Summary

The speaker has cut this year’s baseline equity exposure from “100% stocks plus puts” to 60%-65%, and is personally at 65%; the core issue is not turning bearish, but a change in the risk regime. Last year, economic and earnings expectations were being revised higher in tandem, and buying puts in a low-volatility market allowed investors to use the correlation spike and volatility surge during a selloff to fund additions. This year, policy uncertainty could prolong any correction, options are more expensive, and upside is more limited: “the size of the tail risk is completely different from last year.” Even at a full 100% allocation, he would consider using calls while keeping cash on hand to navigate policy shifts.
Selling everything or rotating all U.S. equities into China and Europe is not the speaker’s answer either. The U.S. retains economic and technological advantages, and he expects the market to remain “steadily rising, but probably not by much,” with a more volatile path. China and Europe offer short-term rebounds and valuation repair, not the kind of long bull market suited to buy and hold: “there is no long bull market in a deflationary economy.”
Underlying U.S. inflation is still declining gradually; the real uncertainty is tariff policy, not the current macro data. Core CPI was only about 11bp above expectations, and after incorporating PPI, the speaker believes PCE could move toward 2.5%, or even 2%, in a no-tariff-shock baseline. Nonfarm payrolls are averaging roughly 170,000-180,000, wage growth is moderate, and immigration is expected to fall from a peak of 3M-3.5M to 700,000-1M this year, reducing the risk of another acceleration in rents.
Reciprocal tariffs look more like a starting point for negotiations: the short-term shock is smaller than an outright announcement of universal tariffs, but the eventual tax burden will still rise. The speaker expects China to face another 10% on top of existing tariffs, bringing the cumulative increase under Trump to 20%, with additional duties on critical sectors including semiconductors, pharmaceuticals, autos, and steel and aluminum. The full package could lift the effective tariff rate by roughly 4.5 percentage points; he also cited a further 4.5%-5% increase for other regions, ultimately adding 0.4-0.5 percentage points to PCE. “Tariffs probably won’t blow up the market,” but the negotiations will keep generating negative news.
The rates market is now much closer to fair value than it was in January, and a 10-year Treasury yield around 4.5% also looks near fairly priced. If tariffs push the PCE path from 2% toward roughly 2.5%, the policy rate could remain above 3%, but “rate hikes are impossible.” The market is pricing roughly two cuts, landing in late 2025 to early 2026; the speaker does not yet see a sharp rise in rates or severe liquidity outflows before year-end.
U.S. earnings have not collapsed, but the market is moving from two straight years of upward revisions to pre-pandemic-style normalized growth, so expensive valuations no longer justify a full allocation. EPS is expected to rise about 13% for large and technology companies and about 8% for mid-sized companies, but fourth-quarter 2024 earnings revisions have already turned negative—at minimum, a sign that expectations for 2025 and 2026 are no longer as high. 2025 earnings could still grow 10%-11%, just not at the post-pandemic pace.
DeepSeek has pushed the trading narrative into phase three of AI applications, creating a potential opportunity in software and applications, but it is not a trade to hold for long. NOW and Snow charge based on usage, while Microsoft can charge per seat. Lower training costs will drive experimentation and continue lifting demand for data centers and chips, but user habits and productivity gains take time to materialize, creating the kind of internet-bubble mismatch in which “market expectations have already risen, while actual usage has not.” More certain healthcare names such as TEM and PANT, as well as defense, are already very expensive. China’s rebound should likewise be treated as a tactical trade: the repair is not over, but the speaker believes “it will definitely stop before April.”

Deep dive

1. Still bullish on U.S. equities, but only willing to carry 60%-65% risk exposure

Last year’s premise was that the U.S. faced no recession risk, while earnings, earnings expectations, and earnings revisions all moved higher. The market rose in a straight line and the Sharpe ratio was very high; the opportunity cost of giving up 40% exposure was far greater than the cost of protection.
The puts were a correlation trade: correlations between individual stocks were low and volatility was cheap on the way up. Once the market fell, “all stocks became correlated at one,” the puts could multiply in value, and the proceeds could be used to add exposure.
This year’s policy risk cannot be priced with precision, and any correction could last longer. Options are also more expensive and upside is more limited, leaving a market that may rise steadily but with large swings along the way—and making drawdowns harder to control.
The speaker is 65% allocated to U.S. equities and holding cash. Selling everything would mean giving up the U.S.’s economic and technological advantages; if he were to run a full 100% position, he would favor calls while preserving liquidity for sudden policy changes.

2. Inflation remains contained as labor and immigration return toward pre-pandemic norms

Core CPI was about 11bp above expectations, prompting only a limited market reaction. Combining CPI and PPI suggests PCE could come in slightly below expectations; the normal path remains a gradual decline from roughly 2.8% toward 2.5%, and eventually even 2%.
Recent nonfarm payrolls have averaged about 170,000-180,000, while the speaker said the unemployment rate had fallen by roughly 0.25 percentage points and wage growth remained moderate. Rent is a lagging indicator, and already-high interest rates have eased rental pressure; there is not yet evidence of the conditions needed for a broad rebound.
Immigration rose from roughly 1M a year before the pandemic to 3M-3.5M, is now around 1.7M, and is expected to fall below 700,000-1M in 2025. That could deliver a negative impact of roughly 30-40 points to GDP, but it does not amount to a recession.

