Like I pre-warned you in the Weekly Forecasts 28/2026, I will now pause the service offering of GnS Economics in its totality for three months. This is for the simple reason that we currently lack the manpower to run our services with sufficiently high quality.

When we started in 2012, as a side-project, there were four people (associates) behind our analyses. We agreed from the get-go that I would be the public face of GnS Economics, but the work behind my writing was always a group effort, with two exceptions. Between November 2022 and June 2023, and since late February 2026, when Mate announced his sudden departure, I have been the sole contributor to our outlet.

Our service offering is simply not meant to be run by one man (or a woman), albeit briefly. I did all the work for the second time, during the spring, and almost burnt out (again) as a result. I also noticed that the quality of our analyses and forecasts started to dwindle simply because I was over-extended. So, pausing our billing until another team is gathered is also just for you. I don’t want anyone paying for our analysis if I consider it sub-par with no clear improvement in insight.

This team-gathering is what I have been engaged in since the start of the year. It has run into some unexpected issues and hence delays, but it is still ongoing. So, GnS Economics will return. There’s no question about that. The only uncertainty concerns the exact timing of our return.

I hope that a new team will be assembled within the next three months, allowing us to resume normal operations. I will keep you posted on all developments.

Like I also noted on Friday, I will transfer all our annual subscribers to my newsletter. You should have started to receive welcome emails. If you do not receive one shortly, reply/write to me here and I will add you.

During the break, in addition to writing my newsletter (which will center on the needs of our investment company), I will concentrate on finalizing my book, on some academic projects, and on deepening my statistical understanding of our forecasting models. I will especially look deeper into our failed U.S. GDP forecasting model but will probably also start to build a recession-likelihood model. I will write about these in my newsletter.

I wish you an excellent fall. Remember to be on the lookout for a re-escalation in the Middle East, an onset of a US recession, and a market crash.

Best,

Tuomas

P.s. In the case of something truly drastic occurring, we will return to comment with limited service offerings also during this pause.

Disclaimer:

The information contained herein is current as of the date of this entry. The information presented here is considered reliable, but its accuracy is not guaranteed. Changes may occur in the circumstances after the date of this entry, and the information contained in this post may not hold true in the future.

No information contained in this entry should be construed as investment advice nor advice on the safety of banks. Neither GnS Economics nor Tuomas Malinen can be held responsible for errors or omissions in the data presented. Readers should always consult their own personal financial or investment advisor before making any investment decision or decision about the banks in which they hold their money. Readers are solely responsible for the risks associated with using this post.

Readers must assess the risks and legal, tax, business, financial, or other consequences of their actions. GnS Economics and Tuomas Malinen cannot be held i) responsible for any decision taken, act, or omission or ii) liable for damages caused by such measures.

The views of Tuomas are his own. They may or may not be endorsed and supported by the partners and staff of GnS Economics.

1. There has been a development in the U.S. bank lending.

2. We look for the (statistical) breakpoints in the U.S. bank lending series.

3. Forecasts indicate a slowing down of U.S. business lending, indicating an approaching downturn.

This week, we will begin a more in-depth analysis of U.S. bank lending data. We introduce you to the concept of breakpoint analysis and how it can affect the long-run dependency and cointegration relationship(s) between variables. Therefore, the will be a rather technical report, which is required for more robust future forecasts.

Before this primer on breakpoint analysis, we update the U.S. bank lending data. They indicate that the bailout of the private credit sector would be ending. Our forecasts add to the analysis by indicating a downturn in business lending and hence the U.S. business cycle ahead.

I also need to give you a heads-up about an announcement I am planning to give at the start of next week. Most likely, I will announce a pause in our services due to the inner upheaval our company is experiencing. I will provide our yearly subscribers with paid access to my newsletter for the duration of the pause. But let’s return to this at the start of the week.

Tuomas

A development

It now starts to look like our hypothesis on the bailout of the private credit sector was correct, based on recent updates to the U.S. bank lending data. On April 10, we noted that:

The private credit sector has been in the grips of a crisis, with, for example, massive withdrawal requests hitting the sector. Thus, we may have been witnessing a de facto bailout of private credit sector through bank lending. What we mean by this is that it is possible that major U.S. banks have been increasing lending to companies to pay off the loans acquired through credit funds (etc.) back to them. This is speculative, but the fact is that especially major banks are in the hook with the credit funds and other entities in the Private Credit space.

We speculated that this should lead the smaller U.S. banks to shy off from Private Credit, while major U.S. banks, who were deeply involved in the sector through the credit funds they created, would need to grow their lending to “bail out” the funds in trouble. With the latest patch of data, reaching the end of July, the trend is exactly what we have been seeing.

That is, the growth of lending by small banks to non-depository institutions (NDFIs) stalled in early March and has turned slightly negative during the past four weeks. Interestingly, also the lending by large banks to NDFIs has started to ebb, while it is also too early to say whether this is a hiccup or a clear break in the trend. However, on Monday, Tuomas noted that the bailout seems to have succeeded, at least as far as the stress in the repo markets is concerned. This is indicative that what we are seeing in the lending data reflects the end of the bailout, but does the observed change in the lending activity reflect something more?

When we observe the monthly business lending and lending to NDFIs, the (likely) changes in trends become more visible.

Based on Figure 2, both business lending and lending to NDFIs are slowing down, with business lending possibly peaking and turning down (our forecasts indicate that this would come; see below). These trends are indicative of a downturn approaching the U.S. economy. However, examining the annual growth rates of business loans issued by U.S. commercial banks reveals a different perspective.

Annualized, the business lending by U.S. banks have seen a very strong increase during the past five months, while monthly growth rates have been falling heavily into the summer. Moreover, business lending of small banks is growing steadily, which is indicative of a robust growth in the SME (small and medium-sized enterprises) sector.

That said, we need to acknowledge that commercial and industrial loans issued by small U.S. banks have had the tendency to rise into a recession and even increase within a financial crisis before ebbing (this trend also applies to large banks). Figure 5 emphasizes the point.

So, what can we make of the above, in addition to the bailout of the private credit sector by U.S. commercial banks, possibly closing its end?

Not much, really. The developments we list above may indicate a continued expansion of the U.S. economy or a recession approaching. To gain some more insights, we need to turn into forecasting.

Break points in the U.S. bank lending series

We now take up breakpoint analysis of the time series of U.S. bank lending (a feature available in EViews 14). The aim of this exercise is to improve the forecasts of US bank lending, which have suffered from forecasting model instability and exploding confidence intervals.

How the breakpoint analysis can help us in forecasting is by identifying the possible (likely) breaking points in the long-term (cointegration) relationship between our variables of interest. As you probably remember, cointegration analysis is about finding a stationary relationship between two or more I(1) nonstationary variables. Essentially, we are seeking linear combinations between two (or more) stochastic trends.

As you probably remember, shocks are naturally a defining feature of unit root processes, but what we are interested in are shocks that may alter the cointegration relationship(s) or “shake” it. This is because if there are breaks in the stationary (equilibrium) relationship between I(1) nonstationary variables, it affects forecasts and their confidence intervals (the error structure). It may bias the forecasts, but the shock, almost certainly, increases the mean squared errors, widening the confidence intervals, thus increasing the uncertainty of the forecasts. Moreover, if the shock breaks the stationary equilibrium relationship, the confidence intervals can start to exhibit “unit root behaviour,” i.e., explode. If we can identify and remove such one-time shocks, we will be able to produce more accurate forecasts (presumably).

That is why we turn to breakpoint testing with regard to the U.S. bank lending series. Bank lending is the driving force of the economy, as it fuelled the growth of small and medium-sized enterprises up until at least Private Credit started its phenomenal rise. This is actually one break point in the analysis that follows.

We explained the basics of breakpoint unit root analysis in Weekly Forecasts 26/2026. Essentially, we can have the test run statistics for two different assumptions about the break or an “outlier”. The first is called an additive outlier, while the second is called an innovative outlier. Simplified: the former tests for a break in the trend or slope of the series, while the latter tests for a break in the autoregressive dynamics of the series. The actual testing procedure is naturally (much) more nuanced, statistically.

Read more

1. Global downturn looks to be here.

2. Forecasting OECD’s leading indicators is complicated by common factors.

3. Three yield curves indicate that the U.S. is very close to a recession.

This week we first update OECD’s leading indicators. They indicate that the downturn in the global business is likely to be upon us.

Then we provide you with something of an introduction to common factors in the analysis of cross-section (e.g., countries) of time series. Their analysis is something that was just emerging when I was in academia (I left late 2017). They are crucial, because left unchecked, they can bias forecasts in a very detrimental way.

Common factors can be analyzed with EViews 14, but the problem we encountered was that not all programs used to analyze them, written by EViews users, work in its latest version (14). Thus, we need to work our way around this issue by using two- or three-step estimation methods. This takes time, which is why we cannot present these Factor-Augmented Vector Autoregression forecasts here.

We end with an update to the yield curves of the U.S. They convey a unified message that the U.S. economy has entered a phase with a high risk of recession.

Tuomas

Has the global downturn arrived?

We can now confirm a few trends that we speculated about in May. Figure 1 presents OECD’s leading indicators for China, the G-20, Japan, four major European countries, and the U.S.1

First, Figure 1 confirms that the downturn of the four major European countries (France, Germany, Italy, and the U.K.) has commenced. Secondly, it shows that the Chinese economy’s downturn has continued, but the decline has slowed in the past two to three months. This is indicative of an approaching increase in stimulus by Beijing and hence another upturn in the Chinese economy. We currently consider that this cyclicality of the Chinese economy reflects Beijing’s efforts to controllably cool and de-lever the economy. Because debt is still growing much faster than GDP, this strategy is not working (see, e.g., the 2024 analysis by Tuomas).

Moreover, the U.S. economy continues on an upward trajectory, while the pace of acceleration is slowing, which is indicative of a downturn ahead. The global business cycle (G-20) has moderated, which indicates an imminent downturn (which can be gradual). Figure 2 puts the recent developments in historical perspective.

The only recent development that “registers” in the historical data is the downturn of China. It is very close to reaching the Ukrainian War bottom (in the spring of 2022), which is (also) indicative that Beijing will shortly ramp up the credit machines once again.

Historically, the global business cycle has followed that of China since the late 1990s. For this relationship to hold, the G-20 economies would need to see a notable slowing down, or, conversely, China would need to see a major economic boom. As the latter is relatively unlikely, we consider that all major economies and hence the world economy are on the brink of a major downturn, which is likely to turn into a recession and even to a global economic depression.

Let’s now turn into forecasting. Like Tuomas noted in the introduction, EViews provides us with an important observation, which is likely to determine how we need to approach the forecasting of the leading indicators in the future.

Common factors in leading indicators

The latest patch of updates from the OECD included major revisions to the data. Figure 3 presents the revised data with our May forecast.2

Read more

1. U.S. inflation collapses, temporarily.

2. More clarity for time series properties of U.S. inflation measures.

3. U.S. inflation looks to re-accelerate unless peace is found quickly in the Middle East.

Weekly Forecasts returns from a summer break with some notable improvements. We acquired the latest version of EViews, a statistical analysis and forecasting program.

The program is an old companion of mine, as I used it for almost a decade while in academia. It is widely used in the economist community because of its relatively simple user interface and programmability. The former means that you can do preliminary data analysis fast, and the latter that you can use it also for in-depth analyses and forecasts.

