A few weeks back I linked to a Louis Woodhill article that showed stocks fell very sharply after the Fed announced interest on reserves, and also fell very sharply the two times the Fed raised the IOR in late October and early November. Given the unusual size of the stock market declines, it is extremely unlikely these changes could all have been due to chance. Now Cameron Blank has graciously allowed me to reprint an entire blog post where he fully documents just how powerfully monetary policy news seemed to impact US stock prices during late 2008.
Many of my critics, even some sympathetic to my overall argument, have found it hard to believe that monetary policy was the problem in late 2008. It seems (so they argue) that the financial crisis was the focus of everyone’s attention. I agree that economists were not paying attention to monetary policy, but there’s no doubt in my mind that markets were.
BTW, the TIPS spreads for December 1st that he reports are distorted by a change in the type of TIPS being used, so he is right to be skeptical of that observation.
Everything that follows this sentence is from Cameron’s blog; for ease of presentation I do not indent as is usually done with quotations:
(A more concise version of this post with fewer examples is available here)
Here I present evidence that markets responded sharply to monetary policy during the late 2008 financial crisis.
Perhaps most notably, I find that markets tended to cheer traditional monetary policy (policy aimed at increasing aggregate demand) and snub fed policies aimed at fixing credit markets. This certainly flies in the face of the mainstream explanation for the recession.
The Data
As a general rule, during late 2008, increases in stock prices were correlated with increases in nominal yields, inflation expectations, and expected future Fed funds rates. This isn’t surprising. If the correlation between expected future fed funds rates and other asset prices breaks down however, it should give some kind of an idea of impact of monetary policy.
Note: This means that it is possible that monetary policy was still the cause of certain declines in equity prices and that although fed funds futures fell they didn’t fall enough. The data I use only makes use of days in which the markets moved in “opposite” directions and so my case may actually be stronger than it looks.
October 6, 2008
The S&P, nominal rates, and inflation expectations all tumbled while the dollar rose 1%. Basically all of the 4% decline for the day took place immediately after the market opened. The likely cause?
Scott Sumner has long argued this policy greatly sharply tightened monetary policy during late 2008. So even though the expected FF rate fell (although note that 3 month treasury yields rose slightly), perceived policy was much tighter and markets responded.
It seems likely these events were what they were responding to. Stocks fell dramatically, nominal rates actually rose (but especially at the short end of the curve) along with expected fed funds rates, while inflation expectations plummeted.
The selloff seems like exactly the type of reaction one might expect if markets thought the Fed would do what was necessary to prevent NGDP from falling the 13th, only to realize it was hopelessly focused on the financial crisis (and not aggregate demand) two days later. And why would a Fed Chairman ever say continued weakness was certain? If that was the case shouldn’t they ease policy?
October 20, 2008
The 4.7% rise in the S&P during October 20 serves as more proof that the U.S. faced an aggregate demand problem more than a financial one. Stocks rallied as Bernanke endorsed a second fiscal stimulus plan. Now admittedly, markets may have been reacting simply to a bigger fiscal stimulus with a better chance of passing, but it could also be argued that Bernanke’s support reassured markets the Fed wouldn’t offset the fiscal stimulus.
My guess is it was a mix of both, but still the rally that day certainly serves as more evidence that the primary issue in late 2008 was low AD.
Again, even though 3 month treasuries and fed fund futures signaled slightly lower future fed funds rates, overall policy was likely seen as tighter. Even though this policy helped the Fed increase its balance sheet they were essentially trading AD enhancing policies for ones which helped financial markets. Equities, nominal rates and inflation expectations all fell sharply while the dollar rose sharply.
October 28, 2008
One of the strongest pieces of evidence that monetary policy mattered a great deal during late 2008.
Equities, nominal yields and inflation expectations all rallied sharply while 3 month treasury yields and FF futures fell. Had this been some non-monetary shock, FF futures and 3 month treasury yields should have risen.
I expected that there may have been some bailout or financial news that pushed up markets, but when I looked there was nothing of the sort. News that day noted that…“Stock analysts struggled to make sense of the gains” but also acknowledged that the Fed may be partly behind the rally.
October 30, 2008
A smaller, but similar market reaction as October 28th.
Again all the signs of easier expected monetary policy boosting AD expectations show up once again. Equities, oil and inflation expectations rose while fed fund futures, 3 month treasury yields and the dollar fell. CNN attributed the rally to the election, but this was the day of the election not the day after (no results were in yet) so that explanation makes no sense to me.
November 5, 2008
Again a day when the Fed raised the interest paid on reserves. A bad ISM services report also came out that day, but markets didn’t seem to clearly be reacting to that. Note also that fed funds futures actually increased slightly despite a sharp fall in stocks, nominal yields and oil prices (down 8% according to the NYT article).
November 7, 2008
Another rally with seemingly no explanation, except for easier money.
December 1, 2008
Another really notable day when stocks, nominal yields and oil fell dramatically while 3 month treasury yields, fed fund futures and the dollar rose.
(Something bizarre seems to have been going on with TIPS yields data that day but 5 year inflation swaps show a more plausible decline in 5 year inflation expectations)
The Fed surprised markets by cutting more than expected (to 0-0.25%), downplaying inflation risks, and hinting at possible quantitative easing. Stocks and inflation expectations rose significantly and the dollar fell. Nominal rates fell as well (probably in response to potential QE purchases) but overall the response was obviously very positive.
