From TheMoneyIllusion to conventional wisdom

When I began blogging in early 2009 I made a number of claims that seemed almost preposterous.  One of those was the claim that people had reversed causation; the housing crisis didn’t cause the recession, it was mostly a product of the recession.  This seemed crazy, as the housing crisis began well before the recession.

I distinguished between two phases of the housing crisis.  The first phase occurred during 2006-08, and was concentrated in the 4 subprime states.  It did lead to a modest slowdown in growth, and unemployment rose from 4.7% in January 2006 to 4.9% in April 2008.  But nothing severe.  The second phase of the housing collapse was much more severe.  When NGDP started falling rapidly due to tight money, housing prices fell all across America.  This triggered the severe banking crisis of late 2008, and was associated with a worsening of the recession—unemployment reached 10.1% by October 2009.

Day by day, month by month, opinion is imperceptably shifting in my direction.  Indeed so much so that a news article that basically proves my point seems to have elicited little surprise.  “Of course a severe recession would hurt the housing market.”

Let’s consider the original conventional wisdom, and then my heretical view of early 2009.

1.  The conventional wisdom was that we built way too many houses, and that the housing slump was a hangover from this effect.  If true, housing construction over the past 10 years ought to have been way above normal.  In fact, it has been way below normal.  Which leaves the puzzle of why we have so many empty houses.

2.  My view was that the severe recession greatly reduced the demand for housing.  Since very few Americans are homeless (in percentage terms), a reduction in demand for housing should imply an increase in average household size.  And that’s exactly what we’ve seen:

The number of people living under one roof is growing for the first time in more than a century, a fallout of the recession that could reduce demand for housing and slow the recovery.The Census Bureau had projected the average household size would continue to fall to 2.53 this year. Instead, the average is likely to hit 2.63, a small but significant increase because it is a turnabout.

“A funny thing happened on the way to the future” says Arthur C. Nelson, director of the Metropolitan Research Center at the University of Utah. “Household size increased.”

.  .  .

USA could end this decade with up to 4 million excess housing units because of the reversal in household size, he says.

A key factor: “The Great Recession has forced doubling up among both family and non-family members,” Nelson says.

Multi-generational households are on the rise: 49 million, or 16% of the population, live in a home that had at least two adult generations in 2008. In 1980, there were 28 million, or 12%.

According to a recent Pew Research Center report, the growth is due to demographics, cultural shifts and high unemployment.

“I think it’s the young adults,” says Dowell Myers, housing demographer at the University of Southern California. “Residential mobility has slowed down and when it slows down, they’re back in their parents’ houses or living with roommates.”

Household size began inching up in 2005, before the recession, a trend that might have been driven by the real estate boom that made housing unaffordable to many. Now, it’s more likely to be caused by the poor economy.

“There are a lot of trends going on,” Nelson says. Among them:

– Older Americans.They’re moving in with children and grandkids and vice versa. About 20% of people 65 and older live in multi-generational households

– High unemployment. It’s keeping young adults out of the job market and back home with their parents.

“Clearly, a lot of people are not forming households when they’re getting out of school,” says Karl Case, economics professor at Wellesley Collegewho helped create the Standard & Poor’s/Case-Shiller Home Price Index.

It’s not just that people are not buying homes. They’re not renting either, a sign that more people are squeezing into one unit.

“I can document this with my own students,” Case says. “Rental vacancy is the highest it’s ever been.”

– Immigrants. They have higher fertility rates and a cultural acceptance of extended families living together. Despite a decline in the influx of Hispanics since the economy soured, household size inched up.

“It’s going to have huge implications for the housing market,” Nelson says.

If it lasts. Many economists and demographers are convinced that as soon as the recession ends and jobs open up, Americans will return to their old ways. “I see it as temporary,” Myers says.

“The economy is the most important thing,” says Stephen Melman, director of economic services at the National Association of Home Builders. “Projecting lifestyles is a really tricky business.”

Hayek warned that if the Fed let NGDP fall you’d get a “secondary deflation.”  And that’s exactly what we got.   The first (small) part of the housing crash was a necessary adjustment.  The second much bigger part of the collapse was clearly the result of tight money reducing NGDP.  NGDP is the money people have to buy houses—it’s national income.  With less NGDP there will be less demand for housing.  You’ll have empty houses at the same time as people doubling up because they can’t afford houses.  Just as during the Great Depression you had farmers unable to sell their food, and hungry people who couldn’t afford to buy food.

We were a little poorer in 2008 than in 2006 because we misallocated resources into foolish housing construction.  We were a lot poorer in 2010 than 2008 because the Fed made us a lot poorer.

HT:  John Quiggin, Tyler Cowen.

PS.  The recession also reduced immigration, which reduced housing demand even further.

PPS.  Notice the headline of the USA Today article is still reversing causation.