Essentially, Eviews 14 allows us to incorporate some of the very latest developments in statistical analysis and forecasting into our service offering. It has several features that will not just help us to deepen our understanding about the drivers of the economy and markets but also enable us to produce a whole new array of forecasts. I will return to these in due course.

In this report, Eviews shed some clarity to the question should we model the U.S. inflation series as unit roots or not. “Not”, looks to be the answer.

In June, we saw a collapse in the U.S. inflation rates. Many credible arguments can be presented for this to be temporary, a ‘hiccup’. Our forecasts concur with this, somewhat. Unsurprisingly, the path of U.S. inflation looks to depend on the path of the conflict in the Middle East, like before.

Tuomas

Huge miss in inflation forecasts (of “everyone”)

The U.S. June inflation came in strongly negative. The month-over-month change in headline inflation, i.e., the Consumer Price Index, or CPI, came in at -0.4%. This was the largest month-over-month decline in the index since April 2020, when the US economy was in the midst of its first Corona lockdown. The drop broke through the lower 95% confidence interval of our June forecasts.

The MoM core inflation also came in clearly below expectations, at 0%. It too breached the 95% confidence intervals of our June forecasts, albeit less dramatically.

To our defense, practically no one we know of foresaw such a drastic drop in the U.S. headline (nor a zeroing out of core and supercore for that matter). What drove this unexpected decline was gasoline, which fell by 9.7% in June.

There are many reasons to expect the decline to be temporary, which we detail below in our scenario forecasts. For example, the nowcasting model of the Federal Reserve Bank of Cleveland agrees. Their nowcast actually sees the headline picking up notably by the end of the month already.

Essentially, what we saw in June regarding U.S. inflation was, most likely, a hiccup to the downside. A brief relief in an otherwise complicating inflation picture driven by a (temporary) falls in gasoline (-9.7%), energy commodities (-9.5%), and fuel oil (-9.2%). President Trump also pushed these prices down by threatening oil executives. Considering the environment we are heading towards, these tricks and hence low energy prices are unlikely to last.

Like noted above by Tuomas, we now have some new analysis and forecasting tools at our disposal. Let’s thus now turn our heads towards a deeper analysis of U.S. inflation.

Is (monthly) U.S. inflation a unit root process?

The question we have been pondering, and going back and forth, is whether the process driving U.S. inflation measures is a random walk or not. The new program provides us with some important insights and looks to solve our conundrum.

Read more

1. Oil prices in historical perspective.

2. The stochastic trend in the price of oil.

3. Forecasting the price of oil.

This week, before our summer break, we launch our forecasts for the price of oil. This is directly related to the investment company we have set up, but these forecasts are also something we should have probably done already earlier.

We build our forecasting model using the concept we unveiled last week. The two data series we are using, Brent spot price (European crude) and month-ahead WTI (U.S.) crude oil futures contract, bring some surprises. At the end, we are able to build a model that performs satisfactorily (at the statistical level).

Our first, and preliminary, forecasts indicate a price pressure building below the falling oil prices. We also build a scenario forecast, modeling the combined effect of an oil shock and a U.S. recession. Results are interesting.

Weekly Forecasts will now take a break of four weeks. We will pause the billing for this period, starting on the 1st of July, after our summer offer expires on the 30th. This means A) if you’re a free subscriber, you still have time to reclaim our offer, and B) if you are or become a paying annual subscriber, four weeks will be added to your annual subscription, and C) monthly subscriptions will not be charged during the next four weeks (essentially every subscription is pushed forward for four weeks).

I wish you a joyful and relaxing summer!

Tuomas

Oil prices in historical perspective

As you are familiar by now, we start by visually assessing the time series of Brent (European) Crude and WTI (U.S.) Crude benchmark indices. The data we use (from MacroTrends) includes the Brent spot price (European crude) and the month-ahead WTI crude oil futures contract. In essence, what we use are the physical spot price benchmark for North Sea crude in the European/global market (Brent) and the WTI Crude oil futures contract traded on NYMEX (New York Mercantile Exchange). Figure 1 presents their monthly values from January 1979 (WTI) and from May 1987 (Brent) with major shocks until 24 June 2026.

What we first notice from Figure 1 is the co-movement of the two series. This is, of course, as it should be, because both the spot price and month-ahead futures prices of oil should not drift far apart. However, there was a notable divergence between the two soon after the Great Financial Crisis. This is attributable to the shale oil boom of the U.S., which overwhelmed the local demand and takeaway capacity, leading to the unusually long divergence in the two benchmarks.1

We have also marked the drivers, or shocks, behind major price moves in the respective prices of Brent spot and WTI futures. The second oil shock (note that the U.S. government “decontrolled” the prices only at the end of January 1981) and the Kuwaiti crisis led to notable jumps in their respective prices. However, a major (permanent) increase in the oil prices occurred during the global economic boom that emerged after the collapse of the Dot-com bubble and the recession that followed. This pre-GFC rise in the price of oil is mostly attributable to the rapid growth of the developing economies (especially the rise of China) while global production remained stagnant.2 Other factors behind it were the weakening U.S. dollar and increased financial market speculation.

Oil prices saw a major collapse during the GFC, then recovered, only to collapse in the wake of China’s (hidden) recession in 2015 and the shale oil boom coming into fruition (capacity restrictions easing). The first Corona lockdowns naturally led to another drop in the price, and the commencement of the Russo-Ukrainian War caused a spike in the prices.

Figure 1 also depicts the exceptionality of the current situation in the oil markets. First, both spot and futures prices spiked, as they should in the face of the biggest oil shock ever, but then dropped like a stone. Such a price response would be justified only if the global economy had fallen off the cliff, which it clearly has not done, yet.

There’s quite a lot of evidence suggesting the manipulation of the oil markets. For example, a historical gap has emerged between U.S. gasoline prices and WTI (future) prices due to massive short positions. Traders have also been frequently “burned” by the jawboning of President Trump to push the prices lower (insider-trading claims have also been raised). Manipulation of the oil market is thus the likely culprit behind its ‘uncanny’ behavior.

What we have essentially described above is a structure of shocks affecting the oil prices and therefore their statistical dynamics and the (possible) equilibrium relationship. In other words, the multitude of shocks carrying profound effects in the series of Brent spot and WTI futures is something we need to address in our statistical analysis, to which we turn now.

The stochastic trend in the price of oil

As usual, after the visual inspection, we proceeded with testing of the statistical properties of the two series. Figure 2 presents the autocorrelation and partial autocorrelation functions of the Brent Crude (European) spot.

Read more

1. The price of gold and the U.S. stock market exhibit co-movement.

2. Stochastic trends driving prices of gold and silver and the S&P 500.

3. Forecasts and scenario forecasts for the price of gold, the price of silver, and the S&P 500.

This week, we tackled a research question posed by my partner in the investment company we founded this week. I’ll write more about it when the legal proceedings are completed and the company is fully registered. However, as I have noted a few times, the launch of the investment company will turn our forecasting activity toward the markets. This change is also something some of you have asked for.

The question we took upon analysis was, are the gold price and the U.S. stock market driven by the same market forces, most notably liquidity? This question naturally cannot be answered in any depth within just one research paper (and 10 pages), but we can accomplish a good first look at the topic.

Our results imply that this could be the case. That is, there appears to be a joint stochastic trend driving both the price of gold and the S&P 500. At this point, it is impossible to assess with any clarity whether this trend is “liquidity.” It is also somewhat irrelevant from the standpoint of statistical forecasting (emphasis on the word “somewhat”).

Our forecasts indicate that the price of gold should start to recover, but to be honest, so it has a few times. Whether this time is different remains to be seen. However, the forecasted paths of the price of gold, silver, and the S&P 500 seem to already follow a trajectory of a still-mild oil crisis.

Tuomas

Gold and the U.S. stock market

What we have started to ponder is how big of an effect liquidity has on the price of gold. This led us to the question: What is the relationship between gold and the U.S. stock market (the S&P 500)? Something of a surprise awaited us there.

Figure 1 presents the time series of the monthly price of gold and the S&P 500 stock market index.

The co-movement of the series was striking, and something of a surprise to us. This becomes better visible when we take natural logarithms from the two series, which are presented in Figure 2.

The price of gold rises notably going into the first oil crisis, but this is mostly driven by the collapse of the Bretton-Woods System, which restricted the movements in the price of gold. In the second oil shock, the price of gold was already mostly dictated by the market forces,1 and there we see a “safe haven style” of rise the deeper the world entered into the crisis. We end our dummy variable of the second oil shock in December 1980, while the U.S. interest rates peaked at 19.08% (the effective rate) in January 1981, driven by the extremely hawkish stance of the Fed Chair Paul Volcker. Notably, the price of oil peaked just before our dummy ended, implying that gold investors anticipated rate cuts approaching (inflation had started to cool in May and June 1980). But could it imply something else too?

Unsurprisingly, the S&P 500 fell into the second oil crisis, but it still grew during the crisis. It’s a feature of stock markets to increase in value in an inflation crisis, which the oil shock naturally caused. We do not go into the details of this process now, but you can find more in Prepper’s Bunker.

Interestingly, during the first Corona lockdowns, the S&P500 fell while gold rose, but during the first months of the Russo-Ukrainian War, both fell at first. Now, while stock markets started their recovery in early April and have recently reached new all-time highs, the price of gold has started to recover only recently.

What is also shown in Figure 2 (remember that logarithmic values present a growth path) is that the prices of gold actually fluctuate a lot more than the S&P 500 index. This is a somewhat surprising result, at least to those who advocate gold as a safe haven (stable). In other words, the natural logarithmic transformation of the gold price and S&P 500 shows that, historically and outside crises, you have gained a more stable flow of profit (based on asset valuations only) from the U.S. stock markets than gold (note that our figure lacks a dummy for the bursting of the Dot-com bubble in the early 2000s).

Thus, figures 1 and 2 present us with an intriguing research question. Are the prices of gold and the S&P 500 driven by the same underlying factors?

The recent collapse in the price of gold has raised speculation that liquidity factors, for example, could affect gold prices in a manner similar to their impact on the stock markets. This argument may have some merit if we consider how the logarithmic price of gold behaved during the GFC (it rose before the crisis and then stagnated). If so, this would imply that A) the two variables should be driven by a joint stochastic trend (of liquidity, etc.) and B) the inclusion of this statistical equilibrium (cointegration) relationship should improve the forecasts of both series. Let’s begin by checking the first assumption.

Stochastic trends in gold, silver, and the S&P 500

We have to say that this was rather fascinating research to conduct. The conclusion we ended up with is that there’s a joint stochastic trend driving the price series of gold and S&P 500, but such an equilibrium relationship would not exist between silver and S&P 500. The first hint to the latter conclusion comes from a simple visual inspection of the natural logarithms of the S&P 500 and the price of silver depicted in Figure 3.

Read more

1. The U.S. inflation cools but remains elevated.

2. Forecast and scenario forecasts for U.S. inflation indices.

3. How will the war between Iran and Israel end?

This week, we’ll update our forecasts for U.S. inflation and speculate on how the war between Iran and Israel could end. The conflict has now entered another round of escalation, the heightened likelihood of which we warned about two weeks ago. Effectively, the fate of the U.S. and world economies ‘hangs in the balance’ once again.