Conclusion :
Obviously if you are only looking at days in which monetary policy and markets moved it opposite directions it will, by definition, look like monetary policy matters a lot… but the magnitudes are what are impressive. The two biggest increases, as well as the two biggest declines in equity prices during the period seem to be strongly correlated with changes in expected monetary policy.
I think this data strongly supports the belief that monetary policy mattered a great deal in late 2008. It also supports the idea that insufficient AD was a bigger problem than the financial crisis (which were indeed being fed by AD problems). Markets could of course always be wrong, but from today’s standpoint (deflation and high unemployment in despite a relatively normal financial sector) they seem to have been quite correct.
(Data for the S&P was acquired form Google Finance, Fed Fund Futures data was acquired from the Cleveland Fed, Oil data was recovered from daily CNN Market report articles, and all other data was downloaded from FRED)”
I often wonder why, if changes in expected future monetary policy explain much (if not all) of the big swings in asset prices during late 2008/early 2009, why none of the traders whose trades were motivated by these changes seem to ever show up on CNBC/Bloomberg/etc. Granted, lots of these trades were probably high-frequency automated trades, but presumably many investors (individual or institutional) were conscious of the reasons why these automated trades on such a large scale were occurring. If so, wouldn’t you think that a few of these people or representatives of these institutions would chime in every so often, saying what to them is obvious, that monetary policy is and has been way too tight?
Maybe they don’t want to let the lay investor in on their highly valuable information, but if that’s the reason, then I would have to believe that nearly every commentator who ever appears on these channels or writes for major publications is a deliberate liar. It’s not hard for me to believe that most of them are, but nearly all of them? I want to know why it is that it took so long for anyone to lay out the case for more monetary stimulus in one of these forums (Don Luskin being a prominent recent example), if so many traders were in fact responding so dramatically to policy that could be corrected so easily.
TED spread, overnight LIBOR and 3m LIBOR is the data that I miss the most in this analysis.
It is a mistake to say that 3m T-bill rates reflected fed funds rate expectations during the period. They were not. Fed funds rate is risky, T-bills are safe.
Expected FF rates should also be interpreted with caution, as they suffer from the survivorship bias – expected FF rates do not reflect cost of fed funds for those institutions that will default, as the probability of default was high during the period, this is a very significant issue.
October 6 was no different from other days in the early October stock market crash, so I am not persuaded that IOR announcement mattered. In the October 6 announcement, there were two pieces of news – IOR and expansion of monetary base. Bernanke hoped that on the net basis the announcement will increase AD, I think that Bernanke miscalculated and October 6 announcement had zero effect. The stock market crash in early October was caused by the breakdown of the fed funds rate peg, and October 6 was not different from other days in early October. I have written about this previously here: http://themoneydemand.blogspot.com/2010/11/collapse-of-fed-funds-rate-peg-and.html
October 22 and November 5 IOR announcements reduced AD. At the time the Fed funds rate peg was mostly restored – overnight Libor was below the policy rate, but 3m LIBOR was showing that some risk of renewed collapse of the peg remained. By increasing IOR on these days, the Fed indicated that conditions in the Fed funds market were OK, and no further efforts to decrease risk of a fed funds market breakdown were needed, markets were disappointed by this.
I’m really just stealing your NGDP expectations theory and using the Dow Jones as a proxy for NGDP expectations. It’s a simple chart and the relationship looks rather robust.
Ram, I hardly ever watch those shows, but on a rare time I tuned in (in late 2008) I heard Jim Cramer complain that tight money was sharply driving down stock prices (or more precisely a smaller than expected rate cut by the ECB.)
I’m sympathetic to your argument, but the stock movements on the 3 days the IOR was raised must be significant at the 99.9999% level (in combination, not individually.) The declines are 3.85%, 6.1% and 5.27%. Those are huge declines, which occur rarely (especially the latter two.) Perhaps a reader that knows finance could calculate the share of days that declines that big occur, and multiply the three numbers together. My guess; one in a million.
Let’s also remember that these shows cater to viewers, and viewers may have little understanding of monetary policy.
In addition, suppose that bond markets follow monetary policy closely, and the stock market takes its lead from the bond market.
In addition, a relatively small number of traders can move the market if they are sufficiently well-funded.
I do see your point, and have wondered the same thing myself. Indeed that’s one reason why early on I was dismissive of the idea that IOR had much effect–it didn’t seem to be a focus of the markets. But that’s before I saw the data.
123, At least we agree on the big picture, tight money was depressing stocks in late 2008. I’m not sure why you compare October 6th to other days in early October, stock movements are serially uncorrelated. Just because they fall one day, doesn’t make them like to fall the next.
He mentioned October 6 and 7, whereas I think it likely that Fed policy also depressed stocks on October 3rd (when IOR was decided but not publicly announced) and October 8th (when the Fed cut rates a measly 50 basis points.) So tight money can explain a good bit of the early October crash.
I just finished reading Lord’s of Finance and I tried reading it with a Sumnerian perspective. It’s a great book by a great author, but even Liaquat Ahamed ultimately blames the Depression on this sequence of events.
Fed cutting interest rates under pressure from Bank of England in order to “recycle” gold back to England.
This cut causes a huge run up in US stock prices. Ultimately, the bubble bursts for no apparent reason whatsoever and a Depression ensues.
