The BoJ has started helicopter drops

Oddly, some economists think that helicopter drops are either beyond the capacity of central banks or highly unlikely. Neither is true – they have already started.

It is now clear that the Bank of Japan is engaging in helicopter drops. Helicopter drops are transfers to the private sector financed by the creation of money (bank reserves) under the direction of the central bank. This is the definition that Milton Friedman uses when discussing actual practical policies in a liquidity trap (in his AER presidential address in 1968) – although he does not use the term “helicopter”. I only differ from Friedman by designating institutional responsibility to the central bank – he does not discuss who should direct the policy, but deems it “monetary policy” because it is financed by base money.

The recent innovation in monetary policy by the Bank of Japan is a clear case of a helicopter drop. Here’s why: central banks, due to quantitative easing (QE), have created huge levels of bank reserves. Bank reserves are essentially electronic money which the banks hold at the central bank. The central bank determines the stock of reserves, usually by purchasing assets (QE).

Because the stock of reserves is so high, central banks pay “interest on reserves” (IOR) to influence market interest rates. For example, when the US Federal Reserve Board raised interest rates in December, it raised the IOR to ensure market rates rose in line with their new target. If the IOR remained below the target Fed funds rate, market rates would never reach target because banks could lend reserves to each other at a rate close to the IOR.

For this reason, it is assumed that when the central bank changes its target interest rate, it will change the IOR in line with its target – just as the Fed has done.

But the BoJ has clearly broken ranks, and made a helicopter drop to banks. It has also shown how easy this is to do. It has done so by introducing three distinct interest rates on reserves: required reserves – which banks must hold – these are paid zero, and are relatively small in quantity; existing reserves – these are now paid 10bps; and a new third tier – a “policy balance” which will be paid minus 10bps. To be clear, the income banks derive from reserves held at the BoJ has been substantially insulated from negative interest rates. Using the BoJ’s terminology, the ‘basic balance’ of existing reserves which pays 10bps, totals ¥210trn. Zero interest rates will be paid on ¥40trn. Only ¥10trn-30trn will receive negative rates. Also – importantly – it appears that this balance of ¥10trn-30trn will be fixed at these levels, despite QE creating ¥80trn of new reserves annually. New reserves will receive zero interest. So QE may well increase the profitability of banks (if the BoJ buys negative yielding JGBs and then pays zero) – a further transfer to the banking sector.

The helicopter drop is the transfer payment that the BoJ is making to banks on their existing reserves, which is unnecessary in conventional monetary policy: it is neither a regulatory requirement nor an interest rate which affects market rates. It is in fact a transfer payment to the private sector (in this case, the Japanese banks) financed by the BoJ – i.e. a helicopter drop.

And the Bank of Japan is explicit about this – it has introduced the new tier of reserves – a “basic balance” – to support the profitability of financial intermediaries.

The BoJ can increase its helicopter drop by raising the IOR on this balance.

So, helicopter drops have started. Now let’s make them transparent and sensible.

About The Author

Eric Lonergan is a macro fund manager, economist, and writer. His most recent book is Money (2nd ed) published by Routledge. He is also a supporter of Big Issue Invest (BII), the investment arm of The Big Issue, and is one of the initial limited partners in BII’s Social Enterprise Investment Fund LP. In a personal capacity, he makes direct investments in social enterprises. He also supports and advises The Empathy Museum.

25 Responses

  1. JP Koning

    One of the effects of the decision to provide a protective tier is a reduction in BoJ profits and a decline in seigniorage flows the government. The more IOR is raised, the bigger the decline in these flows. If the government in turn decides to reduce its spending or increase taxes to make up for the budget shortfall, then helicopter money isn’t really helicopter money. It’s a gift that’s just as quickly removed.

      • Bob

        “A helicopter drop of any sort can be negated by raising taxes.”

        Isn’t that just basic income?

      • t.

        this makes no sense.

        think in reverse. imagine a banking system with policy rates at 10% and inflation at 10%, and huge required reserves remunerated at 0%. cash and checking accounts pay zero–even though opportunity cost is huge.

        in your example, this tax on currency and checkable deposits would all else equal impede inflationary dynamics. But in such a system the hot-potato effect is actually very strong and it is this which determines an equilibrium price level.

        lets say jgbs yield -1% and the boj pays the banks plus 5%, if this policy is credible in anyway it will only be so to the extent that banks are willing to hold an unlimited quantity of this asset as a store of value. anything that is held as a store of value willingly cannot be used to tie down the price level path (except via strong modeling assumption likes the last period exchange of real debt for real goods in the FTPL for example). and yet in this example you tie down a higherprice level by the manifest irresponsibility of this policy via the negative carry implying an explosive path for the future monetary base. sorry, not going to work.

        by the way, your EZ helicopter drop wont work for the same reason. lets just assume in the future i can buy retail government bonds with my stimulus check. well those are a good asset today, they dominate the return to like assets for nowand can be exchanged for higher yielding assets tomorrow (if yields are never higher tomorrow then they dominate in return at EVERY period, and therfore cannot be used to tie down the price level–this is actually a big problem for nirp advocates like Miles Kimball, he needs policy rates to be positive in the future but there is no guarantee his policy will get us there.)

