October 26, 2009
The National Saving Identity: Private Saving, Household Saving, and Rebalancing
The National Saving Identity states:
CA ≡ (T-G) + (S-I)
Where CA is the current account, (T-G) is the consolidated government budget balance, and (S-I) is the private sector saving-investment balance. Figure 1 depicts the profound shifts that have occurred in these components (normalized by nominal GDP).
Figure 1: Net government saving (blue), net private saving-investment balance, (red) and current account (green), all normalized by nominal GDP. NBER defined recessions shaded gray; assumes latest recession ends 2009Q2. Source: BEA, GDP 2009Q2 3rd release, Tables 3.1, 4.1, 5.1.
Note that I've omitted the statistical discrepancy which makes these items add up exactly.
How much of the recent shift in the net private saving is due to changes in personal saving (as opposed to corporate behavior)? Actually quite a bit. Of the 2.6 ppts shift in net private saving since 08Q1, 2.9 ppts is accounted for by the shift in personal saving.
Figure 2: Net private saving (pink), and net personal saving, (teal). NBER defined recessions shaded gray; assumes latest recession ends 2009Q2. Source: BEA, GDP 2009Q2 3rd release, Table 5.1.
How persistent will this shift in the personal saving rate be? This is the big question, in terms of the rebalancing issue (keeping in mind that the national saving identity is a tautology). Deutsche Bank provides an interesting set of calculations, which indicates how long it will take to hit the 20 year average net wealth/disposal personal income ratio.
Chart 6 from Hooper, Slok, Dobridge, "U.S. Consumer Balance Sheet Adjustment: Half Way Done," Global Economic Perspectives (Deutsche Bank, Oct. 7, 2009) [not online].
Peter Hooper, Torsten Slok and Christine Dobridge write:
To try to gauge historical norms that households may aim for we appeal once again to average values that have prevailed over time. The 20-year average of household net worth is 533% of income. On this basis, net worth has returned about half way to its historical norm from the low reached in Q1. Chart 6 shows two prospective paths of adjustment back to the 20-year average, a 3-year path and a 5-year path. To follow these paths, we assume that households use half of their saving to pay down debt, and the other half to purchase assets. We also assume that income grows at 1% a year and asset values grow at the same rate. In order to rebuild wealth in three years then, households would need to raise their saving rate to 7% immediately and to 8% by 2012. In order to rebuild wealth in 5 years, however, households would need only a 2% to 3% saving rate. The saving rates implied by this wealth calculation are lower than the rates implied by the debt calculation. This is because net worth has risen since Q1 because of the rebound in the stock market. Net worth-toincome looks to have been about 500% in Q3; households have already made good progress towards their wealth target.
This set of calculations suggests at least a few years of relatively muted consumer behavior. The key factor is the rate at which households seek to reestablish their target net worth/income ratios.
It's interesting to contrast this perspective with that the "Blame it on Beijing" view, which holds Rest-of-World excess saving as the driver. I believe that when considering the US economy -- which is about three times as large as that of China (according to IMF WEO data) -- one can reasonably argue that what happens here is at least as important as what happens in East Asia (in contrast to some observers, I take Chairman Bernanke's recent speech, focusing on raising US national saving, as a welcome return to thinking about the primacy of US factors [speech text]).
Posted by Menzie Chinn at October 26, 2009 05:17 PMdigg this | reddit
"We also assume that income grows at 1% a year and asset values grow at the same rate."
Therein lies the crucial issue of the current account.
Household savings compete with private sector business savings, but private sector business savings rates move in the opposite direction of asset prices:
Earnings are retained to the degree that internal investments have a return in excess of the cost of capital for the firm, which moves in the opposite direction of asset prices, at least on a long term basis. Therefore business savings rates will fall "voluntarily" only if asset prices adjust downward. They can always fall involuntarily, due to a collapse in revenue, but in this case the response of the business sector is to liquidate labor up until their savings rate matches the cost of capital. If they cannot do this, the entire firm liquidates.
But the assumption here is that *both* household savings and asset prices are going up. This can only happen if G-T + NX is increasing at least as fast as the desired household savings rate.