Tags:

 
 
 

56 Responses to “From TheMoneyIllusion to conventional wisdom”

  1. Gravatar of David Pearson David Pearson
    15. September 2011 at 06:15

    You comment reminds me of the first time the global central banks responded (as they did today) to an acute banking system liquidity shortage . 2008? No, August, 2007.

    Why was the global banking system experiencing the beginnings of a run during a mild recession?

  2. Gravatar of johnleemk johnleemk
    15. September 2011 at 06:32

    Scott, this is quite possibly one of the best posts you’ve written to date. Makes the same points you were making in 2009, but in much more concise and snappier form.

  3. Gravatar of W. Peden W. Peden
    15. September 2011 at 06:41

    Do announcements of monetary policy affect asset prices?

    http://www.bbc.co.uk/news/business-14928774

  4. Gravatar of Scott Sumner Scott Sumner
    15. September 2011 at 06:57

    David, The global banking system was experiencing difficulties because of all the bad sub-prime loans, among other things. My point was that those difficulties were not a causal factor in the recession, tight money caused it.

    Thanks Johnleemk.

    W. Peden. I’d say so!

  5. Gravatar of Jon Moen Jon Moen
    15. September 2011 at 06:57

    We seem to focus on price indices while neglecting measures of quantities in the housing markets, as your point 1) implies. Are there useful measures of housing quantities out there to compare to the price indices?

  6. Gravatar of Britmouse Britmouse
    15. September 2011 at 07:37

    That quote about compound interest being the most powerful force in the universe was wrong: the central bank press release is the most powerful force in the universe.

  7. Gravatar of Morgan Warstler Morgan Warstler
    15. September 2011 at 07:42

    1. Please DESCRIBE what the Fed would have DONE DIFFERENTLY 2002-2008 to keep the “small” correction from needing to happen.

    YOUR THEORY SAYS we wouldn’t experience the first bit, please get out your pencil and tell me what changes the Fed would have made when.

    2. On the large correction – old people are SUPPOSED to move back in the homes of their kids. That’s what we want baby boomers to do, consume less.

    If we raised taxes on old people and reduced them on young people, w’ed get the same positive change.

    The point is, you can’t FIAT in the moral conclusion that “real growth” assumes we want to keep sponsoring Seniors, we want to cut seniors.

    SO IF, you want to cut Seniors too… you are going to get the housing compression anyway.

  8. Gravatar of John John
    15. September 2011 at 07:53

    Scott,

    I’m interested in where you get these Hayek arguments concerning NGDP. I’ve read his “Prices and Production,” “A Tiger by the Tail,” “The Constitution of Liberty,” and other essays of his like “The Use of Knowledge in Society” and can’t recall a single mention of the term NGDP.

    Also, I’m curious if you could clarify your reply to Bob Murphy earlier. I think he was trying to say that right now the Central Bank would have to do a lot of unconventional policy options in order to meet your standards which might have a lot of unforeseen consequences you don’t seem to have thought through. It seems like you side-stepped him by saying that with higher NGDP, the Fed would have to sell assets and suck out liquidity. But how do you boost NGDP right now without unconventional policy, and what are some of the possible negative consequences of such policies (moral hazards, distorted price signals, etc)

  9. Gravatar of Bababooey Bababooey
    15. September 2011 at 07:55

    That USA Today article is from May 2010 and reports that household size “is projected to increase slightly in coming years”, and Quiggan/Tyler report (linking the same old article) that household size is back to 1990 levels, but that’s all wrong.

    In November 2010 the Census reported household size was 2.59, down from 2.62 in 2000, the date of the last Census, and 2.63 in 1990. Multi-generational households appear to be increasing, but shrinking married households apparently offset that trend.

  10. Gravatar of David Pearson David Pearson
    15. September 2011 at 08:13

    “those difficulties were not a causal factor in the recession.”

    Those difficulties led to downgrades of the AAA-rated collateral backing the banking system’s short term funding. This was a causal factor in the recession. It began in late 2007, not in the summer of 2008.

  11. Gravatar of Nick Rowe Nick Rowe
    15. September 2011 at 09:42

    Ironically, I’ve just finished writing a post where I say that recessions being a monetary phenomena is now the mainstream view.

  12. Gravatar of Luis H Arroyo Luis H Arroyo
    15. September 2011 at 09:47

    Scott, As you know, I don´t agree with your vision. I think that a run for liquidity not only dry real markets; It downgrades -as Pearson says- all the assets. This, in turn, amplifies the crisis and the uncertainty which make much more complicated the recovery.
    I can not separate the financial economy from real one. The fall of assets prices are more determinant than that of prices of goods. They lead the rest simply because they are more much speedy and influent. They embody several years of income that now are lost.

