How Trade Barriers Reach Inflation, Companies and Markets

A barrier reaches a household through the price of the import and of whatever competes with it, a company through its input bill or its selling price depending on which side it stands, and a market through expectations. Expectations move first of the three. Each step passes on only part of what it received, so a tariff rate is the start of the chain rather than its size.

The kinds of barrier that exist, whether charged at the border or imposed as a non-tariff restrictionA trade measure that works through a rule rather than a charge: a quota, a licence requirement, a standard an import must meet. The forms it takes are covered where trade barriers are covered., how a duty is charged and who ends up carrying it are settled where trade barriers are covered. How much of a cost increase turns into a price increase, and how a basket of prices is put together into a single reading, are settled under inflation and prices. How a claim gets priced off what people expect to happen is settled under monetary policy, where the same arithmetic appears with a rate decision in place of a duty.

One barrier leaves the port and travels three separate roads until it lands on a shelf, on a profit line and on a traded price. The interesting thing about that journey is not that it happens. The interesting thing is how much of the move goes missing at every junction, and how badly a reader misjudges the destination by never checking.

Every place, business and figure below sits inside the Republic of Sankhya, an invented economy that trades with an invented partner called Marut. Sankhya grows onions cheaply and makes machines expensively, so it sells onions and buys machines. The one event traced here is a 20 per cent duty on an imported onion grading machine whose landed costWhat an imported item costs by the time it has actually arrived and can be collected: the seller's price plus freight, insurance and anything charged at the border. How it is built up is covered where trade barriers are covered. is Rs 50,000/-. A grading machine of this kind is a small bench machine rather than plant, so Sankhya buys 2,50,000 of them in a year.

What are the three routes out of one trade barrier?

A barrier is one measure, but it does not arrive at one address. The measure splits into three roads out of the port, and almost every muddled argument about barriers comes from watching one road and reporting it as the whole journey.

The first road goes to a household, through prices. Something that used to arrive at one cost now arrives at a higher one, and part of that eventually shows up on a shelf. The second road goes to a company, through its accounts. A business that buys the item sees a bigger bill; a business that makes something similar sees a weaker rival. The third road goes to a market, through what people expect. Anything with a traded price is repriced the moment opinion about the barrier changes, and opinion changes long before customs does anything.

The three routes are almost always listed household first and market last, and they arrive in exactly the opposite order. A market reprices in the weeks a barrier is being argued about, sometimes before any measure is drafted. A company sees it inside a quarter or two, as soon as the next consignment is invoiced. A household sees it slowest of all. A shelf price is only revised when a producer decides to revise it, and the producer is working through stock bought at the old cost.

How does a barrier reach a household's prices?

Start at the port. The first fraction is subtracted there. Sankhya charges 20 per cent on a machine with a landed cost of Rs 50,000/-, so the duty is Rs 10,000/-. The duty itself is arithmetic and nobody argues about it. The whole question is what happens next. The buyer and the seller now have to decide between them who carries the Rs 10,000/-, and pass-throughThe share of a cost increase that actually turns into a price increase for the next person along. The share is measured rather than assumed, and it is covered under inflation and prices. is the word for however that argument comes out.

Three settlements are worth drawing. If the whole duty is handed on, the buyer pays Rs 60,000/-. If none of it is, the buyer keeps paying Rs 50,000/- and the Marut exporter absorbs the entire Rs 10,000/- out of its own margin to keep the sale. The case traced here sits between the two: 60 per cent is handed on, so the buyer's price becomes Rs 56,000/- while Rs 4,000/- comes out of the Marut exporter's own margin. The duty is identical in all three settlements and only one of them puts the whole of it on the buyer. A duty rate on its own therefore tells nothing about what any price does.

Now keep going. The machine is not what anybody eats. Belavali Grading Works buys these machines to grade and pack onions, and the machine is a cost inside a price that a Sankhya household actually pays. Two more fractions sit between the port and the shelf, and both of them are large.

