The Graham Act Waiver: Measuring a Tariff That Did Not Take Effect

What was announced

On October 16, 2026, the US President waived Graham Act duties and enumerated sanctions. Specified existing sanctions remain; relief lasts until further action.[1]

The economic meaning depends on the comparison. Relative to facing an additional tariff, a waiver can limit a company’s burden. But a business that was not already paying that tariff does not acquire an equal amount of accounting profit. A prospective cost and an expense already incurred appear differently in both earnings and cash flows.

There can be effects without an actual tax payment if firms had already changed their plans. A waiver might alter decisions to hold back orders, assess another supplier or shorten quotation validity. A company that already expected relief, however, receives less surprising information. The importance of an announcement and the amount of new information it conveys to each firm need not coincide.

Choose the right starting point

Consider an importer continuing its present trade, continuing while bearing an additional charge, or moving to another supplier. These are three possible future choices, not three observed histories. A difference can be defined between post-waiver trade and trade under a hypothetical tariff, but the hypothetical amount cannot be measured as an observed result.

A mistaken baseline can overstate the effect. Comparing observed trade after a waiver with a scenario in which firms keep paying a large charge without adjusting produces a large difference. Yet purchases, suppliers and selling prices could change if the charge actually applied. Assuming no behavioral response includes trades that might never have been chosen in an account of avoided costs.

Nor does unchanged revenue immediately before and after imply zero effect. The waiver may have preserved orders that otherwise could have been lost. Testing that possibility requires evidence about customer purchasing plans, cancellation terms and preparations to switch, as well as flat sales. Preserving revenue and adding revenue are distinct possible outcomes.

Prior expectations determine the size of the change

A company’s response depends on its prior expectations. A firm that regarded the charge as nearly unavoidable might substantially revise its sales plan after a waiver. One that expected relief may simply execute its existing contracts. Moving into the same legal position can therefore produce different revisions to business plans.

Expectations need not mean only management’s retrospective impressions. Dated quotations, procurement decisions and customer communications may record what firms anticipated before the announcement. Linking those records to orders can identify revisions more precisely than relying only on explanations given afterward. It also avoids assuming that every firm assigned the same probability.

Market responses likewise cannot be mapped mechanically to the apparent scale of the policy text. Prior pricing, other simultaneous information and trading hours matter. Converting a related company’s share-price rise entirely into future profits attributable to the waiver requires ruling out other explanations. Market prices incorporate views about the future; they are not realized orders.

National energy purchases and a product’s origin

In the October 2, 2026 statutory compilation, section 113 addresses US imports from defined countries purchasing Russian-origin energy or facilitating sanctions evasion. Section 115 provides a waiver subject to a national-interest certification and explanation to Congress; section 117 sets separate conditions and congressional review for termination. The compilation’s reference date differs from the date of the new order.[2]

Energy purchases and sales to the United States need not involve the same company. An exporter’s lack of direct trade with a Russian seller cannot by itself explain all implications of a country-based mechanism. Conversely, corporate headquarters do not alone determine where individual goods are imported from. Combining country, legal entity and product into one unit obscures the scope.

For a group operating factories in several countries, applying one rate to total sales is an inadequate estimate. The destination, product, production location and import period must be aligned. This establishes the denominator for assessing exposure rather than recommending a route around restrictions. A country’s total exports cannot be translated directly into a particular firm’s loss.

Keep taxes, profits and policy benefits in separate accounts

Price negotiations separate a change in the importer’s tax payment from a change in the exporter’s profit. Removing a charge the buyer expected to bear does not imply that the seller can raise its price by the same amount. The buyer may regain budget capacity, final prices may fall, or relative prices against competing products may change. Assigning the whole benefit to one firm requires evidence about where contract prices actually adjust.

Corporate profit also differs from the economy-wide benefit. Orders returning to one supplier may increase its revenue while removing opportunities from the firm that expected to replace it. If global demand is unchanged and orders merely shift, adding company-level gains overstates the overall result. The starting point is to distinguish additional production from production transferred between firms.

There may nevertheless be resource savings if checks and preparations needed solely to replace a trading partner become unnecessary. Preparatory spending already incurred is not automatically refunded, however. Avoiding future expenditure differs from finding another use for something already paid for. Tracking avoidance, recovery and redeployment separately broadens the assessment beyond changes in sales.

Related foundation: Revenue, profit and cash flow。Distinguish revenue recovery from available cash.

Deferred orders and lost orders return differently

A buyer that deferred an order may return to the same supplier once conditions permit. A buyer already committed to a long contract elsewhere cannot respond in the same way. Even if the original supplier becomes attractive again, the current contract and product-approval process matter. Comparing recovery speeds solely from the policy date overlooks how far decisions had progressed beforehand.

