How do you measure chaos?

Text-based Economic Policy Uncertainty indices have shot up. Markets are good at pricing known economic risks, but a policy regime that keeps changing is beyond the usual econometric models. Erratic executive orders on trade now move corporate cash flows more than consumer demand or productivity growth do.

How uncertainty gets counted

The best-known measure, built by the economists Scott Baker, Nicholas Bloom and Steven Davis, counts newspaper articles that mention the economy, policy and uncertainty together, and scales them against the total volume of coverage. A companion trade index, developed by researchers at the Federal Reserve Board, applies the same idea to articles about tariffs and trade policy. Neither index predicts anything by itself. Their value is that they turn a vague sense of confusion into a series that can be compared across decades and set against investment, hiring and asset prices.

The source of this spike is plain. In his first days back in office the president declared a national emergency on the southern border, ordered a sweeping review of trade policy and said he was considering 25 per cent tariffs on Canada and Mexico from the start of February, along with new duties on China. None of those threats has yet been written into an order, and their timing and scope keep shifting from one public remark to the next. For an index that counts articles, a policy announced by comment, retracted and restated produces far more coverage than a single formal decision would.

When correlations break

Quant funds and risk-parity desks hedge using historical covariance matrices. When trade policy can change over a weekend by social media post, the old factor correlations stop holding. Safe-haven assets behave erratically, and cyclical stocks drift away from the usual leading indicators. Measures built from newspaper coverage and regulatory filings put trade uncertainty at levels seen only during the 2020 pandemic and the 2008 banking panic.

The technical problem is the estimation window. A covariance matrix built from the past two or three years of returns reflects a period when trade policy was stable and the main driver of markets was interest rates. If tariffs now move currencies, industrial shares and inflation expectations together, the historical matrix understates how much a supposedly diversified portfolio can lose on a single announcement. Models that re-estimate quickly react to the change but also overreact to noise, so managers have to choose between being late and being jumpy.

Paying for protection

Faced with that much ambiguity, managers put off long-term investment and hoard cash. In the options market the caution shows up as steeper implied volatility skew in sectors exposed to cross-border trade. Buyers are paying more for protection against a sharp fall in carmakers, industrial exporters and retailers that depend on imports than against a broad market decline.

Cash hoarding has a measurable cost. A company that delays a factory or a new supplier contract waits for the policy picture to clear, and the orders it would have placed with equipment makers and builders disappear from the data for as long as the delay lasts. Research using the policy uncertainty indices has repeatedly found that firms most exposed to government decisions cut investment and hiring more when uncertainty rises. For investors, the effect shows up first in capital-goods orders and in the guidance of industrial companies, well before it appears in GDP.

What happened in 2018 and 2019

The first trade war is the natural precedent. Trade policy uncertainty rose sharply in 2018 as tariffs on steel, aluminium and Chinese goods came in waves, and it rose again in 2019 when talks with Beijing stalled and fresh duties followed. Business investment in America stalled during 2019, and the Federal Reserve cut rates three times that year partly in response to the drag from trade. The lesson for this episode is that uncertainty on its own, before most tariffs are in force, can slow capital spending enough to show up in growth.

Who pays for the noise

The losers are companies with long supply chains and large foreign sales: carmakers with plants in Mexico, machinery firms selling into China, and retailers that import most of what they sell. Domestic service businesses, from utilities to regional banks, are largely insulated. In currencies, the Mexican peso and the Canadian dollar carry the most direct risk, while the dollar tends to gain from the demand for safety that uncertainty creates. Options sellers collect richer premiums, which suits volatility funds as long as the threats are not carried out all at once.

The case against reading too much into it

Text-based indices measure how much journalists write about uncertainty, and a president who makes news constantly will push them up whether or not policy changes. Broad equity volatility has stayed well below crisis levels, and credit spreads remain tight, which suggests markets see much of the noise as negotiation. Those points are fair, but they concern the level of the index more than its effect: even if many threats are never carried out, companies deciding where to build a factory must plan for the ones that might be. The damage shows up in postponed investment, which no volatility index captures directly.

What would lower the readings

The signals to follow are the February deadline for Canada and Mexico, the departmental trade reviews due in early April, and whether tariff threats begin to arrive as formal orders with clear timetables. A written schedule, even an aggressive one, would reduce uncertainty more than any softening of tone. As long as trade rules depend on the president’s mood, investors will mark down every cross-border revenue stream, and companies with big foreign sales should expect lower multiples for as long as the uncertainty lasts. The discount will probably widen through the spring, with industrial exporters and importers of consumer goods trading below domestic peers until the reviews are published.