The first is how much of the finished price the machine accounts for. On the assumption used here, machine cost is 15 per cent of the price of a packed quintal of onions, the rest being onions, labour, transport and everything else. A machine price 12.00 per cent higher therefore lifts the packed price by 12.00 times 0.15, or 1.80 per cent. The second is how much of a household's spending goes on packed onions in the first place. On the assumption used here, that good carries 8 per cent of the Sankhya consumer basket. The contribution to the headline reading is therefore 1.80 times 0.08, or 0.144 percentage pointsThe unit for the gap between two percentages. The gap between a reading of 6.700 per cent and one of 6.844 per cent is 0.144 of these, and calling that a 0.144 per cent rise would mean something quite different., printed as 0.14 anywhere the convention is two decimals.

The chain of four fractions is worth pausing on for a reason that has nothing to do with onions. Not one of the four numbers depends on how expensive the machine happens to be. Halve the machine price or multiply it by ten. Each step is a share of the step above it rather than an amount, so every line of the table reads exactly the same. A chain of fractions carries no scale of its own. The absence of scale is what lets it be worked out for one machine and then applied to a whole country without redoing anything. Scale only enters when the last percentage is turned back into money, which is the next step.

Put a rupee amount on the contribution. Points are easy to shrug at. Take one invented Sankhya household spending Rs 25,000/- a month across the whole basket. The barrier costs that household Rs 36/- a month, or Rs 432/- over a year. Not Rs 5,000/- a month, the figure a 20 per cent headline sounds like it means. Rs 36/-. The four fractions between the duty and the shelf are the entire difference between those two numbers, and each one of them is a real quantity somebody has to estimate rather than a rounding convenience.

The monthly budget of Rs 25,000/- was picked on purpose, and the reason is a small lesson in itself. A chain this thin can land between two rupees. The same four fractions run against Rs 10,000/- of monthly spending give Rs 14.40, and Rs 14.40 has to be carried in paise or quietly rounded. Quiet rounding at the very end of a chain built to show attenuation is precisely the wrong place to lose precision. Rs 25,000/- was chosen because every rung of the ladder stays a whole rupee at that base: Rs 5,000/-, then Rs 3,000/-, then Rs 450/-, then Rs 36/-. The percentages are the same at any base, so nothing about the teaching changes with it. The base changes only whether the arithmetic can be read without a footnote about decimals.

Why does a barrier reach people who never buy the import?

Here is the part that catches people out, and it has nothing to do with imports at all. Think about a street with two vegetable sellers. One of them raises prices. The other one buys from a different supplier and faces no new cost whatsoever. The price across the street was the only thing keeping that seller honest, so the second seller now has room to raise prices too. Nothing changed in their own costs. Their competition got dearer, and that was enough.

Sudhanvi Machine Works makes grading machines inside Sankhya. Sudhanvi is a high cost maker, and Sankhya imports most of its machines for exactly that reason. Sudhanvi can build one for Rs 48,000/- and sell it for Rs 49,500/-, undercutting the Rs 50,000/- import by a whisker and earning Rs 1,500/- a machine for the trouble. Then the barrier arrives on the import and lifts it to Rs 56,000/-. Sudhanvi's costs have not moved by a rupee. Its competing domestic goodA home-made item close enough to an import that buyers treat the two as alternatives. Which goods qualify, and how close is close enough, is covered where trade barriers are covered. is now competing against a much dearer rival, so it lifts its own price to Rs 52,500/- and still sits Rs 3,500/- under the import.

Now scale it. Sankhya buys 2,50,000 imported machines a year, a whole machine import bill of Rs 1,250 crore, and makes 1,50,000 at home on the figures used here. The imported ones cost buyers Rs 6,000/- more each, or Rs 150 crore. The home made ones cost buyers Rs 3,000/- more each, or Rs 45 crore. Total extra bill Rs 195 crore, of which Rs 45 crore, or 23.08 per cent, was collected by producers who paid no duty at all and faced no new cost. Almost a quarter of what this barrier costs Sankhya buyers is money the barrier never touched, and no amount of studying the import figures will find it.

