Summary

If tariffs land on April 2 at the bearish end of expectations, the most likely initial reaction is a “relief rally”—but that is a window to cut risk, not evidence that the risk has cleared. The speaker’s short-term framework is: Trump’s remarks drive one move, the policy announcement drives another, and the market reverses when implementation begins; a 15%-18% effective tariff is already fully priced, while a grace period, negotiations and exemptions could deliver a positive surprise.
Tariffs have gone from a low-probability tail risk at the start of the year to the central variable capable of pushing the odds of a U.S. recession to roughly 50%. He cited Goldman Sachs estimates including an effective tariff of about 10%, with a 5% base case and 15% risk scenario; his rough calculation is that Goldman’s Q4 growth forecast of about 1.5% could fall to roughly 1%, with the actual recessionary hit appearing only in Q2 or Q3.
The paradox of Trump’s tariff policy is that America’s consumption structure does not support bringing low-margin manufacturing back home. Chinese auto manufacturing may generate only about 5% margins, versus as much as 80% for FSD, 80%-90% for chips overall and roughly 60%-70% for IC design; globalization left the high-margin profits with U.S. companies, so tariffs will initially hit those multinational profit engines rather than automatically revive manufacturing.
Tariffs will simultaneously weaken growth and deliver a one-off price shock, creating a perfect storm in which the economy deteriorates just as the Fed’s ammunition is exhausted. The speaker does not see tariff inflation as a wage-price spiral; he cited Goldman’s inflation forecast of about 3.5% and said even another 5 percentage points of effective tariffs might not add much to headline inflation. The Fed may cut once in July, but a larger move may have to wait until year-end unless tariffs, DOGE and immigration policy combine to produce a deep recession.
The first phase for equities is “valuation first,” followed by “earnings,” and the latter may not be led by mega tech, which has already been re-rated. Semiconductor valuations are reportedly at decade lows, with forward P/E around 20x; NVIDIA could still grow more than 50% this year and about 20% next year. If the economy weakens, financials, industrials and services may be more vulnerable than mega tech.
To express a recession view, the speaker prefers buying CDS, shorting credit or shorting Hong Kong stocks rather than chasing already battered U.S. tech. CDX and credit spreads have only returned to levels near last August, and another 100-200 bps of widening in a recession would be “entirely normal”; his Hang Seng put position is small but has broadly covered his U.S. equity losses over the past two weeks.
China may still be significantly overpriced, while long-duration Treasuries are not viewed as an ideal recession hedge. China faces tariffs of 39% plus another 25% under his framing, while tax revenue had already fallen about 3.8%-3.9% in January and February, before tariffs took effect. He expects the yield curve to continue steepening: “don’t just buy TLT,” while CDX or a Hong Kong short may offer a more direct recession hedge.

Deep dive

1. The Sell-Off Ahead of Tariff Day Has Already Front-Run the Policy

The speaker revisited his mid-March call: CTA buying, quarter-end pension-fund rebalancing and vol-control strategies had largely exhausted their selling, while risk appetite was near an extreme low—conditions that should have supported a modest rally into month-end. The real event to avoid was the final tariff policy on April 2 and the March economic data due afterward.
But Trump’s repeated references to reciprocal tariffs, an additional 25% levy on countries importing Venezuelan oil, and potential “secondary tariffs” on Russian oil pushed the market into a sell-off 3-4 trading days early. Under his framing, China already faces 39% plus another 25% in tariffs, making the amount of relief potentially exchanged in a TikTok deal “irrelevant.”
The short-term trading framework is a three-step sequence: “Trump tweets, the market drops or rallies; the policy comes out, and it makes another move; once the policy is implemented, the market reverses.” Canada and Mexico tariffs followed a similar path, so the first reaction after April 2 may not be another leg down.

2. Bearish Expectations May Under-Deliver, but Tariff Fixation Has Changed the Full-Year Outlook

The speaker acknowledged that he is making “a fairly major adjustment”: at the start of the year, he believed tariffs, DOGE and deportations of illegal immigrants would have to combine to create a “blood-in-the-streets” tail risk, and thought the probability was low. Now, Trump’s fixation on tariffs alone may produce a much more severe growth downside than he had initially expected.
He cited different Goldman Sachs estimates: one puts the effective tariff at about 10%, while another uses a 5% base case and a 15% risk scenario. Market consensus has already reached 15%-18%. Tariffs on autos and auto parts are effectively certain, while chips, metals and levies on 55 countries are also considered high probability; he said those 55 countries account for 80%-90% of the relevant global trade relationship with the U.S.
His blunt revision was: “The simplest explanation is usually the right one: when someone tells you they are going to do this, they are stubborn enough to actually do it” (“大道至简，人家告诉你我要这么干，他就是铁了头真的要这么干”). If diplomatic disputes are all converted into tariff instruments, the 15% risk scenario is “very likely” and can no longer be dismissed as mere negotiating leverage.
Still, an announcement is not the same as immediate enforcement. An executive order could provide roughly a one-month grace period, followed by negotiation, back-and-forth, withdrawals and exemptions. With expectations already “extremely bad,” any final rate below 15%-18% would be enough to trigger a relief rally; the key variables are the implementation date and durability.

