Channel checks and earnings prep: the bar before the print
Next quarter estimated from suppliers, customers and peers, set against consensus, with a prep sheet three days before each print and a post-mortem after.
Request access| NVIDIA revenue (customer) | +85% YoY | lag 2q |
| Alphabet capex (customer) | +107% YoY | lag 2q |
| Alphabet revenue (customer) | +24% YoY | lag 1q |
| Microsoft capex (customer) | +84% YoY | lag 2q |
Consensus revenue is 3.77B (40.7% YoY, 23 analysts) while our chain-implied is 3.33B, a -11.5% gap (z=-3.25) with 5pp growth backtest error that beat naive in 75% of 12 quarters. The real bar is a 3.33B print (or 3.17-3.50B range), as positive customer revenue growth (AMZN +19.6%, META +28.0%, MSFT +17.7%, GOOG +24.2%) and read-across are already incorporated yet still yield the downside gap vs consensus.
- 1Q3 revenue actual vs 3.33B chain-implied and 3.77B consensussettles the -11.5% chain-consensus revenue gap and backtest accuracy
- 2EPS print vs 1.83 consensus and magnitude of beattests continuation of the +4.9% to +25.0% historical EPS surprises
- 3Commentary on demand/capex flows from Alphabet, Microsoft, Amazon, Meta, CoreWeavevalidates positive chain signals and read-across transmission to Vertiv
Next quarter, from the chain
Customers' capex and revenue, suppliers' sales and peers' prints are turned into an estimate of a company's next-quarter revenue. Each driver is tested walk-forward on past quarters using only what was public at the time, and kept only if it beat simply assuming growth continues. The result is set against consensus as a variant gap with its own error band.
Honest by construction
Where no supplier, customer or peer beats the naive forecast, rejo says “no chain signal”. Where consensus implies a step change (a divestiture or acquisition the chain can't see), the gap carries a caution.
The prep sheet
Three days before a print of a company you follow: consensus and its revisions, the surprise history, the chain's estimate, what linked companies already reported this season, and for each thesis what a beat or a miss does to its probability. A short memo sets the real bar and what to listen for, citing only those inputs.
The post-mortem
After the print, the actual is scored against consensus and against the chain's bar, and the track record of the method grows quarter by quarter.