  13. Gravatar of W. Peden W. Peden
    15. September 2011 at 09:54

    Luis H Arroyo,

    “They (asset prices) lead the rest simply because they are more much speedy and influent.”

    And much larger. National income is a small percentage of net national assets or net national transactions. I like economists that focus on assets; for example, Tim Congdon once defined inflation as “too much money chasing too few goods AND ASSETS”, and Milton Friedman said that money is a “special topic within the theory of capital”.

  14. Gravatar of Luis H Arroyo Luis H Arroyo
    15. September 2011 at 10:00

    Yes thanks, Peden I agree. Can you say me any link on the authors you quoted? Have you any blog?

  15. Gravatar of James Davies James Davies
    15. September 2011 at 10:12

    The housing bubble was not that we built too many houses, it was that house prices were way out of line with what people could *actually* afford, and were thus unsustainable. These are two very separate arguments, correct? And the mispricing of housing was not just concentrated at subprime. Subprime was just the first to show the cracks in the system.

    When house prices adjusted, as they had to once the “affordability” loans turned out to be unaffordable, as they relied on refinancing as housing prices went up at an unsustainable rate, then the loss in wealth from that price adjustment in housing and the subsequent attempts to repair balance sheets explains the recession. Granted, monetary tightness as people were fleeing to safety made things worse than they needed to be, but how can a strictly monetary story explain the above?

    How does reverse causation explain these phenomena?

  16. Gravatar of StatsGuy StatsGuy
    15. September 2011 at 11:15

    Every time you start to feel victorious, I should start worrying about my investments… Just saying.

    I think you’re still underestimating many things – role of the reserve currency debate, commodity input inflation, strength of developing economies accelerating commodity prices and thus forcing Fed to lower real growth to defend against inflation, role of financial leverage, serious and perhaps unsolvable moral hazard problems inherent in poorly regulated banks, lack of transparency in unregulated banks (dark pools), massive demographic shifts, true liquidity problems, huge distributional impact of endogenous money…

    It’s a massive leap of logic from “NGDP targeting would have made us better” to “2008 was mostly (80%) caused by nominal problems not real and most of the real recession happened in 2007”. I’m going to say the truth is somewhere between 0% and 80%.

    However, from your perspective, you should stick to saying 80%. Why? Because the way the modern media manages public debates is to pick the two extremes and let them fight. So, you’re better off being an extreme if you want exposure. So, by all means, you should stick to your argument.

    I hope you win approximately 40% of the debate. 🙂

  17. Gravatar of Oscar Oscar
    15. September 2011 at 11:23

    And The Economist discusses NGDP targeting again. They still don’t seem convinced though.

    http://www.economist.com/node/21529029

  18. Gravatar of GT2211 GT2211
    15. September 2011 at 11:29

    Started to respond prior to class, didn’t get finished and when I came back I see James Davies has already made my point. As loans started going bad, housing prices started falling you would still expect see a decrease in demand for housing because housing is not as an attractive or riskier investment and banks are having a harder time securing funds do the uncertainty and risk on their balance sheets.

  19. Gravatar of W. Peden W. Peden
    15. September 2011 at 11:37

    Luis H. Arroyo,

    I don’t have a blog. I don’t think I will have a blog until a long time into the future, because I begin a masters course in a few weeks and I will be very very busy.

    I can’t remember where that Friedman quote comes from. The Tim Congdon quote comes from his collection of essays titled “Keynes, the Keynesians and Monetarism”. Another good book by him that focuses on the role of assets in an economy is “Money and Asset Prices in Boom & Bust”. He also writes regularly for Standpoint magazine; this is my favourite article by him-

    http://www.standpointmag.co.uk/the-unnecessary-recession-features-june-09-tim-congdon-gordon-brown-alistair-darling?page=0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C5

  20. Gravatar of W. Peden W. Peden
    15. September 2011 at 11:38

    That’s actually page 6. Page 1 is here- http://www.standpointmag.co.uk/the-unnecessary-recession-features-june-09-tim-congdon-gordon-brown-alistair-darling?page=0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C0%2C0

  21. Gravatar of grcridlan grcridlan
    15. September 2011 at 12:24

    Morgan Warstler: (1) “Sumner’s” (Sumner would no doubt credit others with the particular ideas for which they are responsible) theory does not predict that the real shock (the correction of 2006-2008) was avoidable by monetary means. His theory predicts that if NGDP growth remained constant, the financial crisis need not have caused a major recession and continued collapse in the housing sector. (2) Who’s this “we,” kemo sabe? I want everybody to be able to consume more, including aging baby boomers. That’s what growth is. Certainly, household size can change for social as well as economic reasons. But we don’t have a household size “goal”; it isn’t “better” for people to move in with their parents or children, unless they decide to do it on their own, without us giving them economic incentives like a terrible economy caused by NGDP stagnation.

