Saturday, November 22, 2008

Roubini's list of doom

Forbes lists Roubini's list of 20 (where many are very similar) reasons why consumption is set to decline over the next few years. I am not ready to call on consumption declining permanently yet, but unless policy makers get their acts together on this financial crisis, there may be no other alternative. From Forbes (I added commentary below some of the items):
"Consider the many severe negative factors affecting consumption. One can count at least 20 separate or complementary causes that will sharply reduce consumption in the next several years:

1. The U.S. consumer is shopped-out, having spent for the last few years well above his means.

RW: Of course this is number one; it's the classic deleverage scenario. I still say that consumers will not cut down permanently until they are forced to, i.e., banks say NO MORE! If consumers are forced to "delever", then we will not be able to separate this effect from the "recessionary" effects.

2. The U.S. consumer is savings-less, as the already low household savings rate at the beginning of this decade went to zero/negative by 2006 and now has to rise to more sustainable levels.

RW: This kind of sounds like number 1. If they reduce their debt burden, then by definition, the consumer must be saving.

3. The U.S. consumer is debt-burdened, with the debt-to-disposable-income ratio having increased from 70% in the early 1990s to 100% in 2000 and to 140% in 2008.

RW: Okay, so number three is very similar to 1. and 2. (just 1 reason so far).

4. Not only are debt ratios high and rising, debt-servicing ratios are high and rising, too, having gone from 11% in 2000 to almost 15% now, as the interest rate on mortgages and consumer debt is resetting at higher levels.

RW: 'Tis true, as long as mortgage rates set higher. Let's see what the FDIC and the Treasury come up with here.

5. The value of housing wealth is now falling by over $6 trillion, as home-price depreciation will soon be 30% and reach a cumulative fall of over 40% by 2010. Recent estimates of this wealth effect suggest that the effect may be closer to 12%-14% rather than the historical 5% to 7%. And with home prices falling over 30%, about 40% of all households with a mortgage (or 21 million out of 50 million who have a mortgage) will be under water (negative equity in their homes) with a huge incentive to walk away from their homes.

RW: Finally, the wealth effect. There is not a lot of impressive supporting evidence that a house is, in effect, an ATM card. But I suspect there is a significant wealth effect that will be estimated as soon as enough data is accumulated - this loss in housing wealth is unmatched in modern history.

6. Mortgage equity withdrawal (MEW) is collapsing from the $700 billion annualized in 2005 to less than $20 billion in the second quarter of this year. Thus, with falling housing wealth and collapsing MEW, U.S. households cannot use their homes anymore as ATM machines.

7. The value of the equity wealth of U.S. households has fallen by almost 50%, another ugly wealth effect on consumption.

RW: This will cycle out, as it always does. Unless equity markets do not return to their previous levels within a few months to a year. It took years and years for the equity markets to return to their peak values following the Great Depression.

8. The credit crunch is becoming more severe as the recent second quarter flow of funds data and the Fed Loan Officers' Survey suggests: It is spreading from sub-prime to near prime to prime mortgages and home equity loans; and from mortgages to credit cards, auto loans and student loans. Both the price and the quantity of credit are sharply tightening.

RW: This is clearly a stress point now, but what happens when the Fed's $1.3 trillion and growing starts to seep out of the banking system as money. That's right - more loans.

9. Consumer confidence is down to levels not seen since the 1973-75 and 1980-82 recessions.

RW: That usually happens in a recession.

10. Real wage growth and real income growth have been stagnant in the last few years as income and wealth inequality has been rising. And now with GDP and real incomes falling, real consumption will fall sharply.

RW: I don't really understand what stagnant real income and wages have to do with income and wealth inequalities. Actually, as long as the rich keep on consuming, you will see no change in aggregate consumption even if a portion of the population reduces their spending.

And GDP and real incomes always fall in a recession - again, consumption suffers in a recession! Nothing new here.

11. The Fed is reaching zero-bound on interest rates as the economy gets close to deflation, given the slack in goods, labor and commodity markets. Deflation means consumers will postpone consumption as future prices are lower than current prices, as real rates are positive and rising and as debt deflation increases the real value of households' nominal debts.

RW: This is back to the financial crisis. I have argued that deflation is not a sure thing - actually, far from it. And furthermore, consumers will not alter their behavior until deflation becomes embedded into expectations. As I argue in this post, consumers expect 6.9% over the next year, which is well above zero.

12. Employment has been falling for 10 months in a row and the rate of job losses is now accelerating. In the last recession in 2001, which was short and shallow (eight months from March to November 2001, with a cumulative fall in GDP of only 0.4%), job losses continued all the way until August 2003--with a job-loss recovery and a total cumulative loss of jobs of over 5 million from the peak. In this cycle, job losses have, so far, been "only" slightly over a million, while labor market conditions are severely worsening based on all forward-looking indicators such as initial and continuing claims for unemployment benefits. Massive job losses and concerns about job losses will further dampen current and expected income, and further contract consumption.

RW: This is certainly going to weigh in on consumption. Better believe it!

13. Tax rebates of over $100 billion failed to stimulate real consumption earlier in 2008. Only 25% of the rebate was spent as U.S. consumers are worried about jobs and needed to use funds to pay off their credit cards and mortgages. The tax rebate was supposed to boost consumption all the way through September 2008: In reality, real retail sales and real personal spending rose only in April and May, while starting in June and all the way into July, August, September, October and now the holiday season, real retail spending and real personal spending have been down month after month. Thus, another general tax rebate would be as ineffective as the first one in boosting consumption.

RW: That's what happens when you send a check to overburdened consumers - they save it.
This does not necessarily imply that consumption is going to decline further (although it likely will over the fourth quarter of 2008 and the first quarter of 2009), just that the second stimulus plan should not be a stupid rebate program.