This seems contradictory to everything you have written.
I know you have written that you have done a lot of research on this time period. Is there a quick answer to why this explanation is wrong?
Scott, I think that whole early October crash was caused by the tight monetary policy that was getting tighter every day. Bernanke managed the feat of surprising markets with tight monetary policy every day. You can look at it this way. On every day in the early October crash, markets were expected to close at a gain of daily cost of equity. This tiny expected daily gain can be decomposed to two scenarios – most likely Bernanke does nothing dramatic, and stocks drop by couple of percent as AD gets lower, and there is a slight chance of Bernanke stopping the crisis by doing something really big, in this case markets are expected to rally 10%. You get something that looks like a serial correlation if the expected daily returns are asymmetric.
During the crash monetary policy was getting tighter every day. You can see it in the chart of 3m LIBOR – it measures expected opportunity cost of reserves during the next three months. You can see it in the chart of TED spread, unless offset by additional monetary stimulus or other unrelated positive news, higher TED spread means AD is lower. One day it changed – in October 13, 2008, Bernanke did something really significant.
If Jim Cramer really thought at the time that monetary policy was too tight, what do you think changed his view since then? I don’t watch CNBC much at all, either, but the few times I have, I’ve noticed their leading commentators ridiculing the idea that monetary policy has anything to do with recent improvements in economic conditions and forecasts. I’d be surprised if Cramer was pushing hard for that view and yet his colleagues were continuously ridiculing it. Half the time what I hear on these programs, from hosts as well as guests, is that the Fed is behind the curve, continuing QE2 “even though” the economy is recovering, as if QE2 is just the execution of a prior promise by the Fed having nothing to do with recent improvements.
Scott, usually monetary policy shocks are well defined discrete events, such as FOMC decisions, important speeches, releases of economic data etc. This was not the case in September-October 2008, when the out-of-control fed funds rate market was generating new shocks on a daily basis.
I think you would paint a more complete picture of market reactions if you included other events. As we know now, Citi was close to closing its doors in November, 2008. This was a result of margin calls (haircuts to collateral) that all TBTF firms were receiving at the time. Imagine the impact that a Citi or Morgan Stanley failure would have had on the financial system, and you will know what traders feared when they sold down stocks in big increments.
BTW, the daily volatility at the time was exacerbated by the interplay between the CDS and equities markets. Those who sold protection on TBTF institutions were not able to buy the CDS back or short the bonds — the markets were frozen. To hedge, they went into the equities markets and indiscriminately sold large baskets of stocks, pummeling the broad indices. In short, there was a lot going on, and the interplay between financial system failure and bail out was at the forefront of trader’s minds. So yes, monetary policy played an important role, but it was at best half of the equation.
BTW, there are a number of “Financial Crisis Timelines” on the internet that provide some context. Here’s what I was able to glean from them:
In the last week of September, the crisis spread to Europe, which saw five major institutions either seized or bailed out. Prominent amongst these was Fortis, which threatened the whole European system.
On October 3, TARP legislation passed, but the initial mechanism for buying “toxic” bank assets was both unclear and thought to be ineffective and/or insufficient.
On October 6 and 7, the Fed and the FDIC announced moves to shore up the Commercial Paper market and increase the ceiling on deposit insurance. This was in response to the CP market freezing up in the days prior. It was thought at the time that a non-functioning CP market would endanger industrial credit and therefore impact the real economy. The Fed also announced, in conjunction with global central banks, an expansion in the TAF lending facility for banks — this, in response to widespread fears about bank collateral calls. In Europe, the German government was forced to bail out Hypo Bank again, incredibly, just days after the first bail out proved insufficient. Further, there was talk of bank runs in Ireland and Hong Kong, as well as in the U.S., which is why the FDIC moved on that day. On October 8, the NY Fed announced its first, $37b bail-out of AIG (worries over AIG had roiled the CDS market as I commented above). The UK also announced a $500b bank rescue package. On October 9, the IMF announced a series of rescue steps. On October 10, the Asian markets opened down 10% on fears of global banking contagion. The weekend of October 11, the G7 met and announced a series of coordinated rescues.
On October 11, IMF-Head Dominique Stauss-Kahn said, “Intensifying solvency concerns about a number of the largest U.S.-based and European financial institutions have pushed the global financial system to the brink of systematic meltdown.”
On Oct. 13, the Treasury announces that they will use TARP funds to recapitalize the banking system rather than buy mortgages. The UK also announces a recapitalization program. The Dow rises 900 points.
On October 14, the Fed announces swap lines with foreign central banks in an attempt to stem the dollar liquidity shortages at European banks. (It was thought that a shortage of dollar liquidity was causing European institutions to dump U.S. assets, including stocks).
I could go on for the rest of the month, but this timeline is clear: the world faced a generalized run on the liabilities of the banking system, and governments were in a desperate race to get ahead of market expectations on bail outs. Of course, monetary easing was also important, but it is clear from the timeline above that the financial crisis was very much at the top of policymakers’ (and the markets’) minds.
Liberal Roman, That wasn’t my take on his argument. It’s been a while, but I thought he blamed the central banks for being too tight after 1929.
He quotes Gov. Strong on not spanking all one’s children just because one has misbehaved. That’s a critique of the Fed’s tight money policy of 1929-aimed at stopping the stock market boom.
In the Spring I did a lot of very long posts on my view of the Great Depression.
