In this note, we use FRB-US to simulate a very strong real GDP growth rate in Q3. Compared to the baseline recently circulated, the forecast is not materially different because the model already expected a very strong growth followed by a contraction in Q4. As an additional exercise, we simulate a set of different shocks and compute the impulse response functions (IRFs) relative to the baseline. This exercise offers a list of “rule of thumbs” (given where the US economy is right now) to assess how the economy could react to a given shock.
The current baseline
The current baseline already projects a very strong GDP growth in Q3. On August 3, we circulated a FRB-US update (note here) showing the new baseline. The updated baseline (which, as usual, we labelled as the “inconsistent” FRB-US simulation) is shown in Figure 1 together with the June SEP. We do not discuss again the baseline and how/why it differs from the SEP because we have already done it on August 3. Nevertheless, given the current discussion about a possible (super) strong GDP figure in Q3, we point out that FRB-US expects real GDP growth (QoQ, ar) at 3.9% in Q3, followed by a (small) contraction in Q4. The path of real GDP growth (QoQ, ar) is shown in Figure 2. This put the 2023 Q4/Q4 growth rate at around 2%, as shown in the upper-right panel in Figure 1. Put it simply: if Q3 turns out to be very strong, there would be no big surprise for the model.
(For the record: we did not discuss the strong projected Q3 figure on August 3 because it looked a bit too strong back then. Ironically, it seems that the model might have a point..)
Figure 1. SEP forecast (blue line) and “inconsistent” FRB-US simulation (orange line)
Note: Real GDP growth and core inflation are expressed as YoY. Core inflation is core PCE price inflation. The blue line shows the latest SEP. The orange line shows the “inconsistent FRB-US” forecast (the current baseline), that is the model-based forecast removing the “add-factors” put by the Fed staff to match the latest SEP.
Simulating higher growth
Higher growth would not materially alter the baseline. Figure 2 shows real GDP growth (QoQ, ar) under the baseline presented in Figure 1 (the orange line). Figure 2 also shows the path of real GDP growth projected by FRB-US under the assumption that Q3 will print at 4.5% (the green line). In the model, this is achieved by the use of “add factors”, distributed across GDP components. The main takeaway is that the profile of real GDP growth is nearly identical after Q3. This happens because growth is already projected so high in the current quarter that the model does not take much signal from a higher print and continues to expect a sharp deceleration.
Figure 2. Real GDP growth (QoQ ar, %) in the baseline (orange) and higher growth scenario (green)
The path of the FF rate would also be similar to the baseline. Figure 3 shows the evolution of the variables of interest comparing the baseline (the orange line) with the “higher growth” scenario (the green line). The bottom line is that the evolution of the variables is not materially different. For instance, in both simulations the FF rate peaks in 2024Q1: at 5.76% under the baseline, and at 5.83% under the “higher growth” scenario. Therefore, unless one assumes that real GDP growth will print at (say) 6%+ ar, or the baseline we have circulated at the beginning of the month is still valid.
(If the reader is puzzled by the fact that core PCE price inflation in Figure 3 is nearly identical across scenarios, we have explained the mechanics here)
Figure 3. “Inconsistent” FRB-US simulation (orange – “baseline”) and higher growth scenario (green)
Note: the figure shows the evolution of the variables of interest under the current baseline (orange line) and under the “higher growth” scenario (green line). The baseline is also labelled “inconsistent” FRB-US forecast.
Shocking the economy with other shocks
In this section, we present the FRB-US impulse response functions (IRFs) in deviations from the current baseline. The idea is to offer the reader “rules-of-thumb” to keep in mind when thinking about shocks that could hit the US economy right now. All IRFs are expressed relative to the current baseline (last quarter in sample is Q2). IRFs are VAR expectations-based (pings.prg simulations).
Here are the shocks:
A 1pp increase of the FF rate. This shock lowers the output gap by about 0.5pp, while the impact on core PCE price inflation is small. Figure 4 below shows the IRFs of the output gap and core inflation, together with the simulated shock (right panel).
Figure 4. IRFs of output gap and core PCE price inflation to a 1pp shock to the FF rate
A 1pp increase of government spending. This shock raises the output gap by about 8 tenths and core inflation by a touch. IRFs here.
A 1pp increase of the equity premium. This shock (which in the model is associated with a 20 percent drop in market wealth) lowers the output gap by about 4 tenths. IRFs here.
A 6$ move in crude oil. This shock lowers the output gap by about 3bps, while the impact on core inflation is null. IRFs here.
A 1pp increase of multifactor productivity growth. This shock increases the output gap by 1 percentage point and lowers core inflation by 2 tenths (to rounding). IRFs here.
A 1pp increase of the term premium (10-year). This shock lowers the output gap by 7-8 tenths and lowers core inflation somewhat. IRFs here.
Implications for the FOMC and conclusion
A strong Q3 does not imply a different Fed. The conclusion of this exercise is that according to FRB-US, a strong real GDP growth print in Q3 would not alter the path of the FF rate relative to the current baseline. Core PCE price inflation continues to be projected around 2½ percent at the end of the medium-term. According to the model, the Fed is very careful, especially because the unemployment gap remains close to zero at the end of the medium-term. Overall, it is difficult to exclude that core PCE price inflation will return (on average) back to target without additional slowdown relative to the (already weak) baseline.