The distinction also affects order backlogs. A surge consisting of accumulated deferred purchases may not continue into the next period. New customers may bring different prospects for continuity, but they too might merely be replenishing inventories temporarily. Separating returning customers, newly acquired customers and shifts in order timing clarifies the nature of a recovery.

Where capacity cannot expand quickly, returning orders can affect delivery times for other customers. Buyers releasing deferred demand may reserve slots ahead of ordinary purchases. A supplier with spare capacity, by contrast, may handle the same increase at little additional cost. This is why the assessment needs to examine allocation of capacity among customers as well as the seller’s revenue.

Historical research is not an estimate of this waiver

In a September 2019 FEDS Note, Federal Reserve Board researchers measured trade policy uncertainty using newspaper and earnings-call text and examined its relationship with investment and economic activity. Their estimates concern that historical episode, not the present waiver.[3] The useful perspective here is that firms can change decisions before a tax rate changes.

A decline in an uncertainty indicator alone does not establish better business decisions. Coverage may simply have diminished, or a firm may have abandoned entry rather than resolved the problem. Frequency of discussion and usable commercial options differ. The reason a firm stops discussing the issue needs to be read alongside its actual plans.

Nor does a large negative historical effect imply an equally large positive effect from this waiver. Halting investment can be quick, while restoring equipment and staffing takes time. Better policy information alone may not restore lost customers. Whether deterioration can simply be reversed along the same path is a question for the circumstances of individual businesses.

Make the before-and-after comparison meaningful

One approach compares the same firm’s US orders with orders from other markets. Shared changes in global demand may then be partly accounted for. But differences in products, seasons, currencies and customer industries prevent an automatic comparison. Even if the waiver changed only one market’s policy conditions, the other conditions need not have been identical.

Comparison firms should not be selected only after their outcomes are known. Choosing companies with large recoveries risks attributing product strength or customer composition to policy. Selecting them using pre-waiver destinations, products and order trends, including firms that did not recover, helps reveal the conditions associated with different outcomes.

The outcome measure must also be consistent: orders, shipments, imports and revenue are different stages. Goods need not clear customs in the month an order returns. Combining a shipment increase with revenue recorded later can make one recovery look like several. Establishing the sequence of one transaction and dating each stage also shows where the response stalls.

Related foundation: Lead-lag relationships and causation。Why an earlier-moving indicator need not be a cause.

Separate more transactions from a different transaction mix

Higher export value can reflect more units of the same goods or a shift toward more expensive products. A higher average unit value does not establish price increases for individual products: removing cheaper goods changes the average. Linking policy to selling prices requires comparisons of matching quality and contract terms, with volume and composition tracked separately.

The number of trading firms also matters. Expansion by an established large supplier differs competitively from a smaller firm remaining in a market it considered leaving. Aggregate values can obscure the latter. Tracking continuing, newly formed and discontinued buyer-seller relationships reveals whether participation broadens.

More small transactions are not automatically an improvement. They might result from buyers splitting unstable large contracts into shorter orders. Higher counts accompanied by smaller quantities, shorter terms and lower renewal rates may not indicate better prospects. Distinguishing broader participation from fragmentation requires evidence about the persistence of trading relationships.

Related foundation: Read economic data。Align quantities, periods and revisions.

Energy procurement and export orders move on different timetables

When a mechanism depends on national energy procurement, understanding exporters also requires examining energy decisions separately. Changing fuel sources involves product characteristics, transport, compatibility with facilities and existing contracts. An individual exporter’s wish to secure orders does not change those conditions on the same day.

Exporters and energy purchasers may use the time afforded by a waiver differently. The former may settle customer sales plans while the latter compare procurement options. Preparation by both does not establish an eventual change of supplier. Expanding the options that can be prepared and committing to one option are different outcomes.

The timing gap limits claims linking a short-term export recovery with a long-term policy change. Returning orders do not establish a new energy-purchasing policy. A different fuel mix may reflect seasonality or price differences. Even movements in the same direction require records and participants’ explanations linking particular decisions to particular actions.

Making commitments without a fixed end date

A fixed grace period and a measure lasting until another decision provide different calendars for business planning. The latter does not promise unchanged conditions until a date of the firm’s choosing, but nor does it establish a definite near-term expiry. Linking contract review dates to the frequency of new information makes assumptions clearer than inventing a duration.

An inability to commit for a long period does not always halt trade. Businesses with demand determined frequently may already use order periods close to their policy review horizon. Orders requiring dedicated equipment or design changes may instead need years of sales to recover costs. The useful comparison concerns how many sales a project needs to support its expenditure, rather than a guessed number of waiver days.