We will first go through the latest inflation figures from the U.S. and reflect upon our models. This reflection leads us to return to the vector error correction, or VECM, model, simply because it has the best track record (thus far).

Our baseline forecasts indicate that even if the war ended now, U.S. inflation would remain elevated for the period of the next 12 months. Our scenario forecasts, assuming a deepening oil crisis and a re-worsening of the supply chain shock, paint an inflation crisis. As things currently stand, this outcome looks worryingly likely.

We end with speculation about how the war between Iran and Israel could end. We analyze this because the conflict between Iran and Israel is behind our current predicament. We see two endings to the conflict, whose implications for the world differ radically.

Remember our summer offer.

Tuomas

U.S. inflation cools but remains elevated

U.S. inflation, unsurprisingly, continued on its elevated trajectory, while the month-over-month (MoM) inflation cooled a bit. The Consumer Price Index (CPI) increased by 0.5% (from 0.6%), core inflation by 0.2%, and super-core inflation by 0.3% month-over-month. CPI was just like economists expected, while core inflation came in cooler. The annualized figures were 4.2%, 2.9%, and 3.4%, respectively.

Our last month’s forecasts were 0.3% (0.27) for CPI, 0.3% (0.27) for core, and 0.4% (0.39) for super-core. These were all rather clear misses even though our models did correctly anticipate the direction (cooling).

We are now in a position to assess the accuracy of different forecasting models we’ve been trying out. As you may remember, we have “flip-flopped” between vector autoregression, or VAR, vector error correction, or VECM, models. This has been due to our changing underlying assumptions regarding the statistical properties of the time series. VAR is used to estimate a system of stationary variables, while VECM is used to estimate (cointegrated) (I) non-stationary variables.

Last month we noted that:

Autocorrelation functions (not shown) of the three U.S. inflation measures, CPI, core inflation, and supercore, still show very high persistence, with the autocorrelation of, e.g., the MoM CPI series dying down only after 80 lags (months). It is the characteristic of a unit root series that the memory of the series is “infinite” or very close to it. This “infiniteness” is manifested in the autocorrelation function as a very persistent autocorrelation (that does not seem to “die out”). This is what the autocorrelation function of MoM CPI essentially shows. Moreover, the autocorrelation of core inflation does not fall under the 5% statistical significance even after 100 lags.

However, unit root tests reject the I(1) non-stationarity of all three series with a 5% level of statistical significance (but not at a 1% level of statistical significance). The Johansen Trace Test also indicates that all three series would be stationary.

In other words, statistical tests provide evidence to support both of the hypotheses, i.e., whether CPI, core inflation, and super-core inflation would be stationary or I(1) non-stationary. In such situations, it is best to rely on real-world observations (if they’re available). That is, to observe how the different model assumptions performed in forecasting the future path of the variables in question.

Figure 1 presents our (baseline) forecasts for U.S. CPI from March vs. the actualized values, which we found to have performed the best (out of the baseline models). The March baseline model, assuming no crisis, was (unsurprisingly) unable to forecast the major spike caused by the war in March, but its forecasts for April and May were very accurate.

The problem we had with these forecasts was that they included an error. The CPI and core-inflation data changed from not seasonally adjusted to seasonally adjusted in 2025 (the exact month escaped from us). Despite this, the model was able to produce reasonably accurate forecasts, also short-term. While this accuracy may be just a statistical coincidence (driven by the error), we want to give the VECM modeling framework another go. Therefore, we used it to generate the forecasts presented below.

Outside modeling issues, it is now clear that U.S. inflation is picking up. It is (very) likely that the U.S. inflation will either stay at the current elevated level (if peace in the Middle East is found quickly) or re-accelerate (if not). Our forecasts below emphasize these two scenarios.

Forecasts for U.S. inflation

We now present two sets of forecasts for U.S. inflation. The first is a “baseline” forecast assuming no crisis. That is, the model does not include any crisis dummies. In the second set, we assume that the U.S. economy faces both a continuing oil crisis and a supply chain shock. The latter is a naturally more plausible scenario considering the renewed escalation in the Middle East. Figure 3 presents the forecasts for the three U.S. inflation measures produced by a VECM model.1

Read more

1. Restructuring the forecasting model of U.S. real gross domestic product (GDP).

2. Updated forecasts for U.S. real GDP.

3. A preliminary look at the forecasting of the GDP of the Eurozone.

This week, we’ll dive relatively deep into the realm of statistical analysis. What we will present to you is, first, an update and an extension to our GDP forecasting model. We retested the assumptions behind our model and adjusted it accordingly. This was done to check whether data updates and revisions had changed something, and they had.

It now looks like the “correct” cointegration rank of our system of three equations (variables) would be one, instead of two we used previously. While this may sound like statistical jargon, such assumptions do affect the statistical estimator and hence the forecasts it produces. Still, the “corrected” model finds that the U.S. economy would enter something of a mini-collapse in this quarter.

Because the high-frequency data we have on the U.S. economy does not support such a scenario, we consider it relatively unlikely. However, the model’s primary objective has consistently been to predict turning points, not to accurately forecast the growth rate of the U.S. economy (we have not yet even tried to calibrate it to be accurate). Hence, while the collapse is unlikely to occur now, we wait with interest to see whether this quarter will be the turning point for the U.S. economy.

Lastly, we take a preliminary look into the forecasting of the real GDP of the Eurozone. The construction of the model is met with several challenges, which we start to map here. Thus, we do not provide any forecasts for the development of the economy of the Eurozone yet.

I also like to remind you of our summer campaign, which offers our annual subscription with 20% off.

Tuomas

Restructuring the forecasting model for U.S. GDP

Subscribe now

In this section, we detail how we used the autocorrelation functions and the error correction framework to check and partly remodel our forecasting model. In practice, we sought the smallest number of lags that was able to control for autocorrelation and partial autocorrelation but which also produced an appropriate error correction presentation of the underlying long-run dynamic (cointegration). Per Grok Expert:

This means that if the auxiliary regressions (you can refer to these as “preparatory regressions”) used to assess or uncover the nature of the underlying cointegration relationship are biased, then the estimated coefficient for the cointegration vector (beta) and hence the error correction presentation will also be biased. From Grok Expert:

This is the very basic presentation of the ECT. What the lag structure of the auxiliary regressions affects is the estimated presentation of the last term (vector) beta. In other words, if this estimate is biased, so will the error correction term. See the full summary from the footnote.1

Testing for the correct lag structure in the VECM is not easy or simple. We applied a two-staged method. We observed both the autocorrelation functions and the error correction term (EC) term.

To remind you, error correction terms are such that, if cointegration between two (or more) I(1) nonstationary variables exists, they present the path of the series returning to the equilibrium defined by the (stationary) cointegration relationship. If an EC term, for example, exhibits an upward or a downward trend, or “wanders,” it is either not well defined (incorrectly estimated, see above) or it captures a deterministic trend or stochastic trend process. This feature can also be used to assess whether the lag structure of the (auxiliary) regressions is correct and to assess the rank of cointegration in the system.2

We started by re-running the unit root and cointegration tests for the updated values of the three variables: the U.S. (seasonally adjusted) Real Gross Domestic Product (RGDP), Gross Private Domestic Investments (GPDI), and Loans and Leases in bank credit (LL) from Q1 1947 until Q1 2026. They brought upon one change. The Johansen Trace Test now indicated that there would exist one stationary cointegration vector between the three variables instead of the two it found the last time.

To test this change further, we took natural logarithms from all three series to smoothen their distribution and linearize the series.3 The extreme values (peaks and drops) make it more difficult to find a correct lag structure, as the extreme values usually carry a long-time effect in the series, which can require a (considerably) longer lag structure to control for what would be needed without them. “Over-lagging” leads the model to lose the power to correctly identify statistical restrictions. With the Johansen Trace Test, this implies that the test can, for example, over-identify the number of cointegration vectors.

The Johansen Trace Test, run with the (natural) logarithmic values of RGDP, GPDI, and LL, indicated that there would exist one (stationary) cointegration vector between the three variables. In other words, the test implied that three variables would share the same stochastic trend. Thus, we took the cointegration rank of one as the basis of the VECM.

Next we ran several estimations with the three-variable VECM system, assuming different lag structures. We started with the smallest amount of lags suggested by BIC (Bayesian Information Criterion), which were unable to account for the autocorrelation. We started to grow the number of lags and, at the same time, we observed how the error correction term behaved. If it, e.g., exhibited a trend, we changed the lag structure.4 Figure 1 presents the error correction term of the VECM model we ended up using.5

The error correction term of our model behaves as it should, mostly. It fluctuates around zero without any clear time structure or a trend, while it “wonders” a bit. Unit root tests find it to be stationary, but only barely (the ADF test rejects the unit root in the series at the 5% but not at the 1% level). This warrants further study, later.

The exercise presented in this first section should make our forecasting model more reliable, statistically speaking. The extensive retesting indicates that our model is (more) sound or, more precisely, that the underlying assumptions seem to be correct. This naturally does not guarantee that the model is good, i.e., that the forecasts it produces are accurate, which is only something time can tell.

Let’s now turn to the forecasts our updated model provided.

Forecasts for U.S. GDP

Subscribe now

In practice, the only things that truly change in our model are the assumed rank of cointegration in our system of three equations and the lag structure. These changes do not change the implications of our forecast much, as you will notice below. Note that we do not provide scenario forecasts this time because we want to see how our updated baseline model performs and what comes from the Middle East.

Let’s start by checking how the Q4 forecasts of our baseline model performed. Figure 2 presents our forecasts from February 23.

Like you notice, the realized value of U.S. real GDP went straight through the lower (95%) confidence interval of our forecast. This simply means that our model failed to forecast the short-term development of the U.S. GDP. This has been the tendency of our model since Q3. That is, the model tends to exaggerate the change in the U.S. real GDP, but like we noted above accuracy is something we have not aimed with it, yet. We have to emphasize, though, that our oil crisis and financial crisis scenario forecasts got close to the actualized GDP print (the second estimate for Q1 2026 GDP was $24152 billion). Keeping this in mind, what does the model see coming next?

Figure 3 depicts our updated forecasts using the model constructed above.

The model still sees the Q2-Q3 slump in U.S. GDP followed by a strong recovery. What makes this result so perplexing is that, while the model has gotten the magnitudes of U.S. real GDP growth mostly wrong in the past, it did correctly foresee the acceleration of the U.S. economy during Q2 and Q3. It has also consistently forecasted a slump for the U.S. economy for Q2 and Q3 this year.

This consistency is what perplexes us. Oftentimes one sees an econometric forecasting model to anticipate a down- or upturn but postpones its onset with new or revised data. The prediction of our model of the Q2-Q3 slump of the U.S. economy has not changed through several revisions, some of which changed the GDP figures notably, and through four new data points (Q2, Q3, Q4, and Q1).

However, while the U.S. consumer is hurting, for example, the Aruoba-Diebold-Scotti Business Conditions Index, which we have been using as a basis for our nowcasts of U.S. GDP growth, does not show such a collapse. It is thus likely that our model is exaggerating the magnitude of the downturn. Hence, the real question becomes whether it has correctly identified the downturn starting in this quarter? Based on its consistency, we are keen to think so, but the only way to find out is to wait.

