Thirty tiles, two questions

Two questions, thirty tiles

Every tile on a macro page is answering one of two questions: is money getting cheaper or dearer, and is capital moving toward risk or away from it. The dollar, US yields, crude, credit spreads and foreign flows are five different ways of asking them.

That is the whole map. Hold it in mind and a screen of thirty numbers stops being thirty things to track.

The reason it works is that, reduced to its arithmetic, a share price has two moving parts: the cash a business is expected to produce, and the rate at which those future rupees are discounted back to today. Nothing in global macro reaches an Indian company any other way. It either changes what the company will earn, or it changes what a rupee of future earnings is worth now — and the discounting half is the one most people under-weight, because it is invisible in the quarterly results.

So the test for any tile is a sentence, not a chart: state the channel by which this number reaches an Indian company's cash flows or its discount rate. If you cannot finish the sentence, the tile is decoration for you today, however important it is in principle. The five channels below are the five sentences worth learning.

Green means helpful, not higher

Before anything else, the convention that trips up nearly everyone on their first visit.

A tile's colour shows the effect on Indian equities, not the direction of the number itself. A falling dollar index shows green. A rising dollar index shows red. Crude down is green; crude up is red. The colour is a translation, and the thing it translates into is always the same: is this, mechanically, a tailwind or a headwind for Indian shares.

Two tiles make this genuinely counter-intuitive rather than merely unfamiliar. US Treasury yields going up shows red, even though a higher yield is usually reported as good news for savers. And USD/JPY going up — a weakening yen — shows green, because a weak yen means the borrow-cheap-yen-and-buy-risk trade is intact. That is the opposite of the reflex that a falling currency is bad news.

The specific mistake to avoid is reading a row of green as a row of rising prices. It is not a price screen. It is a screen of already-interpreted signals, and the interpretation is ours — which is the second thing to know about the page, and the subject of the section on the score.

The five channels

Here is the map in one table. Each row is a channel, and the third column is the sentence that has to be true for the tile to matter to you.

The rest of the page — gold, copper, the Nikkei, the DAX, S&P and Nasdaq futures — is context and confirmation. Useful, but each of those is a second read on a channel already covered rather than a sixth channel of its own. Copper says something about global growth; the Nikkei says something about the yen. Neither reaches India by a route the five above have not already described.

Channel one — the dollar

Most of the world's cross-border borrowing is written in dollars, which makes the dollar's price a global setting rather than one currency pair among many.

When the dollar strengthens, three things happen at once to an emerging market. Dollar debt gets harder to service in local-currency terms. The return a foreign investor earns in India has to survive conversion back into a currency that is now worth more, so the same rupee gain buys fewer dollars. And India's import bill — crude, electronics, edible oil — is denominated in the currency that just went up.

None of that is a forecast about where the dollar goes. It is the channel, and the channel runs in both directions: a softening dollar loosens the same three constraints. What you can take from the tile is the direction of pressure on the other four channels, which is why the dollar sits at the top of the page.

It also carries the largest weight in our composite score, 0.20 — and that is a judgement we have made, not a measured constant. It reflects our reading of what has mattered most to Indian equities. Someone with a different reading would weight it differently and would not be wrong on the arithmetic. Treat the number as a stated opinion with a formula attached.

USD/INR is on the same channel but tells you something narrower: how much of the global move has already been transmitted into the rupee. EUR/USD is a cross-check, since the euro is the largest single weight in the dollar index — a move in the index that EUR/USD does not mirror is usually telling you about the yen or sterling instead.

The four channels inside this one channel, and the leg that catches people out — that a foreign investor takes the market's fall and the rupee's fall multiplied together, while a domestic investor takes only the first — are worked through in the dollar index and emerging markets.

Channel two — the rate the world discounts against

A bond's yield is the annual return a buyer earns holding it to maturity at today's price — the arithmetic is in yield to maturity explained. The US 10-year Treasury yield is the one almost every other asset in the world is priced relative to. Two mechanisms run off it, and they are separate.

Discounting. A share is a claim on cash a long way out. Raise the rate those distant rupees are discounted at and their present value falls — and it falls most for the businesses whose earnings are furthest in the future, so the present value of a richly valued growth company moves more on a given yield change than that of a mature cash generator. This is the same arithmetic that makes a long-dated bond fall further than a short-dated one for the same change in rates, set out in duration and interest-rate risk. Equities have a duration too; it is simply never printed on the factsheet.

The gap. Foreign capital allocated to India is, among other things, taking the difference between what Indian assets yield and what the same money earns at home, less the currency risk. Narrow that gap by raising the US end and the trade is worth less. Whether it is worth enough is a question about the rate at which risk is priced, not a question we answer here — the underlying idea of what a rate is and what moves it is in interest rates explained, and both mechanisms are traced from the US yield to the Indian share price in US yields and Indian share prices.

The 2-year yield is the market's read of the policy path; the 10-year blends that with growth and inflation expectations; the 30-year adds the compensation demanded for holding duration at all. The 2s10s curve is just 10-year minus 2-year, and an inverted curve — the shorter yield above the longer — means the market is priced for rates to be lower later. That is a statement about what is priced, which is not the same as a statement about what will happen. We report the level and the sign. We do not forecast from it, and neither should the page be read as doing so.