Flexible contracts have a price. Preparing repeated small quantities can cost more than a production run, and buyers may concede terms to secure supply. Using shorter contracts as a sign of anxiety therefore requires examining lost discounts or efficiency. Duration alone does not distinguish sensible adaptation from a deterioration in trade quality.

Other conditions remain after one barrier recedes

When a new burden is waived, preparations for that burden can be reconsidered. Weak demand, product competitiveness, existing sanctions and the counterparty’s ability to pay do not all improve at once. If orders fail to return, identifying the binding condition is more useful than immediately concluding that the waiver was meaningless. A small response is unsurprising where another condition was decisive from the outset.

This is a way to identify decisive conditions, not to assemble an ever-longer list. Clearer supply-side policy information may do little for orders where customer budgets are shrinking. For customers that take time to replace an approved product, the same policy change may first matter at renewal. The observation period should follow the circumstances of the business.

Even firms winning new orders may not earn higher profits. Returning to a market can require discounts or higher selling costs. An order recovery provides evidence of access, while profitability requires a separate assessment. Distinguishing expanded buyer choice from preserved seller margins clarifies where the benefit goes.

Counterargument: firms may already have expected relief

A strong case for a limited effect is that firms expected a waiver from the start. Few plan revisions after the announcement would then be unsurprising, and the large gap against a full-tariff scenario would not represent newly acquired value. The relevant comparison includes the sales plan actually used just before the decision, rather than only the harshest possible scenario.

Support for this argument would include long-term orders already fixed on unchanged terms before relief and no subsequent change in price or duration. Records showing a buyer explicitly deferring an order for policy reasons and resuming that same project after the waiver would weaken the claim of complete anticipation. A changed reason for the same decision answers this question more directly than quantities alone.

Different prior expectations can also produce a small average response. One firm may upgrade plans, another leave them unchanged and a third face falling demand. A modest aggregate response does not establish that the announcement mattered to no one. Differences associated with prior expectations can identify where it conveyed new information.

What happens when a buyer regains budget capacity?

A buyer regaining funds reserved for a possible charge has options beyond purchasing more of the same product: buying something else, reducing borrowing, lowering selling prices or retaining cash. Examining the demand channel therefore requires attention to spending allocation, not only sales of the goods concerned.

Retail inventories matter for transmission to households. Stock on which the additional tax was never paid cannot be assumed to fall in price by the amount of a waived charge. Any price effect would instead operate through expected replenishment costs or competition. Explaining consumer prices requires distinguishing historical costs from the outlook for the next purchase.

Benefits from having more suppliers may appear in delivery choices, smaller minimum orders or preserved specifications rather than price alone. These conditions need to be assessed using actual quotations and deliveries. Unchanged prices with continuing sales can still warrant examination of benefits through quality or delivery reliability.

Orders using existing capacity and orders requiring new investment

Sales opportunities handled with existing capacity may translate into shipments relatively quickly. Work requiring new equipment, certification or customer-specific designs cannot begin merely because a firm wants to sell. Upfront expenditure depends on the expected durability of the opportunity. Immediate shipments and expansion of future capacity are separate decisions.

A firm reconsidering deferred investment still needs evidence beyond sales prospects before approval: equipment lead times, staffing and competing uses of factory capacity, for example. No immediate rise in capital spending is not automatically distrust of policy. Distinguishing renewed planning, equipment orders, payments and operation makes the reason for a delay more concrete.

Nor should every subsequent investment be counted as an effect of the decision. Some equipment may already have needed replacement or serve other markets. If management cites the waiver, that explanation should be read alongside the plan’s date and intended uses. Such comparisons can test whether more specialized investments are more sensitive to the policy outlook.

Align the exporter’s currency with the customer’s

Currency movements can overlap with policy effects when assessing US export contracts. An unchanged dollar selling price can produce different revenue in the currency used to pay the exporter’s costs. Unchanged exporter prices can also imply different burdens for non-US customers. Payment and reporting currencies must be fixed to separate operating improvements from currency translation.

Even a currency move coinciding with the announcement has several possible explanations. Interest-rate expectations, other trade policies and global risk appetite may act simultaneously. Before interpreting appreciation as an assessment of this waiver, examine similar economies and other transactions in the same currency. One time series provides limited identification.

At company level, sales and profitability can move in opposite directions. Returning US orders may coincide with appreciation of the currency in which costs are paid, compressing local-currency margins. Cheaper imported materials could offset part of that effect. Mapping the currencies of materials, labor and selling costs alongside revenue helps explain differences between firms.

Related foundation: Exchange-rate drivers。Separate business sales from currency movements.