Forecasting the Eurozone real GDP

Subscribe now

Let’s now take a quick first look at forecasting the real gross domestic product of the Eurozone. The macroeconomic variables available for the Eurozone differ a bit from those of the U.S. While the real GDP is essentially calculated the same way everywhere, Eurostat does not currently provide gross private domestic investments. Hence, we rely on gross fixed capital formation, which is a broader measure including, e.g., military spending.

The gross fixed capital formation, or Investment series is “messy,” and there are several options for how to calculate it: unseasonal, seasonal, calendar, etc. We selected a series that is both seasonally and “calendar” adjusted. The latter “smoothens” the effect of national holidays, etc., which are automatically removed from, e.g., the FRED database of the Fed St. Louis. This is the first time we heard about such, and we consider it problematic in the sense that it makes the series more artificial by essentially guessing (statistically) what the effect of these holidays is on the gross fixed capital formation.6

Figure 5 presents the time series of the Eurozone’s real gross domestic product and gross fixed capital formation.

The similarity with the relationship between the U.S. (seasonally adjusted) Real Gross Domestic Product and Gross Private Domestic Investments and Figure 5 is obvious, but the relationship of the U.S. variables looks much tighter. The likely reason for this is that private investments tend to be more productive than government (federal) investments, which are included in the gross fixed capital formation series, and thus contribute more to the GDP (hence the co-movement).

This leads us to the first problem in our forecasting setup for the Eurozone’s GDP. While the ADF (Augmented Dickey-Fuller) test finds that the two variables are unit root processes with a drift, the Johansen Trace Cointegration test does find them to be cointegrated. What this implies is that the two series would be driven by different stochastic trend processes.

The above indicates that we cannot estimate the two variables as a system, because adding two non-cointegrated I(1) integrated processes into the same model leads to a spurious regression problem (more on this sometime later). We could go around this by first-differencing the variables, which removes the stochastic trend process. This would counterproductive for our aim, because it would remove the long-run driving force of the series, i.e., the stochastic trend.

However, the most important thing for you to understand about statistical analysis is that essentially nothing can be proved to be “right or wrong”.7 Hence, more research is needed to know whether the above results hold true. As this report is already several pages long, we’ll leave this for the future.

Share

Disclaimer:

The information contained herein is current as of the date of this entry. The information presented here is considered reliable, but its accuracy is not guaranteed. Changes may occur in the circumstances after the date of this entry, and the information contained in this post may not hold true in the future.

No information contained in this entry should be construed as investment advice nor advice on the safety of banks. GnS Economics, Tuomas Malinen, or other authors cannot be held responsible for errors or omissions in the data and information presented. Readers should always consult their own personal financial or investment advisor before making any investment decision or decision on banks they hold their money in. Readers using this post do so solely at their own risk.

Readers must assess the risks and legal, tax, business, financial, or other consequences of their actions. GnS Economics, Tuomas Malinen or other authors cannot be held i) responsible for any decision taken, act, or omission or ii) liable for damages caused by such measures.

1

Summarization by Grok Expert:

1. Lending to the private credit sector by large U.S. banks accelerates, while lending by small banks stagnates.

2. Forecasts indicate a steady (rapid) growth of lending to the private credit sector (NDFIs).

3. (Preliminary) forecasts imply an acceleration of the bailout of Private Credit by large U.S. banks after a brief “summer pause.”

I am glad to start by announcing our traditional summer sale campaign. This year it runs a little over a month until the 30th of June. We offer an annual subscription to our newsletter with 20% off. I urge you to take this offer up, as exciting things are coming for us during the next 12 months. :)

This week, we’ll update the forecasts for lending by U.S. commercial banks and develop something new. By the looks of it, the bailout of the private credit sector, which we speculated to be occurring in early April, is still ongoing, but there’s also something more into it. We suspect that the data center boom has begun to manifest in the bank lending activities of U.S. banks. The question naturally is, for how long does it last?

Our beginning-of-April-forecasts for business lending and lending to non-depository financial institutions by U.S. banks hit their mark well despite some revisions. We continue with the same model, which foresees continued growth of lending.

We also start to experiment with a model predicting the development of NDF lending by small and large U.S. banks. While just a first (preliminary) attempt, the forecasts provide an interesting speculation on the near-future development of bank lending flowing into the private credit sector.

Tuomas

What is happening with U.S. bank lending?

The lending of U.S. commercial banks has continued strong. Figure 1 presents the loans to non-depository financial institutions (NDFIs) from small and large U.S. banks.1

While the large (top 25 by assets) U.S. commercial banks keep pushing money into the NDFIs, small banks continue to shy away. The growth of lending to NDFIs by small banks has stalled in a way we have not seen since 2022. While there have been deeper drops (like in 2025), the ‘stagnation’ of NFDI lending by small banks is starting to become notable, which feeds into our hypothesis of an ongoing bailout of the private credit sector. Because it is mostly the larger banks that are “on the hook” with Private Credit issues emerging there, they should provide most of the “bailout,” while smaller banks would shy away. This is exactly what we are seeing.

However, there may also be something else going on. Figure 2 presents business (commercial & industrial) loans issued by U.S. banks.

What is now emerging is the growing business lending activity of the small U.S. banks. To emphasize the point, Figure 3 presents the four-week moving average growth of business lending by small and large U.S. commercial banks.

What we observe from the above figure is that in addition to a major increase in the four-week moving average of growth of business lending by large U.S. banks, small banks have also started to increase their lending notably during the past few weeks. This latter part is indicative of growing business activity among small and medium-sized enterprises (SMEs) of the U.S. We consider this to be indicative of the data center boom. For example, Bloomberg noted that data center capital expenditure (CapEx) was close to $750 billion already at the end of March. This is also visible in the U.S. freight, which is growing rapidly currently.

Such developments make it likely that the U.S. economy will not see the plunge our forecasting model has been anticipating since mid-June. Naturally, it would have been impossible for the model to see a major data center boom to manifest during the past few months. That said, we have yet to see the impact of higher gasoline prices and accelerating inflation on the U.S. economy. So, let’s see what comes.

Forecasts for business lending by U.S. banks

At the beginning of April (with data reaching the end of February), we had two “competing” forecasting models. The one forecasted a gradual increase in lending to NDFIs and business lending, while the other forecasted more drastic movements for the series. The difference between these models was the lag structure. The first one used shorter lags, while the latter had a (much) longer lag structure. Figure 4 presents the latter, i.e., the model that foresaw the drastic developments with a longer lag structure.

We now have a “winner,” and it is not the model that produced the above forecast. Figure 5 presents the forecasts of the former early April model with (smallish) revisions to the data.

Read more

1. Developments in U.S. inflation.

2. Oil crisis scenario forecasts for U.S. inflation.

3. Oil crisis + supply chain shock scenario forecasts for U.S. inflation.

I discovered a data error from our U.S. inflation forecasts. For some reason, some of the CPI (Consumer Price Index) data was unadjusted for seasonality, while most was adjusted. About a year ago, we made comparisons with non-seasonally adjusted and seasonally adjusted CPI data. The data error most likely arises from there.

The above means that basically all of our U.S. inflation forecasts have been biased or erroneous in a statistical sense. Although the error is likely small because the same process drives both seasonally adjusted and not seasonally adjusted series, there has been one nonetheless. This discrepancy may also explain why our CPI forecasts have been missing their mark, while our core inflation forecasts have been more accurate.

My apologies.

In this week’s weekly forecasts, we naturally correct this error. Our model changes a bit as a result, as do our forecasts. Our baseline scenario forecasts, assuming no shock, see U.S. inflation cool rapidly but continue at an elevated level for the next 12 months.

Our oil crisis scenario forecasts see the U.S. inflation falling first and then continuing on its upward trajectory. Our scenario forecasts, assuming also a supply chain shock, foresee the U.S. inflation heading to the ‘stratosphere,’ with a projected annualized change in the CPI reaching 17%. The likelihood that the U.S. enters this last “apocalyptic” scenario naturally grows by the day the traffic through the SoH remains strictly limited, and the continuation of the “kinetic” phase of war would practically guarantee its manifestation.

Tuomas

Recent developments in U.S. inflation

Below you find a summarization of U.S. inflation by Grok Expert. It gathered the data from the U.S. Bureau of Labor Statistics, Trading Economics, and CNBC.

— —

The most recent U.S. inflation data is the Consumer Price Index (CPI) for April 2026, released by the Bureau of Labor Statistics on May 12, 2026.

• Headline CPI (all items): Rose 0.6% month-over-month on a seasonally adjusted basis (down from +0.9% in March) and 3.8% year-over-year (up from 3.3% in March). This was the highest annual rate since May 2023 and slightly above consensus forecasts.

• Core CPI (excluding food and energy): Increased 2.8% year-over-year (up from 2.6% in March) and 0.4% month-over-month.

Key drivers included a sharp rise in energy prices (+17.9% year-over-year, largely from gasoline), while food prices rose a more moderate 3.2%. Shelter and other services also contributed to the annual pickup, though some categories like new vehicles and medical care saw monthly declines.

For context on the Fed’s preferred gauge, the Personal Consumption Expenditures (PCE) price index (latest available: March 2026, released April 30) showed headline PCE at +3.5% year-over-year and core PCE around +3.2% (both in line with expectations at the time).

The most recent U.S. Producer Price Index (PPI) data—for final demand—is the April 2026 report, released by the Bureau of Labor Statistics on May 13, 2026.

• Headline PPI (final demand): Rose 1.4% month-over-month on a seasonally adjusted basis (the largest gain since March 2022, up from a revised +0.7% in March and +0.6% in February). On an unadjusted basis, it increased 6.0% year-over-year (the largest 12-month rise since December 2022, up sharply from 4.3% in March).

• Core PPI (final demand less foods, energy, and trade services): Advanced 0.6% month-over-month (largest since October 2025) and 4.4% year-over-year (largest since February 2023).

• A broader core measure (excluding only food and energy) rose 1.0% month-over-month.

Key drivers were broad-based but led by energy and services:

• Final demand goods rose 2.0% (more than three-quarters from energy prices +7.8%).

• Final demand services increased 1.2% (nearly 60% of the overall headline gain), with trade services margins +2.7% and transportation/warehousing +5.0% (e.g., truck freight).

Intermediate demand also showed pressure: processed goods +2.7%, unprocessed goods +4.1% (with unprocessed energy materials +9.2%).

This PPI report shows a clear reacceleration in wholesale inflation—mirroring the April CPI pickup—driven heavily by energy costs but with broadening services and core pressures. It came in well above consensus expectations (0.5% m/m forecast) and has raised concerns about sticky inflation feeding into consumer prices.

In short, inflation reaccelerated in April 2026 after cooling earlier in the year, primarily due to energy costs, though core measures remain lower than headline. Upstream inflation surged in April 2026 after earlier moderation, with both headline and core measures hitting multi-year highs.

— —

So, like we forecasted, the U.S. inflation continued to pick up speed in April, but not by so much as our (partially erroneous) forecasts anticipated. Tuomas already commented on the figures and analyzed the success of our forecasts. We add here just that the big jump in the PPI indicates that inflation pressures are building rapidly again (we are about at the level of May 2021). In other words, this inflation shock is just starting, even if the war would end tomorrow and the Strait would be fully opened.

U.S. inflation forecast: Baseline

Like noted above, we will divide the forecast into two segments. In the first, we present our non-scenario or baseline forecasts. In them, we let the model do its work without adding, or “forcing,” any crisis dummies into it. In the second segment, we build scenario forecasts assuming first an oil crisis and then the combination of an oil crisis and supply chain shock.

