April twenty-eighth, two thousand twenty-five to May second, two thousand twenty-five

I. Foundation

resistance to change and hence greater dependability for the results shown in the past.

A definitive answer to the question, “What really is a safe-haven asset?”

Many trades are labeled as safe-haven trades; it seems that anything can be called a safe-haven play, but in fact they are not really safe havens—Most people’s idea of a safe-haven asset is still essentially a risky asset, or merely a low-beta risky asset. Why did Notes create the foundation section? To make good market forecasts, you first need an accurate understanding of the market and causal reasoning—not “guesswork”!

If you take the 10% worst-performing days for U.S. stocks over the past 15 years (roughly, days when the daily decline exceeded 1.1%; this figure is also easy to remember), extract them, and aggregate the corresponding performance of the assets then called “safe havens” in global markets, would yield something very interesting (unfamiliar to most people) result:

Chart: “Which assets are truly safe havens?”

1) The Australian dollar is a typical high-beta risk currency, while the euro is relatively stable. So when markets seek safety, AUD is sold and EUR is bought relatively, causing EURAUD to rise;this currency pair is the most canonical safe-haven trading proxy and gauge。

Chart: The most canonical proxy pair—EURAUD

2) Next, we have to talk about gold—Gold really is not a stable safe-haven asset, although... many people, when they buy gold, have to find a reason for their actions and then say they are buying it for safety. Essentially, they are putting the cart before the horse and have not understood the asset’s true characteristics at all; when they get it right, they call it their “judgment skill,” and when they get it wrong and lose money, they just cry..

Gold’s characteristics are highly diverse, At times it behaves like a risk asset, at times like a way to measure fiat liquidity from global central banks, and at times like a safe-haven asset. From a portfolio allocation perspective, it can be said to have some safe-haven value, such as low correlation. But if we look only at its trading performance during stock-market crashes, it truly does not meet the standard for a short-term “safe harbor.”

Chart: To survive in the market, don’t make excuses or rationalizations for your positions—you’ll die a miserable death!

You need to have a reason before you act

rather than acting first and then trying to rationalize your behavior

Chart: Actually, this round of gold trading had little to do with safe-haven demand

After reading this Notes, everyone will be able to judge easily the true level of those who talk at length about narratives.

3) One interesting key change to point out: go back to the first chart and you can see EURUSD posted another extremely sharp upward inflection during this round of declines in U.S. stocks—a move that actually reflects a clear change in trading behavior in today’s market.Previously, The euro has risen against the U.S. dollar when equities fell only twice, once in the early days of the pandemic, and once amid unusual trading activity in 2015/2016, when German bond yields turned negative.

Chart: EURUSD showed very unusual behavior

Friends who know the Notes well, please say that line with us:

...If everything is normal, an anomaly simply cannot occur; and if an anomaly does occur, then something is already abnormal..

Works every time..

From the current situation, the shift in global safe-haven currency from the U.S. dollar to the euro is a real phenomenon. As confidence in the sustainability of the U.S. fiscal deficit and the credibility of its economic policies collapses, on days when U.S. stocks plunge, capital naturally withdraws from U.S. assets and the dollar and moves to other large, stable capital markets—such as Europe.

Going forward, as the scale of European debt issuance increases(Given Germany’s relatively low starting debt level, this could actually benefit the euro),and the possibility of Joint EU Defense Bonds, this will most likely further consolidate the euro’s position as a destination for sovereign wealth management capital.

There’s something fascinating about this: Germany and the United States are both issuing debt, yet the interpretations and outcomes differ. Everyone is playing a relative beauty contest. Even if neither looks good, being the better-looking one is enough.

Note: Notes 250428 briefly revisits Wednesday’s Bank of Japan meeting (with observation ranges for the yen and euro) and gives the euro’s observation range. If a breakout is confirmed, we’ll enter.

4) Then there are U.S. Treasuries: after the pandemic, amid high inflation and debt problems, bonds lost their sound safe-haven characteristics.

5) Silver, Bitco—don’t overthink it; they’re typical high-risk assets.

A simple aggregation; naturally, other factors also have an impact, but it explains a lot. Let this Notes post build up on its own.

2025.04.29

A study of the seasonal effect of Sell in May

5 itself is the S&P’s eighth-best month of the year, It usually starts to weaken around midmonth, then rebounds as Memorial Day approaches (May 26, 2025):

Chart: S&P seasonality (focusing on 5)

Among seasonal effects, Sell in May is arguably one of the best-known. Its The trading logic is to sell stocks at the start of 5 (and stay out of the market for the following 6 months) and buy them back on 10 31.

