Economy to company: the top-down approach
The answer
Top-down analysis means working from the economy to the industry to the company, so that you know which outside forces decide this company's results before you judge the company itself.
The useful version is not forecasting the economy. It is measuring exposure: finding, in the company's own annual report, exactly how much its profit moves when a rate, a currency or a commodity price moves.
Why this costs you money
Here is the version of top-down analysis that loses money, and it is the common one.
You read that interest rates are expected to fall. You conclude that banks and housing will do well. You buy a housing finance company. Two things then go wrong.
The forecast was wrong. Rates did not fall, or they fell later, or they fell and then rose. Professional economists working full-time with better data get the direction of rates wrong regularly. You will not do better.
Even if the forecast was right, the company was the wrong one. Falling rates help a lender whose loans are priced for a long fixed term and whose borrowing re-prices quickly. They hurt a lender in the opposite position. The 2 companies sit next to each other in the same sector list on your screener. Nothing on the screen tells you which is which.
The cost is a specific kind of loss. You were right about the world and still lost money, so you learn nothing from the experience. You conclude the market is irrational, and you do the same thing again next year.
There is a second, quieter cost. You get the number right and misread what it means. Revenue grows 12% in a year when the company raised prices 11%. That is not growth. It is inflation passing through, and volume was flat. Somebody who does not separate price from volume will pay a growth multiple for a company that is not growing.
How it works
The chain has 4 links, and the whole method is refusing to skip any of them.
Link 1: the macro variables that actually reach a company
Only a handful of economy-level numbers reach a company's profit and loss statement directly. Learn the route each one takes.
Interest rates. They arrive twice. Once through the company's own interest cost, which depends on how much of its debt is floating rate and when it must refinance. Once through demand, if the product is bought on credit — houses, cars, capital equipment. A company with no debt can still be badly damaged by rates if its customers borrow to buy.
Inflation. It arrives as input cost first and selling price second, and the gap between the 2 in time is the whole story. A company that can raise prices within a month is barely affected. A company on annual contracts absorbs a year of cost increases before it can respond.
The exchange rate. It arrives 3 ways: on revenue earned abroad, on inputs bought abroad, and on debt borrowed abroad. A company can have all 3, and they can partly cancel. That cancellation is called a natural hedge, and it is either real or it is not, depending on whether the amounts and the timings match.
Real demand growth. Gross domestic product growth is the total size of the economy's output. It matters most for companies selling things people postpone in bad years: cars, steel, travel, advertising. It matters least for companies selling things people buy anyway: soap, medicine, electricity.
Commodity prices. These are an input for some companies and the entire selling price for others. For a producer, the commodity price is revenue and almost all of the profit swing.
Link 2: how an industry converts a macro move into a profit move
The same macro move produces very different results in different industries, and 2 structural features explain most of it.
Operating leverage is the share of costs that do not change with volume: factories, rent, salaried staff. High fixed costs mean a 10% fall in volume produces a much larger fall in profit. An airline, a hotel and a cement plant have high operating leverage. A trading business has almost none.
You can estimate it from the accounts. Indian statements disclose cost of materials consumed, employee benefits expense, finance costs, depreciation and other expenses as separate lines. Materials are mostly variable. Depreciation and a large part of employee cost are fixed. The larger the fixed share, the more violently profit will move.
Financial leverage is debt. It multiplies whatever operating leverage already did. A company with high fixed costs and high debt does not have 2 problems. It has 1 problem that arrives twice as hard.
Industry supply is the part everybody forgets. Demand is the story people tell. Supply is usually what decides the price. When a commodity price is high for 2 years, every producer announces new capacity. Those plants arrive together, 3 years later, into weaker demand. That is why cyclical industries lose money at the moment their end-market finally looks healthy.
Link 3: where the company sits inside the industry
Two questions, both answerable from the report.
Is it a price taker or a price setter? A price taker sells a product identical to everybody else's, so it accepts the market price. A price setter has something customers will pay extra for. In a downturn the price taker's revenue falls with the market and the price setter's does not.
Where does it sit on the cost curve? In a commodity industry, the producer with the lowest cost per tonne survives every downturn and buys the assets of the ones that do not. Cost per unit is often disclosed, or can be derived from cost of materials consumed divided by volume, which many Indian companies publish in the management discussion.
Link 4: the sensitivity numbers the company already publishes
This is the part that turns opinion into measurement, and it is the reason to open the annual report rather than an article.
Companies are required to disclose their market risk. The disclosure usually includes an actual sensitivity table: if the currency moves 5%, profit changes by this amount. If interest rates move 100 basis points, which is 1 percentage point, finance cost changes by this amount.
Those tables convert "the rupee is weak" into a number. They are the difference between a view and an estimate, and most investors have never opened one.
