PMI and the high-frequency dashboard

The last module left us with a problem. GDP, the official growth number, is late and provisional — it tells you about a quarter that already ended. If you want to know how the economy is doing right now, waiting weeks for the official figure is no help at all.

So analysts built a workaround: a dashboard of fast, frequently-released signals that stand in for activity before the official data is ready. A factory survey out within days of the month ending. Tax collections published monthly. Trucking permits, electricity generation, car sales, fuel demand, digital payments — each a small window onto the economy, each available quickly. Stitched together, they give a real-time read that GDP can only confirm later.

This is genuinely useful, and it is genuinely easy to over-read. This module teaches you what these fast indicators are, why they are faster, and — the harder lesson — what their speed costs you in reliability.

The single most-watched fast indicator is the PMI — the Purchasing Managers' Index. It is not a count of output; it is a monthly survey of businesses asking whether things like new orders, production and hiring got better or worse than the previous month. The answers are boiled down to one number around a pivot of 50: above 50 means more firms report improvement than deterioration, below 50 means the reverse. There is a separate PMI for manufacturing and for services, and because it is a survey rather than a tally, it can be published within days of the month closing — long before any hard data.

Around PMI sits a whole panel of high-frequency indicators — fast, frequently-released proxies for activity used to read the economy before the official series arrive. The important ones for India, each a different window:

• GST collections — monthly goods-and-services-tax receipts, a proxy for the rupee value of transactions across the economy.
• E-way bills — the electronic permits generated when goods above a value move between places; a fast read on trade and logistics volume.
• Power demand — electricity generation and consumption, which rises and falls with factory and commercial activity.
• Auto sales — monthly vehicle dispatches and registrations, a classic pulse of consumer and rural demand.
• Fuel consumption — diesel and petrol offtake, tracking transport and industrial use.
• UPI and digital payments — the volume and value of digital transactions, a near-instant read on spending.

None of these covers the whole economy. Each is a partial, sometimes quirky proxy. But together they arrive weeks ahead of GDP and let you form a view of the present while it is still the present.

The right mental model is a dashboard of gauges, each reading one part of the machine, feeding a rough picture of how the economy is running now. Analysts sometimes call this a "nowcast" — an estimate of the present rather than a forecast of the future. GDP arrives later and confirms, or corrects, what the dashboard suggested.

Two features of the dashboard decide how to use it. The first is that most of these are coincident indicators — measures that move at the same time as the economy, describing the present rather than forecasting the future. They tell you what is happening now, quickly. They do not tell you what happens next; a firm dashboard this month can roll over the following month. The second is that each gauge is partial and noisy: one month of PMI, or a single GST figure flattered by higher prices, can mislead. The dashboard is trustworthy as a pattern across several gauges over several months, and unreliable as a single reading on a single month.

That is why the discipline is to read the dashboard as a whole and as a trend. When PMI, power, e-way bills and fuel are all pointing the same way for a few months, the signal is worth something. When one gauge jumps for a month while the others are flat, it is usually noise wearing the costume of news.

Take one strong-looking signal and read it honestly. illustrative

GST collections come in sharply higher than the same month a year ago, and the headlines call it proof of a booming economy. Before accepting that, a careful reader runs two adjustments. First, prices: GST tracks the rupee value of sales, so if prices rose over the year, collections rise even if the quantity of goods sold was flat. Part of the "boom" may simply be inflation passing through the till. Second, the base: if the year-ago month was unusually weak, this year's growth rate looks large for that reason alone, with nothing to do with current strength. After both adjustments, the strong jump may soften into something ordinary — or it may hold up, in which case it is more convincing. Either way, the nominal headline was not the answer; it was the starting point.

Now widen the view. A single firm GST print means little on its own. But if, over three months, GST is firm and PMI is above 50 and power demand is rising and e-way bills are climbing, the gauges are corroborating one another, and the picture of a strengthening present becomes hard to dismiss. That corroboration across independent windows is the real value of the dashboard — not any one number, but several partial views agreeing.

The high-frequency dashboard is the best tool there is for reading the economy in real time. Its speed comes at a price, and the price is exactly where beginners get caught.

It cannot forecast. These gauges are coincident — they describe the present and can turn without warning. The dashboard helps you judge, roughly, where in the cycle you are now; it does not tell you where the cycle goes next, and honest reading keeps those two apart. A strong "now" is not a promise about "next."

It cannot be trusted as a single number. One month of any gauge is noisy, and several are nominal — flattered by prices and base effects. A rising collections figure is a claim about value, not volume; strip out inflation and the base before believing the boom. Read the trend across several gauges, not the spike in one.

And it cannot tell you what the market will do. A widely-followed dashboard is watched by everyone, so its signal is often already in prices by the time you read it. Understanding the present is not the same as having an edge on the future. Use the dashboard to know where things stand — not as a set of buy and sell buttons.

The dashboard's speed makes it seductive, and seduction is where the errors live. The common ones:

The habit to carry: treat the high-frequency dashboard as a set of windows on the present, read together and over time, and never as a forecast or a trigger. It answers "how is the economy doing right now?" better than anything else available — and it deliberately does not answer "what should I do about it?"