The “Lehman moment” of 1931
I’ve been reviewing the editor’s proofs of my Depression manuscript (don’t ask) and I suddenly felt a sense of deja vu. Britain left the gold standard on September 20, 1931 (the crises are always September or October—1929, 1931, 1937, 1987, 1992, 2008, etc). And then everything started falling apart.
But not right away. Stocks in the US did fall on September 18-19 as rumors of the devaluation leaked out, but then stocks actually rose modestly over the next few days. And then the US stock market crashed.
Why? And why the delay?
Now flash forward to the Lehman failure. Stocks fall in September 2008 due to Lehman, but nothing disastrous. Then the stock market collapses in the first 10 days of October, and zig zags much lower up until March 2009.
Why? And why the delay?
It turns out that the dynamic was very similar on each occasion. The initial shock in 1931 was worrisome, but not by itself a major factor. Then a few days after the British left gold a run on the dollar began (fear of devaluation), and massive gold and currency hoarding commenced. Gold and currency were the dual media of account. Massive hoarding means a massive increase in demand for gold and cash. Econ 101 says an increase in the demand for any asset makes that asset more valuable. But the nominal price of the media of account cannot rise. So the only way for them to increase in value (in real terms) is for the price level to fall. And commodity prices did plunge during this period.
In 2008 the Lehman event was worrisome, but not catastrophic. Then it gradually became apparent in early October that demand for liquidity was soaring. Even worse it became clear that the world’s major central banks were not willing to cut interest rates sharply enough and/or supply enough liquidity to prevent NGDP expectations from plunging. And despite Bernanke’s insistence that the Fed should never let the US fall into a Japanese-style liquidity trap, we did. Even worse, the Fed adopted IOR in early October and stock prices crashed. Then they raised IOR a few weeks later and stock prices crashed again. Then they raised IOR a few weeks later and stock prices crashed again.
In both cases the delayed stock crashes and severe output declines and severe financial distress all had the same cause—falling AD caused by bad monetary policy. The 1931 event was actually slightly more excusable—the gold standard was a constraint on policymakers. Fiat money banks have no excuse at all. Especially fiat money central banks that continued to raise and lower interest rates over the next few years, and hence were able to avoid the zero bound trap. I.e. the ECB.
Tags:
16. December 2013 at 06:49
Scott,
Can you please describe the precise mechanism through which falling AD causes a stock market crash AND a bank run/crisis. For a moment, as if you were to clearly write out a math model (which I know you hate).
And can this mechanism be tested? In other words, can there exist some empirical test to see whether the stock market crash of 2008 and Lehman’s fall and the subsequent financial crash were in fact caused by falling NGDP. What would a researcher have to find in order prove your point.
I’m thinking like a Pragmatist, William James stile.
Regards,
Ilya
16. December 2013 at 06:51
Also, what about Bear Sterns, which fell before the falling AD.
16. December 2013 at 07:14
“Even worse, the Fed adopted IOR in early October and stock prices crashed. Then they raised IOR a few weeks later and stock prices crashed again. Then they raised IOR a few weeks later and stock prices crashed again.”
In my morning reading I noticed a very interesting debate between Nick Rowe and Francis Coppola (and several other commenters) on IOR. There’s 177 comments and it took a while to read most of them. But I think it’s especially well worth reading Nick’s comments which presented arguments for why IOR is contractionary and why negative IOR should be considered.
http://pragcap.com/stop-with-the-cutting-ioer-will-increase-lending-madness
16. December 2013 at 07:15
The market indices bottomed on March 9 2009. That same day Fed Chairman Bernanke offered his tepid support to reverse the FASB mark-to-market requirement. Coincidence?
What would it say about monetarism if the 160% increase in the stock market was not a result of QE but a simple change in bank regulations? Does the crisis not show that financial structure matters far more than monetary policy? What mattered most in March 2009 was not the idea of easy money or NGDP or stimulus. What mattered was investor confidence that new money could put into the market without the fear that it would be washed away with the bad. Monetarism will always be secondary to the greater forces of greed & fear.