One practical note: the 2-year tile is sourced from FRED, whose daily series typically publish a session or two behind. A tile that has not moved may simply not have updated. The same applies to both credit-spread tiles.

Channel three — energy, the one that lands in the accounts

India buys the large majority of its crude from abroad, which makes the oil price a domestic macro variable rather than a foreign one.

The chain is unusually direct. All else equal, a dearer barrel widens the current-account deficit, because the import bill rises without any change in what India sells. A wider deficit means more rupees sold for dollars, which pressures the currency. A weaker currency raises the rupee cost of the same barrel again. And fuel costs enter transport and manufacturing across the economy, so measured inflation picks up — the mechanism set out in inflation explained — which in turn narrows the room a central bank working to an inflation mandate has to ease.

How much of a barrel move reaches the price level is a choice, not an identity. Excise duty, the pump-price decisions of the state-owned oil marketing companies and the subsidy line can absorb part of it instead of passing it through, which is why a comparable rise in crude can land mostly in the inflation print one year and mostly in the fiscal accounts another. Who ends up carrying each leg is followed through in crude oil and the Indian economy.

That last link is the one people skip. Crude reaches equities twice: once through company margins, and once through the policy rate. The first is visible in the results of paint, tyre, aviation and logistics businesses, where fuel or crude derivatives are a real line of cost. The second is invisible and larger, because it moves the discount rate applied to every listed company at once.

The trade-off in reading this tile: the same barrel that squeezes an airline supports an upstream producer, and India has both listed. A green energy tile is a statement about the index-level effect our model assumes, not a claim about every sector inside it. Sector-level effects go the other way often enough that the aggregate reading and the stock-level reading routinely disagree.

Channel four — credit and volatility, the early-warning pair

A credit spread is the extra yield a lender demands over a government bond of the same maturity for taking the risk of not being repaid. Investment-grade and high-yield spreads sit on the page for one reason: credit tends to register stress before equity does.

The mechanism is about seniority. A lender is paid before a shareholder, so when the market starts doubting a set of borrowers, the doubt shows up first in what lenders charge them. Equity is the residual claim — last in line — so the same information, taken at face value, bears on it more heavily rather than less. When spreads have widened and share prices have not, what you are looking at is a disagreement, and a disagreement has two readings: lenders are pricing something shareholders have not yet, or lenders are wrong. The page shows the disagreement. It does not resolve it, and nothing here says which side moves. The MOVE index, which measures expected volatility in Treasuries, sits on the page for the same reason, applied to funding conditions.

Note the word "tends". This is a judgement about the usual order of events, not a mechanical certainty, and it fails when the shock originates in equity itself — a single-sector earnings collapse, an index reconstitution, a policy surprise that touches shareholders and not lenders. Credit does not lead everything. It leads the class of event that starts with the price of borrowing. Why high yield moves before investment grade, and why the rate of widening carries more information than the level, are the subject of credit spreads and where trouble shows up first.

The two volatility tiles are a pair worth reading against each other. The US VIX prices expected movement in the S&P 500; India VIX prices it in Nifty options. When India VIX is calm while the US VIX is rising, the domestic market is pricing insulation from something the global market is not. Whether that insulation is real is exactly the sort of question the page raises and does not answer — and for what implied volatility is and how it is read off the chain, see option chain analysis.

Channel five — who is actually buying

The other four channels are theories about what should happen. The flow tile is a record of what did.

Foreign institutional investors and domestic institutions publish daily net cash-market purchases, and the difference between them is the clearest read available on who the marginal buyer of Indian shares was today. A market where foreign money is selling into domestic buying is a different market from one where both are buying, even if the index closed at the same level.

The limitation is in the word "cash". The figure covers the cash segment only. Foreign positioning also runs through index and stock futures, options, and primary market participation, and none of that is in the print. So a large cash sale is evidence of selling in the cash segment and not proof of a reduction in total exposure — the position may have moved to derivatives rather than gone away. This is the most over-interpreted number in Indian market commentary, and the reason is that it is the only one published daily with a rupee sign on it.

In our composite, the flow input saturates at ±₹5,000 crore of net FII cash — anything beyond that counts the same as that day's ₹5,000 crore. That ceiling is a design choice, chosen so that one extraordinary session cannot dominate a cross-asset average, and it has the deliberate consequence described in the next section.

The composite is a summary of a view

The gauge at the top of the page compresses the tiles into a single number between −100 and +100. Here is exactly what it does, because a score you cannot re-derive is a score you should not lean on.

Each scoring tile is converted to a contribution: take its percentage change, divide by a per-input scale, clamp the result to the range −1 to +1, then apply a sign that says whether rising is helpful or harmful for Indian equities. Those contributions are averaged with weights, and the average is multiplied by 100. Inputs whose feed failed drop out of both the top and the bottom of the average, so a partial page still produces a comparable number on fewer inputs.

