Yichuan Wang had a spectacular, wonky post trying to adjudicate a debate about interest on excess reserves that I’ve been having with David Beckworth and Dan Carrol.
I’ve been meaning to write about it for a while, but unfortunately a very lengthy recap is needed first, or it won’t make sense to the new reader. The three of you already familiar with the debate should skip ahead to the next section.
— I first argued that there are risks to lowering IOER that weren’t being considered by some economists who were recommending it. Specifically, such a move could create havoc in money markets that aren’t built to handle negative nominal rates, which would be a possible if not likely consequence. An unlinking of policy from effective rates, a run on money market funds, or chaos in Treasury auctions are some of the possibilities I mentioned — potential consequences of removing what is essentially a safe asset substitute from heavily collateralised short-term lending markets.
Be sure to also read the incomparable Bond Girl on the role of the GSEs, the FDIC’s deposit insurance assessment, and how this would affect federal funds trading; and of course Izzy’s excellent post on delivery fails in Treasury repo markets.
(I’m further worried, like Izzy, that negative nominal short-term rates in a world with a shortage of safe assets would have an unexpectedly deflationary influence. But this suggestion might get me laughed out of some parts of the econo-blogosphere, so here let’s just stick to the problems specific to financial markets.)
— David Beckworth, who is always worth reading on monetary economics, responded that lowering IOER would send a powerful message that the Fed was committed to a permanent expansion of the monetary base. This signaling effect could well raise expectations for economic growth and inflation, leading to higher demand for investment and credit — and banks would then have more opportunities to lend money into riskier ventures. This would leave plenty of collateral for the shadow banking system to function properly and allow repo and other short-term rates to stay positive.
— I countered by reiterating a few points and arguing that even if David is right, the signaling effect might take some time to work, whereas the problems in money markets would more likely emerge abruptly.
— Beckworth acknowledged this, and said that it was possible that lowering IOER might have to be accompanied with a more dramatic, sudden monetary policy regime change — ideally NGDP level targeting — for it to be effective.
None of these effects that I worry about is certain, of course, but given that the benefits of eliminating IOER are also likely to be quite small, I think it’s at the very least a problematic idea. As I wrote a couple of weeks ago, it seems there are better ways for monetary policy to boost demand and raise inflation expectations (and thereby lower expected real rates), most notably by making it clear that the Fed would temporarily tolerate higher inflation while unemployment comes down further; or perhaps some variation of the Evans Rule.
— Dan Carroll’s smart and considered contribution came next.
I found parts of Carroll’s response to misrepresent what I had originally written — not that I think he did so deliberately, as this is complicated stuff and much can get lost in translation, surely my fault as much as his.
But he seemed to think that I’m worried about money market funds, or the bank arbitrage that IOER provides, for their own sakes rather than for the systemic threats their disruption would represent. Trust me, I’m not — I think money market funds need to be regulated further and would shed no tears if their business model, under regulation that would make them safer, turned out to be unsustainable. But in the meantime, they do remain systemically important, and thus we all have to deal with the money markets we have, not the money markets we wish we had.
He also notes that what I wrote was speculative. True, but then anything written on the topic is speculative.
Carroll’s other idea, and a few points about the mechanics
But Carroll did make one intriguing suggestion that deserves a detailed explanation before getting to Wang’s post. I don’t agree with the idea, but it is well worth a discussion for the issues it raises.
Carroll writes that the Fed could sterilise its lowering of IOER by simultaneously selling Treasuries from its balance sheet:
1. The key flaw in [Garcia’s] argument is that it is assuming that the move of eliminating IOER will be unsterilized. The Fed invests the excess reserves into T-bills and other Treasuries. If the excess reserves go away, one would hope that the Fed would sell off assets in response. Indeed, the Fed could simply give banks T-bills when they ask for their reserves, rather than pay a dealer to conduct a transaction.
a. I am assuming that the Fed wants only to incrementally expand the money supply, rather than take an amount equal to all of the excess reserves and dump them onto the economy. ….
The huge balance sheet at the Fed is a political football. By incentivizing the banks to move excess reserves elsewhere, the Fed gets to shrink its balance sheet, a move that is sterilized if the Fed sells its holdings, while covertly engaging in other stimulative moves (such as selling fewer holdings than it should).
I understand this differently and think Carroll might have the relevant mechanisms mixed up, though perhaps I’m misreading him and some of our differences are just semantics.
The Fed doesn’t really “invest” excess reserves into T-bills and other Treasuries. Rather, the Fed originally bought those T-bills and other Treasuries by crediting the dealer banks with reserves that didn’t previously exist (which is why those reserves count as liabilities on the Fed’s balance sheet). This is just how the Fed’s balance sheet is expanded — and there certainly wouldn’t be any obligation on the part of the Fed to sterilise, as Carroll appears to suggest, though this fact on its own obviously doesn’t disqualify his idea.