But G-T will not increase beyond the current level -- most likely it will decrease, and rapidly so. If we assume a linear reduction to, say, a 600B deficit, then that is eventually a 5-6% headwind for household savings + net exports.
If you assume that households savings will remain at their current level or even increase, then that means the current account shifts by 5-6%. Even if you assume smaller decreases in the deficit going forward, it remains that case that the current account must absorb *all* of the household savings increases + government deficit reductions from here on out.
Therefore the savings identity belies the "blame China" characterization of those who focus on NX, since it shows that unless G-T increases from the current level, NX must absorb at least the desired household savings increase.
The fact that the Chinese economy is three times smaller only emphasizes how out of hand the situation got, since you would not expect such a small trading partner to be able to drain such a large amount of savings. It was an amazing achievement, and will be much harder to roll-back than had our deficit been with a much larger economy, since the adjustment will be three times as painful for China. That is one problem with identities -- NX must still take the hit, even if China's economy is small.
Posted by: Robert at October 26, 2009 09:09 PM
How about breaking down private savings into the lower/middle class and the rich (along with foreigners too)?
Will you find that the lower/middle class (and recently the gov't) have gone into debt (denominated in currency) to the domestic rich and the foreign rich because of negative real earnings growth on the lower/middle class and really positive earnings growth for the domestic rich and foreign rich?
Posted by: Get Rid of the Fed at October 26, 2009 10:06 PM
"How much of the recent shift in the net private saving is due to changes in personal saving (as opposed to corporate behavior)?"
How about the lower/middle class have stopped going further into debt while the domestic rich are still making excess savings?
See this link:
Posted by: Get Rid of the Fed at October 26, 2009 10:12 PM
Robert, any chance that the high wage locales like Canada, Japan, and Europe will take the NX "hit" instead of china?
Posted by: Get Rid of the Fed at October 26, 2009 10:16 PM
"This is because net worth has risen since Q1 because of the rebound in the stock market. Net worth-toincome looks to have been about 500% in Q3; households have already made good progress towards their wealth target."
Ignoring wealth/income inequality? Do the people with the most debt own the least amount of stock?
Posted by: Get Rid of the Fed at October 26, 2009 10:19 PM
The twenty year average net worth of 533% net income is way high. Look at the chart. A quick glance shows that value is distorted by the internet and housing bubbles. Not a single year prior to say 1998 exceeds this level. What's the median twenty year net worth?
Posted by: jd at October 26, 2009 11:02 PM
Very interesting analysis ; would you have a link with more details to point to ?
Posted by: vince at October 27, 2009 02:41 AM
Why assume that asset values grow smoothly at the same rate as income? I suppose its a scenario, but it just begs the question about asset values. Household net worth fluctuates due to saving and asset valuation changes. What reason is there to assume that an existing gap will be restored by smooth changes in income, asset values, and saving, and not by more abrupt asset valuation changes? The projection for required saving is as good as the assumption about asset valuation changes. Assuming a smooth trend is like assuming current valuation metrics are "normal" - a very fragile assumption.
Posted by: anon at October 27, 2009 02:58 AM
Note that I've omitted the statistical discrepancy which makes these items add up exactly.
That's okay. This is voodoo anyway. Pretty graphs though.
Oh, if my company buys a building is it savings? If I buy a building to live in is it savings? If my company buys a car is it savings? If I buy a car is it savings?
Posted by: RicardoZ at October 27, 2009 05:56 AM
The US matters, but primarily to the downside. Thus in oil, for example, half of global growth in demand is coming from China, the rest from the non-OECD countries. OECD oil growth (leaving aside the current recession) is negligible. So on the upside, China is the price-maker, the US is the price-taker. That's Part 1 of the US becoming a small open economy.
On the other hand, on the downside, the US is the price-maker, because China's oil demand has not collapsed, but had stopped growing. On the other hand, US demand collapsed, and this collapse drove prices until, say, mid summer. So, for Part 2, to the downside, the US is still a large open economy. It matters.
Blame on Beijing is not monolithic view, it is rather looking for a root cause for the rate of change. To the upside, China is the primary driver of demand and its liquid reserves, I would suspect, drive the upside (ie, speculative) change in prices.