What it tells you, and what it does not
It tells you the size of the outside risk you are carrying. After 10 minutes you can say: a 10% currency move changes this company's profit by roughly 6%. That is a fact about your holding, not a forecast.
It tells you which news matters. Once you know a company's revenue is 70% domestic and its debt is all rupee-denominated, a headline about the dollar is not about your holding, and you can ignore it.
It does not tell you where the economy is going. Nobody knows. Macro forecasting has a poor record, including among people who do it professionally.
Sensitivity tables assume everything else stays still. They are calculated by moving 1 variable. In a real downturn, demand, the currency and rates all move together, and the combined effect is usually worse than the sum of the individual lines.
Hedges expire. A company hedged for the next 6 months is not hedged. It has postponed the exposure, and the disclosure will say for how long.
The decision rule
Never buy a company because of a macro view. Buy a company you have already judged on its own merits, then check whether a macro move you consider likely would break it.
If a company's profit disappears on a 15% move in 1 outside variable that it does not control, that is a leveraged bet on that variable, whatever the business does — unless the exposure is genuinely hedged for longer than your holding period, in which case check when the hedge ends and treat that date as a risk date.
Try this now
Five minutes, 1 holding, 1 annual report. You are going to compare what management says about the world with what the numbers say.
- Open the latest annual report of 1 holding. Find the chairman's letter or managing director's message near the front. In a US Form 10-K there is usually no such letter, so use the shareholder letter in the separate annual report, or go straight to Item 7.
- Read it once and write down every external factor it names: inflation, interest rates, monsoon, government policy, oil, the dollar, demand recovery. There are usually between 3 and 6.
- Write down every claim about performance the letter makes. Phrases such as "strong growth", "margin expansion", "robust demand", "improved efficiency". Count them.
- Now go to the financial statements and check each claim against a number. Did revenue grow faster than last year, or slower? Did operating margin rise or fall? Did volume rise, or only price? Many Indian reports give volume in the management discussion.
- Finally, find the market risk disclosure. In an Indian report it is a note near the end called financial risk management, with sub-headings for currency risk, interest rate risk and commodity price risk. In a 10-K it is Item 7A, Quantitative and Qualitative Disclosures About Market Risk. Write down 1 sensitivity number: how much profit changes for a stated move.
What you should see. Count how many claims in step 3 the numbers in step 4 did not support. In most annual reports the answer is not 0. The usual pattern is a letter that describes a difficult year as a year of building for the future, and a set of numbers that shows margins fell. That is not dishonesty. It is the job of a chairman's letter, and knowing that changes how you read every one you open afterwards.
From step 5 you should get 1 sentence you did not have before, such as: "a 1 percentage point rise in interest rates costs this company roughly ___ of profit." Keep it. You will use it every time rates move.
Three real cases
1. Countrywide Financial, 2007 to 1 July 2008 (United States) — the macro was the business Countrywide was the largest mortgage lender in the United States and reported years of strong, growing profits. Its earnings depended on 2 outside conditions: that house prices kept rising, and that it could keep borrowing short-term money to fund long-term loans. Both were macro variables, and neither appeared as a risk in the profit line while they held. When house prices fell and short-term funding markets closed in 2007, the company could not refinance itself. Bank of America agreed to buy it in January 2008 and completed the purchase on 1 July
- No fraud was needed for the equity to be destroyed. The business model was
a bet on 2 macro variables, and the accounts looked excellent right up to the point where the bet lost.
2. Jet Airways, 2018 to 17 April 2019 (India) — costs in dollars, revenue in rupees Jet Airways was one of India's largest airlines. Aviation turbine fuel, aircraft lease payments and much maintenance spending are priced in or linked to US dollars. Ticket revenue was mostly in rupees. During 2018 crude oil prices rose sharply and the rupee weakened to about 74 per dollar in October 2018. Both moves pushed costs up while competition kept fares down. The airline had also carried heavy debt for years. It stopped flying on 17 April 2019 after failing to arrange funding. The currency and the oil price did not cause the failure on their own. They arrived at a company with high operating leverage and high financial leverage, which is what converts an ordinary macro move into an ending.
3. Peabody Energy, 2011 to 13 April 2016 (United States) — capacity added at the top of a cycle Peabody was the largest private-sector coal producer in the world. Coal prices were high in 2010 and 2011, driven by demand from China. In 2011 the company completed a large acquisition of Australian coal assets at prices that reflected those conditions. Coal prices then fell for several years as supply arrived and demand slowed. The company filed for Chapter 11 bankruptcy protection on 13 April 2016. This is the standard shape of a commodity cycle. The industry expands when prices are high, the new capacity arrives when prices are low, and the companies that borrowed to expand do not survive the gap.
The question that resolves it
A novice looks at a company and asks: what is the economy going to do?