What determines the size of the monetary base? As with any other institution’s balance sheet, the Fed dictates that its liabilities (plus capital) equal its assets. The Fed’s assets are predominantly Treasury and mortgage-backed securities, most of which have been acquired as part of the large-scale asset purchase programs. In other words, the size of the monetary base is determined by the amount of assets held by the Fed, which is decided by the Federal Open Market Committee as part of its monetary policy.
It’s now becoming clear where our story’s going. Because lowering the interest rate paid on reserves wouldn’t change the quantity of assets held by the Fed, it must not change the total size of the monetary base either. …
And a complementary note from Credit Suisse strategists:
Some have suggested that a cut in the interest paid on excess reserves would incentive banks to withdraw excess reserves, but this is not possible. More correctly, the hope would be to incentivize banks to transform a portion of excess reserves into required reserves, which is done by banks extending credit or purchasing some asset that then becomes someone else’s deposit – meaning some excess reserves become required reserves (in aggregate, not necessarily at the lending bank).
The excess reserves could also become currency in circulation. The point here is that banks’ reducing their excess reserves would only change the composition of the Fed’s liabilities, not their size. There’s a difference between measures of money supply (which don’t include excess reserves) and the monetary base (which does).
This is why Carroll writes that in his sterilised version of lowering IOER, the Fed will only be increasing the money supply “incrementally”. In other words, the sterilisation process would allow the Fed to actually decrease the monetary base by selling some of its assets while at the same time increasing the money supply because lowering IOER will hopefully result in a higher money multiplier (as banks are induced to increase lending).
So rather than reducing the number of safe assets, the Fed would be engaging in a kind of safe asset swap — more Treasuries in exchange for lower IOER. The Fed won’t be “pushing on a string” any longer, or not to the same degree that it has been to this point.
Now, there’s another part of this argument that is difficult to grasp. Beckworth and Carroll believe that lowering IOER will send a signal that the Fed is committed to a larger monetary base in the future than markets are currently expecting, which will raise longer-term inflation expectations and drive more economic activity now. Put differently, they argue that lowering IOER sends the signal that the Fed will shrink its balance sheet by less than expected.
But the sterilisation process, of course, shrinks the monetary base. How can it be that shrinking the current monetary base can also send the signal that the future monetary base will be bigger than expected?
Yichuan Wang’s comparison
This is where I’ll excerpt at length from Wang’s post, which does a fantastic job of elaborating on — but stopping short of endorsing — Carroll’s idea:
In this case, the Fed removes IOER, but then partially compensates for the treasuries bought by commercial banks by selling its own stock of treasuries. The drawing looks something like this:
This might seem counter intuitive as the central bank appears to be doing two actions that seem to contradict each other. Yet if we consider the role of expectations, such a policy becomes much more logical. Given that the monetary base has more than tripled since 2008, it should be clear that the market does not expect that expansion to be permanent. According to Krugman, a fully credible expansion of the monetary base in this period and all future periods should directly lead to inflation. Therefore, since prices have not tripled in response to the change in the base, markets must be pricing in the fact that the base expansion will be sterilized by the Fed in the future.
To raise inflation expectations, the Fed must credibly commit to a future base expansion, and, paradoxically, it cannot do so if the monetary base is too large. Therefore, if sterilized IOER reduction is seen as a move to hitting a nominal target, such as higher NGDP, it can still be part of a credible package that restores the nominal target while preserving the shadow banking sector. …
This would expand the supply of safe collateral and address Garcia’s concerns. In addition, giving up on IOER and credibly committing to a permanent base expansion would address Beckworth’s call for a regime shift that would restore trend NGDP growth. …
Another framework in which sterilized IOER makes sense is DeLong’s law, a modification of Say’s law and Walras’ law. …
This framework clearly delineates between what Beckworth, Garcia, and Carrol are proposing. Beckworth argues that lowering IOER directly solves excess demand for money, and therefore go on to solve excess demand for safe assets. But Garcia argues lowering IOER directly increases excess demand for safe assets to such an extent that it overwhelms any reduction in excess demand for money. So while directly cutting IOER reduces excess demand for money, it’s ambiguous whether it reduces the general glut for goods. Carrol’s proposal then comes in the middle, as cutting IOER reduces excess demand in money while sterilization reduces excess demand for safe assets.
In the end, this debate shows not only why the market monetarist focus on expectations is important, but also why an analysis of mechanisms cannot be ignored. All of these arguments for sterilized IOER depend on a credible commitment [to] expansion of the monetary base, so if the market expects the Fed to maintain a 2% inflation ceiling the policy change would still be useless. However, changing the target without being aware of the collateralized world we live in would also be a failure, as we would be twisting the dials in all the wrong directions, endangering monetary credibility.
