Jan Hatzius Friday Night Bomb: "We Are Downgrading Our Real GDP Growth Estimate To 1¾% From 2½%"

Nobody could have seen this coming: "With most of the news on first-quarter growth now in, the GDP “bean count” looks even softer than it did a couple of weeks ago. The most recent disappointments have come on the export side—with trade now set to subtract significantly from growth in the quarter—and from inventories. Consequently, we are downgrading our real GDP growth estimate to 1¾% (annualized), from 2½% previously (and from 3½% not too long ago)." Some other things nobody will be able to predict: Hatzius dropping full year GDP from 4% to 2.25%; Goldman's downgrade of precious metals, Kostin's 2011 S&P 500 price target reduction by 20%, and Goldman getting its New York Fed branch to commence monetizing $1.5 trillion in debt some time in October.

From Goldman: Do Consumers Have Enough Fuel?

  • With most of the news on first-quarter growth now in, the GDP “bean count” looks even softer than it did a couple of weeks ago. The most recent disappointments have come on the export side—with trade now set to subtract significantly from growth in the quarter—and from inventories. Consequently, we are downgrading our real GDP growth estimate to 1¾% (annualized), from 2½% previously (and from 3½% not too long ago).
  • Other indicators still point to solid activity in Q1, but markets have become increasingly concerned about growth in the remainder of the year as well. A key reason for concern is the sharp rise in gasoline prices so far in 2011—nearly 70 cents per gallon—which is siphoning off household income at a run rate equivalent to $100 billion per year. We are adjusting our headline inflation forecasts over the remainder of 2011 to take the surge in fuel prices into account.
  • Despite these higher fuel costs, consumer spending looks to have grown at a 2½% pace in real terms in Q1, and—given strength towards the end of the quarter—is headed for a stronger pace in Q2. An important reason for the resilience: the payroll tax holiday has helped consumers to absorb the increase in gasoline prices over the past few months. (Put another way, higher oil prices have fully offset the impact of the payroll tax cut.)
  • Going forward, our forecasted reacceleration in spending growth still looks possible, but will require a fortuitous combination of circumstances—a modest further pickup in the labor market, gasoline price relief, and a benign asset price environment that encourages consumers to gradually reduce saving.

With most of the news on first-quarter growth now in, the GDP “bean count” looks even softer than it did a couple of weeks ago. The most recent disappointments have come on the export side—with trade now set to subtract significantly from growth in the quarter—and from inventories. Consequently, we are downgrading our real GDP growth forecast to 1¾% (annualized), from 2½% previously (and from 3½% not too long ago).

Other indicators still point to solid activity in Q1, but markets have become increasingly concerned about growth in the remainder of the year as well. Growth-sensitive equities have suffered in recent days, and some forecasters have taken down their expectations for growth later in the year. A key reason for concern is the sharp rise in gasoline prices so far in 2011, which has the American public—and policymakers—on edge. Retail pump prices are approaching their peak levels in the summer of 2008 (Exhibit 1). The extra cost of about 70 cents per gallon, relative to prices at the end of 2010, is siphoning off household income at a run rate equivalent to $100 billion per year—income that otherwise could have been spent on other goods and services.

Despite these higher fuel costs, consumer spending looks to have grown at a 2½% pace in real terms in Q1, and—given strength towards the end of the quarter—is headed for a stronger pace in Q2. Solid growth in consumer spending will be essential if the US economy is to post above-trend growth for the remainder of 2011, as we continue to expect. But will households have enough “fuel” from income growth to sustain such an expansion, especially with fiscal and monetary stimulus reaching their peak?

In the next few pages we look at the prospects for household income growth in 2011 and 2012. We find that the payroll tax holiday has helped consumers to absorb the increase in gasoline prices over the past few months. Put another way, higher oil prices have fully offset the impact of the payroll tax cut. Going forward, a reacceleration in spending growth is possible, but will require a fortuitous combination of circumstances—a modest further pickup in the labor market, gasoline price relief, and a benign asset price environment that encourages consumers to gradually reduce saving.