Over the past 75 years, The recommendation to stay out of the market for six months under Sell in May has had some validity，

because returns over those six months have indeed lagged significantly behind other periods:

Chart: Average monthly change in the S&P over the past 75 years

Chart: 5.1 through 10.31 really are the 6 worst months for U.S. stocks

However,, the Sell in May trading logic actually has two more important nuances,

For professional investors (a consistent requirement for members of the community), these two things are essential to know:

1. The effect remains, but its impact is weakening (especially in recent years)

The chart below shows the difference in returns between the 6-month trading periods of the S&P 500 “Buy in 11” (Buy in Nov, i.e., from 11 to 4) and “Sell in 5” (Sell in May, i.e., from 5 to 10):

The gap was very pronounced in the 1970s—the performance from 5 to 10 lagged markedly. Since the 1970s, the moving average of this gap has gradually narrowed and now stands at just 3.2%,In other words, “Sell in May” still works, but it has indeed become much less effective.

2. The effect is stronger in years of economic downturn

When the market performs poorly at the start of the year, the Sell in May seasonal rule tends to be more reliable.Since 1950, the S&P 500 has fallen by more than 5% through 4 30 YTD only 14 times.

In those years, the subsequent period from 5 to 10 saw an average decline of 6.2%—In other words, a poor start to the year makes a pronounced “Sell in May” effect more likely.

So, in a down year like this one, Sell in May may have a stronger effect.

2025.05.02

II. Insights

Our Viewpoints and ideas on market.

Another macro-level (counterintuitive) but more likely combination of paths

The U.S. session is stuck in an awkward position at 5500. In fact, bears are clearly more active than bulls in the market, but they are also more cautious, That is the paradox of an efficient market: when everyone is waiting for a signal to go short, the signal ends up taking a long time to arrive.Everyone understands that the anticipated decline is likely to happen, but no one is willing to make the first move, because short selling demands far more smoothness than going long; a decline that does not unfold smoothly is very, very bad.

Chart: The market is right at this major technical resistance level, 5500

We are bearish on the U.S. session over the medium term. In the short term, however, a highly counterintuitive but realistically plausible macro path is taking shape—This path is actually more likely and more realistic than a recession, namely, a combination of slowing growth and rising inflation emerging over the next 1-2 months.

The logic is actually quite straightforward: when faced with rising costs, companies will always prioritize price pass-through (The path of least resistance is naturally to pass the costs directly on to consumers),Clearly, this worked very well in 2022, so it is worth trying again this year..

Corporate earnings are nominal, which means that as long as inflation is high enough, reported earnings can still rise even as real demand weakens. If this path plays out, it will most likely be the main reason U.S. stocks can maintain their performance in the short term—earnings and share prices supported by inflation itself.

Of course, this is not good news in itself, We have already seen how a similar situation ends in 2021-2022: amid high inflation, even when earnings figures look good, overall market sentiment gets worse and worse, so even if earnings figures temporarily support share prices, that may only delay the eruption of the real problem.

Chart: Looking back at 21-22, if it happens again, the path would head toward 4500

If this path plays out, a more natural consequence is that the probability of no rate cut in June will be much higher than the probability of a cut.

Another structural issue is that The U.S. economy is services-oriented. In fact, manufacturing now contributes very little to employment, so even if manufacturing is hurt (for example, by tariffs), the short-term impact on overall employment figures will not be particularly severe.

In other words, if headline growth does not collapse immediately, inflation may rise first, forcing the Fed to stay on hold. The greater risk to watch is bond vigilantes risk: if long-term U.S. Treasury yields surge again and the Fed cuts rates rashly at that point, it could trigger even greater turmoil at the long end.The bond market’s basic mechanism is to force policy discipline through selling; once it is seen as abandoning discipline, selling pressure on long-term yields will intensify further.

So even from a financial-stability perspective, especially with bond vigilantes in mind, the Fed is more likely at this stage to wait and see rather than ease.In short, unless growth suddenly falls off a cliff (unlikely, because the data lag), or the market experiences systemic panic, the bar for a Fed rate cut in June is very high.

More likely, inflation will rebound and force the Fed to keep waiting.

Bears wait, market sentiment builds... downside breakout.

2025.04.28

Revisiting that massive rectangle that lasted 4700 days 20 years ago: from TINA to US Exceptionalism, a once-in-a-century election beauty contest (why FX/the dollar is key), and another discussion of value and spreads (why not to chase positions)

A systematic narrative is unfolding—Global Steady Hands (the long-term institutions that have stood firmly by U.S. assets for many years) are indeed steadily reducing their holdings of the dollar, U.S. stocks, and U.S. Treasuries, and this behavior is not driven by news; it is a genuinely systematic, sustained reduction—capital really is leaving the United States (although the pace has now slowed somewhat).