Whether that second effect is smaller or larger than the first is a matter of two quantities and nothing else: how big the home made base is, and how fully home makers use the room. Change one of them and the ranking flips. If Sankhya made 6,00,000 machines a year instead of 1,50,000, the same Rs 3,000/- of extra price would collect Rs 180 crore, more than the Rs 150 crore direct effect. A reader who only ever adds up the import bill would have missed the bigger half.

How does one barrier reach two companies with opposite signs?

Whether a barrier is good for business is a question with no answer in it. Two Sankhya businesses stand on either side of the same measure, and the same Rs 10,000/- duty pushes them in opposite directions with different force.

Belavali Grading Works buys forty machines a year. Its input billEverything a business pays out to make what it sells: materials, bought-in equipment, power, labour. How it is classified and reported belongs with financial statement analysis. for machines goes from Rs 20,00,000/- to Rs 22,40,000/-, so Rs 2,40,000/- more. On revenue of Rs 2,00,00,000/- and operating profit of Rs 24,00,000/-, that takes profit to Rs 21,60,000/- and the operating marginProfit from running the business, before financing and tax, as a share of revenue. How it is built and what moves it belongs with financial statement analysis. from 12.00 per cent to 10.80 per cent. Look at what that does to the arithmetic of attenuation. The duty was 20.00 per cent. The machine price rose 12.00 per cent. Profit fell 10.00 per cent. Profit is the small gap between two large numbers, and a cost increase lands on the gap. So the chain thins out on its way to a price index and concentrates on its way to a profit line.

Sudhanvi Machine Works sells two thousand machines a year at Rs 48,000/- of cost. Its margin per machine goes from Rs 1,500/- to Rs 4,500/-, so profit goes from Rs 30,00,000/- to Rs 90,00,000/-, three times what it was. Same barrier, same year, same country. One business is down a tenth and the other is up two hundred per cent, and no property of the barrier decides which. Position decides it.

What reaches a market, and when?

A shelf price waits for somebody to reprint a label. A traded price is nothing more than today's opinion written down as a number, so it waits for nothing. So the third road is the fastest, and it runs on a clock that has almost nothing to do with the barrier's own dates.

Take two invented claims, one on Sudhanvi and one on Belavali, each priced at 100 in a world where no barrier is expected. Suppose the market judges that a barrier would eventually be worth 150 to the maker and 90 to the buyer. Then the price at any moment is just those two ends blended by how likely the barrier is thought to be. At even odds the maker's claim sits at 125. At 90 per cent, at 145. When the measure is finally confirmed and the odds go to certainty, it reaches 150.

Now count where the movement happened. The whole journey is 50.00 per cent. The bit that happens on the day of the announcement, from 145 to 150, is 3.45 per cent of the price and 10.00 per cent of the total move. Ninety per cent of it was gone before anybody read the notification. A market that barely moves when a barrier is announced is reporting that the barrier was expected, not that it does not matter, and reading a flat price as indifference is one of the most common misreadings in this whole area. The same arithmetic, with a rate decision in place of a duty, is worked through under monetary policy.

The buyer's claim behaves the same way in miniature. The mirror image rules out an easy escape. The buyer's claim falls from 100 to 90 across the whole journey, so on the day of confirmation it goes from 91 to 90: again a tenth of the move, and again nine tenths already spent. Nobody could look at either price on the day and read the barrier's size off it. The timing distortion is a property of expectations rather than of good news or bad news, so it applies to the gainer and the loser identically.