Due to the data error, we needed to run our whole analysis again. The fact that the data series changed from not seasonally adjusted to seasonally adjusted in the middle of the series implies that it brought unexplained stochasticity into the model. This may have fooled the unit root tests and altered the autocorrelation function.

Autocorrelation functions (not shown) of the three U.S. inflation measures, CPI, core inflation, and supercore, still show very high persistence, with the autocorrelation of, e.g., the MoM CPI series dying down only after 80 lags (months). It is the characteristic of a unit root series that the memory of the series is “infinite” or very close to it. This “infiniteness” is manifested in the autocorrelation function as a very persistent autocorrelation (that does not seem to “die out”). This is what the autocorrelation function of MoM CPI essentially shows. Moreover, the autocorrelation of core inflation does not fall under the 5% statistical significance even after 100 lags.

However, unit root tests reject the I(1) non-stationarity of all three series with a 5% level of statistical significance (but not at a 1% level of statistical significance). The Johansen Trace Test also indicates that all three series would be stationary.1

We’ve been “dancing” around this topic, i.e., the stationarity vs. non-stationarity of the U.S. inflation series for some time. At some point, we tried the VAR-GARCH models to control for the heteroscedasticity of the series but returned to the VECM framework earlier this year, when we decided to approximate the statistical properties of the U.S. inflation series with a random-walk process. Back then, the results of testing were more supportive of this assumption, possibly due to the error in the data.2

Due to all the hassle, we return to the basics now. We assume that the U.S. inflation series is stationary (like the unit root and cointegration tests indicate). Hence, we apply the vector autoregression model, or VAR, to forecast the U.S. CPI, core inflation, and super-core inflation. We also did some other adjustments to the modeling, which we explain below.

Figure 1 presents our “baseline” forecasts for month-over-month U.S. inflation measures using a VAR model.3

Read more

More precisely, even our standard forecasts, assuming no crisis, paint a picture of the global business cycle heading down steeply from henceforth.

This is naturally no surprise, considering the economic hit manifesting from the Middle East, which is now progressing at an increasing speed. Moreover, our oil crisis scenario forecast presents a warning of a collapse of the global economy. It indicates that in just 12 months (this one included), the global downturn would reach the depth of the Great Financial Crisis. I consider this scenario to be worryingly likely.

A thought appeared after their publication.

We noted in the free-to-read section that China’s business cycle seems to have bottomed out, driven, most likely, by the strong stimulus push Beijing has run since the start of the year. In terms of OECD’s leading indicators, it looks like this:

There are two main takeaways from the figure above. First is the apparent breaking of the strong upwards momentum of the business cycle of the four major European economies (France, Germany, Italy, and the U.K.). Second is the likely bottoming out of China’s business cycle, as noted above.

I started to ponder, how does the renewed Chinese stimulus and upturn affect the global business cycle?

China was lacking from our forecasts because we needed data going back to the early 1970s to properly model the effect of the first oil shock, whilst China’s data in the OECD’s database range only until early 1992. However, like I have noted also here (see, e.g., this and this), China has been carrying the world economy since 2009. To note, I’ve been a “China bear” since late 2017, but Beijing has surprised me, repeatedly, with their ability to avoid the unavoidable, i.e., the crash of the Chinese economy.

Regardless, the fact that China has (effectively) driven the world economy could affect our forecast on the impending global recession. This is because the effect of the renewed stimulus by Beijing, properly incorporated into our model, could possibly turn the forecast around.

So, what I did is that I ran the forecasts with a shorter time-span with China’s leading indicator included. Figure 2 presents the forecasts for OECD’s leading indicators for China, Japan, and the U.S.1

Read more

1. OECD’s leading indicators show a slowing down of the European economy, while China re-accelerates.

2. Our forecasts indicate an imminent onset of a downturn in the global business cycle, which has the potential to morph into a global depression.

It looks like we have entered a turning point. Despite the recent cautious optimism with regard to the U.S. economy, our forecasts paint a gloomy picture for the world economy going forward. More precisely, even our standard forecasts, assuming no crisis, paint a picture of the global business cycle heading down steeply from henceforth.

This is naturally no surprise, considering the economic hit manifesting from the Middle East, which is now progressing at an increasing speed. Moreover, our oil crisis scenario forecast presents a warning of a collapse of the global economy. It indicates that in just 12 months (this one included), the global downturn would reach the depth of the Great Financial Crisis. I consider this scenario to be worryingly likely.

Western economies are dangerously exposed to the economic hit emerging from the Middle East. This applies especially to Europe, but also to the U.S., like we saw just yesterday. I would take heed.

Tuomas

An update to the leading indicators

Latest updates to the OECD’s leading indicators present an acceleration of the global economy and include smallish revisions to the indicators we have been monitoring. Our oil crisis scenario forecasts correctly anticipated the uptick in both the global and U.S. business cycles but missed the strength of the Japanese economic momentum. Our updated forecasts indicate that a global downturn is upon us.

Let’s start by looking at the recent developments in the grand scheme of things. Figure 1 presents the long time series for China, Japan, four major European nations (France, Germany, Italy, and the U.K.), the G-20 countries, and the U.S. from May 1992 until April 2026.1

The updated (and revised) indicators show three major developments. First, the leading indicator of China has turned upwards from a relatively low level, indicating a recovery of the Chinese economy. Considering the major stimulus push enacted by Beijing during the first months of the year, this is no surprise. Secondly, the leading indicators of both the G-20 and the U.S. keep on heading upwards, signaling a strong economic momentum. Thirdly, the strong upward momentum of the indicator of four major European nations has broken. Figure 2 provides a more detailed look at the recent developments, including the indicator for Japan.

From Figure 2, the turning point in the indicator of the major four European economies becomes clear, like the upturn in China’s (revised) indicator. The continued upward momentum of the G-20, Japanese, and U.S. economies is also clearly visible.

What we can deduce from the above is that 1) the downturn of the European economy, which we noted last week, has commenced and 2) that the global economic momentum continues strong still, likely helped by the major stimulus from China. Figure 3 presents our beginning-of-April forecasts run with data that reached February and the revised actualized value of the leading indicators of the G-20, Japan, and the U.S.

Forecasts: A global downturn is upon us

We apply a similar model setup in our forecasts as we did in the beginning of April. We use the vector autoregression, or VAR, model, and the values of the OECD’s leading indicators of the G-20, Japan, and the U.S. enter in their natural logarithms. We use (natural) logarithms to smoothen the probability distribution of the volatile leading indicator series.2

At this point it looks like we are not heading into a “mild” but a “heavy” oil crisis. Our mild oil crisis scenario assumed that the conflict would have been solved by the end of last month. This obviously has not happened, as the Strait of Hormuz remains under “smart control” of Iran. There’s also no endpoint of the conflict in sight currently, because the views of Iran and the U.S. are so far apart. This also implies that the risk of the conflict re-erupting remains high. This is why we use the heavy oil crisis scenario from early April as a basis for the structure of the dummy variables.

Figure 4 establishes a baseline by presenting the forecasts of our model in a standard, i.e., in a no- oil crisis scenario. Contrary to our previous forecasts, even the standard model now sees a major global downturn dead ahead.3

Read more

1. Economic indicators indicate a collapse approaching the Eurozone and Beijing resorting to major stimulus, again.

2. Updates for forecasts for U.S. GDP in three scenarios: oil crisis, oil + financial crisis, oil + financial crisis + supply chain collapse.

We have not updated the economic indicators since mid-February. This is mostly because of the internal turmoil that hit our company in late February and because the structure of Weekly Forecasts has been somewhat unclear. Now I see that the correct structure has finally started to manifest. It will take some still, e.g., because I am actively seeking to expand our contributor base, but we’ll get there.

This week, we establish what will be a permanent section in our Weekly Forecasts. We’ll update the economic indicators for China, the Eurozone, and the U.S. In addition, we update the scenario forecasts for U.S. real gross domestic product.

In the future, this section will also provide forecasts for the Eurozone GDP. This has been our long-standing aim, which has just gotten postponed. This week, one of our paid subscribers wrote to me in hope of seeing scenario forecasts also for Europe. This pushed me into action, but the Eurozone data has proven to be rather complex to analyze. I may report something about it next week.

The uniform message of this week’s forecasts is that sh*t is about to hit the fan of the global economy. Economic indicators are flashing signs of impending economic hardship, while our forecasts continue to anticipate challenging times for the U.S. economy. The global economy is approaching a precarious period.

Tuomas

Economic indicators

United States, April (March)

Richmond Fed manufacturing: 3 (0)

Empire State manufacturing: 11 (0)

Dallas Fed manufacturing: 19 (7)

Kansas Fed manufacturing: 10 (11)

Manufacturing PMI: 54.5 (52.3)

Services PMI: 51.0 (49.8)

Consumer Confidence: 92.8 (92.2)

U.S. leading indicator; November (September): 97.3 (97.9)

Manufacturing indicators improved quite notably across the board in the U.S. in March. Manufacturing PMI reached a high not seen since May 2022, a year that was also marked by war. The gradual improvement of the manufacturing sector is also visible in the new orders index of the sector.

Consumer confidence edged up, but the increase was miniscule in the grand scheme of things. The mood of U.S. households continues to hover in the depths of Corona lockdowns, while expectations continue to sink.

How we see the situation in the U.S. economy is that the domestic policies of President Trump have given something of a boost to it, which our forecasting model anticipated already in June. We consider the recent uptick in the manufacturing sector to have arisen from a war-related boost, as the government is increasing military spending while firms and households prepare for the approaching shortages. We expect this trend to accelerate to a point where the U.S. populace truly awakens to the dire economic consequences of the war. Based on the Services PMI, this has not happened yet.

Eurozone, April (March)

Manufacturing PMI: 52.2 (51.6)

Services PMI: 47.6 (50.2)

Germany ifo Business Climate: 84.4 (86.3)

The manufacturing PMI of the Eurozone continued strong in April, but the services PMI collapsed. The dive of the service purchasing managers’ index was the deepest since February 2021, i.e., during continued Corona lockdowns.

When interpreting manufacturing PMIs in the current predicament, it is important to consider that the supplier deliveries component is reported inversely. What this means is that longer delivery times appear as positive in the index because they are considered to signal a backlog in orders. However, if supply chain issues cause longer delivery times, like now, this sub-indicator generates a false positive signal. Therefore, in the current situation, services PMIs serve as a more reliable indicator of the economic condition, and they are declining in the euro area.

As a likely reflection of things to come, the Business Climate index of ifo Institute collapsed in April.

This collapse, most definitely, is war-related, but unlike in early 2022, not just expectations but also assessment of the current situation among German firms has collapsed. It is a sign that the German economy is about to enter a recession, followed most likely by the Eurozone, due to the supply chain issues stemming from the war in the Middle East.

China, April (March)

Caixin (RatingDog) manufacturing PMI: 52.2 (50.8)

NBS manufacturing PMI: 50.3 (50.4)

Caixin (RatingDog) services PMI: 52.6 (52.1)

China’s economic momentum looks positive in the light of PMIs. This is no surprise, noting the strong stimulus-push Beijing has enacted this year. The flow of aggregate financing to the real economy, depicted in Figure 4, has started this year in the vicinity of record flows.