    David P., Luis A.: I think the “neomonetarist” response to your position is “Of course that’s true! And if the Fed had been aggressive enough, it would have compensated for that problem.” It is accurate to say that liquidity shortage, failures of collateral, and asset price changes affect the real economy. But the Fed has literally unlimited power to provide liquidity, provide collateral, and reinflate asset prices.

    [Note: by “provide collateral” I do not mean propping up failing institutions, but rather ensuring adequate liquidity to maintain NGDP growth in spite of their failure.)

    Statsguy: I would argue that the Fed has no business defending the price level against commodity price shocks, and that moral hazard is inherent in regulation of banks (and the accompanying implicit government guarantee) not in poorly regulated banks. But that’s by the by. I’m not sure Sumner is necessarily making the drastic point you say he is making. (First of all, a RE price fall is a monetary phenomenon as well as a (ha!) real phenomenon). He’s just saying that to the extent that the recession is the result of the financial crisis, rather than the real factors, the Fed failed, and that he believes that the majority of the pain since late 2008 is the result of these monetary factors.

    I think.

  22. Gravatar of marcus nunes marcus nunes
    15. September 2011 at 12:31

    W Peden
    Great article!

  23. Gravatar of ssumner ssumner
    15. September 2011 at 12:46

    Jon, There is good data on housing construction and sales. We don’t have good data on empty homes, as the data also includes vacation homes.

    Britmouse–especially when they are screwing up and AD is a problem!

    Morgan, You asked;

    “Please DESCRIBE what the Fed would have DONE DIFFERENTLY 2002-2008 to keep the “small” correction from needing to happen.”

    That’s easy. Ban subprime mortgages. Regulated the hell out of F&F. Cut deposit insurance from $250,000 to $25,000.

    John, It’s common knowledge that Hayek supported NGDP targeting, but I don’t have a citation. Perhaps Glasner or Woolsey know–they are better informed than I am. Selgin discussed that fact in an article on the history of NGDP targeting.

    The unconventional tool needed is a promise to do NGDP targeting, level targeting. That will raise NGDP growth expectations, and reduce the demand for base money. You are basically promising to provide a bit more base money in the future, when we are no longer at the zero bound and hence the Fed controls NGDP directly.

    The Fed shouldn’t use the money supply as a tool, it should respond endogenously to whatever base money the public wants to hold at that level of NGDP expectations.

    bababooey, I’m not prepared to argue the point. Perhaps housing size fell from 2000 to 2005, during the boom, and then increased afterwards. But I am nonetheless convinced I’m right. We know that housing construction has been really low over the past 10 years. That means if we’ve got lots of empty houses it is a mathematical certainty that the problem is a demand shortfall. Maybe bigger families, maybe less immigration, mayber more homelessness, but it’s demand. Because we know for sure it’s not too much supply.

    David, The very mild recession of early 2008 was caused by a slowdown in NGDP growth. The severe recession of late 2008 was caused by a decline in NGDP. That’s monetary policy failing to do 5% NGDP targeting, level targeting. The debt crisis had nothing to do with the problem. It is a symptom.

    Nick, I look forward to reading it.

    Luis, You are reversing causation. The recession caused most of the debt crisis. It’s not coincidence that debt crises suddenly popped up everywhere in the world as soon as NGDP fell. Housing, CRE, sovereign debt, pensions, you name it, the cause is too little NGDP. Yes, there are specific problems that would have occurred w/o falling NGDP (Greece, the 4 subprime states in the US, etc.) But the big problem was falling NGDP.

    James, You said;

    “The housing bubble was not that we built too many houses, it was that house prices were way out of line with what people could *actually* afford, and were thus unsustainable. These are two very separate arguments, correct?”

    Yes, but people who argue that housing caused the recession rely on the argument that we built too many houses. Otherwise the problem could be solved by simply printing money.

    Your comments on the housing bubble are factually incorrect. There was no housing bubble at all in many states like Texas. Yet their prices also began falling once NGDP fell in late 2008. Housing prices had to fall in a few areas where they had gotten wildly out of line, but certainly not in all states. They didn’t have to fall in Boston where I live–they fell because NGDP fell.

    Statsguy, You said;

    “I think you’re still underestimating many things – role of the reserve currency debate, commodity input inflation, strength of developing economies accelerating commodity prices and thus forcing Fed to lower real growth to defend against inflation”

    That is my argument!! The Fed wrongly tightened in 2008 to prevent commodity inflation. They should target NGDP, then they wouldn’t have tightened, they would have loosened. And “forcing the Fed???” The Fed has a dual mandate, they should target NGDP. They are breaking the law by targeting inflation.

    I’ve never argued most of the real recession happened in 2007.

    Oscar, Thanks I’ll take a look.

  24. Gravatar of ssumner ssumner
    15. September 2011 at 13:01

    GT2211, I agree that bad loans would weaken the housing market, but that was a regional problem. See also my response to James.