123; You said;
“Scott, I think that whole early October crash was caused by the tight monetary policy that was getting tighter every day. Bernanke managed the feat of surprising markets with tight monetary policy every day. You can look at it this way. On every day in the early October crash, markets were expected to close at a gain of daily cost of equity. This tiny expected daily gain can be decomposed to two scenarios – most likely Bernanke does nothing dramatic, and stocks drop by couple of percent as AD gets lower, and there is a slight chance of Bernanke stopping the crisis by doing something really big, in this case markets are expected to rally 10%. You get something that looks like a serial correlation if the expected daily returns are asymmetric.”
Can’t really disagree, because I made the same argument last year.
Ram, First, I’m not sure Cramer opposes QE2. Does anyone know? And if he does, I don’t know why.
123, I think there were some discrete events, which have been documented by Cameron. Even if you don’t agree about October 6th, you do seem to agree about the two later increases, and the December 75 basis point cut was a discrete event, as was the October 13 event you highlight.
But in general, I agree with your errors of omission view, especially for September/October.
David, I view the events you mention as effects of tight money. But yes, they were also important.
“Of course, monetary easing was also important, but it is clear from the timeline above that the financial crisis was very much at the top of policymakers’ (and the markets’) minds.”
This is precisely the problem, monetary authorities were tightening policy, not easing. Stocks fell because the Fed’s actions taken did not prevent policy from tightening. NGDP expectations probably fell almost every day in early October, and that drove asset prices sharply lower, which hurt bank balance sheets. They were desperately trying to bail water out of the boat, without first plugging the leak.
I would give Cramer a lot more credit than he is given. And its kinda sad the has become a target for the anti-Wall Street crowd. As much as I like John Stewart, he could not have picked a worse scapegoat.
99% of the punditry today falls into two camps.
Optimistic with a general feeling of “we shouldn’t worry about every single pitfall the economy could have”. But these people are still completely unaware of the power of monetary policy.
or
Perma-bears. Who think the Fed is anywhere from evil to impotent.
Jim Cramer is one of the few (the only?) main stream pundit who criticized Bernanke as early as 2007 in his famous tirade and then on and off throughout 2008.
However, his big mistake was thinking that Bernanke’s interest rate cuts in early 2008 would pretty much contain the problem. There he was wrong.
Still, even acknowledging that the Fed could do something already puts him way ahead of everyone else in my book.
“Stocks fell because the Fed’s actions did not prevent policy from tightening.”
The statement is true, but overlooks the most direct link, the proximate cause. Stocks fell because, as Strauss Kahn said, the world banking system was pushed to “the brink of a systematic meltdown.”
In other words, tight money may have caused the financial crisis of October, 2008; but clearly, it was the financial crisis that caused stocks to fall.
“TED spread, overnight LIBOR and 3m LIBOR is the data that I miss the most in this analysis.”
Do you know a place where I can find daily data of these things? I will add them if you do.
“It is a mistake to say that 3m T-bill rates reflected fed funds rate expectations during the period. They were not. Fed funds rate is risky, T-bills are safe.”
Doesn’t that make movements in 3m t-bills more puzzling rather than less? During October 28, for example, equities rose ~11% while 3 month t-bill yields actually fell. If it was predominately a flight to safety issue that still seems bizarre. I admit FF futures are a superior measure though.
“Expected FF rates should also be interpreted with caution, as they suffer from the survivorship bias – expected FF rates do not reflect cost of fed funds for those institutions that will default, as the probability of default was high during the period, this is a very significant issue.”
Is there a reason to believe this would bias the data only on these particular days though? FF futures seemed to otherwise react quite predictably during the period of Mid-September to Mid-December, steadily falling from 2% to 0.32%(these are all expected FF rates for the very next meeting BTW).
Ram,
The media is generally pretty bad at determining the cause of market changes. Take the election day rally for example. CNN attributed the movement to the election… but that doesn’t make any sense. Markets already knew the election was going to be held on that day and nothing particularly surprising happened on that day with regards to the odds(and remember, this was pre-results). During this period the media seemed to be frequently caught off guard unless changes were directly related to financial issues.
The real question is whether markets responded to changes in the perceived stance of monetary policy. The answer seems to be yes.
You’re right. Financial issues were certainly responsible for a lot of the big swings in stock prices(ie Sept 29 when the bailout failed to pass stocks fell almost 9%), but others seem to have little relation to any notable financial events.
Can you find a financial explanation for the December 1 crash? And some of the rallies I mentioned seem to have no financial news of note that day. (October 28 and November 4)
I couldn’t find specific news for those dates. Here is what I remember (I was quite short financial shares): The markets were trying to make up their mind as to whether the various financial rescue plans would actually work. One of the defining features of the sep-dec, 2008 period was that bail out announcements were generally greeted with selling. The consensus thought was that policy makers were simply behind the curve, and that their plans would not work. Nevertheless, on some days the shorts were squeezed as optimism over bail outs took hold.
As someone that was short, its hard to explain how sudden and viscous a short covering rally can be. You see the market drop and drop, supposedly on “natural” selling. Then you realize much of the selling was from shorts piling on with abandon. This “dumb” short money is like gasoline for rallies: strike a match of optimism and their positions go up in flames. They scramble to cover at any price as a result.
So the best explanation I can give you for those rallies, having survived them with my positions intact, is that there were many then that thought shorting shares was easier than shooting fish in a barrel, and realized otherwise.