Chart: The Yale Model, pioneered by Swensen, is also fading

The last systematic outflow happened in 01-02, when the dot-com bubble burst:

Chart: From a broader perspective, after this outflow, U.S. stocks actually traded in a huge rectangle lasting nearly 4700 days

At the time, the dollar was also much stronger (far stronger than now):

Chart: With the dollar index overlaid

In “Market Illusions and Reality,” we wrote:

“One detail is worth noting: after the extreme volatility at the start of 4, market volatility overall has contracted sharply. Not only stocks, but Treasury yields have also largely stabilized. But the dollar remains near its recent lows, far below its level before the tariff plan was announced.If we are looking for evidence of a genuine structural change in the market since the tariff event, the dollar’s performance may be the signal most worth tracking”

https://mp.weixin.qq.com/s/vliMzuQo7E_I6Ssy1Un3Xw

Chart: U.S. stocks began recovering after April seventh, but the dollar continued to fall

After the tariff event, the dollar is the only indicator still stuck at the shock level..

Chart: The confirmation between the dollar and U.S. stocks

After the previous wave of outflows, the dollar also hovered at low levels, the gains in U.S. stocks within the rectangle never received breakout confirmation from the dollar, After that, it was actually the dollar’s breakout that decisively confirmed the renewed breakout in U.S. stocks。

In other words, relative to U.S. stocks, the dollar is actually a more accurate signal for judging the overall positioning of the United States, because its pricing is done more by major institutions.

Go back 24 years: back then, capital from global markets was also rushing into the United States, How did everyone interpret it at the time?

“There is no alternative”… a phrase with a strong historical resonance, TINA..

And today, 24 years later...?

The sons of those who once walked Wall Street have taken over, offering a term that mirrors their fathers’: US Exceptionalism

From TINA to US Exceptionalism, it is essentially the same thing.

Correspondingly,

the collapse of TINA and the end of US Exceptionalism—American exceptionalism—are naturally one and the same.

Chart: Notes 250429 definitively answers the question of what truly is a safe-haven asset

This is the beauty contest of the century, and what the market always trades is the “spread.” Because “value” has a different answer in everyone’s mind, while the “spread” is plainly visible—this was Graham’s final counsel to the world in Security Analysis.

Sell dollars, buy gold, buy non-U.S. currencies—some things are beyond Trump’s control, no matter how much he changes course, because he himself has set this change in motion, and it may well be irreversible.

Looking back at the U.S. business model over the past 35 years, it was essentially a process of continually accumulating a negative NIIP (Net International Investment Position)—that is, many non-U.S. countries around the world sold goods to the United States, received dollars, and then reinvested those dollars in U.S. assets.This made the United States, in essence, a “financial debtor” to the rest of the world: what it owes is a capital position, not physical debt.

And which countries are the true holders of positive NIIP? In other words, who really owns large amounts of U.S. financial assets?

Japan ranks first, with a net position of 3.3 trillion dollars. Second is not China, but Germany.

If we exclude locations whose data are distorted by offshore financial centers and return to the level of actual economies, cleaning the data reveals that Europe’s overall NIIP is very high; in fact, the European region as a whole is one of the world’s strongest sources of capital outflows.

Like the United States, Germany’s business model over the past 30 years has also been very clear: use cheap energy and relatively inexpensive labor to make high-quality products for large-scale export. The result is that Germany accumulated a large surplus, and that money ultimately flowed into overseas asset investments. That is why Germany’s NIIP (Net International Investment Position) ranks so highly.

The Netherlands is similar (mainly its large pension institutions), followed by Norway, China, Canada, Switzerland, and South Korea, all of which are also high on this list.

In fact, because countries have different economic structures, they naturally follow different paths in allocating dollar assets:For example, Germany is led by insurance institutions, including giants like Allianz; Norway by its sovereign wealth fund; and Canada mainly by pension funds. What these differently structured pools of capital have in common is that they all hold large amounts of U.S. Treasuries and U.S. stocks.

Another implication of holding U.S. stocks and Treasuries is that behind every purchase there are always two things—U.S. stocks (+ dollars) or U.S. Treasuries (+ dollars).，Any portfolio of U.S. assets will inevitably include dollar exposure, and dollar exposure is often managed as a separate asset class—This is the core reason the dollar is a stronger positioning signal than U.S. stocks: stocks are the outcome, while the dollar not only moves first but also carries far less noise.

Generally, the dollar is negatively correlated with risk assets: if the stock market plunges or credit blows up, the dollar usually rises and helps protect the portfolio. If stocks rise and the dollar rises too, that naturally also acts as a positive-drift hedge, This means institutions usually do not need to hedge their dollar exposure completely, and in practice that is exactly what they do—

This arrangement is very comfortable when the dollar and U.S. stocks confirm each other (that is, when the United States is in a period of TINA or US Exceptionalism). But if the end of TINA/US Exceptionalism is confirmed, they will have to start rebuilding the dollar hedges they should have held, and may even need to hold more..