  25. Gravatar of Luis H Arroyo Luis H Arroyo
    15. September 2011 at 13:04

    thanks to you, Peden & Scott. Scott, I think that all you say is the fault of an economy centered in flow variables. An economy (at least of cicles) of assets would be more productive.

  26. Gravatar of David Pearson David Pearson
    15. September 2011 at 13:22

    “The debt crisis had nothing to do with the problem. It is a symptom.”

    The debt crisis — a run on liabilities of the shadow banking system — started with downgrades of AAA-rated collateral, well before the fall of 2008.

    The financial crisis had a very well defined mechanism. Gorton has a clear, convincing argument on what that mechanism was. Its worth a read.

  27. Gravatar of Luis H Arroyo Luis H Arroyo
    15. September 2011 at 14:10

    David P., Luis A.: I think the “neomonetarist” response to your position is “Of course that’s true! And if the Fed had been aggressive enough, it would have compensated for that problem.” It is accurate to say that liquidity shortage, failures of collateral, and asset price changes affect the real economy. But the Fed has literally unlimited power to provide liquidity, provide collateral, and reinflate asset prices.
    I agree, of course. It only emphasizes the QE needed, because The fall in asset prices is so destructive.
    I just read the Congdon article, and is very good. He proves that Euro zone is in a great mistake trying to recapitalize banks!

  28. Gravatar of David Pearson David Pearson
    15. September 2011 at 16:17

    Luis,

    The Fed did attempt to provide liquidity. It succeeded up until the GSE failure in August, 2008. The Great Run on Shadow Banks was fought by the Fed through the alphabet soup of liquidity programs beginning in late 2007. These included, the TAF, the PDCF, the CPFF, and Maiden Lane (the Bear Stearns bail out). These programs failed precisely because the Fed was unable to reinflate asset prices: specifically, the prices of ABS, CDO’s, and finally, (GSE-guaranteed) RMBS. It failed to reinflate these because the Fed could do little about falling house prices. It could do little about falling house prices because housing demand was shifting leftward as shadow banks returned to reasonable underwriting practices.

  29. Gravatar of JTapp JTapp
    15. September 2011 at 16:33

    I agree with Pearson about Gorton. It’d be helpful if Scott would frame Gorton’s investigations inside his own, IMO.

    If it wasn’t for the run on the repo market the subprime collapse would not have been catastrophic. How those MBS used as collateral in the repo market came to exist can’t be understood without understanding the loans that were made under the assumption that housing prices would keep rising, counting on a refinancing every 3 years or so. And perhaps that can’t be understood without a “Greenspan put.”

  30. Gravatar of David Pearson David Pearson
    15. September 2011 at 16:39

    W. Peden, Luis,

    Congdon relies heavily on the portfolio balances effect. That effect assumes that risk-asset QE reduces the risk of private sector portfolios, forcing them to rebalance by buying more risk assets.

    I just don’t see how this works. Other than liquidity risk, the Fed cannot cancel any portfolio risk: instead, it just transfers it to Treasury, which transfers it to the private sector (tax payers).

  31. Gravatar of An alternative script – One where interest rates don´t make an “appearance” | Historinhas An alternative script – One where interest rates don´t make an “appearance” | Historinhas
    15. September 2011 at 17:33

    […] evidence above provides ample confirmation that, as Scott Sumner has long argued, both the “greatness” of the housing slump and the “epic […]

  32. Gravatar of Benjamin Cole Benjamin Cole
    15. September 2011 at 19:05

    Brilliant commentary by Scott Sumner. Again.

  33. Gravatar of Luis H Arroyo Luis H Arroyo
    16. September 2011 at 01:24

    Pearson, the Congdon´s article sounds me like the Minsky/keynes desequilibrium: all assets are fixed around the liquidity, and liquidity is not so stable as it is generally thought.
    An investors don´t decide his investment without regarding his tenance of liquidity. The problem is that confidence tends to reduce the demand for money, but when confidence disappears and is replaced by uncertainty, the desired stock of money grows a lot.
    That is how I see it. The fall in asset prices complicates the way of exit, because I suppose that confidence is extremly more sensible to the balance sheets desequilibria.
    In this crisis, the leveraging and asset prices were so high, that to recover from the huge fall in both requires much times, I suppose, than for a simple recession.

  34. Gravatar of W. Peden W. Peden
    16. September 2011 at 04:01

    David Pearson,

    “instead, it just transfers it to Treasury, which transfers it to the private sector (tax payers).”

    In the final analysis, this is mistaken: a country that is willing to engage in QE is willing to underfund government borrowing, which under current conditions doesn’t even involve an inflation tax. That’s perhaps why no-one ever thinks about a central bank going bust.