The weights we use are the dollar 0.20, the US 10-year 0.15, Brent 0.12, FII cash 0.12, USD/JPY 0.10 and the VIX 0.10, with smaller weights on the rest. Every one of those is our modelling judgement about what matters most to Indian equities — a considered view, not a measured constant. They were not estimated from a regression and they are not a property of the world. The regime labels sit at the same status: above +20 we call it risk-on, below −20 risk-off, and the cut-offs are ours too.

Now the part that is genuinely non-obvious, and it follows from the clamp. Because every input saturates at ±1, the score measures agreement across channels, not severity within one. A dollar move three times the size of its scale contributes exactly as much as a move at the scale. So a reading of +45 does not mean "things are strongly positive" — it means a weighted majority of channels are pointing the same way. An extreme move in one channel with the other four flat produces a modest score, by construction. That is a deliberate property, and it is the cost of using an average: you gain a stable summary and you lose the outlier.

Which is why the page carries a separate stress state that bypasses the average altogether. A score at or below −35 triggers it, and so does a one-day VIX jump of 20% or more, or a one-day fall of 1.2% or more in USD/JPY — three more cut-offs that are ours, on the same footing as the weights. An average is designed to be slow, and a carry unwind is not slow — when yen borrowing costs move against a crowded trade, the unwind arrives before a weighted mean of thirty inputs has finished noticing. The global selloff of August 2024, when a yen-funded carry trade came apart, is the episode usually cited. The stress override exists precisely because the score, on its own, would have been late.

The trade-off in the whole design is worth stating plainly: one number is easier to read and tells you less. It cannot tell you which channel is doing the work, which is why the rationale lines under the gauge name the largest contributors. Read those before reading the number.

The heatmap is a window, not a mechanism

The correlation panel measures how tightly two series moved together day by day, on a scale from −1 (opposite directions, every day) through 0 (no linear relationship) to +1 (the same direction, every day) — the Pearson correlation. It computes that on Nifty's daily returns against each driver's daily returns, over a window you choose — 30, 60 or 90 sessions — using the dates the two series have in common.

It is the most misused thing on the page, so the discipline matters.

A correlation is an observation about a window. A channel is a mechanism. The two are different kinds of claim and they age completely differently. "A dearer dollar tightens global funding and that reaches India through allocation, the rupee and import costs" is a mechanism, and it is as true in a quiet quarter as in a violent one. A panel entry of the form "Nifty and the dollar index correlated −0.4 over the last 60 sessions" is a measurement of 60 particular days — the coefficient is whatever those days produced, not a property of either series.

Short windows are noisy, and 30 sessions is short. When a correlation flips sign between the 30-day and the 90-day window, the honest reading is almost always that you have learned something about the window, not about the world. A single violent week inside a 30-session sample can carry the whole coefficient. Widen the window and it disappears — not because the relationship changed, but because one week stopped being a third of the evidence.

So the correct use is narrow and still worth having: take a channel you can already state a mechanism for, and check whether the recent record is consistent with it. Consistency is mild support. Inconsistency is a prompt to ask whether something else has been dominating — which is a question, not an answer. What the heatmap never licenses is the sentence "X is driving Nifty". Correlation cannot establish that, and two assets moving together because both respond to a third thing is the ordinary case in macro, not the exception.

Four ways this page gets misread

1. Reading colour as direction. Green is "helpful to Indian equities", not "went up". A green dollar tile means the dollar fell.
2. Treating the score as a measurement. It is a weighted summary of our view of what matters, computed from our chosen weights and cut-offs. Change the weights and the same day produces a different number. Nothing was measured wrongly; a different opinion was applied.
3. Turning a correlation into a cause. Covered above, and worth the repetition because it is the failure that feels most like analysis while it is happening.
4. Trading the tile instead of the channel. A macro reading tells you about the environment a position sits in, not about the position. It does not generate an entry, and a page that never mentions your holding period or your allocation cannot answer a question about either — that is what asset allocation and goal-based investing are for. The people who lose money to a macro dashboard are not the ones who read it wrongly. They are the ones who read it correctly and then acted as though a directional read had been established when only an environment had.

There is a fifth, quieter one: checking it too often. The tiles refresh every minute, and a channel that moves an allocation does not change every minute. If a long-horizon holder finds a macro page altering a decision daily, the page has become a source of anxiety rather than information — the same trap described in what happens when you stop a SIP in a drawdown.

Where this sits in the app

All of the above is on one screen rather than assembled from ten browser tabs.

FNOTrader's Options Analytics app carries the macro page described here: the cross-asset tiles grouped by channel with a plain-English explanation attached to each, the composite score with its largest contributors named in words, the correlation heatmap across 30, 60 and 90-session windows, central-bank policy rates with the RBI first, and a filtered calendar of scheduled macro events. Every tile opens its own history chart over one month to five years, so a level can be read against its own range instead of against a memory of it.

The weights, scales and regime cut-offs are the ones stated in this article, and they are ours. They are published rather than hidden precisely so that a reader who disagrees can see what they are disagreeing with.

Common questions

Continue reading

More in Macro & Markets · App: Options Analytics · Definitions: glossary · Free tools: calculators · All: every article