An expert looks at the same company and asks: what has to happen in the economy for this company to be in trouble, and how far away is that?
The first question has no reliable answer. The second one has an answer printed in the annual report, and it does not require a forecast.
What would make this wrong
If top-down analysis worked as advertised, then correctly predicting a country's growth would let you pick the right stocks. It does not.
Several honest limits.
Growth and returns are weakly related. Across countries, faster economic growth has not reliably produced better stock returns, partly because growth gets paid for with new share issues that dilute existing owners.
The market prices macro faster than you can read about it. By the time a rate change is in the news, the obvious beneficiaries have usually moved. Exposure measurement still helps, because it tells you what you own. Forecasting does not.
Sensitivity disclosures are approximations. They are prepared on assumptions that may not hold, they cover only financial instruments in many cases, and they often exclude the demand effect, which can be larger than the direct effect.
Some companies genuinely escape their macro. A dominant company in a growing niche can grow through a recession. Treating every company as a function of the economy is as wrong as ignoring the economy.
In India
The macro variables with the most reach. The Reserve Bank of India's repo rate sets the base for most corporate borrowing costs, and a large share of Indian corporate debt is floating rate, so a policy change reaches profit within 1 or 2 quarters. The rupee-dollar rate matters because India imports most of its crude oil, so a weak rupee and a high oil price usually arrive together and push inflation up. The monsoon still moves rural demand for 2-wheelers, tractors, consumer goods and gold.
Where the exposure is disclosed. The financial risk management note, prepared under Ind AS 107, carries currency, interest rate, credit and liquidity risk, usually with sensitivity tables. There is a separate disclosure of unhedged foreign currency exposure, which is the amount of foreign currency debt and payables with no hedge against it. That single number has predicted a great deal of Indian corporate distress.
Schedule III to the Companies Act 2013 forces a useful cost breakdown. Cost of materials consumed, purchases of stock-in-trade, changes in inventories, employee benefits expense, finance costs, depreciation and other expenses are all separate lines. You can calculate operating leverage from the face of the statement, which is harder in a US filing.
Government policy is a first-order variable in more sectors than in the United States. Fuel and fertiliser subsidies, sugar and pharmaceutical price controls, import duties, production-linked incentive schemes and telecom levies all change industry economics directly. The management discussion usually names the ones that matter.
In the United States
The macro variables with the most reach. The Federal Reserve's policy rate sets short-term borrowing costs, and the 10-year Treasury yield sets the discount rate the whole market uses to value long-dated cash flows. A rising 10-year yield mathematically reduces what any long-duration business is worth, regardless of its own results. Consumer spending is roughly 2 thirds of US output, so the employment and wages data reaches most companies. For large US companies, foreign revenue is a major exposure: a strong dollar reduces the dollar value of profits earned abroad.
Where the exposure is disclosed. Item 7A of the 10-K is a dedicated section on market risk, often with a tabular sensitivity analysis. Item 1A, Risk Factors, names macro risks in the company's own words. The notes carry the debt schedule with maturities by year, which tells you when refinancing risk arrives.
The debt structure is different, and it matters. US companies borrow heavily through fixed-rate bonds with long maturities. That means a rate rise reaches a US company's interest cost slowly, only as bonds mature and are refinanced. Investors who assume a rate rise hits profit immediately are usually wrong about the timing.
Segment reporting is generally more detailed, so you can often see which part of a company carries the macro exposure rather than guessing from the total.
Where they differ, and what that tells you
The transmission speed of an interest rate change is different, and that changes what a rate cycle means for your holdings.
In India, a large share of corporate borrowing is floating rate and linked to a bank benchmark, so a policy rate change reaches company profit within about 2 quarters. In the United States, much corporate borrowing is fixed-rate bonds issued for 5 to 30 years, so the same policy change reaches profit over years, as each bond matures. The practical instruction: for a US company, read the debt maturity schedule and ask what rate the maturing debt will be replaced at. For an Indian company, read the split between fixed and floating and assume it arrives soon.
The currency exposure runs in opposite directions for the typical large company. A large Indian company more often imports inputs and borrows abroad, so a weak rupee hurts. A large US company more often earns revenue abroad, so a strong dollar hurts. Both are currency risk. They are not the same trade, and an investor who holds both is less exposed than an investor who holds either.
Policy risk is concentrated differently. In India, policy risk arrives as direct intervention in prices, duties and subsidies, and is sector-specific. In the United States, policy risk arrives more often through monetary policy, antitrust action and tariffs. When you read an Indian annual report, look for which government decision the business depends on. When you read a 10-K, look for which rate or which regulator.
Carry this
- Do not forecast the economy. Measure how much your company depends on it.
- Operating leverage plus financial leverage is what turns a normal macro move into a failure.
- The sensitivity table in the risk note converts an opinion into a number. Read it once per holding.