14. The 1990-91 and 2001 recessions were not global; this time around, the IMF is forecasting a global recession for 2009.

15. The recent rise in inflation--that is only now slowing down--reduced real incomes even further for lower-income households who spend more than the average households on gas, transportation, energy and food. The recent sharp fall in gasoline and energy prices will increase real incomes by a modest amount (about $150 billion), but the losses of real disposable income and thus falling consumption from other sources (wealth, income, debt servicing ratios) are much larger and more significant.

16. The trade-weighted fall in the value of the U.S. dollar since 2002 has worsened the terms of trade of the U.S. and reduced further real disposable income and the purchasing power of U.S. consumers over foreign goods.

RW: Well, as long as there is a financial crisis underway, the U.S. dollar is staying strong. However, a decline in the U.S. dollar improves export income, which will fall in the hands of workers who will spend it. All this is an argument is for reducing the U.S. current account, which was going to happen - eventually - in spite of any recession or financial crisis.

17. With consumption being over 71% of GDP, a sharp and persistent contraction of consumption all the way through at least the fourth quarter of 2009 implies a more severe recession than otherwise. Consumption did not fall even a single quarter in the 2001 recession and one has to go back to 1990-91 to see a single quarter of negative consumption growth. But the worsening balance sheet of U.S. consumers in 1990-91 (debt ratios, debt servicing ratios, employment contraction, wealth effects of housing and stock markets) was much less severe than the current downturn.

RW: We are back to 1.-3: saving more and consuming less to reduce debt burden.

18. Monetary easing will not stimulate durable consumption and demand for residential housing, as demand for such capital goods becomes interest-rate insensitive when there is a glut of capital goods; monetary policy becomes like pushing on a string. In the previous recession, the Fed cut the Fed Funds rate from 6.5% to 1% and long rates fell by 200 bps. In spite of that, capital expenditure in the corporate sector fell by 4% of GDP between 2000 and 2004 as there was a glut of tech capital goods and it took years to work out such a glut. Today, there is a glut of housing, consumer durables and autos/motor vehicles; so it will take years to work out this glut and monetary policy is becoming ineffective to resolve that glut.

RW: This is a bit like number 11. But he is failing to mention the effects that fiscal stimulus may have on capital goods and even consumer durables - yup, that could be very positive....and perhaps just enough to get us through the next year.

19. While policy rates are sharply falling, the nominal and real rates faced by households are rising rather than falling: rising mortgage rates, rising rates on credit cards, auto loans and student loans, together with less availability of credit are severely dampening the ability of households to borrow and spend.

RW: They are rising, but if deflation does occur - which Roubini is proposing - guess what will happen to long term interest rates. Yup, they will fall.

20. To bring back the household savings rate to the level of a decade ago (about 6% of GDP) consumption will have to fall--relative to current GDP levels--by almost a trillion dollars. If all of this adjustment were to occur in 12 months, GDP would contract directly by 7% and indirectly (including the further collapse of residential and corporate capital expenditure in a severe recession) by 10%, an exemplification of the Keynesian "paradox of thrift." If such an adjustment were to occur over 24 months rather than 12 months, you would still have negative GDP growth of 5% for two years in a row with a cumulative fall in GDP from its peak of 10%. (Note that in the worst U.S. recession since WWII, such cumulative fall in GDP was only 3.7% in 1957-58). One can only hope that this adjustment of consumption and savings rates occurs slowly over time--four years, say, rather than two.

Even in that scenario the cumulative fall of GDP could be of the order of 4% to 5%, i.e., the worst U.S. recession since World War II. Note that the cumulative fall in GDP in the 2001 recession was only 0.4%--and in the 1990-91 recession only 1.3%. So, the current recession may end up being three times as long and at least three times as deep (in terms of output contraction) than the last two.

Nouriel Roubini, a professor at the Stern Business School at New York University and chairman of Roubini Global Economics, is a weekly columnist for
Rebecca Wilder


  1. I would suggest that items 2 & 3 are the constituents of item 1: the shopped out nature of the US consumer is due to both the debt burden and the lack of savings. Note, I don't believe that the savings rate will automatically rise until the debt burden has fallen dramatically. Consumers can't afford to save until they have written down some of their more expensive debt (in particular credit card debt). So I don't think you can really say that they are the same thing.

  2. Hi Yoganmahew,

    Thank you for your comments!

    I think that it is just semantics in the language - saving and dissaving, or saving and reducing debt (as long as you are assuming that the consumer is indeed paying down debt).

    1. If you save, then you are accumulating assets and raising the your stock of wealth.

    2. If you dissave, then you are accumulating debt - borrowing against the future - and thereby reducing your stock of wealth.

    3. Therefore, if you start with net debt, then by saving you are paying off that debt until you are no longer in debt, and increasing your stock of wealth along the way. Therefore, you must have been "saving" because wealth is increasing.

    Thanks again for reading and I look forward to hearing from you again.


  3. Hi Rebecca, thanks for the reply. My point is that there are two forces at work - one is saving for the bad times (like now, when one should be using savings to cushion the blow of reduced earning power), the other is high cost debt (in particular credit cards with rolled over balances.

    The savings should be used to even out consumption and also to continue to pay 'vital' debt (in particular mortgage payments) - debt that isn't going to go away any time soon because it is simply too big to pay off in a clump.

    So consumers have double jeopardy - not only do they not have any savings to ease the burden of recession, but they also have oodles of high cost debt.

    While you are undoubtedly right that the net effect is the same thing, in practical terms, they are too different items - each one of them on their own would be a bad thing to go into a recession with - no savings, lots of high cost debt.

    We are facing an extraordinarily similar situation here in Ireland. It is going to be a long winter!