“However, his big mistake was thinking that Bernanke’s interest rate cuts in early 2008 would pretty much contain the problem. There he was wrong.”
I made the same mistake.
David, I think falling NGDP expectations had a huge impact on the stock market. The biggest stock price falls were not associated with the financial crisis in mid-September.
I think you are right about the Fed being behind the curve. The Fed would plan for a problem of X magnitude, and then falling NGDP expectations would increase the problem to 2X magnitude.
“I think there were some discrete events, which have been documented by Cameron. Even if you don’t agree about October 6th, you do seem to agree about the two later increases, and the December 75 basis point cut was a discrete event, as was the October 13 event you highlight.
But in general, I agree with your errors of omission view, especially for September/October.”
Yes, there were discrete events, but the important ones were visible in the LIBOR and Ted spread data.
I think there is no free public source for LIBOR and Ted spread historical data, although there are charts available from Bloomberg (tickers US0003M:IND and .TEDSP:IND).
“Doesn’t that make movements in 3m t-bills more puzzling rather than less? During October 28, for example, equities rose ~11% while 3 month t-bill yields actually fell. If it was predominately a flight to safety issue that still seems bizarre. I admit FF futures are a superior measure though.”
“Is there a reason to believe this would bias the data only on these particular days though? FF futures seemed to otherwise react quite predictably during the period of Mid-September to Mid-December, steadily falling from 2% to 0.32%(these are all expected FF rates for the very next meeting BTW).”
The bias works in two different directions. After Lehman, the default risk steadily increased, and the actual expected cost of FF was much higher than indicated by FF futures. After Ocbtober 13 everything changed. As the Fed started restoring the order in the FF market, default risk started decreasing. Decreases in expected FF rates taken from FF futures market underestimated actual reductions in expected cost of FF.
Welcome to a new blog on the endlessly perplexing problem of monetary policy. You’ll quickly notice that I am not a natural blogger, yet I feel compelled by recent events to give it a shot. Read more...
My name is Scott Sumner and I have taught economics at Bentley University for the past 27 years. I earned a BA in economics at Wisconsin and a PhD at Chicago. My research has been in the field of monetary economics, particularly the role of the gold standard in the Great Depression. I had just begun research on the relationship between cultural values and neoliberal reforms, when I got pulled back into monetary economics by the current crisis.
"Couldn't find the substack and if it's moderated just as well... Voted for Trump and he won. Sorry liberal Scott Sumner who believes in money non-neutrality, which is akin to..."
"Scott, Quick note of thanks. I've hugely enjoyed your blog and the intellectual stimulation I gotten from it. Also it was pleasure getting to know you and your wife. Hope..."
14. January 2011 at 07:10
I often wonder why, if changes in expected future monetary policy explain much (if not all) of the big swings in asset prices during late 2008/early 2009, why none of the traders whose trades were motivated by these changes seem to ever show up on CNBC/Bloomberg/etc. Granted, lots of these trades were probably high-frequency automated trades, but presumably many investors (individual or institutional) were conscious of the reasons why these automated trades on such a large scale were occurring. If so, wouldn’t you think that a few of these people or representatives of these institutions would chime in every so often, saying what to them is obvious, that monetary policy is and has been way too tight?
Maybe they don’t want to let the lay investor in on their highly valuable information, but if that’s the reason, then I would have to believe that nearly every commentator who ever appears on these channels or writes for major publications is a deliberate liar. It’s not hard for me to believe that most of them are, but nearly all of them? I want to know why it is that it took so long for anyone to lay out the case for more monetary stimulus in one of these forums (Don Luskin being a prominent recent example), if so many traders were in fact responding so dramatically to policy that could be corrected so easily.
14. January 2011 at 08:08
TED spread, overnight LIBOR and 3m LIBOR is the data that I miss the most in this analysis.
It is a mistake to say that 3m T-bill rates reflected fed funds rate expectations during the period. They were not. Fed funds rate is risky, T-bills are safe.
Expected FF rates should also be interpreted with caution, as they suffer from the survivorship bias – expected FF rates do not reflect cost of fed funds for those institutions that will default, as the probability of default was high during the period, this is a very significant issue.
October 6 was no different from other days in the early October stock market crash, so I am not persuaded that IOR announcement mattered. In the October 6 announcement, there were two pieces of news – IOR and expansion of monetary base. Bernanke hoped that on the net basis the announcement will increase AD, I think that Bernanke miscalculated and October 6 announcement had zero effect. The stock market crash in early October was caused by the breakdown of the fed funds rate peg, and October 6 was not different from other days in early October. I have written about this previously here:
http://themoneydemand.blogspot.com/2010/11/collapse-of-fed-funds-rate-peg-and.html
October 22 and November 5 IOR announcements reduced AD. At the time the Fed funds rate peg was mostly restored – overnight Libor was below the policy rate, but 3m LIBOR was showing that some risk of renewed collapse of the peg remained. By increasing IOR on these days, the Fed indicated that conditions in the Fed funds market were OK, and no further efforts to decrease risk of a fed funds market breakdown were needed, markets were disappointed by this.