Like we noted in the April Black Swan Outlook, stimulus is likely to continue strong and even increase this year due to the approaching economic shock, which is already emerging in Europe (see above). This (implied) stimulus implies that China could become the best-performing major economy during this year.

That said, it is truly amazing how the command-and-control grip Beijing holds on the banking sector enables China to keep on building debt into the already massive debt pile. This is something that keeps on surprising us. Standard economic models lost their explanatory power a long time ago in the face of China’s semi-authoritarian economic model, which naturally complicates predictions about the sustainability of its debt levels and the potential consequences of the approaching financial crisis.

Some kind of day of reckoning is still approaching the Chinese economy, because there are limits on the level of debt any entity can hold. Let’s see whether the coming global crisis leads it to emerge.

Updated scenario forecasts for U.S. GDP

Now we have information which of our scenario forecasts performed the best for U.S. GDP during Q1 and the winner was… (drum-roll)… the scenario assuming an approaching financial crisis. This can be just a statistical coincidence and the first quarter of U.S. GDP can also change a lot in the second (and third) estimate, like we have seen recently. Thus, we should not make too much of the first estimate also forecasting-wise.

What is of interest to us, is how the emerging supply chain disruption and possible collapse, in the case the war continues and/or the closure of the SoH (Strait of Hormuz) drags on, affects the trajectory of U.S. GDP. We continue to approach this conundrum through scenario forecasting, by providing you with three updated forecasting scenarios: oil crisis, financial crisis, and supply chain collapse.

Figure 5 presents the results of our forecasting model assuming an oil crisis scenario.1

Read more

1. The bailout of the U.S. private credit sector continues unabated.

2. The two statistical dimensions of U.S. repo markets.

3. Is a third bailout channel of Private Credit operating through U.S. repo?

There is a serious problem brewing in the private credit sector. I have been detailing the (major) issues there since late October.

Just two weeks ago, I explained why U.S. commercial banks were so deeply “on the hook” with the issues emerging in the sector. Three weeks ago, we detailed the “hidden bailout” of the private credit sector by large commercial banks of the U.S.

This week we expose more problems in the U.S. financial system while uncovering another likely bailout channel of Private Credit running through repo. We use time series analysis to uncover the statistical drivers of repo rates. They reveal that there are indeed two sets of drivers at play, like I suspected some months ago there would.

Based on these findings, we extend the financial structure of a credit fund I unveiled two weeks ago. It reveals a similar structure that was behind the financial collapse of 2008. The private credit sector is likely to be much more levered than what is generally understood or acknowledged.

Tuomas

The bailout of the U.S. private credit sector continues

Large U.S. banks have continued to grow their lending to the non-depository financial institutions (NDFIs) in a relentless manner in April. Similarly, smaller banks have continued to shy away from the battered sector.

A similar pattern has continued in business lending. Figure 2 depicts the commercial and industrial loans issued by large and small U.S. commercial banks.

Big banks have continued to increase their lending to SMEs (small and medium-sized enterprises), while business lending of small banks remained stagnant (stable). There’s also an interesting pattern emerging, where the business lending of large U.S. banks grows massively during the first week of the month and then cools off.

At this point, we have no knowledge of what would be the reason behind this, but the growth is notable and highly seasonal. During the first week of February, the business lending of large U.S. commercial banks grew by 1.7% and remained subdued for the remainder of the month. The same occurred in March with a growth of 1.5% during the first week, after which the weekly growth rates fell close to zero or even to negative. The first week of April saw a growth of 1.3%. For some reason, U.S. banks thus look to push loans to businesses at the beginning of the month. Whether this relates to the bailout of the private credit sector, e.g., in some accounting-sense, remains unknown at this point.

It thus now looks that both large U.S. commercial banks and the Federal Reserve are engaged in the bailout of the private credit sector. Massive amounts of money (liquidity) are being pumped into the U.S. financial system through them, which shows in the relentless growth of the U.S. money supply.

Furthermore, there may be another hidden bailout occurring through the U.S. repo market.

The two statistical drivers of U.S. repo markets

In the April Black Swan Outlook, we took notice of the repo markets and the Federal Reserve concealing a “black swan” in the private credit sector, continuing the analyses by Tuomas.1 Here we aim to deepen our understanding of the statistical drivers of U.S. repo markets, as it will help us comprehend the nature of the stress involved. Bare with us.

Let’s start by looking at the rates of the Triparty General Collateral Rate (TGCR), the Secured Overnight Financing Rate (SOFR), the Broad General Collateral Rate (BGCR), and the Interest Rate on Reserve Balances (IORB). Figure 3 reveals three important observations on them.

Read more

1. Iran in the ‘Grand Eurasian Chessboard.’

2. Scenario I: The Israel-U.S. bloc win the war against Iran.

3. Scenario II: Iran Victor.

4. Conclusion: “The battle for the world.”

The general misunderstanding within the vast majority of global populace as well as analyst community is that the war against Iran would be simply about the control of Iranian resources. In actuality, and at the fundamental level, it is a battle of the control of Eurasia. Like I noted, when the war started, this is the “first great battle of our time.”

The war is essentially about the survival of the U.S. global hegemony vs. the multipolar world. This rather clear distinction arises from the major role the Middle East holds in the world. This is a war the U.S. has not fought since the Second World War, after which it cemented its role a global superpower.

However, like we noted in the ‘Grand chessboard revisited: the Iranian play’ to fully grasp what is at stake in the Middle East, requires scenario analyses. That’s what we set out to do this week.

Below, you find two different end-scenarios for the Iran - U.S. (Israel) war. In the first, the U.S.-Israeli bloc is victorious, while in the latter Iran prevails. Expectedly, the outcome of these scenarios, for the world in general, differ greatly.

Tuomas

Iran and the ‘Grand Chessboard’

Before entering into the scenarios, we shortly summarize the role of Iran in the ‘battle for Eurasia.’ We can consider everything starting from the miscalculation of both the capabilities and the role of Iran.

In the Weekly Forecasts 12/2026, we noted the following:

What Zbigniew, quoted above, probably through some of the miscalculations by the U.S. intelligence/political scientist, failed to see in his masterful analysis was the rise of Iranian military might and determination during the 2000s and 2010s. This is now changing the Eurasian ‘chessboard’ through an Iranian play.

The person whom we referred to was Dr. Zbigniew Brzezinski, a political scientist who served as National Security Advisor in Jimmy Carter’s cabinet. He published The Grand Chessboard: American Primacy and Its Geostrategic Imperatives in 1997, where he outlined the U.S. strategy in the world, which concentrated on controlling the Eurasian landmass through Europe.

What we added to this in the Introducing the Grand (Unified) Geopolitical Theory was war. We outlined a ‘militarized strategy’ of the U.S. in Europe, following the “moves” envisioned by Zbigniew.

All NATO needed to do was attempt to expand into Georgia and Ukraine, which would compel Russia to respond. Pushing these efforts far enough would trigger a military response. It is likely that the Russia-Georgia war, fought in August 2008, was a test for the Pentagon and NATO leadership to see whether the strategy would work, and it did. It implied that Russia would respond to a similar threat in Ukraine with force. Moreover, when such a point was reached in Ukraine, all Washington needed to do was launch a propaganda campaign labeling Russia as an imminent threat to “freedom” and to Europe. With the help of selected European leaders, the campaign would, most likely, bring about a separation of Europe from China and Russia.

This was where Eurasia found herself in the morning of the 27th of February. However, after that almost everything has changed.

From the Weekly Forecasts 12/2026:

What we like to detail here is the way the likely victory of Iran is likely to alter the “chessboard” of Eurasia. Drawing a complete picture of what we’re heading (hurtling) towards naturally requires scenario analyses to which we’ll return later. At this point, it is important to acknowledge that the main Eurasian pieces are still not in play, fully, in the Middle East. These include the European assets, which we do not consider to play a major role, unlike China, North Korea, and Russia, which we expect to make a notable contribution, if the ground war commences.

At that time, March 26, we were expecting some form of ground campaign against Iran. This actually most likely occurred, like noted by Patricia Marins just this week:

On April 3, Iran rejected a 48-hour ceasefire proposal delivered by intermediaries, arguing that the United States had badly misjudged Iran’s military capabilities.

Just two days later, the United States and Israel attempted an incursion with hundreds of special forces into Iranian territory. The troops were targeted by Iranian air defenses, forced into an emergency landing, and subsequently destroyed. Iran had defeated an attempt to invade its territory against two nuclear powers. This event certainly alarmed the White House, which was planning a major incursion into the Isfahan complex with extensive air support and hundreds of special forces. Even if they could not seize the uranium, they would bomb the site again and claim that the uranium had to be destroyed in the process. This is how unachieved victories are drafted in conflicts.

Iran repelled this incursion and may also have repelled a second attempt just before the White House started its push for a ceasefire (more on that another time). In other words, Iran has been able to repel the combined military might of Israel and the U.S., which can be considered a major achievement by itself.

Moreover, a coalition between China, Iran, and Russia is forming, which is something Zbigniew explicitly warned against in The Grand Chessboard (p. 116).

However, a coalition allying Russia with both China and Iran can develop only if the United States is shortsighted enough to antagonize China and Iran simultaneously.

This is what President Trump has managed to create with a likely heavy cost for the U.S.

We are now in an extended ceasefire, which we do not expect to last simply because so much is at stake. For the stakes to become visible, we need to formulate two scenarios with opposite sides winning.

Scenario I: The U.S.-Israel axis emerges victorious

In this scenario, which we currently consider the least likely one, we assume that both covert and symmetric (kinetic) warfare employed by the U.S. and Israel leads to a defeat of Iran. This enables them to extract heavy concessions from Iran in the peace negotiations.

While we do not assume that the U.S. and Israel would be able to enact an actual regime change (effectively instating Mohammad Reza Pahlavi), they would be able to create enough internal disorder for some type of “cannibalization” to occur in the Iranian leadership. This would lead to Balkanization of Iran down the line.

Read more

1. Geopolitical risk forecasts for Europe (by Sakari Linden).

2. U.S. inflation exploded in March.

3. A near-worst-case scenario for U.S. inflation.

This week we’ll concentrate on U.S. inflation and provide you with a “near-worst-case” scenario for it, which at the time of writing looks scarily likely. However, we begin with a geopolitical risk forecast of Europe by Sakari Linden.

Sakari is a Belgium-based geopolitical analyst who is a former political advisor in the European Parliament. He outlines the worrying trajectory Europe is on, concerning the stance of European leaders towards Russia. He argues that, at the current trajectory, a war between Europe and Russia looks inevitable. Moreover, he argues that, again, at its current trajectory, the war would erupt between this year and 2028.

Then we move into U.S. inflation. There was something of an explosion in the U.S. Consumer Price Index (CPI) in March. It was actually so big that even our supply chain collapse scenario was unable to capture it fully.

The “error” in our previous scenario forecasts for U.S. inflation was that we did not combine the oil crisis with the supply chain collapse. We also did not correctly model the severity of the current shock. In this report, we set out to fix these issues.

Our modeling implies that in the case of an extended closure of the Strait of Hormuz and the “hot war” restarting and continuing for a few more weeks, U.S. inflation would reach double digits by September. At the end of the forecasting period (March 2027), U.S. inflation would break a record with a 17% annual change in the CPI.