Still not sold
Wang’s analysis is truly excellent, and I recommend the whole post as a wondrously clear discourse on how the shadow banking system has complicated the lives of monetary policymakers in unprecedented ways.
And the idea of a sterilised reduction of IOER certainly seems like the best of all worlds: more collateral for the shadow banking system, positive short-term rates, a stimulative move that increases lending and aggregate demand, plus the Fed ends up with a smaller balance sheet from the sterilisation. Woot!
But as the cliche goes, if something seems too good to be true…
If IOER has to be lowered, then a sterilised reduction is probably better than an unsterilised reduction, but I’m unconvinced that either is a good idea.
As Wang notes, the efficacy of the idea still relies on the signaling effect translating into higher aggregate demand. And for this to happen, the Fed must convince the markets that it is no longer treating 2 per cent as an inflation ceiling, but as a symmetric medium-term target. Otherwise the idea is “useless”, as Wang writes. Though we’d go a step further and say that it also still has risks that aren’t justified by its likely modest benefits.
— The NY Fed explained why simply lowering the IOER was unlikely on its own to be enough incentive to change banks’ lending decisions. Their analysis didn’t consider the signaling effect, an important omission that we’ll discuss next, but there are other reasons. Part of it has to do with banks’ lending standards: these have happily been loosening for everything but mortgages — but it is a gradual loosening, and lowering IOER would probably only speed it up marginally. This is especially the case when you consider this next point, which we hadn’t made before, also from Credit Suisse:
Banks are more concerned about capital retention and asset size to comply with new capital regulations that either have been mandated or are in the process of being finalized. Evidence of this risk aversion can be found in aggregate bank data, in which total assets have continued to climb (in no small part to excess reserve balances increasing), while risk-weighted assets have remained stable.
Again this doesn’t mean that lowering IOER wouldn’t increase lending, only that it’s uncertain and might not be by much.
— Just how much sterilisation of this kind by the Fed would be realistic? There is no way to know, of course, what amount of the banks’ excess reserves would go towards backing lending versus how much would simply chase Treasuries if IOER were eliminated. But there are currently some $1.7 trillion in excess reserves. The Fed certainly wouldn’t have to sterilise all of it to keep short-term rates positive, but the numbers would likely be very high. Which…
— … could muddy the signal on which the mechanism depends in the first place. Let’s reproduce one of Yang’s comments from above:
To raise inflation expectations, the Fed must credibly commit to a future base expansion, and, paradoxically, it cannot do so if the monetary base is too large.
This would be an unbelievably delicate balancing act on the part of the Fed. It would have to lower the base by such an amount, and in such a way, as to signal a commitment to a higher future base without the markets misinterpreting the shrunken balance sheet as itself a form of Fed tightening. Got all that? Remember that this essentially would be a kind of reverse quantitative easing, which means the signal already has a pretty good chance of becoming muddled. Not tomention a big element of what’s-the-point-ism to it, as IOER is itself already a kind of sterilisation, as Izzy wrote.
— But let’s say the signal itself is unambiguous. My earlier response to Beckworth still stands, which is that the signal could take some time to work through the system, especially given the reluctance of banks to accelerate lending and of companies to accelerate borrowing and investing — while the threat to money markets could still materialise very quickly, especially if the Fed encounters unanticipated logistical problems for a massive sterilisation. In fact, Beckworth’s second response to me also still stands — which is that this move would probably have to be accompanied by a full-scale monetary policy regime change for it to have a chance at producing an immediate and significant rise in expectations.
— If you’re going to sterilise a monetary policy operation to protect collateral, then sterilised QE, using reverse repos, seems a much better idea. So is an MBS-Treasury swap, or even an MBS Twist. These don’t risk the threat of zero or negative rates on money markets, though the signaling problem remains.
— And even if the collateralised parts of the shadow banking system would stay unharmed, Bond Girl’s point about losing control of federal funds trading (by definition uncollateralised) would also still hold, and this can lead to other, future complications in the Fed’s communications regime.
Bottom line… There are better, safer ways for the Fed to raise demand and inflation expectations (and let’s not even start on fiscal policy). It should stick to those.
Many thanks to David, Dan, and Yichuan for this very engaging thought experiment. But I continue to think that eliminating IOER is better done later, once NGDP has been accelerating for a while, not as a mechanism that’s meant to spur this acceleration.
A response to market monetarists on IOER – FT Alphaville
IOER, negative rates, and Ben – FT Alphaville
lncrementalism is the killer – Macro and other market musings
A practitioner’s perspective on eliminating IOER – Symphonic Chaos
How cutting IOER could potentially affect federal funds trading – Self-evident
Interest on excess reserves: an illustrated investigation – Synthenomics