GDPNow was -1.6%, the St. Louis Fed News Index was +4%, IHS-Markit was -2.0%, and Goldman Sachs was 0.5%. Bloomberg consensus is +0.4% (all q/q SAAR).
figure 1: Real Advance GDP Growth Rate (black), Atlanta Fed GDPNow (red line), St. Louis Fed News Index (cyan line), IHS Markit (sky-blue triangle), Goldman Sachs (blue+), Bloomberg Consensus (pink open square), All in %, with NBER-defined peak-to-valley decline dates shaded in gray. sal. Source: BEA (via ALFRED), Atlanta Fed, St. Louis Fed (via FRED), IHS-Markit, Goldman Sachs, Bloomberg, and NBER.
In the three-year sample (apparently a non-representative sample given the impact of the pandemic), GDPNow is skewed upwards, while nowcasts based on the St. Louis Fed news are skewed downwards, by 0.8 and 0.9 percentage points, respectively. On the other hand, the RMSFE of the former is about 2 percentage points, and the RMSFE of the latter is about 5.9 percentage points. If you drop in Q1-Q3 2020, GDPNow has a smaller downward deviation than the St. Louis Index, and RMSFE is smaller (1.5 vs. 2.5 percentage points).
Having said that, recall that these nowcasts are intended to be estimated ahead of time. The 7/28 version of 2022Q2 is pre-released with estimates for many components of GDP. Estimates will be revised, sometimes significantly, especially in the annual benchmark (which happens in September instead of July this year) and the five-year comprehensive revision (see discussion in this article) postal). Then, positive readings can eventually become negative and vice versa.