It is precisely the start of this contrarian trade that creates a negative trading logic for the dollar as a whole, institutions that have not sufficiently hedged their dollar risk will have to start selling dollars steadily. This is a genuine rotation: if others are selling, you have no choice but to follow suit—

That is why we say the key to this major transformation is not in the stock market or the bond market—it is in the foreign exchange market.

Chart: WeeklyWrap 250427

This is also the core source of our bearish view on the dollar in Weekly Wrap 250427.

Long gold, short the dollar—it is destined to become the core conviction of global macro investors. I hope everyone truly understands this today.

Chart: Notes 250425 The (only) reason to sell gold (and every trade)

Do you remember why we made a large bet on gold?

“Structural buying by sovereign institutions”

The same naturally applies to the dollar: a wave of structural selling by sovereign institutions (including right now) is on the way. This is the holy grail of macro trading. Sovereign institutions hold nearly 32 trillion dollars in U.S. assets, and Japan alone has a huge pool of 3 trillion dollars, not billions, but trillions. That is an enormous figure, meaning that even adjusting only a portion of it would take a very, very long time; even hedging and adjusting just 1% of the asset allocation would generate enormous capital flows.

So, to answer another question from a community reader: why did we choose to take profits instead of rushing to chase either a short or a long position?

Once the conviction to Sell the US takes hold, it becomes a long-term structural theme, not something reversible through short-term policy.While policy remains active, there is no need to chase 1% and 2% volatility at the risk of a short-term short squeeze.

This will be an allocation trade cycle measured in years; we can enter whenever we want.

As I keep saying:

2025.04.30

When will valid data arrive

The labor market is almost certainly weakening, but what really matters is the hard data for 5 and 6 (to be released in 6 and 7, respectively).In other words, Data released before 6 do not truly account for the impact of Trump’s Liberation Day; instead, front-loading driven by uncertainty will distort the overall data, The data are not truly “valid.”

Although hiring freezes and stalled investment do slow economic growth, the first wave of impact may instead be “front-running”—that is, a rush to buy goods before tariffs take effect. Therefore, In the short term, a series of consumer data releases could instead show “false prosperity,” creating a market illusion that is disconnected from the real fundamentals.

That is why, in the current environment, economic data such as retail sales and personal consumption expenditures (PCE) have little reference value and may even be completely distorted. When consumers face clear expectations of imminent price increases, they often choose to bring forward and concentrate their purchases, while companies stockpile and replenish inventories as much as their capacity allows.

When goods subject to higher tariffs finally reach the market, the purchasing data will show a gap—because 1) consumers will already have bought everything they need; 2) the goods will no longer be attractively priced.Although price elasticity and substitution effects may provide some marginal cushioning, the overall transmission of the shock cannot be fully absorbed.

Before June, another expected (and indeed observed) development is a divergence between hard data and soft data (that is, sentiment)after investors, consumers, and businesses all fell into a state of paralysis because of policy uncertainty, This divergence between soft and hard data is the main source of market fatigue—and starting next month, we will truly see the pause period expire after 90 days, along with the release of a series of genuine economic indicators (hard data).

This means the next three months will be crucial. Volatility will truly begin to expand, and the period that will determine the annual trajectory of the U.S. dollar, U.S. equities, U.S. Treasuries, global foreign exchange markets, and global commodities markets is approaching—whether bullish or bearish, the true trading window is arriving, and this will also definitively set the tone for this round of tariffs.

2025.05.01

III. Market Pulse

Our latest Comments on the economic and market environment

A brief follow-up on Wednesday’s Bank of Japan meeting (with the yen observation range)

Over the weekend, I mentioned that last week we gradually closed several trend positions in the currency market, especially in the yen; we are flat and waiting.

Chart: The key observation range is 140-144

This meeting is quite important. So far, the market has barely priced in any clear outcome, but Given that the Nikkei is currently in the middle of its range, inflation remains above target, USDJPY is stable, and there is substantial gamma cushioning between 140 and 135，It is also possible that the Bank of Japan will choose to make a small rate hike during this window, in step with U.S. policy.

Chart: The Nikkei 225 is also in the middle of its range

Additionally, Even if the meeting ultimately makes no changes, shorting USDJPY near the 144 level is still a reasonable setup.

However, we would not specifically trade such a narrow range; this is just a brief indication of our basic view.

A weakening pattern does not mean the trend has ended, but consolidation may become the dominant rhythm until new information with a material impact emerges.

Chart: Finally, the euro trading range as a bonus