  35. Gravatar of David Pearson David Pearson
    16. September 2011 at 05:33

    W. Peden,

    I’m not sure if I understood your comment. Are you saying the Fed can operate with a negative net worth indefinitely? Or that the Fed can always fund any Treasury shortfall?

    Chronic money-financed deficits are a foolproof way out of a liquidity trap. They also cause the central bank to lose control of the price level, more often than not in presence of a large supposed output gap. We have plenty of examples of this.

  36. Gravatar of David Pearson David Pearson
    16. September 2011 at 05:37

    BTW, I should also note that my discussion of portfolio risk assumes there is no demand for marginal bank reserves. When there is, risk-asset QE is unnecessary. When there isn’t, risk asset QE is ineffective (in reducing duration or beta risk).

    The thing about money-financed deficits is that they create marginal demand for bank reserves as the PSBR boosts total credit. That is why they can always “work”. The problem is that demand for marginal bank reserves typically derives from inflation hedging decisions that are harmful to investment and aggregate supply. Again, there are plenty of examples of this (Brazil in the 90’s, for instance).

  37. Gravatar of David Pearson David Pearson
    16. September 2011 at 05:39

    Luis,

    I agree QE always reduces liquidity risk. Liquidity risk is the smallest component of equities risk; if liquidity premia are “normal”, it would take an enormous amount of equities QE to reduce aggregate private sector portfolio risk. Besides, this is not the claim Congdon is making: he believes the Fed “takes out” beta risk from portfolios, causing investors to increase it afterwards.

  38. Gravatar of K K
    16. September 2011 at 06:43

    David Pearson: “Other than liquidity risk, the Fed cannot cancel any portfolio risk: instead, it just transfers it to Treasury, which transfers it to the private sector (tax payers).”

    I sort of agree with you. Except… You have to ask what is the sensitivity of the wealth of different private agents to changes in the wealth of government. In other words, who are the owners of government? So the government removes the risk from the owners of capital. Changes in the value of the assets will now directly impact the both future levels of taxes *and* services, so poorer members of society are now more long those assets (and richer ones correspondingly less so). Lets assume the assets are positive beta. It seems reasonable to assume that reducing the beta of the private owners of capital would cause them to want to invest more, which is good. But increasing the beta of consumers who are already fearful for their own employment (i.e. feeling long beta) might make them consume less. So it depends on the sensitivity of each groups actions to their beta *and* on the sensitivity of the economy to investment vs consumption. At the very least, anyone arguing in favor of positive portfolio effects ought to be wary of potentially offsetting impacts.

  39. Gravatar of W. Peden W. Peden
    16. September 2011 at 06:52

    David Pearson,

    “Chronic money-financed deficits are a foolproof way out of a liquidity trap.”

    If so, then any talk of “risk” being attatched to the private sector from QE is misleading. Any government willing to let a central bank engage in QE is willing to engage in underfunding in the unlikely event that the central bank gets into trouble.

    “They also cause the central bank to lose control of the price level, more often than not in presence of a large supposed output gap. We have plenty of examples of this.”

    The central bank never has strict control over the price level anyway. It always has indirect control under inflation-targeting, however, by changing the level of AD. It has neither more nor less control over the price level when the PSBR is underfunded.

    I’m still not sure you’ve got Congdon’s point, which is just Fisher’s old hot potato effect.

  40. Gravatar of David Pearson David Pearson
    16. September 2011 at 08:10

    You argue the private sector will incorrectly quantify the risk of future inflation resulting from Fed risk-asset losses. This is a plausible argument, but needs fleshing out. What is the mechanism? Information costs?

    You are right on Congdon. I view the private sector as holding a portfolio of claims on future net cash flows. I don’t see how the Fed can affect the risk of that portfolio, other than liquidity risk. Its very simple, in a way. What am I missing that Congdon is saying? If he is assuming QE produces inflation, then he is assuming the outcome of the process, but missing a mechanism for getting there. Also, he seems to explicitly talk about the Tobin effect, which, again, assumes the Fed can extinguish beta. How?

  41. Gravatar of Luis H Arroyo Luis H Arroyo
    16. September 2011 at 08:56

    P & P
    I agree QE always reduces liquidity risk. Liquidity risk is the smallest component of equities risk; if liquidity premia are “normal”, it would take an enormous amount of equities QE to reduce aggregate private sector portfolio risk.
    Well I don´t know why LR would be normal in any ciscunstance. And I don´t see that how risk premia would be independent of liquidity risk.
    If liquidity demand rise, Price of all assets falls. Can we think about an asset with negative beta risk in these ciscunstances? I don´t know

  42. Gravatar of W. Peden W. Peden
    16. September 2011 at 09:53

    David Pearson,

    As I see it, the key factor is the pursuit of profit by holders of assets. Selling assets for cash isn’t done to use up time on a lazy afternoon, but because that cash (which it is not desirable to hold) is to be used to acquire other securities. That equilibriates things at the level of the firm, but not at the level of the financial system. The financial system can only attain equilibrium after a change in the composition of private sector cash holdings by increasing the number of assets i.e. increasing the quantity of money.