“Although he doesn’t endorse the full suspension of mark to market principles, he is open to improving it and provide “guidance” on reasonable ways to value assets to reduce their pro- cyclical effects”
http://en.wikipedia.org/wiki/Mark-to-market_accounting#Effect_on_subprime_crisis_and_Emergency_Economic_Stabilization_Act_of_2008
16. December 2013 at 07:34
There were two problems in October 2008.
First, the IOR was too low relative to the fed funds rate. During normal times this is not a big deal, but implementing the Friedman rule was a very good idea at that time when market interest rates where much higher than the fed funds target.
Second, fed funds rate target was set too high.
IOR regime itself is very helpful. At the end of 2011, greek default did not became Lehman II as a result of two 3 year LTRO operations – a policy instrument that uses IOR and is so powerful that ECB is afraid to use again.
16. December 2013 at 07:43
“…Britain left the gold standard on September 20, 1931 (the crises are always September or October””1929, 1931, 1937, 1987, 1992, 2008, etc)…Then a few days after the British left gold a run on the dollar began (fear of devaluation), and massive gold and currency hoarding commenced…The 1931 event was actually slightly more excusable””the gold standard was a constraint on policymakers…”
Krugman has recent post on a new paper by Nicholas Crafts:
http://krugman.blogs.nytimes.com/2013/12/15/if-only-it-were-the-1930s/?_r=0
“…First, his chart:
[Graph]
He also notes that while many European countries had lower debt ratios in the 30s than they do today, the UK actually had a substantially higher ratio “” and even more so after World War II. How did it deal with this debt? Not through the recipe currently being imposed in Europe, of fiscal austerity and internal devaluation. Instead, the UK relied on a cheap-money policy that produced low interest rates and moderate inflation “” “financial repression” “” with the central bank “subservient” not just to the government, but to government debt-management policies…”
There’s nothing particularly wrong with what Krugman says, and indeed the primary focus of the Crafts paper is debt reduction. But in his Vox article Crafts says the following about his graph which Krugman reproduced in his post:
“…Figure 1 contrasts the growth experience of the Eurozone since 2007 with two groups of countries, namely, the sterling bloc and the gold bloc after 1929. The former devalued in 1931 and experienced an early and quite rapid recovery. The latter stayed on the gold standard till the bitter end and even in 1938 had only just regained the 1929 level of real GDP. Sadly, the Euro area is following a trajectory that looks rather too reminiscent of that of the gold-bloc countries in the 1930s…”
You will search in vain in Krugman’s post for any mention of what the “sterling bloc” and “gold bloc” depicted in Crafts’ graph refer to.
The sterling bloc of countries included Japan, Norway, Sweden and the UK. They devalued early in 1931 and recovered from the Great Depression first. The gold bloc included Belgium, France, Italy Netherlands, Poland and Switzerland, and with the exception of Belgium didn’t devalue until 1936 and recovered from the Great Depression last.
In short the graph clearly shows the effectiveness of monetary policy at the zero lower bound, something Krugman goes out of his way to conceal these days.
P.S. Krugman also mentions (and links to) an earlier paper by Crafts, of which he says:
“In an earlier paper Crafts argued, in effect, that high government debt levels may even have been a sort of advantage, in that they made it possible to credibly commit to inflationary policies.”
But Krugman fails to note that one of Crafts main points in that paper was that UK monetary policy was effective at the zero lower bound despite fiscal consolidation.
16. December 2013 at 07:48
123,
IOR has complicated the situation for the Fed. Traditionally, IOR was on excess reserves, the Fed pays IOR on ALL reserves. IOR on required reserves represents compensation for a rise in capital requirements. Payment for excess reserves has complicated the nature of quantitative easing as banks have less incentive to lend, even as reserves increase. As Jeff Hummel has noted, this is on par with the Federal Reserve raising reserve requirements in 1937.
http://www.federalreserve.gov/monetarypolicy/reqresbalances.htm
16. December 2013 at 08:06
The problem in October 2008 was that the market believed the true value of bank assets was much lower than claimed. And for good reason! The market value of bank assets in 2008 was much lower than claimed. The problem was not IOR. It was not NGDP. It had nothing to do with monetary policy. The problem was the insolvency of the banking system as a result of fraudulent loans and investments and basic bad lending practices.
How did that happen? Trust me, it was not because of bad monetary policy.