Tuomas

Geopolitical Risk Forecast: Europe slides towards a continental war

By Sakari Linden

Ukraine has been in crisis since February 2014. The conflict escalated after Russia invaded Ukraine on February 24, 2022, after Ukraine’s armed forces started a massive attack against rebels of two Russian-speaking regions of Ukraine, Donetsk and Lugansk, from February 16, 2022, onwards. Currently, in April 2026, Russian troops occupy about 20% of Ukraine.

Geopolitics can be used in order to explain past events and to forecast future developments. Both the western powers and Russia have good reasons not to give up their grip on Ukraine. On the one hand, Anglo-American theories of Eurasian geopolitics put emphasis on the containment of Russia, the primordial interest of the US to stop a coalition between Germany and Russia, and the goal to isolate Russia from Eastern Europe.

On the other hand, the Northern European Plain was historically used by European powers to conquer Russia. This historical fact has affected deeply Russian way of thinking of its security. It also explains Russia’s determination to act in its western neighborhood.

Russian leaders have set their hopes in the “Spirit of Alaska” i.e. potential for US-Russian rapprochement after presidents Trump and Putin met each other on August 15, 2025 in Anchorage, Alaska. Unfortunately, historical patterns and current geopolitical dynamics do not favor a peaceful solution to the conflict.

The problem from the point of view of the “Spirit of Alaska” is that the question is not only about Ukraine. There is a much bigger battle going on at the world level between the unipolar world model based on the hegemony of the United States, on the one hand, and the emerging multipolar world model promoted mainly by China and Russia, on the other. The US has started to defend aggressively its hegemony against its challengers by launching wars in Venezuela and Iran and targeting China and Russia’s maritime energy routes.

The status quo does not satisfy anyone on the battlefield. The EU cannot admit a defeat in the war that has already become one of its own. The defeat would result in the EU losing face and authority in front of its member states, which would release the centrifugal forces of the EU system and thus start the break-up process of the EU. Russia, for its part, is seeking better protected borders against the western threat, which the current frontline does not provide. Russia has an incentive to reach for the Dnieper River and eventually for Odessa, which would enable Russia to cut Ukraine off from the Black Sea coastline.

While it is true that the US does not have sufficient resources to wage war in all three global theaters of war, in Europe, the Middle East, and East Asia, anymore, it does not mean that the US would be willing to decrease the level of pressure against Russia. In the new burden-sharing model between the US and the EU, the EU invests significantly in its military capabilities, but in a way that implements the geopolitical goals of the United States. The EU’s role in this new equation is to exert military, economic, and political pressure on Russia and to strive to bring Ukraine into the sphere of influence of the Collective West.

The presence of continued pressure can be seen in Ukraine too. Rather than finding a peaceful solution to the conflict, the Western powers have preferred escalation in Ukraine. Examples of escalation on the Western powers’ part are Ukraine’s Western-backed drone attacks against Russia’s strategic bombers in June 2025, as well as against President Putin’s residence in December 2025.

Recently, Ukraine has used airspaces of the Baltic States to attack Russian targets. The UK is also threatening to start to seize Russian ships at the seas. These are the most potential ways that could escalate the conflict and expand it into a continental war between the EU and Russia.

Moreover, the recent victory of the Tisza party in the Hungarian elections will likely remove obstacles from EU’s further financial support to Ukraine, as well as from Ukraine’s EU membership, which could escalate the war in Ukraine.

Ukraine is at the center of great power interests to such an extent that restoration of peace seems highly unlikely. The US fights for its superpower status, and the EU needs outside enemies in order to preserve its unity, to avoid break-up, and to further its common defense plans. Therefore, a direct war between Russia and the EU looks to become inevitable between 2026 and 2028.

U.S. inflation goes “there”

We saw a massive month-over-month jump in the U.S. Consumer Price Index in March. Figure 1 depicts this 0.9% jump and March figures of core inflation and super-core inflation alongside our oil crisis scenario forecasts.

We can safely assert that our modeling structure, assuming just a repetition of the 1970s oil crisis, failed to account for the severity of the current shock. To our defense, the massive 0.9% MoM jump in the CPI did fit within the 95% confidence intervals of our oil shock scenario forecasts (just barely).

We get much closer of the realized figure of the CPI, when we compare it to our supply chain collapse scenario. Even that scenario was not fully able to account for the massive jump, though, while in this scenario the actualized is safely within the 95% confidence intervals (the point estimate of our scenario forecast for March CPI MoM change was 0.55%).

Read more

1. Developments in lending to the Private Credit.

2. Forecasts for U.S. business lending and lending to NDFIs.

This week we concentrate on Private Credit. Yesterday, I published an entry that detailed the “hidden” explosive growth of loans to the non-depository financial institutions, which consist mostly of credit funds and other “creatures” of the private credit sector. I actually came across the data while I was writing this.

What we uncover here are notably developments occurring in the lending of U.S. banks. We consider them to be symptoms of an ongoing bailout of the private credit sector by major U.S. banks.

Our forecasts present two paths ahead for U.S. bank lending. The other is more lenient, while the other presents a shock approaching the sector, which could lead to tighter lending standards and increased risk for borrowers. At this point, we cannot fully assess which is more reliable. The “shock model” has better statistical properties, but the scenario it paints is extreme.

Patricia will take a break from writing to us this week, while I report our recent email discussion later today. Another geopolitical expert will soon join us to write semi-regularly. I will introduce him when his first piece is published.

Have a great weekend!

Tuomas

Are there signs of cracks in lending to the U.S. private credit sector?

We start our analysis by assessing the recent developments in lending of U.S. commercial banks. They show clear developments or changes from our last comprehensive look at the sector at the end of November. Let’s begin by assessing the development of lending to non-depository institutions by U.S. commercial banks.1

Lending by U.S. commercial banks to the NDFI’s has continued unabated since our last check in November. There has been a smallish slowing down in the lending activity during the recent weeks, but nothing drastic. However, business loans issued by U.S. banks reveal a very interesting recent development.

From late January on, the business lending activity of U.S. banks saw a massive increase in the magnitude we have not seen since June 2022 (when U.S. inflation peaked).2 This is interesting and could be speculated to indicate a notable pickup in business activity of U.S. companies. However, the devil is in the details also on this one. Figure 3 presents the split of business lending between large and small U.S. commercial banks.

Figure 3 shows that most of the growth in business lending has come from large U.S. banks, while the growth of the business lending of smaller banks has remained steady. If we were to expect a notable pick-up in U.S. business activity, we should see the business lending of smaller banks increase because they fund especially the local small- and mid-sized companies, SMEs, which are the backbone of the U.S. economy. We are not seeing this. So, what is going on?

Read more

1. The divergence in OECD’s leading indicators between China and the rest of the word keeps on growing.

2. Two oil crisis scenario forecasts paint an abysmal picture for the global economy going forward.

Essentially everyone (“and his uncle”) is trying to assess the effect of the closure of the Strait of Hormuz and the destruction of oil and gas infrastructure in the Middle East. Analysts’ estimates range from “apocalypse is coming” to “the president got this under control.” Like we have noted several times, the latter of these is most likely to be badly misguided, while we’ve been cautiously leaning towards the former. Cautiously, because there’s a clock with this war. The longer it lasts, the worse the economic hit will be.

This Easter week, we continue to assess the effect statistically. Two weeks ago, we assessed the effect on the U.S. economy. Here we analyze the effect of the emerging energy crisis on the global business cycle with the help of the OECD’s leading indicators and statistical scenario modeling.

Our forecasts imply a drop in the global economic activity rivaling that of the Global Financial Crisis approaching. Before moving into the actual scenario forecasts, we update the recent developments in OECD’s leading indicators.

Tuomas

The divergence diverges, again

The divergence in OECD’s leading indicators between China and the “rest” kept on widening in February. The business cycles in Japan, four major European nations, the G-20 countries, and the U.S. have now headed in opposite directions from that of China for about a year.1

The months, when this divergence from the business cycle of China started, are as follows:

• January-February 2025 with four major European nations;

• May 2025 with Japan;

• June 2025 with the U.S. and the G-20.

We have speculated on the source of this divergence several times (see, e.g., this) and will not continue with that here. Our updated forecasts (not reported) indicate a faster convergence of the business cycles of China and the G-20 than our previous forecasts from February. However, it is unlikely that the model is capable of seeing the oil and likely supply chain shocks that are currently approaching. For this, we need scenario analysis and forecasting.

The effect of an oil shock to the global business cycle

What we are going to do next is to build two scenarios to model the effect of the oil or energy shock to the global business cycle. Like previously, we use the information from 1970s oil shock for our scenarios. In other words, we harvest information on the effect of the 1970s oil shock on OECD’s leading indicators to forecast their behavior in the shock that is now emerging.

Some of you may question the plausibility of this approach. That is, some may wonder how information from 50 years ago would help us understand what’s going to come next. This is where the “magic” of statistical models and statistical information comes into play.

The statistical model we are using (VAR) assimilates both the information from the effect of the oil shock in the 1970s and the development of the leading indicators since. It does not just mechanically add the information from the 1970s into now but brings the information forth into the model and thus “today” through the dummy-variables.2 The dummies store the effect of the 1970s crisis, and the model mirrors the information they hold against the most up-to-date information (data) to construct the forecasts.

The problem we have with the OECD data is that not all the country series of leading indicators reach the beginning of the 1970s. This forces us to exclude China from our analysis, because its data reaches only the early 1990s.

To keep the model simple, we concentrate on assessing the effect using three leading indicators: Japan, the G-20, and the U.S. We include them with the assumption that the “Japanese miracle” and the U.S. as the leading economy of the world before the GFC were the main drivers of the global business cycle approximated by the leading indicator of the G-20 in the 1970s.3

Figure 3 presents the leading indicators of the two countries and the G-20. We mark October 1973 as the starting month of the first oil crisis following the onset of the Yom Kippur War and the oil embargo in that month. We link the end of the crisis to the official end of the U.S. recession in March 1975. The effects of the oil crisis, naturally, continued beyond March 1975, but we believe that this period contains the most relevant information for our analysis of the economic impact. We can also see from the the figure below that the world economy had already started to slow down before the oil embargo hit. However, we consider this to have been a part of normal business cycle fluctuation (and not, e.g., due to expectations of the war approaching) and thus exclude it from our analysis.

Read more

1. Grand (Unified) Geopolitical Theory in the Middle East.

2. After 26 days of war, most of Iranian military arsenal is still intact (by Patricia Marins).

3. The Iranian play altering the Eurasian “chessboard.”

We are approaching a crucial turning point in the war in the Middle East. First President Trump issued and then backed out of his 48-hour deadline based on good news on “negotiations” that never happened. Now, we are running with another deadline, which ends on Saturday. In the worst case, this deadline will mark the beginning of a new Vietnam War with dire consequences for the world.

This week, we’ll provide you with information on both the larger play behind the conflict in the Middle East and updated information on the capabilities of Iran. We start with the extension of my Grand (Unified) Geopolitical Theory to the Middle East. It shows how President Trump’s administration is failing in a crucial long-term play for the U.S. I assert this to occur only due to the incompetence of his administration.

In the second piece, Patricia will detail and update the military capabilities of Iran. Her analysis shows that any ground assault by the U.S. is likely to not just fail but to lead to a high number of casualties, especially on the U.S. side.