Posted by: Steve Kopits at October 27, 2009 06:15 AM
Nice piece! I think it shows why a surge in consumer demand--and therefore "a return to the good old days"--is unlikely for some time. Frankly, I hope consumers are permanently smarter!
Posted by: Terry at October 27, 2009 06:27 AM
Is it me or does chart 6 showing the average ratio of net worth to income seem biased by the stock market bubble and the housing bubble?
Posted by: Chris at October 27, 2009 07:45 AM
In defense of the "Blame it on Beijing" wing, even if the Chinese economy is smaller, their savings rate is ridiculously high. The pool of savings is quite large when contrasted to the miniscule US savings rate.
Posted by: Buzzcut at October 27, 2009 09:08 AM
Buzzcut, is china's savings rate high, or is it wealth/income inequality?
"At no point did an ordinary Chinese person decide to send so much money to America. In fact, at no point was most of this money at his or her disposal at all. These are in effect enforced savings, which are the result of the two huge and fundamental choices made by the central government.
One is to dictate the RMB�s value relative to other currencies, rather than allow it to be set by forces of supply and demand, as are the values of the dollar, euro, pound, etc. The obvious reason for doing this is to keep Chinese-made products cheap, so Chinese factories will stay busy. This is what Americans have in mind when they complain that the Chinese government is rigging the world currency markets. And there are numerous less obvious reasons. The very act of managing a currency�s value may be a more important distorting factor than the exact rate at which it is set. As for the rate�the subject of much U.S. lecturing�given the huge difference in living standards between China and the United States, even a big rise in the RMB�s value would leave China with a price advantage over manufacturers elsewhere. (If the RMB doubled against the dollar, a factory worker might go from earning $160 per month to $320�not enough to send many jobs back to America, though enough to hurt China�s export economy.) Once a government decides to thwart the market-driven exchange rate of its currency, it must control countless other aspects of its financial system, through instruments like surrender requirements and the equally ominous-sounding �sterilization bonds� (a way of keeping foreign-currency swaps from creating inflation, as they otherwise could).
These and similar tools are the way China�s government imposes an unbelievably high savings rate on its people."
IMO, the fed LOVES this scenario (it also helps produce wealth/income inequality in the USA) and is cooperating with china.
Posted by: Get Rid of the Fed at October 27, 2009 09:52 AM
Thanks. It tells the story reasonably for household investment and savings and likely trajectories. But I do get the sense that corporate investment is going to lag for a long long time.
Posted by: david at October 27, 2009 11:20 AM
I'm curious whether anyone has looked at wealth/income ratios as a function of age, and whether demographic trends appear in the data...
Posted by: Demos at October 27, 2009 11:22 AM
Get Rid of the FED,
Do you really not understand that China's savings rate has actually been driven by the FED? If not why the name?
Posted by: RicardoZ at October 27, 2009 12:39 PM
member commercial banks do not loan out savings (just like they don't loan out excess reserves) savings held within the cb system are lost to investment that mistake was responsible for the beginning of stagflation
Posted by: flow5 at October 27, 2009 02:03 PM
RicardoZ said: "Do you really not understand that China's savings rate has actually been driven by the FED? If not why the name?"
Can you expand on that because I think we are "on the same page"?
Posted by: Get Rid of the Fed at October 27, 2009 02:33 PM
Very nice analysis, and I am particularly glad to see you say "what happens here is at least as important as what happens in East Asia (in contrast to some observers)". We clearly are the ones who have to fix the problems.
The 500% (or whatever) net worth to income number seems like a good starting point, but it needs to be refined in a dynamic model that contains age demographics and income groups. Both factors will increase the savings rate - my hypothesis. Its pretty clear that as people age they save more, and the US has a large cohort of baby-boomers who will increase their saving rate and their target wealth number over the next few years, regardless of the stock markets performance. Additionally, low income families with large credit card debt have really been burned in recent months. My hypothesis is that low income families of any age will increase their savings rate to first reduce their credit card debt and eventually have a savings cushion in the bank. Anecdotal evidence suggests to me that this is happening. Put the two of these factors into the initial analysis, and savings will be higher than it currently suggests.