14. January 2011 at 08:35
FWIW, I have made my own rebuttal to Cowen’s Zero MP theory by showing the simple relationship of the stock market trend to the unemployment rate:
http://dorsetnaga.wordpress.com/2011/01/14/zero-mp-workers-no-its-not-different-this-time/
I’m really just stealing your NGDP expectations theory and using the Dow Jones as a proxy for NGDP expectations. It’s a simple chart and the relationship looks rather robust.
(I’m a trader, not an economist.)
14. January 2011 at 08:39
Ram, I hardly ever watch those shows, but on a rare time I tuned in (in late 2008) I heard Jim Cramer complain that tight money was sharply driving down stock prices (or more precisely a smaller than expected rate cut by the ECB.)
I’m sympathetic to your argument, but the stock movements on the 3 days the IOR was raised must be significant at the 99.9999% level (in combination, not individually.) The declines are 3.85%, 6.1% and 5.27%. Those are huge declines, which occur rarely (especially the latter two.) Perhaps a reader that knows finance could calculate the share of days that declines that big occur, and multiply the three numbers together. My guess; one in a million.
Let’s also remember that these shows cater to viewers, and viewers may have little understanding of monetary policy.
In addition, suppose that bond markets follow monetary policy closely, and the stock market takes its lead from the bond market.
In addition, a relatively small number of traders can move the market if they are sufficiently well-funded.
I do see your point, and have wondered the same thing myself. Indeed that’s one reason why early on I was dismissive of the idea that IOR had much effect–it didn’t seem to be a focus of the markets. But that’s before I saw the data.
123, At least we agree on the big picture, tight money was depressing stocks in late 2008. I’m not sure why you compare October 6th to other days in early October, stock movements are serially uncorrelated. Just because they fall one day, doesn’t make them like to fall the next.
He mentioned October 6 and 7, whereas I think it likely that Fed policy also depressed stocks on October 3rd (when IOR was decided but not publicly announced) and October 8th (when the Fed cut rates a measly 50 basis points.) So tight money can explain a good bit of the early October crash.
14. January 2011 at 08:45
Dirk, Thanks, I’ll try to do a post.
14. January 2011 at 09:12
scott,
I just finished reading Lord’s of Finance and I tried reading it with a Sumnerian perspective. It’s a great book by a great author, but even Liaquat Ahamed ultimately blames the Depression on this sequence of events.
Fed cutting interest rates under pressure from Bank of England in order to “recycle” gold back to England.
This cut causes a huge run up in US stock prices. Ultimately, the bubble bursts for no apparent reason whatsoever and a Depression ensues.
This seems contradictory to everything you have written.
I know you have written that you have done a lot of research on this time period. Is there a quick answer to why this explanation is wrong?
14. January 2011 at 09:30
Scott, I think that whole early October crash was caused by the tight monetary policy that was getting tighter every day. Bernanke managed the feat of surprising markets with tight monetary policy every day. You can look at it this way. On every day in the early October crash, markets were expected to close at a gain of daily cost of equity. This tiny expected daily gain can be decomposed to two scenarios – most likely Bernanke does nothing dramatic, and stocks drop by couple of percent as AD gets lower, and there is a slight chance of Bernanke stopping the crisis by doing something really big, in this case markets are expected to rally 10%. You get something that looks like a serial correlation if the expected daily returns are asymmetric.
During the crash monetary policy was getting tighter every day. You can see it in the chart of 3m LIBOR – it measures expected opportunity cost of reserves during the next three months. You can see it in the chart of TED spread, unless offset by additional monetary stimulus or other unrelated positive news, higher TED spread means AD is lower. One day it changed – in October 13, 2008, Bernanke did something really significant.
14. January 2011 at 09:38
If Jim Cramer really thought at the time that monetary policy was too tight, what do you think changed his view since then? I don’t watch CNBC much at all, either, but the few times I have, I’ve noticed their leading commentators ridiculing the idea that monetary policy has anything to do with recent improvements in economic conditions and forecasts. I’d be surprised if Cramer was pushing hard for that view and yet his colleagues were continuously ridiculing it. Half the time what I hear on these programs, from hosts as well as guests, is that the Fed is behind the curve, continuing QE2 “even though” the economy is recovering, as if QE2 is just the execution of a prior promise by the Fed having nothing to do with recent improvements.
14. January 2011 at 14:26
Scott, usually monetary policy shocks are well defined discrete events, such as FOMC decisions, important speeches, releases of economic data etc. This was not the case in September-October 2008, when the out-of-control fed funds rate market was generating new shocks on a daily basis.
14. January 2011 at 14:51
Scott,
I think you would paint a more complete picture of market reactions if you included other events. As we know now, Citi was close to closing its doors in November, 2008. This was a result of margin calls (haircuts to collateral) that all TBTF firms were receiving at the time. Imagine the impact that a Citi or Morgan Stanley failure would have had on the financial system, and you will know what traders feared when they sold down stocks in big increments.
BTW, the daily volatility at the time was exacerbated by the interplay between the CDS and equities markets. Those who sold protection on TBTF institutions were not able to buy the CDS back or short the bonds — the markets were frozen. To hedge, they went into the equities markets and indiscriminately sold large baskets of stocks, pummeling the broad indices. In short, there was a lot going on, and the interplay between financial system failure and bail out was at the forefront of trader’s minds. So yes, monetary policy played an important role, but it was at best half of the equation.
14. January 2011 at 16:05
BTW, there are a number of “Financial Crisis Timelines” on the internet that provide some context. Here’s what I was able to glean from them:
In the last week of September, the crisis spread to Europe, which saw five major institutions either seized or bailed out. Prominent amongst these was Fortis, which threatened the whole European system.