One honest wrinkle before leaving the market road. Sudhanvi's profit tripled and its traded claim moved only half. Belavali's profit fell 10.00 per cent and its claim moved exactly 10.00 per cent. A traded price is a view on many years and on how long any shelter lasts. A profit figure is one year of it. The two need not match. How a stream of future results turns into one price today is a subject in its own right and is not attempted here.

Why is the first effect rarely the whole effect?

Everything traced so far is the easy part, and it is worth being blunt about that. The direct contribution of 0.144 points is the one number in this whole area that can be computed from four fractions and checked by somebody else. Being checkable is exactly why the direct contribution dominates the commentary, and exactly why stopping there is a mistake.

Three larger things sit behind it. The competitive effect, already worked above, put Rs 45 crore on machines that paid no duty. The input cost effect runs through every business that uses the item, and each of those businesses then makes its own decision about how much to hand on, so the chain of fractions starts again at every stage. And then there is retaliationA partner responding to one country's trade measure with a measure of its own. Which forms it takes and how disputes over it are handled is covered where trade barriers and trade agreements are covered.: if Marut answers with a barrier of its own on Sankhya onions, the Sankhya onion grower's selling price falls, and that reaches a fourth set of people through a route the machine duty never touched.

The measurable effect and the important effect are not the same effect, and an account that computes the first and stops has done the arithmetic that was available rather than the work that was needed. None of that makes the 0.144 points wrong. The three larger effects make the 0.144 points a floor. When somebody says a barrier will add a certain amount to inflation, the honest reading is that the direct route adds that much and the other three routes are unstated.

What can a reader actually trace, honestly?

Three effects, three very different levels of confidence, and the useful discipline is to say which is which before quoting any of them.

The direct price contribution is computable. Four fractions, multiplied, giving Rs 36/- a month and 0.144 points. Two people working from the same four assumptions get the same answer, and where they disagree the fraction they disagree about can be pointed at. The direct contribution is a genuinely checkable number.

The competitive effect is estimable, not computable. The Rs 45 crore above is real arithmetic sitting on an assumed response: 1,50,000 home made machines and a Rs 3,000/- price rise. An assumed response of Rs 1,500/- instead halves it. The effect can be bounded and its direction stated, and it cannot be pinned.

The market effect is visible only afterwards, and not always then. The move happens on expectations, so by the time the measure exists most of it has already been priced, and nobody can say in advance how much was expected. Reporting an estimate and a computation in the same tone of voice is how a defensible chain of arithmetic turns into an indefensible forecast.

What does an analyst do with all this?

Watch the order of operations. Most people do the opposite. Somebody covering a set of businesses through a barrier does not start by asking how big the measure is. The first question, for each business on the list, is which side of the measure that business stands on: buying the restricted item, making something that competes with it, selling into the partner that may respond, or none of the three.

The sorting is cheap and it fixes the sign. Only afterwards does size come in, and size needs three more things: what share of that business's costs or revenues the item is, how much of any change it can hand on to its own customers, and how much of the move was already in the traded price before anyone asked. Getting the sign wrong is unrecoverable and getting the size wrong is only imprecise, so the sorting comes first and the arithmetic second.

A lender does the same sorting for a different reason. A borrower whose input bill has just risen 12.00 per cent on one line has less room before its payments get tight, and a borrower whose margin per unit has tripled has more. Both are the same barrier arriving at the same lender's desk from opposite directions, and what the barrier did to the loan book is only ever a list.

A household can run the same discipline on itself without any of the apparatus, and this is where the everyday version lives. Somebody hearing that a duty has gone on imported cooking equipment has two useful questions and no third one. Do I buy the thing, or something made with it, and how much of my month goes on that? If the honest answers are no and very little, the correct expectation is a number like Rs 36/- rather than a number like Rs 5,000/-, and the difference between those is not optimism. The difference is four fractions that somebody did or did not bother to multiply.

The Republic of Sankhya, Marut, Belavali Grading Works and Sudhanvi Machine Works are invented.
Educational material. Not advice on any investment, tax, budget or market position.