The Berkeley Economic History Lab is blogging a lot of its recent working papers, and they’re a goldmine of great stuff. Here’s Richard Sutch writing in October this year, whose The Liquidity Trap, the Great Depression, and Unconventional Policy: Reading Keynes at the Zero Lower Bound basically recovers an important idea from the General Theory and Keynes’ practice during the Depression.
Sutch’s gloss of Keynes is that an important way in which the zero lower bound constraint bites is that there is always a term-structure of interest rates, rather than anything like a single economywide rate of interest. As a result, even if short rates hit the ZLB or even go negative, a large segment of the yield spectrum will still be significantly positive. This of course has some consequences for the debate about when Keynes broke with the Wicksellian idea of a single market interest rate that might deviate from a full-employment natural rate.
He argues that Keynes micro-founded this on differences between the risk profiles of borrowers and lenders. Borrowers and lenders both face the risk that whatever enterprise is being financed will fail and the loan won’t be paid off. Borrowers stand to lose whatever security is put down for the loan, while lenders stand to lose the difference between the security and the principal (i.e. their risk is fundamentally about estimating how much security is enough). In theory, arbitrage should transmit lower rates at the short end along the whole curve, because if you can borrow for a year and roll it over cheaper than you can borrow for 5 years, you will.
But here’s the problem; lenders bring their own idiosyncratic risk to the table. Each event of refinancing brings with it the risk that potential lenders have become illiquid, a so-called sudden stop. This always exists unless the life of the loan matches the life of the asset exactly, and it is an attribute of lenders, not borrowers. Therefore, long-term credit comes at a premium, and in a sense what is “long” is defined in relation to the typical life of capital investments.
Therefore, it’s quite possible for the policy rate to hit zero or even theoretically drive through the ZLB, while a large proportion of the universe of credit still has significantly positive real interest rates. This implies that unconventional policy of some sort – perhaps a combination of QE and an “Operation Twist”-like effort to target long rates, or direct fiscal reflation – would be needed and that’s what the man concluded.
An example of the sudden stop would be another of their papers, Olivier Accominotti and Barry Eichengreen’s The Mother of All Sudden Stops: Capital Flows and Reversals in Europe, 1919-1932. In this one, Accominotti and Eichengreen have literally discovered a trove of historical documents in an archive. It’s a catalogue of major capital-raising exercises in Europe in the 1920s and 1930s, covering the major financial centres and most of the second tier as well. The conclusion is that the rolling financial crisis starting with Creditanstalt in 1931, defined as a sudden stop of international lending followed by capital flight, was driven by volatility in the stock market – it was, in fact, the Great Crash and its lesser crashes that did it. The correlation with volatility in world equities was much higher than with any economic variable in the countries affected.
An example of policy would be Eric Monnet’s Financing a Planned Economy: Institutions and Credit Allocation in the French Golden Age of Growth (1954-1974). This one comes from Paris School of Economics – surely the fac Piketty these days – and you can tell because it’s crunchy with empiricism. Monnet has constructed a database of lending registered with the Banque de France that provides series into very detailed industrial sectors, and another one of firms’ operating results based on tax returns, going through what sounds like epic pain to match the excisemen’s classification up with the central bankers’ and further with the national statistics. The key result is that the change in the state-directed, or as he would put it, state-influenced lending was very strongly correlated with internal rates of return, implying that the system worked well as an allocator of capital.
He’s also done a lot of qualitative work to understand how the French financial sector worked at the time. It was a lot more complicated and subtle than the caricature of being directed by the government, and it evolved over time. To begin with, a lot of lending really was directed by government and issued by the finance ministry, mostly to large capital projects in infrastructure and heavy industry. With time, the heavy lifting moved to a new layer of specialist lenders who faced projects in manufacturing, housing, and tourism. Influence rather than control was very much the point. The key financial product was long-term lending of 5 years plus.
There’s much more stuff in there – the fall of the USSR in a trade perspective, equities and anti-Semitism, Ottoman and Austrian administration and their long-term effects on growth.