The announcement window and the window for observing trade

A short market-response window requires checking prior information, exchange hours and competing announcements. Daily closing prices combine other news with movements already under way. Finer timing is not always better: one transaction can look large in a thinly traded stock.

Actual trade requires a longer observation window because orders precede production, shipment and customs clearance. No change in the following week’s imports does not establish no business response. An early import increase could also consist of cargo shipped before the decision. Attribution requires tracking whether the transaction began before or after the announcement.

The windows answer different questions rather than competing. Market responses concern revised expectations; transaction records concern actual behavior. A large price response followed by little trade invites comparison between expectations running ahead and difficulties in execution. A small price response with substantial trade changes may reflect anticipation or effects concentrated in particular firms.

SG Group View: assess avoidance, resumption and persistence separately

SG Group assesses three stages: whether an anticipated burden was avoided, whether work held back by it resumed, and whether that work became a repeatable trading relationship. Jumping from the first institutional change to lasting economic effects omits bargaining, customer choice and production capacity.

Evidence differs by stage. Avoidance calls for prior estimates and response plans; resumption for orders and performance in the same project; persistence for repeat orders and renewals. Revenue alone cannot explain all three. Where prior plans are unavailable, observable subsequent behavior supports a sounder assessment than a strong numerical claim about avoided burdens.

This interpretation weakens when other explanations account for expectations, behavior and outcomes. Orders returning before the waiver may reflect demand or another policy already at work. Similar improvements among affected and unaffected firms increase the weight of common factors. Retaining cases that do not fit, rather than only successes, is essential to revising the assessment.

Three paths for subsequent trade

On the first path, deferred projects resume and further orders follow. Repeat purchases of the same specifications, supplied without unsustainable discounts, support a link from policy change to persistent trade. Renewed customer relationships with stable pricing and performance terms provide stronger support than the size of one shipment.

On the second path, deferred purchases cluster after the announcement and then return to their earlier pace. Completing delayed work can still be beneficial, but extending the initial growth rate across a year overstates future sales. Distinguishing ordinary requirements from catch-up purchases helps identify this path.

On the third path, orders barely change despite a different policy outlook. Relief may have been expected, customers may already have switched, or demand and product conditions may dominate. The task is to distinguish prior decisions from remaining constraints rather than assign one cause to a small response. All three paths can coexist across firms and products.

Track what changes within the same project

Useful evidence follows the same project before and after the announcement. Revised quotation validity, customer terms, required sales volumes or investment approvals can trace transmission into business plans. Company-wide revenue alone, potentially including growth elsewhere or acquisitions, offers a narrower basis for isolating this effect.

Customs data need to link product, origin, value, quantity and period. Composition can change within a classification, so higher unit values alone do not establish higher prices. Greater detail can improve comparability but leave fewer transactions and more volatile monthly observations. Precision and stability both matter.

Further official action should be assessed by which provisions or measures it actually changes. More commentary is different from changed application. Following published documents, their dates, effect and scope is more useful than assigning speculative dates to future monitoring decisions. The period used to assess firms can then be adjusted to actual changes.

What remains unresolved for the international economy

An international aggregate requires matching gains and losses over the same period. Preserved US exports may reflect orders transferred from elsewhere rather than additional global production. Lower buyer costs can coexist with changed income distribution if sellers reduce discounts to other customers. Moving from company cases to global benefits requires these relationships.

Post-decision statistics also struggle to recover costs firms already incurred in anticipation. Cancelled projects and suppliers never selected may leave no customs record. Studying shipped goods alone misses trades that never began. Combining transactions with prior plans helps, although firms do not disclose every plan externally.

These gaps need not halt every assessment. Orders, quantities and continuity can be examined within the observable scope. But extrapolating benefits from observed firms to all firms, or totaling hypothetical losses as actual savings, sacrifices precision. A bounded assessment allows new evidence to change the conclusion without minimizing the significance of the decision.

Test the waiver’s value through the work that follows

Begin by distinguishing a burden that did not take effect from a refund of costs already paid. Then examine prior expectations, revised decisions and work that continued. Connecting expectations, resumed transactions and persistent relationships provides a route from the policy document to economic activity.

For firms and investors, the framework does not require choosing optimism or pessimism. Flat sales may include valuable preserved trade, while rising sales may be temporary clustering. Aligning destination, product, timing and customer relationships explains why the same headline number can support different conclusions. Subsequent forecasts should follow those details rather than the force of the initial headline.

SG Group places greater weight on tracing changed decisions than on displaying a large gap against the harshest assumption. Did the waiver confirm what was already expected, restart deferred work, or sustain subsequent orders? Answering that question allows corporate profits, buyer options and global production to be assessed on their respective evidence.