Also, it is economic malpractice to ignore that commodity bubble that occurred in the first 9 months of 2008. Energy prices hit record levels in the summer of 2008. Not only was this a drain on consumer finances but this made it politically unacceptable for the central bank to argue for monetary easing. Fact is there was no simple monetary answer to what was happening in the economy in 2008. To argue in hindsight that there was is either hubris or ignorance.
16. December 2013 at 08:29
How does price level fall? Is it due to a slowdown in velocity associated with hoarding?
16. December 2013 at 08:38
JoeMac, It’s not easy to prove. Tthe best way is to look for signs that stocks responded negatively to news of falling NGDP expectations. You can look at market responses to government data announcements, for instance. Glasner showed that stocks fell in rseponse to falling TIPS spreads during the period after 2007. Unfortunately we lack a NGDP futures market. If we had one this would all be much easier to show. You can also look at the market reaction to monetary stimulus or tapering. Japan is a good recent example, where the decision to raise the inflation target to 2% seems to have sharply raised stock prices.
The mechanism is that when the economy is depressed rising NGDP tends to raise corporate profits, because wages are sticky.
Mark, Thanks for that info.
Dan, Markets immediately incorporate new information, or at least do so fairly quickly. So the effect should have shown up quickly, and it’s hard to believe the entire 160% gains was caused by Bernanke’s decision.
123, My hunch is that markets would respond positively to a cut in IOR, which increases the demand for base money.
Mark, Good point on the 1930s, I’ll try to do a post at some point.
Dan, You said;
“The problem in October 2008 was that the market believed the true value of bank assets was much lower than claimed. And for good reason! The market value of bank assets in 2008 was much lower than claimed. The problem was not IOR. It was not NGDP. It had nothing to do with monetary policy. The problem was the insolvency of the banking system as a result of fraudulent loans and investments and basic bad lending practices.
How did that happen? Trust me, it was not because of bad monetary policy”
Trust you? When you have not addressed any of the claims in my post at all? When you have not even shown any evidence that you understand the claims I’m making? When you simply charge ahead with the same old myths and fallacies with no supporting evidence?
No, I think I choose not to “trust you.”
I trust economic analysis of data based on well established theory. That’s what I do here.
16. December 2013 at 08:40
Dan, See my new post for a more thorough rebuttal of your views.
16. December 2013 at 08:42
Scott,
When you say, “The mechanism is that when the economy is depressed rising NGDP tends to raise corporate profits, because wages are sticky.”
….are you saying that Lehman’s and the financial sector crashed because NGDP lowered their profits. What about Bank Runs, how would falling NGDP cause bank runs?
16. December 2013 at 08:44
James Hamilton has a new post that shows the problems that have arisen for the Fed trying to control short term interest rates;
http://www.econbrowser.com/archives/2013/12/federal_reserve_3.html
Though he stops short of accepting the inevitable; that the Fed should give up trying.
16. December 2013 at 08:55
Also,
I once asked Nick Rowe on his blog the following question… “Let’s say we KNOW that had the Fed not allowed NGDP to fall in the summer and fall of 2008, then there would have been no Lehman Brothers crash, financial crisis, and Great Recession. That would logically mean that the NGDP crash caused Lehman Brothers crash and the ensuing financial crisis. This argument is the practical implication of Sumner’s view, that the fall in NGDP caused the Great Recession. But I don’t understand the following…. according what mechanism could that happen? According to what mechanism would the fall in GDP have caused Lehman’s and the financial system to crash? What’s the model?”
He answered with… “For example: if a fall in NGDP causes a recession, that causes unemployment, and unemployed people are more likely to default on their mortgages.”
Do you agree with him? This is different from your answer, I believe.
16. December 2013 at 09:52
Scott,
A year and a half ago, the ECB has cut the IOR to zero. The Fed should do the same too.
16. December 2013 at 10:46
JoeMac, Bank runs are an even better example. The biggest fall in NGDP was 1929-33, when it fell nearly in half. Bank runs increased sharply because with less nominal income it was harder to repay nominal debts. When there is a wave of debt defaults, it tends to cause bank failures. A fear of bank failures causes bank runs.