We end with a notion of how the war against Iran is altering the global geopolitical “chess board.” What we are currently watching can very well turn out to be the end of U.S. military and financial hegemony, like we noted last week. This is also the main reason why the war is likely to continue for some time still; empires never go down quietly nor peacefully. We end our report with some hedging recommendations.

We are also extremely glad to announce that Patricia has agreed to become our weekly contributor :). Her insightfulness will provide critical information on the rapidly changing geopolitical landscape.

Tuomas

Implications of the Grand (Unified) Geopolitical Theory for the Middle East

In January we introduced the Grand (Unified) Geopolitical Theory Tuomas has been gradually constructing since his revelation about the Russo-Ukrainian War in September 2022. The main building block of the theory has been the observation that it was the superpower-structure, in other words, the battle between the U.S. (NATO) and the Soviet Union (Warsaw Pact) that fostered the long period of calm in Europe. When it ended, we entered into the period of never-ending wars run by the U.S. military machine.

The collapse of the Soviet Union led to the failing of the superpower-structure, which had dominated the world since the Second World War. Many still consider that the creation of the supranational-structure, enshrined in the supranational entities, like the United Nations and the International Monetary Fund, was behind the long period without major wars in Europe and relative calm elsewhere. However, recent developments reveal that it was actually the superpower-structure that enforced the period of (relative) peace and prosperity in Europe and the ability of the supranational structure to function (somewhat) properly in the world. The failure of the one superpower, the Soviet Union, caused the ambitions of the other superpower (the U.S.) to be unleashed, leading us to the current predicament.

The breaking of this proverbial military “stalemate” created a condition for ever-increasing U.S. influence in the world. At first it seemed like liberty, but the reality turned out as something very different. Wars started to appear across the Western hemisphere, including, for example, the first Gulf War (1991); the first Chechen War (1994-1996); NATO’s (terror) bombing campaign of Serbia in 1999; the Global War on Terrorism, launched after the September 11 (2001) terrorist attacks; and, finally, the still ongoing Russo-Ukrainian War, launched on February 24, 2022 (or 17?), and war against Iran. Taking them separately, the wars could be viewed by some as wars of “freedom” or “security,” but with a high likelihood, all the wars since the collapse of the USSR are just pieces in the (grand) geopolitical chessboard. Let us now elaborate on this in more depth.

Dr. Zbigniew Brzezinski, a highly influential political scientist who served as National Security Advisor during Jimmy Carter’s presidency, envisioned a roadmap for U.S. global dominance after the fall of the Soviet Union. His “grand plan” was published in his magnum opus: The Grand Chessboard: American Primacy and Its Geostrategic Imperatives, in 1997. One passage from the book basically laid out the plan in Europe, which we outlined in January. It also links Europe to the “fate” of the Middle East.

It follows that a wider Europe and an enlarged NATO will serve well both the short-term and the longer-term goals of U.S. policy. A larger Europe will expand the range of American influence-- and, through the admission of new Central European members, also increase in the European councils the number of states with a pro-American proclivity-- without simultaneously creating a Europe politically so integrated that it could soon challenge the United States on geopolitical matters of high importance to America elsewhere, particularly in the Middle East.

What are these “geopolitical matters of high importance” to the U.S. in the Middle East? Dr. Brzezinski details them as the following:

Eurasia is the globe’s largest continent and is geopolitically axial. A power that dominates Eurasia would control two of the world’s three most advanced and economically productive regions. A mere glance at the map also suggests that control over Eurasia would almost automatically entail Africa’s subordination, rendering the Western Hemisphere and Oceania geopolitically peripheral to the world’s central continent.

This is to say that to control Eurasia, with its vast mineral, economic, and population resources, is to control the world. The map does not lie.

After the fall of the Soviet Union, Dr. Brzezinski asserted the U.S. leadership should concentrate its power to control this landmass:

This huge, oddly shaped Eurasian chess board -- extending from Lisbon to Vladivostok -- provides the setting for “the game.” If the middle space can be drawn increasingly into the expanding orbit of the West (where America preponderates), if the southern region is not subjected to domination by a single player, and if the East is not unified in a manner that prompts the expulsion of America from its offshore bases, America can then be said to prevail.

With the “middle space,” he means Europe, with the “southern region” he refers to the Middle East, and with “East” he refers to China and her neighboring countries, including Russia, as her landmass essentially covers the width of the “chessboard.” These were the developments Dr. Brzezinski noted the U.S. should work against.

But if the middle space rebuffs the West, becomes an assertive single entity, and either gains control over the South or forms an alliance with the major Eastern actor, then America’s primacy in Eurasia shrinks dramatically. The same would be the case if the two major Eastern players were somehow to unite.

Finally, any ejection of America by its Western partners from its perch on the western periphery would automatically spell the end of America’s participation in the game on the Eurasian chessboard, even though that would probably also mean the eventual subordination of the western extremity to a revived player occupying the middle space.

All the actions after the fall of the Soviet Union up until the current war against Iran can be seen as, more or less, calculated actions to establish the U.S. hegemony over Eurasia (and thus the world) following the advice of Dr. Brzezinski. Like explained in the first part of the Grand (Unified) Geopolitical Theory, the EU and especially the war in Ukraine were, most likely, “moves” to establish full control of the Middle (Europe). The main aim, as somewhat covertly recommended by Zbigniew, was to destroy the power of France and Germany to exert their influence over Europe and North Africa. In other words, to “neuter” them through the EU. Considering the speed Europe has sunk itself in the wake of the onset of the Russo-Ukrainian war, we can consider these resounding successes from Washington, until now.

The process the administration of President Trump has put in motion with Iran, which we detail below, goes completely against the advice of Dr. Brzezinski. This is probably because of the hubris that seems to have taken over the Trump administration as well as Israeli, neocon, Zionist, and even Evangelical factions affecting it.

Iran is currently altering the ‘chessboard’ and emerging as something much more than just a “geopolitical pivot,” like she was labeled by Dr. Brzezinski.1 We can expect even more, because Iran still has major firepower in store.

Iran: 26 Days of War and a Still Largely Operational Arsenal

By Patricia Marins

After almost one month of war, Iran’s military arsenal remains largely operational across air, land, and sea domains. This stands in stark contrast with the narrative touted by the White House.

Read more

1. The U.S. inflation picture in February.

2. Forecasts for U.S. inflation assuming no oil shock scenario.

3. Forecasts for U.S. inflation under an oil shock and supply chain collapse.

The war in the Middle East rages on. Patricia published an excellent analysis on the worrying trajectory of the conflict yesterday. If you have not, I urge you to check it out. It is free to read.

This week, we’ll continue to assess the economic effects of the war, concentrating on the U.S. Most of the effects arise from the closure of the Strait of Hormuz. Writing about it now is even a bit surreal considering that I warned of the (dire) economic effects of the closure already in October of 2023. The interesting recent twist in it is that “Petroyuan” can break the closure relatively soon, as it looks like Iran allows passage to ships that stay away from the conflict and buy their oil with Renminbi. This, i.e., the beginning of an end of the petrodollar, is essentially what we speculated would occur last week.

Another positive development is that rumors of serious infighting within the Trump administration have started to appear. According to them, Vice President James David Vance would be strongly opposing the continuation (and deepening) of the war. My current thinking is that we could see an end to the conflict by the end of April. Let’s see (I’ll detail them later).

Regardless, an economic shock is already on its way and the only question is, how bad will it get? Reports of high prices at the pump in the U.S. and even fuel shortages emerging in Asia have started to emerge. The path of inflation is dictated by heavy uncertainty.

We will provide a path through the uncertainty with the help of two scenario forecasts detailing the possible paths of U.S. inflation for the next 12 months. We start with standard forecasts, which include no specific information on the approaching shock. These forecasts act only as a benchmark, as the shock is already emerging, and it will get worse, even if the war ends tomorrow. The more serious scenarios encompass an oil (energy) shock and disruption of global supply chains. Naturally, the latter of these paints a very detrimental path for the U.S. economy going forward. Our assessment is that, if the war continues, we will start to move into the supply-chain collapse scenario after next week.

Tuomas

The state of U.S. inflation

Let’s first check how our inflation forecasts from the past month fared. In February, the Consumer Price Index (all items, all urban consumers, U.S. city average, seasonally adjusted), or CPI, rose by 2.4% annualized, 0.3% month-over-month (MoM). The core inflation rose by 2.5% annualized and 0.2% MoM. The “super-core,” i.e., services less energy costs, rose by 2.9% annualized, 0.3% MoM.

Our forecasts, from 17 February, indicated that the CPI would increase, month-over-month, by 0.3%, the core inflation by 0.4%, and the super-core by 0.3% in February. In other words, our forecasts accurately predicted the rate of change of CPI and supercore inflation, but they failed to foresee the deceleration of core inflation. Every realized figure was within the 95% confidence intervals, though. Figure 1 presents our February forecasts and the realized value for the MoM change in core inflation with 95% confidence intervals.

Given the accuracy of our forecasts for the CPI and super-core inflation, it’s plausible that the deviation from the core is merely a random statistical event. What this means to say is that no forecast can be assumed to be 100% accurate, but there’s a stochastic variation in the forecasts based on the assumption of the underlying probability distribution. All statistical forecasts are of the form: estimate +/- the margin of error, where the “estimate“ or “point estimate” is the term for the actual forecasted value and the margin of error is the assessment of the model on the fluctuation of the forecast based on the underlying distribution and the predetermined threshold. In economic sciences, 95% is generally used as a threshold for “statistically significant” observation (and forecast), while, for example, in physics, much tighter thresholds, like 99.999%, are oftentimes used. This difference arises from the fact that stochasticity (randomness of observations) in measured physical phenomena is generally much smaller than in economics. Just think of it as a difference between the actions of men vs. laws of physics. Thus, at this point we do not make any major changes to our forecasting model, because the miss in our core inflation forecast can be just random statistical variation.

But where is U.S. inflation heading? Figure 2 presents the recent monthly changes in the three key inflation metrics.

What we have seen, since the shock the onset of the war in Ukraine created in 2022 started to ease about a year ago, is a stabilization after a relatively long period of heavy fluctuation. The problem is that this is about to change abruptly due to the shock arriving from the Middle East. We do not know its length and depth, though, because it is determined by political decisions of leaders concerning the Gulf region and their impacts on global markets. We naturally specialize in addressing such uncertainty through scenario forecasting. Let’s first provide the “baseline” by updating our standard inflation forecasts.

Forecasts for U.S. inflation in the no-oil-shock scenario

We extend the forecasting model from February to also include the (monthly) supercore inflation, following the update (and correction) made by Tuomas. This modification diminished the mean squared errors of the model somewhat, which implies that the super-core is providing information that is helpful in modeling and forecasting the system of three variables: CPI, core inflation, and supercore inflation. Otherwise, we use the same setup as in February.

We approximate the monthly U.S. CPI, core inflation, and super-core inflation series with a stochastic trend process. That is, we assume, based on the results of the Augmented Dickey-Fuller (ADF) test indicating that the series contain unit roots, that the series are driven by stochastic trends. The Johansen Trace Test with updated data indicates that two stationary cointegration vectors would exist between the three variables (using a 5% level of statistical significance). We then applied the Vector Error Correction (VECM) to estimate the model and forecast the path of the three inflation measures for the next 12 months.1 Figure 3 presents the forecasts it yielded.2

Read more