Posted by: Mike Laird at October 27, 2009 02:44 PM
Is there any reason to believe that households target net worth to 550% of DPI? The data does not seem to be mean reverting. In fact it seems to be that it has been in a 35 year uptrend which probably not coincidentally marked the last really brutal stock market collapse.
Is there any reason to give this report credence?
Posted by: Chris of Stumptown at October 27, 2009 03:25 PM
I understand that averages of past behavior are often the only objective gauges available to estimate future behavior, but I think this one is particularly dubious. I'd be more interested in what non-economists have to say about the zeitgeist. My impression is that this recession ironically hit America at a time when its consumerist culture was peaking, and the last remnants of resistance to it in popular culture were dying out. There is a natural reaction that is due to come as baby-boomers start reaching retirement age, and fear of old age in poverty spreads among those who are today in their 40s and 50s, but I don't see that starting for another several years.
Posted by: Tom at October 28, 2009 04:11 AM
Get Rid of the FED,
The FED financed the import boom of the early new century. As dollars were transferred to China to purchase goods the Chinese had to find a resting place for their dollars. Their savings rate appeared to be huge because they were "investing" in US t-bills and Fannie and Freddie bonds.
The statistics of "savings" are a fantasy. They do not include real savings such as investments in plant and equipment, but do include investment related to governmental actions that are actually social spending. As with so many government statistics the "savings rate" is a Keynesian illusion that actually reports nothing but a conglomeration of figures with little meaning.
The FED has allowed the dollar to tank with much of this problem feeding the Chinese "savings" rate (not to mention the Real Estate and credit crisis). As China begins to wean itself off of US trade its reported "savings" rate will decline, while China itself invests in real savings continuing its climb to economic prominence.
Posted by: RicardoZ at October 28, 2009 05:32 AM
Get Rid of the FED,
It appears that you agree with the monetarist money cranks who believe that money can somehow create real goods and services out of thin air. When the money supply is increased it initially distorts the market and allows the misallocation or resources but just as the Japanese are not richer than the US because it takes more yen to buy something than one dollar, tomorrow's dollar that will buy half of what today's dollar could buy does not make an American richer. The dollar is an identity and changing its value only distorts and diminishes its usefullness.
Floating currencies do nothing but destroy by misallocating resources to monetary hedges. They create nothing.
Posted by: RicardoZ at October 28, 2009 05:45 AM
RicardoZ said: "The FED financed the import boom of the early new century. As dollars were transferred to China to purchase goods the Chinese had to find a resting place for their dollars. Their savings rate appeared to be huge because they were "investing" in US t-bills and Fannie and Freddie bonds."
It seems to me that the chinese savings rate is also high because the gov't/central bank does most of the "saving" while the workers are not getting paid enough (see my Oct 27 9:52 a.m. post).
It also seems to me that the fed was able to prevent price deflation from china's wto entry by getting enough lower and middle class people in the USA to go further into debt using housing as the asset to back up loans (the import boom was financed with debt, which is actually future demand brought to the present).
Posted by: Get Rid of the Fed at October 28, 2009 09:34 PM
RicardoZ said: "It appears that you agree with the monetarist money cranks who believe that money can somehow create real goods and services out of thin air."
Not really. I am trying to say that the central banks should NOT use debt (future demand brought to the present) to prevent price deflation from positive productivity growth and/or a cheap labor shock.
Posted by: Get Rid of the Fed at October 28, 2009 09:43 PM
@Chris of Stumpton
Great point. The vast majority of the graph in Chart 6 is under the mean line. (I rather suspect that a median line would have been a lot lower.) The part that is aove the mean line is two giant asset bubbles: the late 90's dot-com bubble and the mid 00's housing bubble. Is it really meaningful to include those periods when trying to decide on the level of assets with which households feel comfortable? There is a lot of anecdotal evidence that people went with negative savings rates during these two booms because they felt richer. In particular, the housing bubble was characterized by home equity lines, reverse mortgages, and the like.
Posted by: Maximum Liberty at October 29, 2009 04:08 PM