Moderate Income Growth, with Risks from Taxes and Oil

Our US economic forecast envisions personal income from wages and salaries, assets, and transfers should grow at roughly a 5% nominal rate through most of 2011 and 2012.

1. Wages and salaries should grow at a 4%-4½% clip. This assumes payroll growth in the 200,000 range (just about a 2% annual growth rate—see Exhibit 2), a small increase in hours per worker, and growth of about 1½%-2% in wages per hour (similar to recent growth in private sector average hourly earnings or the Labor Department’s employment cost index; see Exhibit 3). Ultimately, it’s labor income that is needed to fuel a self-sustaining expansion, and this is more important than ever now that fiscal policy is turning towards restraint.

2. Asset income—weak interest, but growing dividends and business income. A continued low-rate environment should dampen interest income, but the rebound in the economy should lead to further gains in dividend income and small business income (proprietors’ income). We envision this component growing at a 3-5% nominal rate through 2012.

3. Transfer income will be more restrained. Aside from the annual cost-of-living increase in Social Security in early 2012, which should be more robust next year due to higher headline inflation this year, transfer income should grow relatively slowly. In particular, unemployment benefits should dwindle as individuals find jobs or exhaust their extended benefit eligibility.

Modeling each component of income growth separately suggests some downside risk to asset income and transfer income relative to our current forecasts, but potential upside risk to our current numbers on wages and salaries. Overall, we see some small downside risk to our current disposable income forecast, perhaps about half a percentage point. If we assume trend-like headline inflation of 1½%-2% over remainder of 2011 and 2012, these calculations would imply real disposable income growth in the 3% range.

Exhibit 4 illustrates the recent paths of wage and salary income, disposable income, and real disposable income with our forecasts (in shaded area) through the end of 2012. The spike in mid-2010 is due to the labor market improvement in that period, which in turn was partly the result of temporary Census hiring. The data for recent months show clearly the offsetting effects of the payroll tax holiday and rising gasoline prices. Wage and salary growth (dotted line) has been reasonably steady in the 3%-4% range, while disposable income (the gray line) has accelerated to more than 6% annualized with the cut in payroll taxes. However, in real terms (black solid line), there has been no acceleration in income, as higher headline inflation has absorbed the increase in nominal aftertax income.

As for the future, there are two main risks to a “steady as she goes” income path. The first—fittingly, given that it’s tax day—is the increase in payroll tax rates slated for the beginning of 2012, when the partial payroll tax holiday expires. This will decrease households’ after-tax income by roughly $110 billion (about 1%), clearly visible as a drop in income growth in early 2012 in Exhibit 4. Of course, it’s possible the payroll tax cut will be extended—next year is an election year, after all—but right now there is no call to do so either from Democrats or Republicans.

The second risk is the path of commodity prices; continued increases in gasoline prices in particular would pose a serious threat, especially if they occurred alongside a reversion to the higher payroll tax rate. Households currently devote 3.6% of their income to gasoline, on average, so a 10% shock to gasoline prices is worth 36bp on real disposable income growth. This is only a “first-round” effect, and leaves out any feedback into employment (i.e. if lower spending caused companies to become more cautious on hiring, that in turn could affect future spending) or via other sectors of the economy.

The bottom line: we see modest downside risk (unfortunately, a phrase we have been using a lot lately) to our household real disposable income forecasts in 2011 and 2012. The best chance for exceeding our forecasts is either a substantial acceleration in the labor market and/or a large drop in gasoline prices.

Will Consumers Loosen the Purse Strings?

Our forecast has real consumer spending growth at a brisk 4% pace in Q2 and Q3, decelerating to 3.5% late in the year and to 3% by late 2012. Given the more modest path for real disposable income discussed in the previous section, this implies a drop of somewhere between one and two percentage points in the household saving rate by the end of 2012. This would be a meaningful loosening of the purse strings, though mild by the standard of either of the last two economic expansions.