On October 3, TARP legislation passed, but the initial mechanism for buying “toxic” bank assets was both unclear and thought to be ineffective and/or insufficient.
On October 6 and 7, the Fed and the FDIC announced moves to shore up the Commercial Paper market and increase the ceiling on deposit insurance. This was in response to the CP market freezing up in the days prior. It was thought at the time that a non-functioning CP market would endanger industrial credit and therefore impact the real economy. The Fed also announced, in conjunction with global central banks, an expansion in the TAF lending facility for banks — this, in response to widespread fears about bank collateral calls. In Europe, the German government was forced to bail out Hypo Bank again, incredibly, just days after the first bail out proved insufficient. Further, there was talk of bank runs in Ireland and Hong Kong, as well as in the U.S., which is why the FDIC moved on that day. On October 8, the NY Fed announced its first, $37b bail-out of AIG (worries over AIG had roiled the CDS market as I commented above). The UK also announced a $500b bank rescue package. On October 9, the IMF announced a series of rescue steps. On October 10, the Asian markets opened down 10% on fears of global banking contagion. The weekend of October 11, the G7 met and announced a series of coordinated rescues.
On October 11, IMF-Head Dominique Stauss-Kahn said, “Intensifying solvency concerns about a number of the largest U.S.-based and European financial institutions have pushed the global financial system to the brink of systematic meltdown.”
On Oct. 13, the Treasury announces that they will use TARP funds to recapitalize the banking system rather than buy mortgages. The UK also announces a recapitalization program. The Dow rises 900 points.
On October 14, the Fed announces swap lines with foreign central banks in an attempt to stem the dollar liquidity shortages at European banks. (It was thought that a shortage of dollar liquidity was causing European institutions to dump U.S. assets, including stocks).
I could go on for the rest of the month, but this timeline is clear: the world faced a generalized run on the liabilities of the banking system, and governments were in a desperate race to get ahead of market expectations on bail outs. Of course, monetary easing was also important, but it is clear from the timeline above that the financial crisis was very much at the top of policymakers’ (and the markets’) minds.
14. January 2011 at 16:10
Liberal Roman, That wasn’t my take on his argument. It’s been a while, but I thought he blamed the central banks for being too tight after 1929.
He quotes Gov. Strong on not spanking all one’s children just because one has misbehaved. That’s a critique of the Fed’s tight money policy of 1929-aimed at stopping the stock market boom.
In the Spring I did a lot of very long posts on my view of the Great Depression.
123; You said;
“Scott, I think that whole early October crash was caused by the tight monetary policy that was getting tighter every day. Bernanke managed the feat of surprising markets with tight monetary policy every day. You can look at it this way. On every day in the early October crash, markets were expected to close at a gain of daily cost of equity. This tiny expected daily gain can be decomposed to two scenarios – most likely Bernanke does nothing dramatic, and stocks drop by couple of percent as AD gets lower, and there is a slight chance of Bernanke stopping the crisis by doing something really big, in this case markets are expected to rally 10%. You get something that looks like a serial correlation if the expected daily returns are asymmetric.”
Can’t really disagree, because I made the same argument last year.
Ram, First, I’m not sure Cramer opposes QE2. Does anyone know? And if he does, I don’t know why.
123, I think there were some discrete events, which have been documented by Cameron. Even if you don’t agree about October 6th, you do seem to agree about the two later increases, and the December 75 basis point cut was a discrete event, as was the October 13 event you highlight.
But in general, I agree with your errors of omission view, especially for September/October.
David, I view the events you mention as effects of tight money. But yes, they were also important.
14. January 2011 at 16:15
David, You said;
“Of course, monetary easing was also important, but it is clear from the timeline above that the financial crisis was very much at the top of policymakers’ (and the markets’) minds.”
This is precisely the problem, monetary authorities were tightening policy, not easing. Stocks fell because the Fed’s actions taken did not prevent policy from tightening. NGDP expectations probably fell almost every day in early October, and that drove asset prices sharply lower, which hurt bank balance sheets. They were desperately trying to bail water out of the boat, without first plugging the leak.
14. January 2011 at 16:28
I would give Cramer a lot more credit than he is given. And its kinda sad the has become a target for the anti-Wall Street crowd. As much as I like John Stewart, he could not have picked a worse scapegoat.
99% of the punditry today falls into two camps.
Optimistic with a general feeling of “we shouldn’t worry about every single pitfall the economy could have”. But these people are still completely unaware of the power of monetary policy.
or
Perma-bears. Who think the Fed is anywhere from evil to impotent.
Jim Cramer is one of the few (the only?) main stream pundit who criticized Bernanke as early as 2007 in his famous tirade and then on and off throughout 2008.
However, his big mistake was thinking that Bernanke’s interest rate cuts in early 2008 would pretty much contain the problem. There he was wrong.
Still, even acknowledging that the Fed could do something already puts him way ahead of everyone else in my book.
14. January 2011 at 16:31
“Stocks fell because the Fed’s actions did not prevent policy from tightening.”
The statement is true, but overlooks the most direct link, the proximate cause. Stocks fell because, as Strauss Kahn said, the world banking system was pushed to “the brink of a systematic meltdown.”
