A question that comes to mind: can affordability create jobs? If Fannie and Freddie buy their own mortgage-backed securities (MBS), the impact on the economy would be like a past quantitative easing (QE) operation.

In several studies, the Fed estimated that quantitative easing increased bank lending, especially mortgage and commercial. Regions with banks heavily exposed to mortgage bonds (MBS) purchases saw stronger employment gains.

As those purchases lowered the mortgage rate by 30 to 50 basis points, the positive effect of mortgage refinancing on consumption, the “affordability effect,” translated into a ~0.08 to 0.15 percentage points reduction in the unemployment rate.

Indeed, since QE1 in 2009, employment improved, especially in areas of the counties where mortgage refinancing affected the local economy (Figure 1 from the study). Thus, macroeconomically, the President’s quest for affordability may affect non-farm payrolls in the future.

Figure 1: MBS purchases effect on employment (%)

Source: Federal Reserve

To that end, a $200 billion MBS program is akin to Fed operations seen in 2009-2013. While that was a different time, with the unemployment rate materially higher (8% to 10%) and the economy suffering from consumer and bank balance-sheet deleveraging, the impact on the payrolls “break-even rate” can nonetheless be the same.

The break-even point, minimal payroll to keep the unemployment rate stable, has fallen to a range of 25 to 45K, according to various Fed estimates. Historically, during 2009-2013, when most MBS purchases occurred, the average payroll was 147K, and the break-even was around 100-125K.

While it would require a material increase in payroll growth of at least 150-225K/month to get to a 100K+ break-even from current low levels, the impact of $200 billion MBS purchases will nonetheless somewhat contribute.

That said, between 2009 and 2013, the Fed bought ~$1.89 trillion in MBS, with an average impact of ~0.1-0.15% on the unemployment rate, and payrolls reversed from -700K in 2009 to an average of 150K.

The buying capacity of Freddie and Fannie portfolios is ~138bn and ~106.7bn, respectively, while the Fed’s reinvestment of MBS runoff is around $190 billion annually ($15 to $25bn/month), which is used to reinvest in Treasuries. These de facto offset the buying by Freddie and Fannie, albeit not entirely, as Treasury yields are also affected.

Thus, the historical relationship between mortgage rates and unemployment rates is valid; mortgage rates decline, and eventually unemployment rates can follow (Figure 2).

In the pre-market, MBSs are a touch tighter in spread, homebuilding stocks are up, and consumer discretionary stocks are rallying by 2%, reflecting triangulation on the next presidential intervention.

Figure 2: Mortgage rates and unemployment rates (%)

Source: Freddie, Fannie, BLS

For the payroll number, the range is 70K-155K, with the lowest at 23K at the current break-even rate. Leisure, hospitality, and other services are seeing a decent rebound since the shutdown.

Hence, there could again be an upside surprise because most job indicators, ISM/regional PMI, hiring intentions, layoffs, claims, and openings point to modest labor momentum. The FedWatch NFP model “predicts” 144K job gains, the higher end of the range (Figure 3).

For this report, however, the focus is on how the recent productivity surge is affecting labor force participation, which has been rising since September, despite AI’s impact and the government shutdown. Labor participation may increase again, which could put upward pressure on the unemployment rate for good reasons.

That said, the AI replacement factor—companies reducing workforce by implementing AI —has been averaging 5K job losses and has affected the non-farm payroll break-even. On the other hand, data centers have, on average, added 6 to 9K jobs since the summer, according to the Census estimates.

The NFP is expected to be decent, which solidifies the likelihood of the Fed skipping the January meeting.

Figures 3: NFP prediction

Source: BLS, FedWatch

Figure 4: Productivity impact.

Source: BLS