3. Rebuilding U.S. Manufacturing Conflicts With America’s Own Consumption Structure

The speaker’s causal chain is straightforward: over the past 20-30 years, the U.S. outsourced low-end manufacturing while retaining Wall Street finance and Silicon Valley innovation. U.S. companies captured the high-value-added profits, using intellectual property, global influence and dollar power to support household consumption and fiscal transfers.
He used the auto industry to break down the profit pool: Chinese vehicle manufacturing may generate only about a 5% margin, or even a negative margin; FSD can reach 80%, chips overall 80%-90%, with manufacturing around 50% and IC design roughly 60%-70%. “Chinese workers do enormous amounts of production, but they don’t make money.” The entities that actually became wealthy and consumed were U.S. companies and households.
The U.S. represents only 5% of the world’s population but accounts for roughly 30% of consumption, making it impossible to bring all low-margin production back onshore. Even if supply chains leave China, they are more likely to move to India, Vietnam or Africa. China, Japan and South Korea joining forces as the “secondary side” would still not change the power of the demand side: “The secondary side never has bargaining power with the principal.”
That is the most direct capital-markets implication of tariffs: they reduce U.S. demand and block that demand from flowing through to the rest of the world. The first casualties are U.S. multinationals whose profits depend on globalization, not U.S. workers who immediately benefit. Manufacturing has yet to return, while high-valuation global companies have already absorbed the valuation discount.

4. Recession Will Arrive in Two Stages, While Inflation Ties the Fed’s Hands First

The speaker divides the decline into two lags. The first trades recession expectations, with global internet, high-tech and AI companies taking the initial valuation hit. The second waits for tariffs to enter the economy in earnest, when the actual recession and earnings decline are priced; that could happen in Q2 or Q3.
Using a rough conversion, he said Goldman’s Q4 growth forecast of about 1.5% could fall to roughly 1%. If tariffs are fully implemented and trigger a recession, the probability could rise to about 50%. Conversely, if the final measures are materially softened, the current adjustment may remain largely an April expectations shock.
Tariff inflation looks more like a one-off price pop than a persistent spiral: an imported product that originally cost RMB10,000 might jump to RMB16,000 after tariffs, but overseas producers could subsequently cut production costs to RMB8,000, preventing prices from continuing to rise at the same pace. The risk is that the shock arrives rapidly in Q2 or Q3; he cited Goldman’s inflation forecast of about 3.5% and said even another 5 percentage points of effective tariffs might not add much to overall inflation.
The Fed may therefore become less hawkish than before, but it cannot immediately deliver aggressive easing. The speaker guessed that it “might” cut once in July, with more forceful action deferred until year-end, after tariff inflation fades. His summary: “You’ve damaged the economy, but the Fed’s ammunition is gone,” a perfect storm for risk assets such as the Nasdaq and Bitcoin.

5. Any Rally Should Prioritize Cutting Financials and Industrials; Mega Tech May Not Lead the Second Leg Down

The speaker drew an analogy with 2018: the S&P 500 fell roughly 25% in Q4 2018, initially led lower by large tech. But after “valuation first, earnings second,” mega tech became relatively defensive, while companies tied to the real economy suffered more in the second half.
His rationale is not that tech earnings are fully immune, but that valuations and growth provide a buffer. Many semiconductor names are reportedly trading at decade-low valuations, with forward P/E around 20x. NVIDIA could still grow more than 50% this year and about 20% next year, while companies such as Marvell could still post 70%-80% growth over the next two years.
Microsoft, Google, Meta and Amazon may slow their data-center investment, but he believes their growth next year remains relatively secure—not necessarily enough to destroy the earnings outlook for the related chipmakers. If a rally follows tariff implementation, the speaker would continue cutting exposure, prioritizing financials, industrials and services rather than mechanically selling mega tech that has already been heavily re-rated.

6. Credit, China and the Curve Are the Trades He Believes Remain Underpriced

Equities have already fallen, but CDS have lagged materially. Even after their recent rise, credit spreads are only near last August’s levels, nowhere close to recession pricing. The speaker’s preferred expression is to buy CDS, short credit or focus on CDX, because a 100-200 bps widening during a recession would be “entirely normal.”
“The rest of the world is what’s really overpriced,” especially China. His Hang Seng put position is small but has broadly covered his U.S. equity losses over two weeks. He believes the “East rises, West declines” narrative driven by DeepSeek overlooks ultimate U.S. demand and the targeted nature of the tariffs.
His data point is that roughly 50% of China’s GDP growth last year came from exports, by his estimate. Yet tax revenue had already fallen about 3.8%-3.9% in January and February, before tariffs were implemented, while non-tax revenue grew about 11%. His sharpest analogy was: “If the U.S. catches a cold, the world’s heart disease is China’s heart disease.”
On rates, the curve should continue to steepen as recession and inflation expectations coexist. Hence: “Don’t just buy TLT.” Long-duration bonds are not an ideal U.S. recession hedge. If forced to choose, he still prefers CDX, buying credit protection or shorting Hong Kong stocks.

Darren Farber - The Business of Defense - [Invest Like the Best, EP.417]E186 | The U.S. Stablecoin Issuance Wave: How Blockchain Is Reshaping Traditional Finance