    “You argue the private sector will incorrectly quantify the risk of future inflation resulting from Fed risk-asset losses.”

    I’m not sure that I am. As I said, I can only follow what you’re saying if we’re talking about a central bank operating for a government that doesn’t have monetary sovereignty.

  43. Gravatar of David Pearson David Pearson
    16. September 2011 at 11:51

    W. Peden,

    To put it in your terms: when the Fed purchases risk assets, someone is left holding a risk-free asset, and someone is given a new claim that is exposed to corporate cash flow volatility (through the claim on their future tax payment). The first person may go out and buy equities, the other will sell equities. The aggregate effect on private sector risk is zero.

    How could it be otherwise? The Fed is just a bank that happens to have a monopoly on producing bank reserves. That bank can extinguish liquidity risk because it funds itself with irredeemable deposits. However, it can still be insolvent. Do you dispute that? If not, then recognize a bank at risk of insolvency cannot add value to the private sector by buying risk assets funded with interest-paying debt.

  44. Gravatar of David Pearson David Pearson
    16. September 2011 at 11:56

    To be clear: I still think you could make a case for QE having an effect on private sector risk. It would involve modeling how private actors fail to recognize the effect of Fed risk asset purchases on their portfolio of claims.

    From personal experience, I can tell you that private actors in countries where the CB has put their balance sheet at risk are definitely aware of the implications. The more duration risk a CB takes, for instance, the less duration risk the private sectors wants.

  45. Gravatar of K K
    17. September 2011 at 00:29

    David Pearson: “It would involve modeling how private actors fail to recognize the effect of Fed risk asset purchases on their portfolio of claims.”

    You don’t have to do anything near as radical as that. The assumptions that are required in CAPM in order for the equilibrium to be independent of agent preferences and initial endowment are *quite* strong. Multi-period models generally require market completeness to make individual preferences go away. Models with ongoing trading (incomplete markets) are usually representative agent models. Your assertion that government simply passes the risk back to the “tax payer” is equivalent to claiming that we live in a representative agent economy. There are myriads of reasons to expect that we don’t: finite lifespans, different preferences, agent dependent restricted acces to borrowing. We don’t even share the same probability measure, ie we don’t agree on the correct model of the world we are living in. The odds that our economy satisfies the conditions for independence of initial endowment are nil.

  46. Gravatar of W. Peden W. Peden
    17. September 2011 at 03:23

    David Pearson,

    “To put it in your terms: when the Fed purchases risk assets, someone is left holding a risk-free asset, and someone is given a new claim that is exposed to corporate cash flow volatility (through the claim on their future tax payment). The first person may go out and buy equities, the other will sell equities. The aggregate effect on private sector risk is zero.”

    Then what? Now we have another person (the seller of the equities) with the cash asset.

    “How could it be otherwise? The Fed is just a bank that happens to have a monopoly on producing bank reserves. That bank can extinguish liquidity risk because it funds itself with irredeemable deposits. However, it can still be insolvent. Do you dispute that? If not, then recognize a bank at risk of insolvency cannot add value to the private sector by buying risk assets funded with interest-paying debt.”

    If central banks are ultimately backed by the government, then you need an additional assumption to prove that they are ultimately backed by the taxpayer- namely, that the government is unwilling to recapitalise a central bank via underfunding.

    That said, if I understand K’s post correctly (it’s not that it’s unclear, it’s just that it’s an aspect of economics with which I’m not very familiar) you seem to be making some rather heroic assumptions about microeconomic behaviour. Empirically, I’m not convinced that we should take any statement that the central bank doesn’t at least have a short-run influence on financial asset prices seriously.

  47. Gravatar of K K
    17. September 2011 at 05:02

    W. Peden: “you seem to be making some rather heroic assumptions about microeconomic behaviour.”

    Right. I only stated some fairly general conditions under which I’m certain the nature of the representative agent is not independent of the distribution of wealth. In fact, as far as I know, the state of the art is that independence of initial endowment has only been shown where all agents have exactly the same borrowing and lending costs *and* they all have CARA or CRRA utility functions. (It is, of course, implicitly understood that we all agree on the model of the world and all have access to and can process all the available information). If this is not the case, then you would expect *any* redistribution of wealth to change the equilibrium prices. Heroic indeed.

  48. Gravatar of David Pearson David Pearson
    17. September 2011 at 06:09

    W. Peden,

    On your first comment, that person holding the riskless asset (cash) prefers to hold it rather than equities. It is not a “hot potato”.

