that lending schemes allow CBs to ease policy along three axes: the price of credit, the maturity and the credit risk. In extremis, as I have argued many times, they can make perpetual loans at zero interest rates. What is important about these administered interest rates is that central banks are setting interest rates *independently* of the policy rate which influences market rates. The BoJ is doing this through tiered reserves and the ECB through the interest rate on TLTROs. All these schemes as Nick Edmonds points out are - to my knowledge - collateralised (I'm not sure about the BoJ's lending scheme). But that is not really relevant. CBs are currently run with very significant credit risk and with huge potential for balance sheet losses on QE-related holdings - in fact if they *succeed* they will make huge losses. It seems to me that you have not refuted in any way the power of CB lending policies. Your main point seems to be that they should not get involved in selecting private sector credit risks - its not their job and they are ill-equipped. On this point, there is unlikely to be disagreement, although current practice (primarily the corporate bond buying often ECB) already exposes them to this. I would prefer they *optimise* their potential 'loss' and make transparent zero-coupon perpetual loans to all adult citizens. What you rightly point out is that the current system is one of opaque transfers and subsidies.

    1. Btw, all conventional monetary policy is an opaque system of transfers with distributional consequences - so that is definitively not the dividing line between the remit of politicians and central banks!

    2. My point is that governments already engage in practices to selectively lower interest rates for certain borrowers (loan guarantees, export funding, etc.). This is hardly a new area. In the euro area, it might be one of the few ways forward, as I noted above. Otherwise, it is not new policy space for the consolidated government.

      And who cares if 1% (for example) of borrowers get a lower rate? Why would you expect that to have a measurable economic impact? And if the programme is big enough to lower the average interest rate faced by the private sector, why not just lower the policy rate, since that seems to be the objective of the exercise?

      The corporate bond buying is a risk, but the central bank can hide behind the magic of diversification. Since the central bank is only one buyer, other bond buyers are being presumed to do the credit analysis. (Index funds are another source of "dumb money" for the credit market, and people do not spend too much time complaining about the lack of credit analysis -- although that might be a mistake.)

      My feelings on this are somewhat mixed. The central bank needs to act like a bank towards private banks -- and hence, assess credit risks. However, I have my doubts about extending this to the non-bank private sector.

      Sure the effects of monetary policy are opaque, but the instrument used (the policy rate) is transparent.

    3. Thanks Brian - that's all clear. Imagine we all bank with the central bank and we are in a liquidity trap - near zero rates and insufficient aggregate demand. How could the CB raise demand and fend off deflation? Very simple - by running a negative net interest margin (in fact this is what the Chinese used to do): raise deposit rates and cut lending rates ... These levers allow a central bank to raise nominal spending at will. The fact that the BoJ, ECB and BoE have introduced two interest rates and are willing to move them separately is a major institutional break with past practice, and analogous but very different to fiscal subsidies - it is financed by base money, implemented by independent CBs and aimed at meeting an inflation target. I agree that what they have done so far is relatively insignificant, but that is not the point: they now have an extremely powerful tool.

    4. The idea that CB can raise nominal spending at will with this tool is entirely speculative.
      The few instances of transfers that have been studied (bush tax cuts, Alaska fund transfer) suggest that at best you can get a temporary boost. What's the point of a one time boost if potential growth doesn't change? And this was done at a time when the risk free rates were much higher so it's not clear how they can be comparable to today.

      Also what is the evidence of a shortage of credit supply that this policy will fix? There is a lack of credit demand and is unlikely to change.

    5. A one-time boost to growth is useful if the economy is near recession. Recessions tend to take out capacity, so they could reduce potential growth. (However, there is the "creative destruction" angle that weak firms need to be periodically culled.)

      At present, this is not too much of an issue for the U.S. However, the euro area economy (at least the periphery) is on life support, so these ideas are more attractive over there. If and when the US falls into recession, then the call for helicopters will be loud.

      And yes, I see no shortage of credit availability. Institutional investors are desperate for any viable lending opportunities to pick up yield.

    6. (My previous comment was addressed to "Unkown".)


      If the CB is running a negative net interest margin, what exactly is the advantage of "funding with base money"? They are still losing money.

      This is not really costless. If the central bank is no longer profitable, it can no longer pay dividends to the government. This then damages the Treasury's ability to maneuver within the current institutional framework. The Treasury might be forced to tighten policy, cancelling out the alleged stimulus provided by the central bank.

    7. I guess that the CB can have unlimited ammunition - just figure a way to give (as a gift) money a way. That certainly can be interest rate margin, HM or anything.

      I see the point with the ECB because fiscal fetisism in Europe but what is the point with the Fed. The fiscal side is usually called upon when money (wealth and well-being) needs to be redistributed.

      I think to avoid recession one needs to implement policies that benefits those with higher propensity to consume. The CB is not well posed in that sense.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.