3. Tariffs will lift prices, but the reciprocal framework turns the shock into a prolonged negotiation

The speaker sees reciprocal tariffs as a short-term positive: the message is essentially, “Why don’t you all cut your tariffs,” suggesting Trump values negotiation rather than an immediate, across-the-board shock to the economy and markets.
A universal tariff could still be introduced before April, while reciprocal measures may not be implemented until before the end of April. China is expected to face another 10%, taking the cumulative increase under Trump to 20%; semiconductors, pharmaceuticals, autos, steel, and aluminum could also face additional rates as critical industries.
A standalone 10% increase on China would, in the speaker’s estimate, lift the U.S. effective tariff rate by roughly 1.5 percentage points and CPI by 10-15bp. Applying all of the proposed tariffs could raise the effective tariff rate by about 4.5 percentage points. He also cited a doubling of China’s tariff rate and another 4.5%-5% increase for other regions, ultimately adding 0.4-0.5 percentage points to PCE, using Goldman Sachs’ methodology.
This would push the inflation endpoint from 2% toward roughly 2.5%—enough to keep the policy rate above 3%, but not enough to restart rate hikes. Two cuts and a 10-year yield around 4.5% already look close to fair value; the main risk is a steady stream of negative news during the negotiations.

4. Earnings are still growing, but negative revisions show high-growth expectations normalizing

EPS is expected to grow about 13% for large and technology companies and about 8% for mid-sized companies. The issue is not current earnings, but the fact that fourth-quarter 2024 earnings revisions turned negative after two straight years of upward revisions. That at least shows expectations for 2025 and 2026 are no longer as high, though it does not necessarily mean the market must correct.
The speaker expects 2025 earnings to deliver 10%-11% low-double-digit growth, but immigration, revenue, and corporate expansion are all reverting toward pre-pandemic conditions. “The 2023 and 2024 situation cannot continue forever.”
With valuations still expensive, the market may continue to rise modestly, but this is no longer a high-Sharpe bull market. The investment conclusion is to reduce exposure rather than leave the U.S., and not to change the core buy-and-hold allocation because of short-term rebounds in China and Europe.

5. AI enters the application trade, with expectations likely to rise before productivity materializes

The first 2 phases involved building databases and related infrastructure and buying chips. After DeepSeek, the market began trading phase three: more companies experimenting with AI applications, while software and SaaS companies actively reposition their narratives around AI.
Models such as NOW and Snow depend on actual usage of AI services, while Microsoft can charge per seat. DeepSeek convinced the market that training and distillation costs could be very low and that everyone could experiment, but monetization will ultimately depend on real usage.
Wall Street tends to price new technologies on a “fast first, slow later” log curve. AI applications, however, require users to build new habits gradually, while productivity gains and actual usage have not yet materialized. The speaker compared this with the internet bubble: “market expectations have already risen,” while productivity and actual usage have yet to catch up.
With the fourth quarter’s economy still solid, application companies could post decent earnings and “tell a lot of stories,” creating a short-term opportunity. But “AI applications definitely take time,” so the trade should not be held for too long. More certain areas in healthcare, including TEM and PANT, as well as defense, are already extremely expensive.

6. China and Europe are upside risks, not long-term fundamental reversals

Europe’s catalyst is a ceasefire in Ukraine. If it happens, it could deliver a wave of gains; if investors go long in advance and the event fails to materialize, “you could get trapped.” Even if it does happen, the speaker expects a sudden rebound followed by renewed weakness.
China’s catalysts include the private-sector symposium, the President’s participation in the symposium, M1 returning to positive growth, and signs that the monetary easing deployed since last September may finally be taking effect. The market had feared harsher tariffs and compressed valuations; the initial 10% increase instead triggered a recovery. But the official emphasis that “private enterprises are extremely important” is almost identical to the framing in 2018.
The speaker believes the real constraint on Chinese manufacturing is not U.S. tariffs but land and property. Companies can continue operating at negative gross margins, driving bank lending; they can use loans to buy land, repay loans, and profit through property. Once land prices can no longer rebound sustainably, investment in new products, equipment, factories, and land loses its economic rationale, leaving only “ever-intensifying competition” and continual cost-cutting.
The deeper difference is the household balance sheet. By the speaker’s figures, disposable income in the U.S. private sector is above 80%, versus about 40% in China; debt costs are also twice as high in China and unevenly distributed. Policy statements cannot directly repair the balance sheet, so this cycle is merely “a quick rise, a quick valuation repair, and then a slow continuation of the decline.” The speaker believes the current recovery is not yet over, but will definitely stop before April, adding that he would revisit the issue next week if anything else changes.

Ravi Gupta - AI or Die - [Invest Like the Best, EP.411]Macro Conversations 75: Ne Zha 2, DeepSeek and Spring Festival Economy