    I’m also not familiar with what K discusses, but it seems a general critique of Ricardian Equivalence (is that right?). This is fine. I am not trying to argue that QE will not have an impact on asset prices under any conditions; I’m urging its proponents to specify those conditions, and then we can have a real discussion about whether they apply.

    As far as what K is saying, I’m not sure why the assumptions have to be so heroic. If a central bank exposes its balance sheet to losses it sacrifices credibility; if it sacrifices credibility, agents will more closely scrutinize its actions; if that occurs, they will see that risk is merely being shuffled around. This is an empirical observation. If the assumptions were “heroic”, why would we observe it?

    In any case, I’m sympathetic to the types of arguments K makes. Unfortunately, I can’t speak directly to his unless he puts them in clearer terms. Is he saying the asset price effect is temporary? Can it be consistently achieved? If a taxpayer now faces duration risk from his tax liability, why doesn’t he shorten duration to adjust the risk of his portfolio?

  49. Gravatar of David Pearson David Pearson
    17. September 2011 at 06:36

    Btw, it seems K would also argue for fiscal stimulus given his post.

  50. Gravatar of W. Peden W. Peden
    17. September 2011 at 07:11

    David Pearson,

    “On your first comment, that person holding the riskless asset (cash) prefers to hold it rather than equities. It is not a “hot potato”.”

    In an economy where cash could not be used to purchase anything, I can see that being the case. However, in such an economy, would anyone want to hold cash?

    I agree that more detail on the reason why QE has an impact on asset prices is needed.

  51. Gravatar of Scott Sumner Scott Sumner
    17. September 2011 at 09:26

    Luis, I don’t follow.

    David and JTapp, I should have said the severe phase of the debt crisis after Lehman. The estimated size of the debt crisis (IMF data) is highly correlated with expected NGDP growth for 2009-10. Falling NGDP always makes debt crises much worse, as it means people have less nominal income to repay nominal debts. I’d never argue that a debt crisis is impossible with NGDP targeting, but at least it wouldn’t cause a recession.

    The Fed was not trying hard to prevent NGDP from falling, indeed they didn’t even cut rates (then 2%) in the meeting after Lehman failed.

  52. Gravatar of David Pearson David Pearson
    17. September 2011 at 09:52

    W. Peden, K,

    Thanks for the interesting discussion. K’s post of 6:43 clarified my thinking: beta risk affects consumption as well portfolio decisions. Both of you also challenged the thesis, which is always a good thing. I primarily take issue with people like Congdon, strident risk-asset QE advocates that don’t seem to test their assumptions, at least not for the benefit of their policy-making audience.

  53. Gravatar of K K
    17. September 2011 at 09:53

    David: “if that occurs, they will see that risk is merely being shuffled around”

    Agreed. That’s not the “heroic” assumption. The assumption that you are implicitly making is that all agents

    1) are identical (this is your mythical “tax payer”); or
    2) are fully able to hedge all of their future contingent consumption and income streams; or
    3) have access to funds at identical terms *and* have very particular kinds of utility functions *and* have complete and identical understanding of the dynamics of their economy.

    Under any of those three conditions there exists a unique equilibrium price of all assets in the economy *independently* of the distribution of wealth.

    If those conditions are not satisfied then changing the wealth endowment will change asset prices. Then, when the Fed buys assets then the wealth distribution in society after those assets change in price will be different from what it would have been if those assets had been with their previous owners. Therefore the future equilibrium prices will be different. Therefore the current prices will be different.

    I hope that’s more clear… 🙁
    have access

  54. Gravatar of K K
    17. September 2011 at 09:55

    Ignore “have access” at the end of previous comment. No idea how that slipped in

  55. Gravatar of David Pearson David Pearson
    17. September 2011 at 14:24

    K,

    Very clear. I understand a low-income household cannot easily “shorten duration”. That is why your comment above on adjusting consumption was interesting. There are many ways that actors can adjust their portfolios of claims on future cashflows beyond using financial instruments. For instance, in Brazil, households would “shorten duration” by immediately spending paychecks, such that store shelves were usually empty right after payday.

    My other comment is that your argument applies to fiscal stimulus as well. In fact, there seems to be little difference between monetary and fiscal policy in that, to affect asset prices, each must violate one of these conditions. I think from the perspective of political economy, extra-democratic fiscal policy carried out by the Fed is dangerous. It is highly likely to put Fed independence at risk if it fails to work as advertized. Again, something Congdon never seems to mention.

  56. Gravatar of David Pearson David Pearson
    17. September 2011 at 14:28

    K,

    Another way to put it: if you substitute “Treasury” for “Fed” in any risk asset QE proposal, such as Congdon’s, there will be little difference in the mechanism or outcome. That is, of course, unless there exists demand for a marginal unit of bank reserves created by the Fed through QE. At the moment there is not.

Leave a Reply