In other words, tight money may have caused the financial crisis of October, 2008; but clearly, it was the financial crisis that caused stocks to fall.
14. January 2011 at 16:48
123,
“TED spread, overnight LIBOR and 3m LIBOR is the data that I miss the most in this analysis.”
Do you know a place where I can find daily data of these things? I will add them if you do.
“It is a mistake to say that 3m T-bill rates reflected fed funds rate expectations during the period. They were not. Fed funds rate is risky, T-bills are safe.”
Doesn’t that make movements in 3m t-bills more puzzling rather than less? During October 28, for example, equities rose ~11% while 3 month t-bill yields actually fell. If it was predominately a flight to safety issue that still seems bizarre. I admit FF futures are a superior measure though.
“Expected FF rates should also be interpreted with caution, as they suffer from the survivorship bias – expected FF rates do not reflect cost of fed funds for those institutions that will default, as the probability of default was high during the period, this is a very significant issue.”
Is there a reason to believe this would bias the data only on these particular days though? FF futures seemed to otherwise react quite predictably during the period of Mid-September to Mid-December, steadily falling from 2% to 0.32%(these are all expected FF rates for the very next meeting BTW).
14. January 2011 at 17:02
Ram,
The media is generally pretty bad at determining the cause of market changes. Take the election day rally for example. CNN attributed the movement to the election… but that doesn’t make any sense. Markets already knew the election was going to be held on that day and nothing particularly surprising happened on that day with regards to the odds(and remember, this was pre-results). During this period the media seemed to be frequently caught off guard unless changes were directly related to financial issues.
The real question is whether markets responded to changes in the perceived stance of monetary policy. The answer seems to be yes.
14. January 2011 at 17:18
David,
You’re right. Financial issues were certainly responsible for a lot of the big swings in stock prices(ie Sept 29 when the bailout failed to pass stocks fell almost 9%), but others seem to have little relation to any notable financial events.
Can you find a financial explanation for the December 1 crash? And some of the rallies I mentioned seem to have no financial news of note that day. (October 28 and November 4)
14. January 2011 at 18:34
Cameron,
I couldn’t find specific news for those dates. Here is what I remember (I was quite short financial shares): The markets were trying to make up their mind as to whether the various financial rescue plans would actually work. One of the defining features of the sep-dec, 2008 period was that bail out announcements were generally greeted with selling. The consensus thought was that policy makers were simply behind the curve, and that their plans would not work. Nevertheless, on some days the shorts were squeezed as optimism over bail outs took hold.
As someone that was short, its hard to explain how sudden and viscous a short covering rally can be. You see the market drop and drop, supposedly on “natural” selling. Then you realize much of the selling was from shorts piling on with abandon. This “dumb” short money is like gasoline for rallies: strike a match of optimism and their positions go up in flames. They scramble to cover at any price as a result.
So the best explanation I can give you for those rallies, having survived them with my positions intact, is that there were many then that thought shorting shares was easier than shooting fish in a barrel, and realized otherwise.
15. January 2011 at 14:44
Liberal Roman, You said;
“However, his big mistake was thinking that Bernanke’s interest rate cuts in early 2008 would pretty much contain the problem. There he was wrong.”
I made the same mistake.
David, I think falling NGDP expectations had a huge impact on the stock market. The biggest stock price falls were not associated with the financial crisis in mid-September.
I think you are right about the Fed being behind the curve. The Fed would plan for a problem of X magnitude, and then falling NGDP expectations would increase the problem to 2X magnitude.
17. January 2011 at 04:49
Scott, You said:
“I think there were some discrete events, which have been documented by Cameron. Even if you don’t agree about October 6th, you do seem to agree about the two later increases, and the December 75 basis point cut was a discrete event, as was the October 13 event you highlight.
But in general, I agree with your errors of omission view, especially for September/October.”
Yes, there were discrete events, but the important ones were visible in the LIBOR and Ted spread data.
17. January 2011 at 05:11
@Cameron
I think there is no free public source for LIBOR and Ted spread historical data, although there are charts available from Bloomberg (tickers US0003M:IND and .TEDSP:IND).
“Doesn’t that make movements in 3m t-bills more puzzling rather than less? During October 28, for example, equities rose ~11% while 3 month t-bill yields actually fell. If it was predominately a flight to safety issue that still seems bizarre. I admit FF futures are a superior measure though.”
T-bills react to FF expectations, as was the case in October 28. But they also rally if the credit risk of FF increases. I don’t agree with the flight-to-safety view, it was a flight to liquidity:
http://themoneydemand.blogspot.com/2010/10/brad-delong-and-flight-to-safety.html
“Is there a reason to believe this would bias the data only on these particular days though? FF futures seemed to otherwise react quite predictably during the period of Mid-September to Mid-December, steadily falling from 2% to 0.32%(these are all expected FF rates for the very next meeting BTW).”
The bias works in two different directions. After Lehman, the default risk steadily increased, and the actual expected cost of FF was much higher than indicated by FF futures. After Ocbtober 13 everything changed. As the Fed started restoring the order in the FF market, default risk started decreasing. Decreases in expected FF rates taken from FF futures market underestimated actual reductions in expected cost of FF.
17. January 2011 at 05:12
@Cameron
Charts:
http://noir.bloomberg.com/apps/cbuilder?ticker1=.TEDSP:IND
http://noir.bloomberg.com/apps/quote?ticker=US0003M:IND