In the week ahead, the US September employment report is the high-frequency data highlight. It is the last nonfarm payroll report before the November 2-3 FOMC meeting, which—barring a significant downside surprise—will decide to begin the gradual reduction of the Federal Reserve's bond purchases. Although Fed Chair Powell has tried to distance policy from any single labor market metric in the past, he clearly identified the next nonfarm payroll report as a key to the tapering decision at his press conference following the September meeting.
He was quick to add that the bar is not very high; the report did not have to be spectacular. Several of his colleagues thought the conditions for tapering had already been met or will shortly. Indeed, the decision to taper could see one dissent, and that is from the Chicago Fed President Evans. Still, he may be persuaded to vote with the others if the statement explicitly recognizes that the conditions for a rate hike are still a ways off.
ADP's private-sector job estimate sorely missed expectations in August. The median forecast in Bloomberg's survey called for a 625k increase, but instead, it reported a rise of 374k. The ADP provides a reasonably good proxy for the BLS estimate over the medium-term but a relatively poor one in the short term. Consider that the official estimate is that the private sector created an average of 512k month through August. The ADP estimate has averaged 480k a month. That is a 32k average monthly mix, which seems rather small.
However, the fit in the short term is not nearly as tight. In the last three months, the ADP's estimate has also averaged 480k. The average BLS estimate of private sector employment over the three months through August is 616k. That average miss over the past three months is four times bigger than the year-to-date gap. The early estimates are for the ADP to show an increase of 450k private-sector jobs in September, matching the 450k median forecast in Blomberg's survey for private-sector job growth, as of October 1. Economists in the survey anticipate the government sector adding about 38k employees.
We have often acknowledged that the nonfarm payrolls report is among the most difficult high-frequency data to forecast. There are few reliable inputs, and the data set itself is volatile. The BLS also appears to find August particularly difficult, with the back-to-school effects and has frequently revised its initial estimate for August. The four-week moving average of weekly jobless claims fell by about 43k in the September survey period compared with the August survey period. Still, there was a jump in the September survey period that put the actual reading slightly above the nonfarm payroll survey week in August. Also, employment in the service sector fell below the 50 boom/bust level in the September PMI for the first time since June 2020.
Other elements of the employment report are expected to improve. The unemployment rate is expected to fall to 5.0% from 5.2%. It finished last year at 6.7%. It was at 3.6% in each month in Q4 2019. One of the reasons that the unemployment rate has fallen quicker than economists expected is that the participation rate remains low. It stood at 61.7% in August, but it had reached that last August after falling to almost 60% when the pandemic first struck. In H2 19, it averaged about 63.2%
The bottom line is that the September employment report is unlikely to detract the Federal Reserve from reducing the pace of the Treasury and Agency MBS bonds in November. We know the tapering will be completed by the end of H1 22. That works out to be an average of about $15 bln slowing a month, but it could be front-loaded a bit to allow proper tapering. Moreover, the November start date corresponds to when the Treasury is expected to announce a reduction in its borrowing requirements.
We anticipated that to preserve the Fed's maximum flexibility, it needed to allow for two rate hikes next year, not one, even though the latest "dots" showed Fed officials split on even one hike next year. This meant that the Fed would need to finish the tapering by the middle of the year. We also thought the market would have to discount a September 2022 move, but it had already priced in a December move. The market has now completely discounted a September hike (fair value with a 25 bp rate hike is about 15.5 bp, where it settled before the weekend).
The December 2022 contract is interesting. If there is a hike in September and then the Fed stands pat, the fair value is 33 bp. If there was no hike in September but one in December, the fair value is about 22 bp. However, if the Fed delivers the first hike in September and another in December, the fair value is 47 bp. Last week, the contract settled at 27.5 bp, suggesting that the market sees about a 20% chance of a second move next year. This seems reasonable. Given our current information set, it is less reasonable to be 100% confident of the second hike. The 2-year yield of 26 bp appears out of line if the December 22 Fed funds futures contract implies 28 bp.
Next week, other high-frequency data include the final service and composite PMI readings, eurozone August retail sales (which likely recovered after a 2.3% decline in July), and August household spending in Japan. However, the data does not appear to be the likely driver of the price action. The key right now is policy, which seems rather indifferent to the data. The ECB is expected to allow the Pandemic Emergency Purchase Progam to wind down at the end of Q1 22. The policy focus in Japan is on the size and content of the fiscal stimulus package the LDP is advocating before national elections, likely around the middle of Q4.
The Antipodean central banks meet next week. At its last meeting, the Reserve Bank of Australia went forward with its tapering (A$4 bln from A$5bln) but extended the buying three months into February 2022. The recent survey data warns that the Australian economy was still likely contracting in September. The August jobs report was considerably weaker than expected (with a loss of 146k jobs rather than an 80k decline as the median forecast in Bloomberg's survey anticipated). Australia reports August trade figures next week too, and this is one of the bright spots. A record trade surplus (A$12.1 bln) was reported for July, boosted by rising prices.
The Reserve Bank of New Zealand is the more interesting of the two. Most likely, it will join Norway in being the first of the high-income countries to raise rates. Growth accelerated in Q2 to 2.8% quarter-over-quarter. It was twice the pace of Q1 and well above the median forecasts for a 1.1% expansion. New Zealand economic data has been consistently better expected. Still, speculation of a 50 bp initial move did not seem realistic, and unwinding may have weighed on the New Zealand dollar. Still, a follow-up rate hike in November is possible depending on the data, but over the next 12-months, the market appears to be pricing nearly five quarter-point rate hikes.
Lastly, we note the progress on the global corporate tax reform. Recent G7 meetings reportedly have resolved some of the remaining issues. An agreement among the OECD/G20 Inclusive Framework members could be announced as early next week. A little more than 130 countries have signed on to a worldwide minimum corporate tax of 15%, and that companies with large internet sales would pay some (20%-30% above a 10% margin) of their taxes based on where those sales occur. Among the holdouts are Ireland and Hungary. At the earliest, the late October G20 meeting in Rome will begin to finalize the agreement, which won't be implemented until 2023. Even though the shift in the US official position was a critical breakthrough in these negotiations taking place for several years, it is not clear whether the US Congress has the will to adopt its own tax code in line with the international agreement. Nor can the Senate's approval of a new tax treaty that entails rules about the reallocation of revenue be regarded as a sure thing.