Yes, unemployment is part of the problem but I’d focus on less nominal income. There were two problems in 2009, falling RGDP (unemployment) and falling prices (deflation.) Both boost debt defaults, which is why NGDP is the best variable to look at. Of course falling NGDP also reduces property prices, which increases debt defaults.
Thanks Patrick.
123, I agree, but the ECB still needs to do much more. They need either a level target of prices or NGDPLT.
16. December 2013 at 11:09
Yes, cutting IOR to zero was only a minor step. They are close to the CPI PLT. They should switch to NGDPLT with slight adjustments so over the long run price stability is preserved.
My guess is that if you asked Bernanke, he would say the switch to IOR was a right decision, but they should have reduced faster and instituted even stronger,liquidity programs sooner.
16. December 2013 at 11:48
Maybe because I was close to it, I saw the crash of 2008 as a slow moving train.
I remember forecasts of an impending housing bust in 2005.
Liar loans were a headline grabbing event in 2006.
Novastar and New Century (sub-prime lenders) went belly up in March 2007.
Merrill Lost $8 billion in August of 2007.
Bloomberg Magazine put “Toxic Waste” as a cover issue in 2007.
Northern Rock collapsed in 2007.
When did the Auction Rate preferred market colapse? Feb 08, as I remember — before Bear.
Bear Collapsed in March of 2008.
Then the ABCP (asset-baked commercial paper) market fell.
and the rest of the shadow bank SIV soon followed.
FNMA was the shocking event in my mind (august)… a much bigger deal than Lehman.
When it comes down to it, sub-prime lending was a blip. The securities tied to sub-prime loans were an order of magnitude larger.
But what really happened was there was a insufficient money available to fund the short term loans of a multitude of institutions. A massive amount money from the shadow banking system dried up, and central bank created money appeared far to late to take its place.
16. December 2013 at 17:31
IOER has been so successful at draining reserves that the Fed’s trading desk is testing an equivalent program for non-bank financial companies. Simon Potter, executive vice president of the New York Fed, discusses it here.
http://www.newyorkfed.org/newsevents/speeches/2013/pot131202.html
“Similar to IOER for banks, an overnight, fixed-rate, full-allotment reverse repo facility with same-day settlement would provide an essentially risk-free investment directly at a fixed rate, in this case to a broad range of non-bank counterparties. By reaching financial institutions that are ineligible to earn IOER, including, for example, money market funds, the facility widens the universe of counterparties that should generally be unwilling to lend at rates below those rates available through the central bank. The facility should also enhance competition in the markets by strengthening the bargaining position of non-bank lenders, which would now be able to place essentially unlimited funds overnight in the Fed’s facility. In this way, the facility is expected to complement IOER, strengthening the floor on the level of overnight rates and tightening the relationship among various money market rates.”
16. December 2013 at 18:53
I say let’s cut IOER by one basis point a month and see what happens….
17. December 2013 at 06:12
123, The ECB does not have a price level target, hence they cannot be close to the PLT. You can invent a PLT, but since neither the ECB nor the markets believes that’s their target, it’s not meaningful. Even worse, it should not be the target, as you acknowledge. So it does no good to describe policy in relation to a target that does not and should not exist.
Doug, In your last sentence you correctly describe the “tidal forces.” All the stuff before is foam on the top of the waves. It’s the tidal forces that matter.
Ron, Thanks for the info.
Ben, I agree, but make it 2 basis points/month.
17. December 2013 at 06:40
Scott, PLT exists to some degree at the ECB. Here is an excerpt from Die Zeit interview with Benoît Cœuré, Member of the Executive Board of the ECB one week ago:
“The ECB is committed to maintaining price stability, by which we mean keeping inflation below, but close to, 2% – and we have achieved this objective. Average inflation since the introduction of the euro is 2.03%. This has contributed to economic stability in Europe. If we were to miss our objective, people would start to question our commitment.”
18. December 2013 at 07:28
123, Sorry, but that quotation clearly says they engage in inflation targeting, not price level targeting.
18. December 2013 at 08:22
Scott, no other major central bank is regularly talking about the average of historical inflation.
18. December 2013 at 16:23
Japanese-style liquidity trap? There I was thinking, liquidity traps don’t exist….(except in the minds of some people and their models).