Investor Behaviour · Volatility · Asset Allocation
The Hinge In Every Forecast
Macro forecasts are conditionals, and the conditions attached to them carry more information than the verdict.
Crude oil rose to USD 90 a barrel in July as fighting in West Asia disrupted traffic through the Strait of Hormuz. Before the month was out, both sides had signalled that a deal was possible and the price had come off its peak. Same commodity, same month, two opposite stories. Anyone who formed a firm view in the first fortnight had it taken apart in the second.
That is the ordinary texture of macroeconomics, and it is why an honest monthly review of the economy is written almost entirely in conditionals. Growth should hold up, provided demand momentum sustains and the monsoon behaves. Inflation should rise on base effects, provided it stays inside the central bank's tolerance band. Pressure on the trade deficit should ease, provided export growth holds and commodity prices fall. Each sentence has a hinge in it.
The hinge is the part worth reading
Most readers skip it. The eye goes to the verdict, and the conditions attached to it get treated as small print. That reads the document backwards. The verdict is the author's central case, which is a guess. The conditions are the list of things that would have to be true for the guess to hold, which is information.
A forecast is a sentence with a hinge in it. The verdict is the guess. The conditions are the information.
Read July's conditions and the short list is the monsoon, the conflict in West Asia, the price of crude, base effects in the inflation series and the pace of government spending. Nobody on earth knows how the first two will turn out, and the review says so. A reader who carries away the verdict alone has carried away the least durable part of the document.
Fast variables and slow ones
Notice which variables did the talking in July. A shipping lane, a barrel of oil, a truce. All of them moved inside a few weeks and can reverse inside a few more. Now notice which ones got a single line each. Total government expenditure grew 11 per cent year on year in the first three months of the fiscal year, with capital expenditure growing faster still. Merchandise exports grew 16.1 per cent year on year in the June quarter, against 1.4 per cent a year earlier. Slow numbers, long consequences, no headlines.
The same month also refuses to point one way:
| Indicator | Unit | June 2026 | July 2026 |
|---|---|---|---|
| Manufacturing PMI | Index | 54.2 | 53.5 |
| Services PMI | Index | 57.4 | 53.3 |
| Power demand | YoY, % | 10.9 | 11.1 |
| Gross GST collection | YoY, % | 13.9 | 15.4 |
| Unemployment | % | 6.6 | 6.2 |
Two survey measures moderated while electricity consumption, tax collections and the unemployment rate all improved. Every reading is accurate. One month is a set of observations, and a set of observations is not a direction.
The practical use of such a document is calibration. It shows what changed, what is being watched, and how firmly anyone is prepared to state a case. Converting that into conviction about the next twelve months asks more of it than it can carry.
A portfolio that needs one branch of the conditional to occur is a position on that branch. A portfolio built to survive either branch does not need the forecast to be right. Reading the conditions is how an investor tells which of the two they own.