To test the plausibility of such a drop in saving, we update our model of the household financial balance. This measure equals after-tax household income less consumer spending and net residential investment. It is a broader measure of households’ financial stance than the saving rate alone. Since households think about home purchases and renovations as part of their spending, we think it makes more intuitive sense, and it also turns out that we can fit models to it with slightly more accuracy. Statistically, the key drivers of the household financial balance are 1) asset prices—higher asset prices are associated with a lower balance, i.e. more spending and investment, 2) credit conditions—with easier credit also associated with more spending, and 3) nominal interest rates—with lower rates typically discouraging saving and boosting spending.

Our model, illustrated in Exhibit 5, is estimated only on data through 2005, but has continued to track actual behavior quite closely since then: the tightening in credit and collapse in asset prices beginning in late 2007 are consistent with the observed sharp rise in the financial balance. However, the latest improvement in the financial environment—particularly the rally in the equity market over the past several months—suggests that consumers may be willing to “loosen the purse strings” at least somewhat in 2011. The forecast for the financial balance in the remainder of 2011 and 2012 is about two percentage points below the current actual level, implying a desire by consumers to spend a greater fraction of their after-tax income. Note this forecast is contingent in part on a continued rise in equity prices (per our strategists’ forecasts) and gradual easing in credit conditions. Flat equity prices would still imply a decline in the household balance, but a somewhat smaller one.

Where does all this leave us? The models suggest that our income and spending forecasts are feasible and internally consistent. But they also suggest a lot of things will need to go right for our optimistic view on spending to become a reality. First, the labor market will need to continue its improvement, and probably accelerate slightly, to provide the requisite income growth. Second, gasoline prices need to stop rising, and ideally retrace at least part of their recent gains, to ensure that income growth passes through into increases in real spending. Third, overall asset values need to rise—i.e. equity price gains need to more than offset modest home price declines—to ensure households feel comfortable loosening the purse strings. Finally, of course, households need to behave roughly in the way our model suggests they should!

Cash Flow Growth is Healthy

One other perspective on households’ spending power is provided by our measure of “consumer discretionary cash flow”. To calculate this, we take the estimates of disposable income from the previous section, net out non-cash income, add cash flow from borrowing or asset sales, and subtract essential outlays for food, energy, medical care, and financial obligations. Finally, we deflate the remaining series using an adjusted core PCE price index.

This approach paints a somewhat more optimistic picture (Exhibit 6). Near term, cash flow grows more strongly than income. The faster growth of cash flow occurs mainly because recent data on credit extension suggest a noticeable acceleration—in particular, nonrevolving consumer credit (auto loans) has grown steadily over the past six months after declining gradually over the prior two years. We expect this positive “credit impulse”—a positive second derivative of credit outstanding—to persist through most of 2011. That in turn would be consistent both with continued growth in consumer spending and a decline in the saving rate (and household financial balance).

A Divergent Impact Across Households

It’s worth noting that the broad macro themes outlined here have very divergent implications across households. Households with high exposure to equity prices—typically those at the top end of the income spectrum—have become more willing to spend as their net worth recovered quickly following the crisis. Households with relatively more exposure to housing, and/or who spend a higher proportion of their income on gasoline—typically those at the lower to middle income brackets—continue to feel considerable pressure to economize.

Using data on relative exposure to gasoline costs from the Labor Department’s Consumer Expenditure Survey, and to asset prices from the Fed’s Survey of Consumer Finances, Exhibit 7 illustrates the hypothetical impact of changes in asset values and oil prices since 2005 on spending by the top, middle, and bottom income quintiles of US households. The concentration of the negative “wealth effect” among higher-income households is consistent with the sharp drop in luxury spending and disproportionate damage to higher-end retailers in the early part of the crisis. Since then, spending at the higher end seems to have recovered more rapidly, consistent with the implication of the chart. (The payroll tax cut had a broadly similar effect across most households, except among the top quintile where it represented a smaller percentage change in after-tax income.)