        If you make the stimulus check non-fungible via some kind of expiry or the requirement to only hold in cash youre going to run into some big problems. leave aside that we cant do Adair Turner’s suggestion because the liabilities of the central bank TODAY are owed to the private sector so it doesnt matter if you expunge their assets-which is a formidable problem.

        Assume instead that the liability is just a treasury account balance that is used to make routine payments. i think in your model you want to imagine these payments to function like cash and getting all the lift out of thinking cash ia fundamentally different than bank reserves (because you are holding the total spending equal in order to be fair). if you want the structure of central bank liabilties to tie down the future price level, so be it, but this is going to be particularly heavy lifting when cash is itself a credible store of value.

        what they should do is actually quite simple, they should just say our balance sheet will continue to grow until we reach a price level target drawn from 2014 until now (just choose a date where inflation index was already below but not well below long term trend)

      • Eric Lonergan

        I’m not sure we can say anything about your example because you have not specified what the lending rate is. There should be a ‘hot potato’ effect already – but money growth doesn’t occur that simply. It depends equally on the demand for credit – and on constraints other than simple price.

        I should be clear that I am not trying to raise inflation. I actually don’t think any of the policies I advocate would raise inflation. I think targeted LTROs at negative rates to the corporate sector would raise corporate profits mechanically, raising the IOR would raise bank profits mechanically – which it already is doing in Japan.

        As for perpetual zero interest loans to households this would raise household net wealth – and the CB can pretty much target demand and household income.

        I don’t buy the CPI expectations model of Krugman, Cochrane et al. It’s clearly the wrong model. The analogy here is a tax cut. My own personal view is that these policies are highly unlikely to cause a material change in CPI because of a highly competitive global microeconomic structure, lots of spare capacity, an likely endogenous supply side improvement, and an acceleration in innovation which is being missed by depressed confidence.

        Remember virtually all macro models nowadays simply work through real interest rates, so raising inflation expectations by *assumption* becomes the only way to grow demand. But this is little better than a tautology with an empirically false premise.

        All the empirical evidence shows tax cuts raise spending to varying degrees. So, obviously, would perpetual, zero, TLTROs. But expectations are not the channel. Not least because at the moment no one has any. They don’t know.

    • t.

      I disagree but sensible. so then given the preeminence of expectations in modern macro—and its there for a reason, we actually have markets to trade all this stuff–what do you think should tie down the future price level?

      we cant all be switzerland–see my switzerland comment on nick rowe’s site under “is/lm” pictures with money.

      • Eric Lonergan

        I think these models are more meaningful in a 1970s & 1980s regime, or in EM countries like Brazil – where inflation is shifting materially and contracts are indexed. That, of course, is when most of these models were constructed. They have since been refined and complicated – but structural change has rendered them obsolete. I really don’t think inflation expectations are that important in the developed world, partly because the microeconomic structure has changed, These models have not been derived from empirically-based premises. They are simply mathematically tractable, and not really describing the real world. Simpler, old Keynesian models are probably more relevant to the current structural regime. And they have more realistic ‘microfoundations’. As for the future price level, there probably is some underlying inflation, but it is not very relevant to decision-making in the context of relative price shifts and changes in quality. It is true that financial markets provide prices for inflation ‘expectations’ – but these are largely set by convention (i.e. 2%) – hence their remarkable stability when adjusted for liquidity and other risk premia. Also these market prices are of little significance to the economy. I do think expectations for growth (optimism/pessimism) are significant. Most people’s ‘expectations’ are a form of myopic extrapolation, heavily anchored to recent direct experience. The institutionalised indexation of the 1970s & 80s is a very different world. I don’t think we need new models, though. I think pre-rational expectations models work fine.

  2. Helicopter engineering

    […] engineering — to the point where Lonergan himself judges that helicopter drops are already going on in Japan, even if nobody has noticed. His point is that even as the Bank of Japan has lowered its policy […]

  3. t.

    not going to get very far without a nominal anchor. why should treasury bonds yield 1% as against 10%? if we listen to miles kimball they will yield zero percent and milton friedmans dream of money becoming a dominant store of value would be realised. i fail to seewhat would tie down the price level were that to hapen though.

    miles thinks its a particular form of the old-keynesian investment accelerator in real interest rate space linked back to nominal space via full employment.

    you seem to discount this channel.

  4. t.

    the price level is tied down by an equation in any macro model, mv=py, the nkpc in conjunction with an interest rate rule, or the last period real value of government debt for example.

    to say aggregate nominal variables, like the wage level, nominal gdp, inflation, etc are determined by real micro processes is non-sensical. it is actually worse than non-sensical; it renders macro vacuous.

    nominal zero coupon bonds trade below par because we expect money to buy less in the future than we do today. thats a good state of affairs. not incidentally it allows you tie down the price level by allowing the medium of exchange to becstrictly dominated in return by a risk free store of value. again, very helpfully.


Leave a Reply

Your email address will not be published.

* Checkbox GDPR is required


I agree

Privacy Preference Center