Main point: Rajan is not going to parrot Warsh’s somewhat flippant thoughts on the role of the balance sheet in monetary policy. Rajan has some interesting thoughts on what we might call moral hazard in the Fed’s provision of central bank liquidity, which do favor a smaller balance sheet. But they do not rise to the level of a macro story than could guide anyone now.
Bloomberg published a story suggesting that Kevin Warsh has won “credibility” by appointing smart people who agree with him to the task forces, especially those related to the balance sheet and communication strategy. That framing is not atypical among our thirsty mainstream media, although it is funny.
More strikingly, in the case of former Fed Governor Jeremy Stein’s role on the balance sheet task force, it is simply wrong. Stein is an advocate of a large balance sheet on financial stability grounds, although he does have very particular views about what the maturity structure should be. He wants the notes to be long enough maturity to flatten the short end of the yield curve, via portfolio balance effects, but not so long that they get the Fed involved in fiscal policy by taking significant PnL risk. Stein may be persuadable, but his priors are not as described by Bloomberg. That he disagrees with Warsh is the credibility enhancing part of this, although my own view is that he is severely mistaken about the importance of portfolio balance.
The case of Raghuram Rajan is a more complicated one. And I would not be surprised if Rajan is viewed as the effective leader of the task force. He previously ran the Reserve Bank of India, and he won great praise around the time of the Global Financial Crisis for having previously picked a fight with Larry Summers over the role that derivatives might play in financial stability. Speaking to Jackson Hole in 2005, he mentioned that derivatives could end up hiding systemic risks. Responding, Larry Summers called him a Luddite, which people do remember.
In this note I will set out in point form what I take to be the more relevant aspects of Rajan’s take on the balance sheet. And in so doing, I will lean heavily on two bits of writing. The first is a transcript from an appearance he made on David Beckworth’s blog back in February. And the second, is his most recent academic (style) paper on the subject, which was presented to Jackson Hole in 2022. As you will see from the transcript, Rajan published earlier papers on this subject as well, but he seems to have imported the early insights into his most recent work.
· Rajan sort of hints that he might accept Warsh’s earlier claim that QE brought “largesse” (Warsh’s term) to a select part of the economy. But that part is very select indeed. And I see no indication that Rajan believes that pursuing a smaller balance sheet would create room for lower interest rates by removing the largesse from the privileged. Indeed, Rajan’s work emphasizes the ways in which QE might impose restraint on bank lending into the real economy and therefore offset whatever stimulus there might be from the presumed depressing effect on long-maturity yields from long-maturity bond purchases.
· Rajan suggests that his interest in this area was initially motivated by his observation that QE did not seem to provide much stimulus. The US economy recovered in the wake of the QE programs, which beats ongoing contraction, he notes. But the recovery was actually slow. Rajan might have joined folks like myself by insisting that QE had little effect on yields beyond its signaling effects, which were detachable. But that is not the approach he takes. Rather, he accepts as plausible that QE might have depressed longer-maturity yields and provided stimulus through that channel, and then wonders whether there may have been some sort of offset.
· And this is where he and his co-authors seem to offer the greatest analytical and empirical contribution to the debate. The details are elaborated in the academic paper linked above. In summary, he finds that banks responded to the rise of their reserves stock occasioned by QE by trying to shed the related increase of their liquidity. They did this by extinguishing term deposits and replacing the required funding with demand deposits. And they also increased their lines of credit to hedge funds, noting particularly those involved in Treasury market carry trades.
· The above had two effects. First, it put the banking system in a position of being more vulnerable to systematic (as opposed to bank specific) liquidity shocks. As a result, the overall system did not behave as if it were more liquid. Second, Rajan suggests that the increased risk that banks took on board via this liquidity channel may have convinced them to take less risk elsewhere, e.g., through lending into the real economy.
· Rajan also suggests that repeated rounds of QE have resulted in “ratcheting up” what appears to be the level of reserves (or more generally the size of the Fed’s balance sheet) that is consistent with an ample reserves regime. And this claim invokes two forms of “momentum.” The first form of momentum relates to bank efforts to shed liquidity which may persist, within banks, beyond the period when the Fed is raising central bank liquidity provision via QE. For example, the prime brokers may not get the memo to stop pushing lines of credit, he suggests. The second form has to do with moral hazard. If banks get the message that the Fed will offset liquidity shocks, such as those that were associated with earlier efforts at balance sheet reduction, then banks will be less sensitive to exposing themselves to systemic liquidity shocks. Accordingly, efforts at central bank liquidity withdrawal may not be as “benign,” to use his term, as people have blithely assumed. Hence the title of his paper to Jackson Hole: Why Shrinking Central Bank Balance Sheets is an Uphill Task.
· An obvious question here is what is to be done, then? Let’s start with the obvious and most practically important. I get the impression that Rajan would be willing to pay some sort of price, in terms of market disruption and short-term economic performance, to get to a smaller balance sheet and to wean the banking system off its morally hazardous assumptions. But his emphasis is on moving slowly and on monitoring how the banks manage the liability side of their balance sheet. I have no idea if he would be comfortable with regulation there, and I note that much of the discussion these days is about how to reduce regulation to reduce bank demand for reserves. (Separate discussion.) But what we do know is that he does not sound like he would be in a rush. And I suspect that he would prefer the disruptions to be kept small. He mentions letting the first bad actor go and then bailing out all the rest – in the event of a liquidity-related accident.
· This perspective fits into my view that the asset side of the Fed’s balance sheet is not where the action is and that the official story that the Fed long told about how QE works was radically overstated. Relatedly, the effects running through the asset side of the balance sheet being reduced should be quite small and overwhelmed by other considerations, such as the large fiscal deficit or any swing in the stock-bond correlation.
· The liability side is a bit trickier and may eventually be relevant to bank asset-liability management and to exotic liquidity spreads in money markets. I just do not think that there is a macro trade related to that. The liability side will be managed carefully, so there is little to do ahead of the event and when it arrives it will largely be of interest to those directly related to the banks. I doubt even Rajan would be interested in causing a major disturbance through this channel. And to repeat, no, he is not just channeling what Warsh previously suggested, mostly while winging it.
